LIBRARY

; fffii’ESiS MlChigan State
Univcrsuy

 

This is to certify that the
thesis entitled
POSSIBLE SYNTHETIC PATHWAYS TO MODELS
OF THE METAL COFACTORS OF NITROGENASE

presented by

Robert Hugo Tieckelmann

has been accepted towards fulfillment
of the requirements for

M.S. Chemistry

 

 

degree in

mam

Major professor

 

 

Date 'l/1/?j/

0-7639

POSSIBLE SYNTHETIC PATHWAYS TO MODELS‘
OF THE METAL COFACTORS 0F NITROGENASE

BY

Robert Hugo Tieckelmann

A THESIS

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

MASTER OF SCIENCE

Department of Chemistry

1978

ABSTRACT

POSSIBLE SYNTHETIC PATHWAYS TO MODELS
OF THE METAL COFACTORS OF NITROGENASE

By

Robert Hugo Tieckelmann

Methods for synthesis of potential models for the
metal cofactors of nitrogenase have been investigated.
These include Fe4S4 clusters coordinated by: (1) metallo-
cenes, Fe4S4(SR)4_n[(SEt)2TiCp2]§2'n)' and
Fe4s4(sn)4_n(szriCp2)§2+n)'; (2) thiomolybdate bridges,
(RS)384Fe4(SZMoSZ)Fe4S4(SR)§-; and (3) a tripod,
Fe4s4(SR)[(s-g-C6H4-NH)3P=O]Z‘. Isolation of analytically
pure crystals has not yet been achieved, but the
thiomolybdate—bridged entity has been observed and par-
tially characterized spectroscopically in solution. Nmr
spectroscopy shows the coordinated thiolate ligands of the
Fe4S4 clusters may be removed upon the addition of a
suitable proton donor. This technique has been used to
force the Fe4S4 clusters to bind the selected ligands.
Results from attempted syntheses of the titanocene-bound
Fe4S4 clusters indicate that formation of polymers, varied
substitution products, and rearrangements hinder prepara-
tion of the desired compounds. These conditions also hin-

der preparation of the thiomolybdate-bridged Fe4S4 clus-

ters. A tripod ligand was synthesized to eliminate these

Robert Hugo Tieckelmann

problems; it may provide another metal cofactor model. A

discussion of future plans is included.

To
George and Betty
for their support, inspiration, and dedication
to the education of their children

ii

ACKNOWLEDGMENTS

I would like to extend my thanks to Bruce Averill for
suggesting this project. His patience and particularly
his guidance on experiment details, interpretation of
their results, and his encouragement in meeting the
requirements necessary for the completion of this degree
are sincerely appreciated.

Drs. Carl H. Brubaker, Jr., Frederick H. Horne,
Thomas J. Pinnavaia, and Gerald T. Babcock deserve my
gratitude for their special advice and encouragement long
before this project's inception.

I would like to thank H. Craig Silvis for his friend-
ship, advice, and helpful time-consuming discussions.

Many thanks to Kathryn Nyland for completing my fig-
ures with the special care and expertise only she could
provide.

A very special thanks to Alice (Dallas) Ridky. Her
organization and patience with my general ineptitude for
organization made reaching the thesis deadline a certainty
instead of a possibility.

Again, I want to express my sincere admiration for my

parents' devotion to my goals.

iii

TABLE OF CONTENTS

List of Figures

A.
B.

INTRODUCTION.

EXPERIMENTAL.

1. Materials and Methods

2. General Preparations.

a.

b.

Synthesis and Purification of Precursors.

Method of Removal of Coordinated Thiolate
Ligand from an Fe-S Tetrameric Cluster.

Method for the Preparation of the
Metallocene-Bound Tetrameric Species.

Method for the Preparation of the-
Thiomolybdate-Bridged Species

Method for the Preparation of the Tripod-
Bound Species . . . . . . . . . . .

3. Analyses.

a.

b.

Micro-Assay for Molybdenum.

Iodimetric Titration of the Tripod
Ligand.

RESULTS AND DISCUSSION.

1. Removal of the Coordinated Thiolate Ligand
from an Fe-S Tetrameric Cluster . .

2. The Metallocene-Bound Species

a.

bis(nS-cyclopentadienyl)bis(ethane-
thiolate)titanium (IV). . . . . .

iv

Page

vi

10
10

13

14

16

17
17
17

19
20

20
30

31

D.

b. bis(n5- cyclopentadienyl)bis(hydrogen-
sulfide)titanium (IV) . .

3. The MoSfi'-Bridged Species

4. The Tripod-Bound Species.
CONCLUSIONS AND PLANS FOR FUTURE WORK .
BIBLIOGRAPHY AND NOTES.

37
44
59
67
71

Figure

ZaEZb

LIST OF FIGURES

The displacement experiment6 effects the
removal of iron- sulfur (Fe- S) dimeric and
tetrameric clusters from the Fe protein.
Results of ligand substitution experiments31
indicate that the Fens“ core has the greatest
affinity for aryl thiol functionalities. The
addition of HMPA and the Tris-HCl buffer
denatures or unfolds the protein. With the
protein in the unfolded state the thiol is
able to enter the active site and begin the
substitution process by protonating the bio-
logical ligand. Large excesses of thiophenol
were employed to shift the substitution proc-
ess equilibrium and favor formation of

FenSn (8— Ph)?

Each figure is a nuclear magnetic resonance
spectrum in d6-DMSO (99%-d). Figure 2a is
the control spectrum of the tetramer,
(64A5)2[Feu8“(8- t- Bu)u]. The resonance at

= 7. 73 corresponds to the aryl protons; the
resonance at 6 = 2.50 corresponds to the sol-
vent impurity peak. Figure Zb is a spectrum
of tetramer and (ouAs)2MoSH [2:1]. The two
resonances present in Figure 2a are visible
(aryl, 6 = 7.72; impurity, 6 = 2.48), how-
ever the desired reaction has not taken place
as substantiated by the lack of "free" t-
butyl- thiol proton resonances (6 m 1.2)"

The electronic spectra display the changes in
the primary absorption peak of
(¢HAs)2[FeuSk(S- t- -Bu)u] upon the addition of
a proton donor (pyridinium chloride). In the
control spectrum (tetramer in acetonitrile,
--—-—) the (¢QAs)2[FegS (S- t- -Bu)u] absorbs at
420 nm initially. On addition of a proton
donor the 420 nm absorption shifts to higher
energy (tetramer and pyridinium chloride in
acetonitrile, ) and absorbs at 400 nm.

 

vi

21

25

Figure Page

4aa4b Each figure is a nuclear magnetic resonance
spectrum in d3-acetonitrile (99%-d). Figure
4a is the control spectrum of the tetramer,
(okAs)2[FeuSu(S—t-Bu)g]. Figure 4b exhibits
the addition of T.1 equivalents of 2,4,6-
trimethylpyridiniumhexafluorophosphate to

1.0 equivalent of the tetramer. . . . . . . . 27
5 Proposed structure for the metallocene-bound
species (a) . . . . . . . . . . . . . . . . . 32

6a§6b Each figure is a nuclear magnetic resonance
spectrum. Figure 6a is the control spectrum
of 1.0 equivalent of ligand, szTi(SEt)2,
and 1.2 equivalents of proton donor, 2,4,6-
trimethylpyridiniumhexafluorophosphate in
dS-dimethylsulfoxide (100%-d). Figure 6b is
a 1.0:l.0:0.55 equivalents ratio of tetramer,
(ouAs)2[FeuSu(S-t-Bu)k], ligand and proton
donor in dS-dimethylsulfoxide. The reso-
nances of either the ligand or the conjugate

base of the proton donor were not shifted . . 35
7 PrOposed structure for the metallocene-bound
species (b) . . . . . . . . . . . . . . . . . 38

8a,8b, Each figure is a nuclear magnetic resonance

and 8c spectrum in dl-chloroform (100%-d). Spectrum
8a is the control spectrum of szTi(SH)2 with
the two resonances in an intensity ratio of
5:1 and identified in units of 6 (delta).
Spectrum 8b indicates that addition of tri-
ethylamine does not effect the removal of the
thiol proton of CpZTi(SH)2, intensity ratios:
5:1 for szTi(SH)2 and 2:3 for triethylamine.
Spectrum 8c indicates that addition of proton
sponge does not effect the removal of the
thiol proton of szTi(SH)2 either, intensity
ratios: 5.1:1, szTi(SH)2 and 12:26z4 for

proton sponge 42
9 Proposed structure of the thiomolybdate-bound
species . . . . . . . . . . . . . . . . . . . 46

vii

Figure Page

103 8 Each figure is a nuclear magnetic resonance

10b spectrum in dG-dimethylsulfoxide (100%-d)
with resonances in units of 6 (delta). Fig-
ure 10a is the control spectrum of tetramer
and thiomolybdate ligand in an equivalents
ratio of 2:1. Addition of a proton donor to
a mixture of tetramer and the thiomolybdate
ligand effects the release of Efbutyl thiol.
Figure 10b is the nmr spectrum of this
phenomena, intensity ratios: 8:26:11 for
aryl:"bound" t-buty1:"free" alkyl thiol (the
pyridine hydrogens are obscured by the large
aryl resonance) . . . . . . . . . . . . . . . 48

11 All of the following optical spectra are
recorded in dry, degassed acetonitrile. In
spectrum A the thiomolybdate absorbs at 318
nm (a 17,000) and 468 nm (8 12,000), in
spectrum B the tetramer absorbs at 310 nm
(a 21,800) and 419 nm (8 16,700). Spectrum
C displays the results of combining 2 equiva-
lents of pyridinium chloride with 2 equiva-
lents and 1 equivalent of tetramer and thio-
molybdate, respectively. The tetramer primary
absorption has shifted to 398 nm, while the
thiomolybdate primary absorption has also
shifted to shorter wavelength . . . . . . . . 51

12 The electronic spectrum of 2:1:2 equivalents
ratio Of (EtQN)2[FegSq(S‘-‘£'BU)L‘], (BtMN)2MCSI+
and Me3pyrHC1 in spectrograde dimethyl-
acetamide (DMA). After initially mixing the
reactants, the length of the absorption
"shelf" increases with time, however subse-
quent attempts at recrystallizing the solid
obtained from this solution resulted in loss
of the absorption feature . . . . . . . . . . 54

13a 5 Each figure is a nuclear magnetic resonance

13b spectrum in de-dimethylsulfoxide (99%-d) with
resonances in units of 6 (delta). Figure 133
is the control spectrum of the solvent, the
resonance at 2.45 is the solvent impurity.
Figure 13b is the same 2:1:2 equivalents
ratio of tetramer, thiomolybdate, and tri-
methylpyridiniumchloride present in Figure
12. Note the shift of the "bound" 3-butyl
resonance from ca 2.58 to 1.32. The other

viii

Figure

14

15

proton resonances (tetraethylamine and tri-
methylpyridine) have been obscured by both
solvent impurity and the "bound" t- -buty1
protons . . . . . . . . . .

Pr0posed structure of the tripod-bound
species . . .

The electronic spectra display the changes in
the primary absorption peak of
(¢kAs)2[FeqSu(S- t- -Bu)u] upon binding the tri-
pod ligand, O= -P(- -NH- C 6H4-o- SH)3. In the con-
trol spectrum (tetramer in DMA ——————) the
(¢qAS)2[FegSg(S' t- BU)q] initially absorbs at
417 nm (e 16 ,7007. Addition of the tripod
ligand shifts the 417 nm absorption to a
lower energy absorption, 435 nm (tetramer and
tripod in DMA ——~——~)

ix

S7

60

63

A. INTRODUCTION

 

As future energy needs focus attention on the dwin-
dling petroleum resources of our planet, research inter-
ests increasingly concentrate on industrial processes
where potential energy savings may be realized. One such
area is nitrogen fixation, or more specifically, the cata-

lyzed reduction of dinitrogen to ammonia. On an indus-

3H2 + N2 + 2NH3 (l)
T = 450°C
P = 350 atm

catalyst = Fe/A1203/K20

trial scale this process is relatively inefficient (yields
520%) and requires large inputs of energy, often in the
form of fossil fuels. The energy requirement is due both
to the elevated temperatures and pressures necessary for
the reaction to take place and to the requirement for
hydrogen, obtained by the high temperature cracking of
fossil fuels. Consequently, a great deal of recent
research effort has been directed toward developing poten-
tial alternative processes. In 1974 alone over 40 million
tons (two billion moles) of ammonia were produced synthet-
ically,1 all by the Haber-Bosch process (1). An entirely

different process occurs in the biosphere. In bacteria

and algae an enzyme, nitrogenase (Nzase), catalyzes a
process that is chemically and energetically different to

the Haber-Bosch process:

lZATP + 6e- + 6H+ + N2 + 2NH3 + lZADP + 6Pi (2)
T = ambient temperature of the cell
P = atmospheric pressure

catalyst = nitrogenase.

This specific and more efficient reaction (typical of most
biological reactions) occurs with the transfer of six
electrons and the dephosphorylation of the magnesium com-
plex of the nucleotide, adenosine triphosphate. The fact
that conservative estimates project that bacteria and
algae produce 120 million tons of ammonia annually coupled
with the high specificity and relative efficiency of the
biological process indicate that this path is worth inves-
tigating. Understanding the catalyst, nitrogenase, is a
necessary prerequisite to success in solving these econom-
ical and petrochemical problems. An in depth understand-
ing of the enzyme's structural and mechanistic aspects is
needed; until recently, research in this area has been
hindered by the extreme oxygen sensitivity of nitrogenase.
Nitrogenase consists of two components: component I
(the MoFe protein) and component II (the Fe protein).2

The Fe protein has a molecular weight of about 60,000 and

is thought to contain two indistinguishable subunits.3

Assays have shown the presence of 4Fe and 4S= (acid labile
sulfur) per molecule. Mossbauer4 and displacement stud-

ies6 (Figure l) designate that the iron and sulfur are in

S The function

.the form of a single tetranuclear cluster.
of the Fe protein in dinitrogen fixation has been deter-
mined to be that of a one electron carrier and a binding
site for MgATP.7 However, the actual site of reduction of

dinitrogen is not the Fe protein,3

but the MoFe protein.
The MoFe protein, larger and more complex than the Fe
protein, has a molecular weight of approximately 240,000;
current data point to an 6282 structure. Assays indicate
m32Fe, ~3ZS=, and 1-2 molybdenum per molecule.3 Unlike
the relative degree of certainty that surrounds the Fe
protein's structure, the distribution of Mo and Fe among
the MoFe protein's subunits are unknown. Spectroscopic
and displacement experiments3 have shown the iron and sul-
fur to be contained in three distinct types of unit:8
(1) P-clusters, which.are tetranuclear species similar to
the unit found in the Fe protein, but with one iron atom
spectroscopically distinct from the others; (2) MEPR clus-
ters containing 6 to 8 Fe and 1 Mo, with the characteris-
tic S = 3/2 EPR of the MoFe protein; and finally, (3) an
Aunknown Fe-S species thought possibly to be dimeric

(FeZSZ) units. Known iron-sulfur species exist in mono,

di and tetranuclear forms and are prepared from simple

Figure l. The displacement experiment6 effects the
removal of iron-sulfur (Fe-S) dimeric and tetrameric clus-
ters from the Fe protein. Results of ligand substitution
experiments31 indicate that the Fens“ core has the great-
est affinity for aryl thiol functionalities. The addition
of HMPA and the Tris-HCl buffer denatures or unfolds the
protein. With the protein in the unfolded state the thiol
is able to enter the active site and begin the substitu-
tion process by protonating the biological ligand. Large
excesses of thiophenol were employed to shift the substi—
tution process equilibrium and favor formation of
FenSn(S-Ph)E'.

H opswwm

ANNE quNflgV
xaamUAQOUmouuuomm voNfiuouompmsu e
.mofioomm woNfimogucxm xamsofi>ohm m ma .AHUYIm amcom

:5
1O ..

mmouxo wfiom OOH

v a a .\\ w u :9
©Lu. m on A 82-3.; is £398 me

-E <m2: wow

 

starting materials (Fe salt, 8:, thiolate). In fact they
form readily and are stable for extended periods of time,
provided their environment is strictly anaerobic. These
iron-sulfur species are examples of true "synthetic anal-
ogues"9 of biological systems. The facile synthesis, pro-
longed stability, and well-documented presence of these
species in nitrogenase form the basis for the investiga—
tion of syntheses of potential "models” of the nitrogenase
metal cofactors.

The presence of molybdenum in the MoFe protein is of
significant catalytic importance. W. H. Orme-Johnson and
coworkers have stated that the molybdenum is "presumed" to
be part of the catalytic step, yet no direct evidence is
available. They add, "no EPR ascribable to the Mo has

been seen."10

However, without molybdenum the enzyme is
inactive. In a recent series of articles, Hodgson,
Newton, and coworkers have employed the technique of

(Extended X-ray Absorption Fine Structure)
spectroscopy to investigate the environment surrounding
the molybdenum of Nzase. The end result was the reifica-
tion of many previous ideas about the environment of the
molybdenum centers in the MoFe protein of nitrogenase.
The EXAFS studies suggest that the molybdenum in nitro-
genase is surrounded by four sulfurs at "close" distances

and approximately three irons at "near neighbor" dis-

tances; the molybdenum also has approximately three more

sulfurs at "less near neighbor" positions. Although the
quantitative aspects of conclusions based on EXAFS are
open to question, the qualitative information will aid in
the postulation and subsequent preparation of a synthetic
model or synthetic analogue of the metal cofactors of
nitrogenase. The report of the isolation of the iron-
molybdenum cofactor (FeMoco) of nitrogenase13 followed

closely by limited characterization data14

(composition:
m8Fe, ~6S=/Mo atom; molecular weight unknown), provided
researchers with a stoichiometry and certain physical
properties on which to base potential synthetic models.
Our efforts have focused on two specific areas.
First was the combination of organometallic molybdenum
complexes (specifically molybdocene derivatives) with
iron-sulfur tetrameric clusters. The second area entailed
designing a synthetic entity whose stoichiometry would
closely parallel that reported for FeMoco. Researchers
working in widely separated laboratories have recently
synthesized and characterized molybdenum-iron-sulfur spe-

15’16 that are very similar; [MozFe688(S-Ph)9]3' vs.

cies
[MoZFe689(SBt)8]3'. Unfortunately these synthetic models
do not mimic the FeMoco's stoichiometry. Lastly, an off-
shoot originating from attempts to solve the synthetic
problems arising in the first two areas has led us to a

potential synthetic model for the P-clusters of the MoFe

protein of nitrogenase.

Knowledge of the biochemical background of nitro-
genase is necessary to effect the design of potential
metal cofactor models. Elucidation of the enzyme's com-
plex mechanism depends, in part, on the study of synthetic
species whose reactivity and structure model particular
aspects of the nitrogenase enzyme. Indeed, solution of
the problems surrounding the dinitrogen fixation process
will not be possible until reasonable models for the metal
cofactor are synthesized. This thesis describes the pres-
ent state of progress in our laboratories toward synthesis

and characterization of these models.

B. EXPERIMENTAL

 

1. Materials and Methods

 

Reagent grade tetrahydrofuran and toluene were
distilled from sodium ben20phenone ketyl. Acetonitrile
and methanol were distilled from CaH2 and Mg(OCH3)2,
respectively. Heptane, hexanes, and toluene were dis-
tilled from sodium, while N-methylformamide was distilled
in 13339 from BaO (bp 85°C, 10 mm Hg). All spectro-grade
solvents (N,N—dimethylacetamide, acetonitrile) were used
without further purification. All solvents, including
deuterated solvents, were degassed by repeated evacuation
and flushing with prepurified nitrogen gas (freed of O2
and H20 by passage through a hot tower containing BASF
catalyst R-3-ll followed by passage through a tower con-
taining Mallinckrodt moisture absorbent, Aquasorb) prior
to use.

Titanocene dichloride was obtained from two
sources: Dr. C. H. Brubaker, Jr. and Alfa Products. The
titanocene dichloride was recrystallized from hot xylenes;
after cooling to room temperature, hexanes were introduced
until crystallization was imminent. All mercaptans were
obtained from Aldrich Chemical Company, Inc. and used
without further purification. Sodium hydrosulfide was

synthesized by a literature method.17 The 2,4,6-

trimethylpyridine hydrochloride was prepared by bubbling

10

gaseous HCl through 2,4,6-trimethylpyridine and washing
the precipitate with diethylether. The 2,4,6-
trimethylpyridiniumhexafluorophosphate was prepared in a
metathesis reaction with 2,4,6-trimethylpyridinehydro-
chloride and ammonium hexafluorophosphate. The product
was checked for chloride with silver nitrate and recrys-
tallized from isopropanol/hexanes. The tetraphenyl-
arsonium salts of thiomolybdate and thiotungstate were
recrystallized by H. C. Silvis, who also prepared the
tetraethylammonium and the tetramethylammonium salts of
these compounds and distilled the N-methylformamide used
in these preparations. The tripod ligand, OP(HN-C6H4-g-
SH)3, was synthesized and purified by Dr. B. A. Averill.
Electronic spectra were obtained on a Cary 17
spectrophotometer; proton magnetic resonance spectra were
meaSured on a Varian T-60 spectrometer; and all melting

points were determined on a Thomas-Hoover apparatus (melt-

ing points are uncorrected).

2. General Preparations

 

a. Synthesis and Purification of Precursors

 

Unless otherwise noted all operations were

performed in prepurified nitrogen atmosphere.

1) Preparation of CpZTi(SCH2CH3)2.
The preparation was performed according to
the methods of Koepf.18 The reaction stoichiometry is:

11

+ ZCH CH SH + ZEt N 3212222»

T1Cl2 3 2 3

sz

+ ZEt NHCl (3)

CpZTi(SCH2CH3)2 3

The product was recrystallized from toluene/heptane. Nmr

58

data were used to characterize the products: Koepfl8

(C82), 6.09(CSHS), 1.24(CH3), 3.12(CH2); found (d6DMSO/
CDSCN), 6.02(C5H5), l.05(CH3), 2.98(CHZ). Intensity
ratios for the peaks were 5:3:2.
ii) Preparation of CpZTi(SH)2.
The preparation was analogous to that of

19

Koepf and Schmidt according to the following reaction:

Et O@20°C

2 .
2 + 2H28 + ZEt3N or toluene'

@ r.t.

 

szTiCl

szTi(SH)2 + ZEt NHCl ' (4)

3
Characterization: mp, Koepf19 150-160°C (decomp), found
ISO-160°C (decomp); nmr, Koepf19 (CDC13), 6.35(C5§5). 3.44
(SH), found (CDClS), 6.35(CSH5), 3.45(SH), both with
intensity ratio of 5:1. This reportedly air stable com-
pound must be protected from both light and moisture dur-
ing preparation and storage.

iii) Preparation of Fe4S4(SR)i' salts.

20 was £01-

The method of Averill, §£_§l,
lowed, except that the minimum volume of solvent necessary

to dissolve the starting materials (approximately 50% of

12

the literature values) was employed. The salt,
(Et4N)2[Fe4S4(S-£-Bu)4], though not previously syn-
thesized, was prepared with slight modifications of the
above methods. I

iv) Preparation of (¢4As)2MS4.
The preparation of (WSZ)21 and (MoSZ)22
followed their respective literature procedures. Progress
of either reaction was monitored by changes in the elec-
tronic spectra. Hydrogen sulfide was passed through the
solution until no traces of the peaks due to M08;
remained. Recrystallization was effected from an aceto-
nitrile/methanol mixture.

v) Preparation of (Et4N)2MS4 and (Me4N)2MS4.

Either M003 or H2W04 was dissolved in an

aqueous solution of Et4NOH or Me4NOH. H28 was slowly bub-
bled through these solutions until absence of peaks due to
the 0x0 species in the optical spectrum indicated comple-
tion of the reaction. Recrystallization was effected from

a water/methanol mixture.

vi) Preparation of the tripod ligand,
OP(HN-C H -g-SH)3.

6 4
This ligand was prepared in three steps
from g-aminothiophenol: (l) protection of the mercaptide
with benzylchloride in basic solution to yield the benzyl-

thioether; (2) condensation of the amine with hexamethyl-

phosphoramide in refluxing p-cymeme; (3) removal of the

13

S-benzyl protecting groups with sodium in liquid ammonia.

b. Method of Removal of Coordinated Thiolate
Ligand from an Fe-S Tetrameric Cluster

 

 

2- 1-
Fe4S4(SR)4 + Me3pyrHCl + Fe4S4(SR)3 +

RSH + Me3pyr + Cl- (5)

A five dram vial equipped with a tight fitting
rubber serum cap and a small magnetic stirring bar was
repeatedly purged through a syringe needle fitted to a dry
nitrogen/vacuum line. Carefully 250 mg (0.17 mmole) of
(64As)2[Fe4S4(S-t-Bu)4] was introduced, and the vial was
repeatedly purged with nitrogen. The same procedure was
followed for the addition of 90 mg (0.34 mmole) of 2,4,6-
trimethylpyridiniumhexafluorophosphate (2,4,69Me3pyrHFP6).
Degassed deuterated acetonitrile was introduced via
anaerobic syringe into the vials (final concentrations:
tetramer 0.05 M, pyridinium salt 0.30 M), and the solutions
were stirred until dissolution was complete. As a check
of the concentration and purity of the tetrameric species
20 ul was removed from the stock solution, diluted to 500
pl with degassed spectro-grade acetonitrile, and the elec-
tronic spectrum recorded. Upon observation of satis-

23 300 pl

24

factory concentration and absorption features,

and 500 pl of the respective tetrameric and pyridinium

14

species were introduced via syringe into another similarly
prepared two dram vial. Mixing was continued for several
minutes, and the reaction mixture was withdrawn from the
vial and introduced into a 5 mm nmr tube. The nmr tube
was previously fitted with an appropriate rubber serum cap
and purged with nitrogen. This is the general procedure
for preparation of samples for optical or nuclear magnetic

resonance spectroscopy.

c. Method for the Preparation of the Metallocene-

 

Bound Tetrameric Species

 

2- .
Fe4S4(SR)4 + nCp2T1(SEt)2 + nMespyrHCl +
. (2-n)-
{Fe4S4(SR)4.n[(SEt)2T1Cp2]}n +

nRSH + nMeSpyr + nCl- _ (6a)

2- .
Fe4S4(SR)4 + nCpZT1(SH)2 + nEt3N +
- (2+n)-
{Fe4S4(SR)4_n(SZT1Cp2)n} +

nRSH + nEtSNH+ (6b)
Reaction (6a). In a 100 ml round bottom flask
equipped with a side arm, 2.6 grams (1.8 mmole) of
(¢4As)2[Fe4S4(S-t-Bu)4] was dissolved in 36 mls of
degassed acetonitrile. The ligand, szTi(SEt)2 (0.5 gram,
1.8 mmole), was dissolved in 6 mls of degassed spectro-

grade dimethyl sulfoxide and then repeatedly degassed.

15

The proton donor, 2,4,6-trimethylpyridiniumhexafluorophos-
phate (0.5 gram, 1.8 mmole), was similarly dissolved in 6
mls of dry degassed acetonitrile. The ligand and proton
donor were sequentially introduced via cannula with subse—
quent degassing. The proton donor was added dropwise over
a period of 20-30 minutes. Stirring was continued over-
night and the resulting crystals were filtered and dried
£11192.

Reaction (6b). A two dram vial equipped with
a tight fitting rubber serum cap and a small stirring bar
magnet was repeatedly purged with nitrogen. Then 75 mg
(50 mmole) of (¢4As)Z[Fe4S4(S-£-Bu)4] was carefully added,
and the vial was again degassed. A similar procedure was
used for preparation of a szTi(SH)2 sample (61 mg, 0.25
mmole). One milliliter of 100 atom percent déeDMSO was
added to the tetramer (final concentration, 0.05 M) and
0.5 ml to the titanocene derivative (final concentration,
0.5 M).

Concentration of the tetrameric cluster was

23

checked by inspection of an electronic spectrum, and

aliquots were added to an anaerobic 5 mm nmr tube equipped
with a rubber serum cap. To the previously prepared nmr
tube were added in the following sequence: 500 pl (2.5 x

'5 moles) of tetramer, 500 pl (2.5 x 10'5 moles) of

5

10

szTi(SH)2, 5 ul (3.5 x 10' moles) of Ft N, and 1 drop

3
(10-15 ul) of TMS. The tube was agitated vigorously on a

vortex mixer.

16

d. Method for the Preparation of the
Thiomolybdate-Bridged Species

2- = ~
2Fe4S4(SR)4 + MoS4 + 2Me3pyrHCl +
4-
(SR)3Fe4S4(SZMoSZ)Fe4S4(SR)3 +

2Me3pyr + 2Cl— (7)

To a 50 ml round bottom flask equipped with a
side arm was added 700 mg of (¢4As)2[Fe4S4(S-£-Bu)4].
Eight mls of previously degassed spectro-grade aceto-
nitrile were introduced via cannula into the flask. The
solution was stirred until dissolution was complete, and
then degassed several times. Stock solutions of
(¢4As)2MoS4 (7 mg in 2 ml CHSCN) and 2,4,6-Me3pyrHCl (25

CN) were also prepared in this manner.
6

mg in 4 mls CH3

With a gas tight syringe 40 ul (2.38 x 10' mole) of the

tetrameric stock solution was diluted to 1000 pl and
introduced into a 2 dram vial equipped with a magnetic

stirring bar and connected to a dry nitrogen/vacuum line.

6

The thiomolybdate solution (300 pl, 1.05 x 10- mole) was

then introduced via syringe with stirring. An aliquot (60

6

ul, 2.37 x 10- mole) of the 2,4,6-trimethylpyridinium

chloride solution was then introduced dropwise via syringe

with stirring.24

17

e. Method for the Preparation of the Tripod-Bound
Species

 

2_ _
[Fe4S4(S-£-Bu)4] + OP(HN-C6H4-g-SH)3 +

{Fe4S4(S-£-Bu)[(S-C6H4-NH)3PO]}2- + SE-Bu-SH (8)

In a 50 ml round bottom flask equipped with a

4

side arm, 0.5 gram (5.8 x 10- mole) of (Me4N)2[Fe4S4(S-£-

Bu)4] was dissolved in 35 mls of acetonitrile. In another
similarly equipped 100 ml flask 0.25 gram (5.8 x 10-4

mole) of OP(NH-C H4-g-SH)3 (tripod) was slurried in 25 mls

6
of acetonitrile. The tetrameric cluster was added to the
tripod slurry via cannula. The stirring was continued
overnight and the mixture was occasionally evacuated to
remove EfBu-SH (bp. 62-65°C) to force the equilibrium to
the right and favor formation of the tripod-bound species.
The acetonitrile was removed at 40-50°C and the subsequent
solid (a fine silky material with metallic lustre) was

dried i§_vacuo. Recrystallization from acetonitrile was

twice attempted.

3. Analyses

 

25

a. Micro-Assay for Molybdenum (aerobically per-

formed in a 20 ml test tube)

 

Required solutions:

 

i) 30% H202

ii) 4 M HCl

18

iii) KI (prepared as needed); 2.5 grams/5.0 mls
deionized H20.

iv) Ascorbic acid, 20% (w/v), 20 grams acid/
100 mls deionized H20. 4

v) Tartaric acid, 40% (w/v), 40 grams acid/

100 mls deionized H O.

2
vi) Dithiol (3,4-mercaptotoluene): 0.5 grams
of dithiol is melted and dissolved in 190 mls of deionized
H20 previously degassed with nitrogen. Add 60 mls of l M
NaOH also degassed with nitrogen. Thioglycolic acid26 is
added dropwise (2-4 mls) until the first faint Opalescence

is observed.

Procedure:

 

i) To the Mo sample is added 100 pl of
deionized H20 and 100 pl of concentrated HZSO4. The solu-
tion is heated slowly at first, then vigorously with care.
This step is performed in an effort to digest the extra-
neous material present and to "free" the Mo. It must be
repeated until the entire sample is soluble in the HZSO4/
H20 solution, therefore when necessary 100 pl aliquots of
a 5:2:1 solution of HCl/HNO3/HBr were added, and the mix-

27

ture heated until clear and one was certain dissolution

was complete.
ii) After cooling the sample, 25 pl of 30% H202
is introduced. The solution is heated with care.

19

iii) After cooling the solution, 7 mls of 4 M
HCl, 350 p1 of the K1 solution and a micro-stirring bar
magnet are added and the solution stirred for 10 minutes.

iv) 500 pl ascorbic acid solution, 250 p1 of
Tartaric acid solution, and 500 pl of the dithiol solution
are added, and the solution is stirred for 10 minutes.

v) 650 pl of iso-amylacetate is introduced,

the mixture stirred for 10 minutes, and the organic layer
pipetted into a cuvette. The M0 complex absorbs at 680 nm

with a molar absorbtivity (e) of 24,400 M'lcm'l.

b. Iodimetric Titration of the Tripod Liggnd

 

Preparation of solutions and the general pro-
cedure were adapted from standard sources.28’29
Preparation of the tripod ligand as a sample
was as follows: A
i) To a 250 ml Erlenmeyer flask was added
0.05 gram of the tripod ligand weighed to :0.1 mg.

ii) 15 ml of N,N-dimethylacetamide and 15 ml
of deionized H20 was added and the resulting mixture
stirred until the solution was clear.

iii) 10 ml of starch solution was added and the

mixture titrated against the standardized iodine solution

until a pale blue color was just visible.

C. RESULTS AND DISCUSSION

 

1. Removal of the Coordinated Thiolate Ligand from an
Fe-S Tetrameric Cluster (Reaction 3)

 

Combination of [Fe4S4(S—£-Bu)4]2' (tetramer) with
thiomolybdate does not result in an observable chemical

30 (Figures 2a and 2b). The desired result would

reaction
be easily detected by nuclear magnetic resonance spec-
troscopy as the free thiolate ligand in the equilibrium,

is

2' +
[Fe4S4(S-£-Bu)4] + L 1

[Fe4s4(s-£-Bu)3L*]1‘ + g-Bu-s’ (9)

The equilibrium apparently favors the reactants;
implying that the proposed ligands have a lower affinity
than g-Bu-S' for the iron-sulfur tetranuclear (Fe4S4)
core. This is significant because the t-butyl-mercaptide
ion exhibits the least affinity, of all mercaptides
examined,31 for the Fe4S4 core. To bind thiomolybdate or
the metallocene ligands to the Fe4S4 core, it is therefore
necessary to force the equilibrium to the right. The dif-
ference between the exchange reaction

2-

2-
Fe4S4(SR)4 + R'SH : Fe4S4(SR)3(SR') + RSH (10)

and reaction (9) is that the incoming ligand in the

20

21

Figures 2a and 2b. Each figure is a nuclear magnetic
resonance spectrum in d6-DMSO (99%-d). Figure 2a is the
control spectrum of the tetramer, (¢MAS)2[FGQSQ(S‘£'BU)Q].
The resonance at 6 = 7.73 corresponds to the aryl protons;
the resonance at 6 - 2.50 corresponds to the solvent
impurity peak. Figure 2b is a spectrum of tetramer and
(6(As)2MoS1+ [2:1]. The two resonances present in Figure
2a are visible (aryl, 6 = 7.72; impurity, 6 = 2.48), how-
ever the desired reaction has not taken place as sub-
stantiated by the lack of "free" t-butyl-thiol proton
resonances (6 g 1.2).

22

AVG

 

 

 

 

 

 

 

d 1 d u

EN opsmfim

\l“

 

 

 

mN opsmwm

 

1 (1314 1 ii»)

JJ

 

 

 

 

23

exchange reaction protonates the species that is

leaving.32

This suggests that the unfavorable equilibrium
of reaction (9) might be driven to the right by adding a
species that can donate a proton when the entering ligand
cannot perform this function. To prevent coordination to
the Fe4S4 core, the conjugate base of the proton donor
selected must be a poor nucleophile. Further, the proton
donor must not be a sufficiently strong acid to destroy
the acid sensitive tetranuclear cluster.33

The donor, 2,4,6-trimethylpyridiniumhexafluoro-
phosphate (MespyrHPF6), was selected for a variety of rea-
sons. Although pyridine has a relatively low affinity for
both iron (11) and iron (III),34 2,4,6-trimethylpyridine
was used to introduce a steric factor that would further
hinder binding during evaluation of this novel technique.
In addition, the report of the structure of the chloro-

2...

tetramer35 (Fe4S4C1 and the lack of characterization

4)
data prompted the use of the hexafluorophosphate anion in
lieu of chloride. Comparison of physical data was not
possible until subsequent publication of the electronic

36 The primary visible

spectra of the halo-clusters.
absorption peak of the tetramer with alkylthiolate ligands
(N420 nm) shifts to longer wavelength and is reduced in
intensity on replacement of thiolate by chloride ion.

This effect was not observed during the course of experi-

ments with donors such as 2,4,6-trimethylpyridine

24

hydrochloride (MespyrHCl). Comparison of spectra (Figure
3) reveals that the primary absorption peak of the tetra-
mer does not decrease in intensity and instead simply
shifts to N400 nm when an equivalent amount of pyridinium
chloride is added. The chloro-tetramer is clearly not
formed under these conditions, however, so the use of
MespyrHCl and MespyrHPF6 are intermixed throughout this

text. The MespyrH+ has a pKa of 7.437

and therefore pro-
vides two orders of magnitude less driving force than the
pyridinium ion (pKa = 5.4) in forcing the equilibrium
reaction (9) in the desired direction. This does not
limit the ability of the Me3pyrH+ to favorably offset the
equilibrium, as observation of the release of free
thiolate in Figure 4 attests. In addition, the repeated
evacuation and purging of the reaction vessel will also
shift the equilibrium through the removal of the more
volatile alkylthiols(E-butylthiol, bp 62-65° at 760 mm).
Further, Me3pyrH+ is a sufficiently weak acid that, under
these conditions, it can aid the coordination of these
bidentate metallocene or tetrathiometallate species with-
out initiating decomposition of the tetramer or the poten-
tial ligands. Also, while Mespyr is a better base than
pyridine, the steric hindrance introduced by the g-methyl
functions should prevent competition between the base and
our least nucleophilic ligand, CpZTi(SEt)2, for the avail-

able sites on the Fe4S4 core.

25

Figure 3. The electronic spectra display the changes in
the primary absorption peak of (¢gAS)2[FegSH(S'£‘BU)q]
upon the addition of a proton donor (pyridinium chloride).
In the control spectrum (tetramer in acetonitrile, —-—-—)
the (¢kAS)2[FeuSu(S'£'BU)H] absorbs at 420 nm initially.
On addition of a proton donor the 420 nm absorption shifts
to higher energy (tetramer and pyridinium chloride in
acetonitrile, ) and absorbs at 400 nm.

 

26

com

m ocean;

ES 4
com

 

_—A_L_L

 

 

27

Figures 4a and 4b. Each figure is a nuclear magnetic
resonance spectrum in d3-acetonitrile (99%-d). Figure 4a
is the control spectrum of the tetramer, (¢qu)2[Fe4$q(S-
E-Bu)p]. Figure 4b exhibits the addition of 1.1 equiva-
lents of 2,4,6-trimethylpyridiniumhexafluorOphosphate to
1.0 equivalent of the tetramer.

28

 

 

 

 

8.0

 

 

l 1
Figure 4a

1 J
Figure 4b

 

 

29

Figures 4a and 4b display the release of the free
thiol when Me3pyrHPF6 is combined with tetramer. Figure
4a shows the nmr spectrum of the t-butyl tetramer [d3-
acetonitrile; 7.70 (C635), 2.66 (bound E-butyl-S)'; 1.92,
2.11, 3.22 (solvent impurities)]; when compared to the nmr
spectrum of a combination of tetramer and 1.1 equivalents
of MespyrHPF6 in Figure 4b [d3-acetonitrile; 6.72 (per),
2.35 (nge), 2.20 (p-Me), 1.36 (free thiol)], the ease
with which the thiolate ligand is removed and observed is
apparent. Integration of the spectrum in Figure 4b
reveals that 1.1 equivalents of thiol have been released
by addition of 1.1 equivalents38 of proton donor,
Me3pyrHPF6.

The utility of this technique will depend on the
following considerations:

a. The ability of the solvent to stabilize the
Fe4S4 core once the ligand is released.

b. The difference in pKa between the leaving
ligand and the proton donor.

c. There is at present no way to limit the
obvious statistical distribution of products that will
result from having four structurally equivalent sites
available for coordination. The Fe4S4(SR)%' was observed
to decompose on standing; consequently, release of
thiolate from more than one site of an Fe4S4 cluster

appears to accelerate the decomposition due to the

30

increased instability of the core. As a necessary part of
the technique, a dilute solution of the proton donor must
be added slowly to a more concentrated mixture of the
potential ligand and tetrameric cluster. This insures the
proximity of the ligand to the coordination sites as they
are made available.

d. The success of any of these potential ligands
as a coordinating entity will be limited by its nucleo-
philicity compared to that of the solvent and of the con-
jugate base of the proton donor. The selection of sol-
vent, ligand, and donor must therefore be made with
careful consideration of each species' relative nucleo-
philic capabilities.

The results of the application of this new tech-

nique follow.

2. The Metallocene-Bound Species

 

The first application of the technique described
above involved dithiolate derivatives of titanocene.
Since certain metallocenes were catalysts for early
dinitrogen reduction experiments39 and Fe4S4 clusters were
documented as being part of the overall mechanism for

3

dinitrogen reduction to ammonia, the combination of the

two entities became the subject of a synthetic investiga-

tion. Also, evidence concerning the role of molybdenum in

10

the nitrogen fixation mechanism, and recent structural

31

evidencell’13

supporting the existence of Fe-S-Mo segments
in nitrogenase, suggested coupling a molybdenum derivative
to an Fe4S4 cluster. The preparation of any Cp2M0(SR)2
species, however, is tedious, requiring a low yield syn-
thesis of oxygen-sensitive szMon followed by conversion
of szMoCl2 (soluble only in liquid S02). Final conver-
sion of the dichloride to the dithiolate involves yet
another low-yield synthetic step. Titanium was utilized
as a substitute for molybdenum to circumvent the synthetic
problems surrounding the Mo analog. The initial plan was
to work out the synthetic details with titanium analogs,
whose preparation is more straightforward and inexpen-

18,19

sive. Since both metallocene dithiolates have shown

the ability to coordinate in a bidentate manner to other

metal centers,40

extension of the synthetic scheme
developed for titanium derivatives to molybdenum was

expected to require only minor modifications.

a. bis(nS-cyclopentadienyl)bis(ethanethiolate)-

 

titanium (IV)

 

2— .
Fe4S4(SR)4 + nCp2T1(SEt)2 + nMeSpyrHPF6 +

Fe4S4(SR)4_n[(S-Et)2TiCp2]§2-n)- +

nRSH + nMeSpyr + nPF6- (ll)

32

Figure 5. Proposed structure for the metallocene-bound
species (a).

33

m opswflm

O.u/ K aim
\\\\\
QU\ :. / V0ml\ll“/\\¢w_ m
M W\mw/\W /~_
T: a /mu._\

34

The szTi(SBt)2, selected because of its
inexpensive starting materials and ease of preparation,
acts as a neutral ligand in reaction (10). As described
above, MeSpyrHPF6 is used to force an equilibrium shift
and increase the likelihood of coordination. Experiments
monitored by nmr and optical spectroscopy revealed neither
a shift of the cyclopentadienyl or bound thiol resonances
nor any change in the primary electronic absorption (417
nm, e = 17,400 M'lcm'l) of the tetrameric cluster. A
carefully designed nmr experiment showed inconsistencies
in the integrations of the important proton-bearing func-
tionalities (Figure 6). Figure 6a (d6-DMSO) is an nmr
spectrum of a mixture of 1.0 equivalent of szTi(SEt)2
[6.04 (CSMS), 2.93 (CMZ), 1.00 (CM3)] and 1.2 equivalents
of Me3pyrHPF6 [7.40 (m-H), 2.56 (g-Me), 2.35 (p-Me)].
Figure 6b (d6-DMSO) is an nmr spectrum of a solution
originally containing 1.0 equivalent of (¢4As)2[Fe4S4(S-£-

Bu)4], Me3pyrHPF , and CpZTi(SEt)2. The intensities of

6
the Me3pyr and tetramer peaks agree with the expected 1:1
ratio, yet only 0.55 equivalents of szTi(SEt)2 are
observed. The resonances of the metallocene ligand were
not altered in position [6.10 (CSMS), 1.02 (CMS), the
(CMZ) quartet was obscured by the bound t—butyl thiolate],
and those of the conjugate base of the proton donor were

not shifted either [6.78 (27H), 2.34 (g-Me), 2.19 (p-Me);
literature“ CDC13, 6.59 (M-H), 2.37 (g-Me), 2.18 (p-Me)].

35

Figures 6a and 6b. Each figure is a nuclear magnetic
resonance spectrum. Figure 6a is the control spectrum of
1.0 equivalent of ligand, szTi(SEt)2, and 1.2 equivalents
of proton donor, 2,4,6-trimethylpyridiniumhexafluorophos-
phate in de-dimethylsulfoxide (100%-d). Figure 6b is a
l.0:l.0:0.55 equivalents ratio of tetramer,
C¢QAS)2[Fequ(S'£‘BU)Q], ligand and proton donor in d6-
dimethylsulfoxide. The resonances of either the ligand or
the conjugate base of the proton donor were not shifted.

36

 

 

 

 

 

 

Figure 6a

L

 

 

 

 

80

Figure 6b

0.0

 

37

Therefore coordination of either the metallocene ligand or
Me3pyr is unlikely. But 0.5 equivalent of free thiol was
released [1.35 (CM3)] in Figure 6b! Proximity of protons
to a paramagnetic species can result in the loss of

splittings42

and shifting and broadening of peaks into the
baseline,43 suggesting that the cluster-bound metallocene
might not be detectable by nmr. This last rationale was
initially accepted as an explanation for the inconsis-
tencies in the proton integrations. It was then dis-
covered that the nmr solutions formed a dark micro—
crystalline precipitate on standing undisturbed for ten
hours. Workup of the products of a preparative scale
experiment resulted in isolation of an apparently micro-

crystalline substance that was insoluble44

and pyrophoric.
Further characterization proved to be impractical. This
suggests that the "missing" titanocene derivative (that
not observed in nmr) is being removed from solution by a
reaction with (or catalyzed by) a small portion of the

iron-sulfur cluster, with formation of the uncharacterized

solid residue.

b. bis(nS-cyclopentadienyl)bis(hydrogensulfide)-

 

titanium (IV)

 

Fe4s4(SR)i‘ + nszTi(SH)2 + nEt N +

3

. (2+n)- +
Fe4S4(SR)4_n(SZT1Cp2)n + nRSH + nEt NH (6b)

3

38

Figure 7. Proposed structure for the metallocene-bound
species (b).

n ouamwm,

40

CpZTi(SH)2, whose binding to the Fe4S4 cluster
was examined for reasons mentioned above, is unique among
the ligands studied in that it is a diprotic acid, for-
mally capable of donating a proton to the departing
thiolate ligand; the removal of the second proton requires
an additional equivalent of base. This will result in the
titanocene derivative acting as a dinegative bidentate
ligand. Upon loss of a proton, the nucleophilicity of
szTi(SH)2 should be enhanced compared to CpZTi(SEt)z;
therefore, the former should have a greater tendency to
coordinate to the Fe4S4 core. This trend was not
observed, however. Much like CpZTi(SEt)2, coordination
studies with the bis(hydrogensulfide) derivative did not
result in the observation of any shifts in the nmr peaks
of either the CSHS or bound Efbutyl; further, no shifts or
changes in intensity of optical spectral features were
observed. Most importantly, a free t-butyl peak, indica-
tive of release of coordinated thiolate, was not observed
on mixing the components. This last point initiated
experiments to abstract one of the thiol protons of
szTi(SH)2 to generate a better nucleophile. Two common
nitrogen bases were used: triethylamine (pr = 3.00) and
proton sponge (N,N,N',N'-tetramethyl-l,8-naphthalene-
diamine, pr = 1.63);45 neither apparently was able to
abstract a proton from CpZTi(SH)2. Figure 8a [CDC13, 6.20
(CSMS), 3.33 (SM), intensity ratio 5:1] is that of

41

CpZTi(SH)2; the proton resonances integrate well but are
shifted slightly upfield from the literature values. In

Figure 8b [CDCl 6.20 (CSMS), 3.33 (SM), 2.53 (CMZ), 1.03

3,
(CES)’ intensity ratios 5:1 and 2:3] where 0.6 equivalents
of triethylamine were added to the titanocene derivative,
there is no shift of either the titanocene or the tri-

ethylamine resonances,46

or is the intensity of the thiol
proton resonance altered. The spectrum in Figure 8c
[CDC13, ca 7.2 (3,4,5,6—M), 6.90 (2,7-M), 6.10 (CSMS),
3.26 (SM), 2.75 (Me) with intensity ratios 5.1:1 for the
titanocene derivative and 12:2.6:4 for the proton sponge]
is that of a solution containing 3.8 equivalents of pro-
ton sponge per titanocene dithiolate; again the chemical
shifts and intensities correspond to those of the
reactants.47

The above data suggest, that of the following

equilibria
CpZTi(SH)2 : CpZTi(SH)S- + H+ pKa > 13.0 (12)
szri(sn)s‘ : szTi(S); + H+ (13)

the first dissociation must be difficult, requiring an
organic base with a pr < 1.0 to remove the first proton.
Since the acidity of CpZTi(SH)2 has not been investigated,

the conditions necessary for the removal of the second

42

Figures 8a, 8b, and 8c. Each figure is a nuclear magnetic
resonance spectrum in dl-chloroform (100%-d). Spectrum 8a
is the control spectrum of szTi(SH)2 with the two reso-
nances in an intensity ratio of 5:1 and identified in
units of 6 (delta). Spectrum 8b indicates that addition
of triethylamine does not effect the removal of the thiol
proton of szTi(SH)2, intensity ratios: 5:1 for
CpZTi(SH)2 and 2:3 for triethylamine. Spectrum 8c indi-
cates that addition of proton sponge does not effect the
removal of the thiol proton of szTi(SH)2 either, inten-
sity ratios: 5.1:1, CpZTi(SH)2 and 12:26:4 for proton
sponge.

43

‘ Figure 8a ’

A 4L; A ‘M A
‘ VVVTr‘ w W W r—w—V vv

Figure 8b

[l
*[Jle ITEM
1 1 111 1 1 1 1 1 I

Figure 8c

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

44

thiol are unknown and could result in the destruction of
the metallocene. However, if the szTi(SH)Z is to bind to
the Fe4S4 core, its thiol protons must exhibit facile
exchange with the organic base and the thiolate ligands of
the cluster.

Unfortunately, neither of the titanocene
derivatives gave any physical evidence for binding to the
Fe4S4 cluster. The experience gained by the study of
titanocene derivatives should expedite the transition to
work entailing molybdocene derivatives. Indeed, the prob-
lems encountered might be due to titanium's chemical
behavior (for example the szTi(SH)2 and the szTi(SEt)2
are both photo-sensitive), and parallel situations may not
occur with molybdenum. Molybdenum's presence in nitro-
genase and its proximity to an Fe4S4 core indicate that
all future synthetic work deal with coupling dithiolate
derivatives of molybdocene to Fe-S tetrameric clusters,
in spite of the difficulties involved in the molybdocene

derivatives preparation.

3. The MoSi--Bridged Species

2- 2-
2Fe4S4(SR)4 + Mos4 + 2Me3pyrHC1 +

(RS)3Fe4s4(Mos4)Fe4s4(SR)3‘ + ZRSH + ZMeSpyr + 2c1’ (14)

45

The isolation and characterization of the iron-
molybdenum cofactor (FeMoco) of nitrogenasels’14 initiated
efforts directed at synthesizing the Mo-bridged species
(Figure 9) in which two tetrameric Fe-S clusters are
joined by a thiomolybdate bridge. The plausibility of
this species as a potential model for the metal cofactor
was enhanced by the subsequent appearance of extended x-

12 which were

ray absorption fine structure (EXAFS) data,
interpreted as showing that the metal cofactor contained a
molybdenum atom surrounded by S-Fe-S units. The con-
ceptual drawing in Figure 9 portrays this structural unit,
including the Mo-S-Fe segment described in Section 2.

The thiomolybdate anion, prepared from simple

reagents in high yield,21’22

can act as a dinegative
bidentate ligand that would bridge two Fe4S4 clusters.

Its affinity for the Fe S4 core was found to be less than

4
that of t-butyl mercaptide (see below). Consequently, to
make coordination feasible a proton donor was required to
release the coordinated thiolate ligand. Figure 10a (d6-
DMSO) is a spectrum of a solution containing a 2:1 mixture
of (¢4As)2MoS4 (thiomolybdate) and (¢4As)2[Fe4S4(S-t-Bu)4]
(tetramer) [7.72 (¢). 2.58 (bound tfbutyl), 3.25 (impur-
ity); intensity ratio calculated for ¢= t-butyl, 80:36;
found 80:3343]; as predicted, the free alkyl thiol (ca.

1.30) is not observed. The addition of one equivalent of

pyridine hydrochloride effects the release of an

46

Figure 9. Proposed structure of the thiomolybdate-bridged
species.

48

Figures 10a and 10b. Each figure is a nuclear magnetic
resonance spectrum in d5—dimethylsulfoxide (100%-d) with
resonances in units of 6 (delta). Figure 10a is the con-
trol spectrum of tetramer and thiomolybdate ligand in an
equivalents ratio of 2:1. Addition of a proton donor to

a mixture of tetramer and the thiomolybdate ligand effects
the release of t-butyl thiol. Figure 10b is the nmr spec-
trum of this phenomena, intensity ratios: 8:26:11 for
aryl:"bound" t-butyl:"free" alkyl thiol (the pyridine
hydrogens are obscured by the large aryl resonance).

49

 

 

 

 

 

 

 

 

 

 

I I T T I T T I
Figure 10a
1 1 J l I 1 1 l
8.0 00

Figure 10b

 

50

equivalent of free alkyl thiol [Figure 10b, d6-DMSO, 7.76
(¢). 2.58 (bound t-butyl), 1.32 (free thiol), pyridine
hydrogens are obscured by the large ¢ resonance; intensity
ratio calculated for ¢:bound t-butyl:free alkyl thiol, 80:
27:9, found 80:26:11].

Since both reactants have unique optical features

49,50

with known molar absorptivities, binding could be

monitored spectrometrically, using two distinct properties

of the system. First, it is known that binding of a metal

ion to two sulfides of MoSi' induces a shift of the

optical features to longer wavelength; coordination of all

four sulfides will afford species in which MoSi' acts as a

bridging ligand and the optical features shift to shorter

49

wavelength. The bridged species lacks terminal Mo-S

entities; therefore the optical spectrum should also lack
bands due to these features. Second, changing the coordi-
nated ligands bound to the Fe4S4 core causes a correspond-

ing shift of the primary absorption peak (417 nm for most

31

alkyl thiol ligands) of the core. An optical experiment

resulted in the spectra shown in Figure 11. Spectrum A is

that of a solution containing 1.0 equivalent of

49

(¢4As)2MoS4 [literature in H20; nm (M'lcm'l x 10’4), 467

(1.2), 317 (1.7), 242 (2.4); found in spectro-
acetonitrile; nm (e not determined), 468, 318, 243].
Spectrum B is that of a solution containing 2.0 equiva-

lents of (¢4As)2[Fe4S4(S-£-Bu)4] [literature50 in

51

Figure 11. All of the following optical spectra are
recorded in dry, degassed acetonitrile. In spectrum A

the thiomolybdate absorbs at 318 nm (6 17,000) and 468 nm
(a 12,000), in spectrum B the tetramer absorbs at 310 nm
(8 21,800) and 419 nm (a 16,700). Spectrum C displays the
results of combining 2 equivalents of pyridinium chloride
with 2 equivalents and 1 equivalent of tetramer and
thiomolybdate, respectively. The tetramer primary
absorption has shifted to 398 nm, while the thiomolybdate
primary absorption has also shifted to shorter wavelength.

SZ

 

 

 

 

 

 

 

LOP+
Each spectrum is shifted
0.15 absorbance unit at
620 nm.

(18-

O£P-

CM:—

(IZF- B

(10 J. 11 A L

220 300 380 60 540 620

4
Mom)

Figure 11

53

-1

spectro-DMF; nm (M_1cm x 10-4

), 417 (1.67), 303 (2.18);
found in spectro-acetonitrile; nm (e not determined), 419,
310]. Mixing of tetramer and thiomolybdate in a 2:1 mole
ratio, followed by the addition of 2.0 equivalents of
pyridine hydrochloride, shifted the 419 nm absorption of
the tetramer to 398 nm but left the 310 nm peak and the
relative intensities almost unchanged (spectrum C of Fig-
ure 11). The lack of any spectral features attributable
to the thiomolybdate (the loss or shift of both the 468 nm
and 318 nm absorption peaks) coupled with the shift of the
tetramer's 419 nm absorption suggested that the desired
species containing bridging SzMoS2 units may have been
formed. There is no evidence for bidentate MUSE-.51 A
preparative scale experiment was therefore carried out. A
search for a solvent or a solvent system suitable for dis-
solving all species resulted in the use of a minimum of
acetonitrile/N,N-dimethylacetamide/N-methylformamide to
dissolve the tetramer and thiomolybdate. The proton
donor, Me3pyrHCl (in acetonitrile), was introduced into
the tetramer/thiomolybdate solution via cannula. Progress
of the reaction was followed spectrophotometrically; the
final electronic spectrum is shown in Figure 12. After a
few hours stirring, the products were precipitated with a
ten-fold volume excess of tetrahydrofuran. Initial
attempts at recrystallization resulted in isolation of the
starting material (tetramer) characterized by optical

spectrum and melting point.

54

Figure 12. The electronic spectrum of 2:1:2 equivalents
ratio of (Eth)2[Feu84(S-t-Bu)q], (EtuN)2MoSu and
MegpyrHCl in spectrograde dimethylacetamide (DMA). After
initially mixing the reactants, the length of the
absorption "shelf" increases with time, however subsequent
attempts at recrystallizing the solid obtained from this
solution resulted in loss of the absorption feature.

55

000

NH chum“;

cpzex
oon
_ _

00v

 

00»
0.0

 

56

Was the molybdenum ever present in the form of a

2-
4 bands

due to reduction of molybdenum to another soluble

bridged complex, or was the disappearance of MoS

molybdenum complex? The micro-assay for molybdenumsz was

employed (with suitable controls), and the results indi-

cated that a significant amount of molybdenum was present

(theoretical in sample: 4.3 x 10'4

4

gram Mo, found in
sample: 1.1 x 10- gram Mo) in the crude unrecrystallized
products mentioned above. In addition, nmr spectra of the
crude product also exhibited a shift of the bound thiolate
ligand resonance (6 = ca. 1.2 vs. 6 = ca. 2.8) to lower
field, indicative of a substantial change in the environ-
ment of the Fe4S4 core (Figure 13). However, efforts to
recrystallize this material have proved futile. The elec-
tronic spectra show shift of the primary absorption peaks
back to values characteristic of starting material, indi-
cating that rearrangements, not decomposition, are
occurring. One possible solution to this problem would be
to bind a tridentate (tripod) ligand to the Fe4S4 cluster,
so as to tie up three of the four iron atoms with a high-
affinity ligand. This would leave only one thiolate
ligand free to participate in substitution reactions (in
addition to providing a potential model for the P-clusters
of nitrogenase). An attached tripod ligand could prevent

polymeric species from forming and restrict rearrangements

during attempts to synthesize both the molybdenum-bridged

57

Figures 13a and 13b. Each figure is a nuclear magnetic
resonance spectrum in d6-dimethylsulfoxide (99%-d) with
resonances in units of 6 (delta). Figure 13a is the con-
trol spectrum of the solvent, the resonance at 2.45 is the
solvent impurity. Figure 13b is the same 2:1:2 equiva-
lents ratio of tetramer, thiomolybdate, and trimethyl-
pyridiniumchloride present in Figure 12. Note the shift
of the "bound" t-butyl resonance from ca 2.58 to 1.32.
The other proton resonances (tetraethylamine and tri-
methylpyridine) have been obscured by both solvent
impurity and the "bound" t-butyl protons.

58

 

 

 

WW

 

 

 

 

 

 

 

1 Figure 13a
1 L l J l l I
8.0

 

Figure 13b

0.0

 

59

and metallocene-bound species. A discussion of the tripod
and biological significance of synthesis of the tripod-

bound species follows.

4. The Tripod-Bound Species

 

2- ..
Fe4S4(SR)4 + O-P(NH-C ~g-SH)3 +

6H4

Fe4S4(SR)[(S-g-C H4-NH)3P=O]2' + 3RSH (15)

6
The synthetic feasibility of utilizing a tri-

dentate ligand to irreversibly replace three thiolate

ligands of an Fe4S4 cluster has been briefly discussed.

In addition, component I (MeFe protein) of nitrogenase has

3 to contain tetra-

been shown by displacement experiments
meric Fe-S clusters (P-clusters) that appear to be tetra-
nuclear species where one iron is found to be distinct

54 Due to the

from the others by Mossbauer spectroscopy.
equivalence of ligand sites on iron-sulfur tetramers,
substitution of less than four thiolates results in a com-
plete statistical distribution of products: mono, bis,
tris, and tetrakis. This phenomenum prevents synthesis of
a model for P-clusters through a simple ligand substitu-
tion reaction with three equivalents of thiophenol. How-
ever, successful binding of a tripod-type ligand to an
Fe4S4 core would provide the first model of this unusual

S4

cluster type. Results of ligand substitution

60

Figure 14. Pr0posed structure of the tripod—bound
species.

62

31

experiments indicate that the Fe S core has the great-

4 4
est affinity for aryl thiol functionalities. This
prompted the attempts at the synthesis of the species
described in the experimental section.' Unlike the ligands
previously discussed, the tripod should Spontaneously and
rapidly displace the coordinated alkyl thiolate from
[Fe4S4(S-t-Bu)4]2-, with concomitant formation of three
equivalents of t-butylthiol.

The tripod ligand proved to be insoluble in most

53 Nmr studies of the combination

common organic solvents.
of the tripod ligand with the tetrameric cluster,
requiring high concentrations, were therefore precluded,
and the emphasis shifted to optical studies. It was here
that the first evidence for reaction (15) was observed.
The primary electronic absorption of the tetrameric clus-
ter shifted from its normal position (417 nm) out to
approximately 435 nm (Figure 15). This behavior is simi-
lar to that observed for the substitution of thiophenol

31

for other ligands on Fe-S tetrameric clusters. However,

the spectrum displays a feature which may indicate that
polymeric forms of the cluster are present:55 a very
gradual rise to the primary absorption instead of a
sharper rise present in the parent compound (Figure 15).
A preparative scale experiment (experimental section)

showed a more significant shift of the primary absorption

to 445 nm, but the long wavelength absorption was still

63

Figure 15. The electronic spectra display the changes in
the primary absorption peak of (¢4A$)2[Feg8g(8't'BU)g]
upon binding the tripod ligand, O=P(—NH-C6Hu-g-SH)3. In
the control spectrum (tetramer in DMA ——————) the
(oyAs)2[Fe(Su(S-t-Bu)u] initially absorbs at 417 nm
(a 16,700). Addition of the tripod ligand shifts the

417 nm absorption to a lower energy absorption, 435 nm
(tetramer and tripod in DMA

 

 

64

mH mesa“;

AEcI
.nxyv

 

 

0.0

Nd

v.0

@d

0.0

0..

 

65

present. However, overnight stirring of a combination of
an acetonitrile solution of tetramer with a slurry of tri-
pod in the same solvent resulted in complete dissolution
of both species, indicating that a reaction had occurred.
Continued stirring produced a fine material, silky or
micro-crystalline, with a metallic lustre. This material
was collected but attempts at recrystallization proved
futile, and the solid eventually decomposed. The possi-
bility of the tripod ligand forming disulfides is clear.
Presence of a disulfide, whether intra- or intermolecular,
could inhibit the formation of the desired product and
result in the occurrence of detrimental side reactions,
perhaps the formation of polymers through rearrangements.
Alkyl mercaptans have the ability to reduce aryl disul-
fides; introduction of a twenty-fold molar excess of £-
butylmercaptan to a pale yellow tripod solution (aceto-
nitrile) resulted in the instantaneous change to a clear
solution. This result coupled with the long wavelength
absorption of the tetrameric product prompted an iodo-
metric titration to determine the extent to which the
disulfide had formed. No further experiments were
attempted because the tripod was found to contain just
short of one-third of the necessary thiol functionalities.
An x-ray structure study of the ligand is currently under-
way in order to ascertain its exact spatial configuration.

Future work will entail altering the basic skeleton of the

66

tripod ligand in an effort to increase its solubility and
decrease its ability to form disulfides. Under considera-
tion is alkyl substitution in positions meta and para_to
the sulfur functionality; this would increase solubility
by making the ligand's backbone hydrophobic. Also, a suf-
ficiently large alkyl group (t-Bu or i-propyl) would
create steric hindrance that could ultimately prevent
formation of polymeric entities.

Solution of the problems surrounding successful
binding of the tripod ligand to the Fe4S4 core will enable
relevant synthetic investigation to continue. The mono—
substitution of the cluster with metallocene derivatives
and bridging of two clusters with thiometallate ligands

are of special interest.

D. CONCLUSIONS AND PLANS FOR FUTURE WORK

A series of experiments has been performed in an
effort to devise pathways to the preparation of compounds
related to the metal cofactors of nitrogenase. As a pre-
lude, the method for release of alkyl-thiolate ligands
coordinated to the iron-sulfur tetranuclear core, neces-
sary for the binding of ligands with potential catalytic
and biological importance, has been examined in detail.

Subsequent research efforts has been directed toward
these goals:

1. The synthesis of compounds that combine metal
complexes of potential catalytic activity (e.g., metallo-
cene dithiolates, tetrathiometallates) with iron—sulfur
clusters that play a role in biological nitrogen reduc-
tion. Although the structure and reactivity of the
desired compounds may not exactly mimic that of any bio-
logical unit, their chemical and physical properties may
shed some light on the properties of similar units in
biological systems; further, such compounds may be of
interest in their own right as potential catalysts.

2. The synthesis of molecules that closely coincide
with the available data concerning the stoichiometry of
the FeMoco. Characterization and subsequent study of
these compounds will help answer some of the questions
involving the structure and role of the metal cofactors
in biological dinitrogen fixation.

67

68

3. The preparation of a cluster bound by a large
tridentate (tripod) ligand; this may provide a model for
the P-clusters of nitrogenase, as well as furnish a means
of avoiding the problems encountered in the preparation of
the above compounds (rearrangements, polymer formation,
multi-substitution).

Work on the metallocene-bound iron-sulfur clusters
has not yet resulted in the isolation of the desired com-
pounds; hence a major revision of the synthetic approach
is necessary. As previously discussed, the reaction of
the Fe4S4 core with molybdocene derivatives, in lieu of
their titanocene analogs, will be examined. The potential
model of the FeMoco, two iron-sulfur clusters bridged by a
thiomolybdate, approximates the stoichiometry (m8Fe, ~65=/
Mo atom) and at least a portion of the proposed structure
(Mo-S-Fe-S segments) of the metal cofactor more closely

15’16 Since

than the two Mo-Fe-S species reported to date.
the thiomolybdate bridged species has only been charac-
terized spectroscopically in solution, efforts to produce
a pure crystalline solid will continue. Included in the
problems that hinder synthesis of these compounds are
those of rearrangements to less suitable species, forma-
tion of insoluble polymers, and formation of multiple
substitution products. The tripod ligand was designed to

minimize these concerns. Insolubility of the tripod and

its formation of intra- and/or intermolecular disulfides

69

have been major problems in its utilization. While the
structure of the tripod is being determined by x—ray
crystallography, the design and synthesis of a more
soluble tripod ligand is being considered. The tripod-
bound species should simplify the synthesis of iron—sulfur
clusters with metallo-ligands by favoring substitution at
only one iron.

Future work with the metallocenes may include binding
the metallocene derivatives of other metals (e.g., Zr, Nb,
Ta, W) to an iron-sulfur tetranuclear core and studying
the effect of changing size and number of d electrons on
the new systems. Since many of these metallocene ligands

39 upon reduc-

are known to bind small molecules [N2, CO]
tion, the generation of lower oxidation states by electro-
chemical and chemical means will be investigated, as will
the reactivity of any of the species produced with CO, N2,
etc. If necessary, the pentamethylcyclopentadienyl
derivatives will be prepared to avoid the n-o rearrange-

56 Of interest

ments common for reduced M-Cp complexes.
with regard to the FeMoco would be the use of thiotung-
state instead of thiomolybdate in the preparation of the
bridged species; the effect of the substitution on the
physical and electronic properties of the metallate-bound
iron-sulfur tetranuclear cluster could be examined. Some

future effort will be directed toward preparation of per-

substituted (Fe4S4LE’) metallocene and metallate

70

complexes. Recently in our laboratory, while this thesis
was being written, H. C. Silvis has been successful in
preparing a tetra-substituted thiomolybdate cluster,
[Fe4S4(SZMoSZ)4]6-, using the methods outlined here.
While of indirect biological significance, examination of
these compounds will contribute to our understanding of
the Fe—Sz-M unit. Four metallo-ligands coordinated to an
Fe4S4 core may also prove to be of catalytic interest.
Other tripod types being considered, exclusive of those
previously mentioned, are those that will bind through an
oxygen terminus (phenol, carboxylate, etc.) instead of
through sulfur. Iron-sulfur clusters are capable of bind-

ing oxygen functionalities,S7

and while their biological
significance is limited, the ability to preferentially
effect other metallo-ligand substitution reactions at one
iron will be of synthetic and catalytic importance.
Finally, isolation of analytically pure crystalline
solids will subsequently entail many physical and chemical
studies. Among studies performed will be an x-ray struc-

ture determination, electrochemical, spectrosc0pic, and

binding studies.

BIBLIOGRAPHY AND NOTES

11.

BIBLIOGRAPHY AND NOTES

 

G. N. Schrauzer, Chemistry, 2g, 13 (1977).

 

A. L. Lehninger, Biochemistry, Second Edition, 1975,
Worth Publishers, pp. 720-722.

 

W. H. Orme-Johnson, L. C. Davis, M. T. Henzl, B. A.
Averill, N. R. Orme-Johnson, E. Munck, and R.
Zimmerman, "Components and Pathways in Biological
Nitrogen Fixation”, Recent Developments in Nitrogen
Fixation (W. Newton, 3. R. Postgate,TL. Rbdriguez-
Barrenco, Editors), Academic Press, New York, N.Y.
(1977).

B. A. Averill, J. Rawlings, and W. H. Orme-Johnson,
manuscript in preparation.

W. H. Orme-Johnson, Ann. Rev. Biochem., ii, 159
(1973).

R. H. Holm and J. A. Ibers, "Iron Sulphur Proteins"
(W. Lovenberg, Editor), Vol. 3, Academic Press, New
York, N.Y. 4

W. H. Orme-Johnson, L. C. Davis, M. T. Henzl, N. R.
Orme-Johnson, E. Munck, and R. Zimmerman, manuscript
submitted for publication.

J. Rawlings, V. K. Shah, J. R. Chisnell, W. S. Brill,
R. Zimmerman, E. Munck, and W. H. Orme-Johnson, J;
Biol. Chem., 253, 1001 (1978).

 

R. H. Holm, Accts. Chem. Res., 12, 427 (1977).

 

W. H. Orme-Johnson, G. 8. Jacob, M. T. Henzl, and
B. A. Averill, Advances in Chemistry Series No. 162,
Bioinorganic Chemistry II, 1977, p. 389.

 

 

S. P. Cramer, K. O. Hodgson, W. O. Gillum, and L. E.
Mortenson, J.A.C.S., 100, 3398 (1978).

71

12.

13.

14.

15.

16.

17.
18.

19.
20.

21.
22.
23.

24.

25.

26.
27.

72

S. P. Cramer, W. O. Gillum, K. O. Hodgson, L. E.
Mortenson, E. I. Stiefel, J. R. Chisnell, W. J.
Brill, and V. K. Shah, J.A.C.S., 100, 3814 (1978).

V. K. Shah and W. J. Brill, Proc. Natl. Acad. Sci.
USA, ii, 3249 (1977).

 

P. T. Pienkos, V. K. Shah, and W. J. Brill, Proc.
Natl. Acad. Sci. USA, lg, 5468 (1977).

 

T. E. Wolff, J. M. Borg, C. Warrick, K. O. Hodgson,
R. H. Holm, and R. B. Frankel, J.A.C.S., 100, 4630,
(1978).

 

G. Christon, C. D. Garner, and F. E. Mabbs, Inorg.
Chem. Acta, 2;, L189 (1978).

 

R. E. Eibeck, Inorg. Syn., ;, 128 (1963).

 

 

H. Koepf and M. Schmidt, Z. Anong.Allem. Chem., 340,
139 (1965).

 

H. Koepf and M. Schmidt, Angew. Chem., ll, 21 (1965).

 

B. A. Averill, T. Herskovitz, R. H. Holm, and J. A.
Ibers, J.A.C.S., 2;, 3523 (1973).

J. L. Carleis, Ann. Chem., 232, 259 (1886).

 

 

Kruss, ibid., 225, 29 (1884).

 

Purity is arbitrarily gauged by measuring the depth
of the valley following the initial tetramer
absorption (N420 nm, e = 17,400). A value that is
>10% of the absorbance value of the 420 nm peak is
acceptable. See Figure l.

The pyridinium salt should be added dropwise.
Approximately 1 drop of solution every 5-10 seconds.
Adopted from Anal. Chem., i;, 2000 (1955) and a con-
tribution from W. H. Orme- ohnson, University of
Wisconsin-Madison.

 

Thioglycolic acidsmercaptoacetic acidEH-S-CHz-COZH.

Clarity of the solution denotes lack of traces of
HCl, HN03, and HBr.

28.

29.

30.
31.

33.

34.

35.

36.

37.

38.

39.

40.

41.
42.

43.

73

Skoog and West, Fundamentals of Chemical Analysis,
Second Edition, 1969} p. 453.

 

Kolthoff, Sandell, Meehan, and Bruckenstein, Quanti-
tative Chemical Analysis, Fourth Edition, 1971, pp.
842 and 859. .

 

A discussion of the phenomena will follow.

L. Que, Jr., M. A. Bobrik, J. A. Ibers, and R. H.
Holm, J.A.C.S., 32, 4168 (1974).

7, 528 (1975).

G. R. Dukes and R. H. Holm, J.A.C.S.,

T. Herskovitz, B. A. Averill, R. H. Holm, J. A.
Ibers, W. D. Phillips, and J. F. Weiher, Proc. Natl.

 

Acad. Sci. USA, 22, 2437 (1972).

 

Cotton and Wilkinson, "Advanced Inorganic Chemistry",
Third Edition, Interscience Publishers, 1972, p. 864.

M. A. Bobrik, R. H. Holm, and K. O. Hodgson, Inorg.
Chem., gg, 1851 (1977).

G. B. Wong M. A. Bobrik, and R. H. Holm, Inorg.
Chem., ;;,’578 (1978).

"Handbook of Chemistry and Physics", 54th Edition,
CRC Press, 1973-1974, p. D-128.

Integrations based on the ¢uAs (7.70) peak of the
tetramer, the free thiol peak (1.36), and the 9-
methyl peak (2.35) of MegpyrHPFs.

J. E. Bercaw, "Reduction of Molecular Nitrogen to
Hydrazine at Titanium and Zirconium", ibid., ;,
p. 25.

A. R. Dias and M. L. H. Green, J. Chem. Soc. (A),
2807 (1971), and all references therein.

 

2,4,6-trimethylpyridine, Stadtler nmr #8015.

R. H. Holm, W. D. Phillips, B. A. Averill, J. J.
Mayerle, and T. Herskovitz, J.A.C.S., 22, 2109
(1974), Figure 1.

Ibid., Figure 2.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

74

Insoluble in (hot or cold) DMSO, DMF, THF, H20,
CHC13, CH3CN, and acetone.

A. J. Gordon and R. A. Ford, "The Chemists Com-
panion", John Wiley and Sons, Inc., 1974, p. 59.

Triethylamine, Stadtler nmr #29, 2.42 (CH2), 0.98

(C53)-

Triethylamine-HBr, Stadtler nmr #268, 3.20 (C32),

1.45 (C33).

Triethylamine-HCl, Stadtler nmr #9397, 3.21 (C32),
1.45 (CH3).

N,N,N',N'-tetramethyl-l,8-naphthalenediamine,
Stadtler nmr #13067, ca. 7.31 (3,4,5,6—H), 6.91
(2,7—5). 2.77 (Me).

Because of the bound tfbutyl proton's proximity to
the paramagnetic FeuSq core its nmr resonance is par-
tially broadened into the baseline. A discrepancy of
three protons in thirty-six is within experimental
error.

E. Diemann and A. Muller, Coord. Chem. Rev., 10, 79
(1973).

 

B. V. DePamphilus, B. A. Averill, T. Herskovitz, L.
Que, Jr., and R. H. Holm, J.A.C.S., 3;, 4159 (1974).

Personal communications from H. C. SilVis.
Experimental section.

The tripod was soluble in DMA, DMF, NMF, HMPA.

E. Munck, Rhodes, W. H. Orme-Johnson, L. C. Davis,

W. J. Brill, and V. K. Shah, Biochem. Biophys. Acta,
400, 32 (1975).

 

 

Private communication, unpublished results, Dr. B. A.
Averill.

Cotton and Wilkinson, "Advanced Inorganic Chemistry",
Third Edition, Interscience Publishers, 1972, p. 761.

R. W. Johnson and R. H. Holm, J.A.C.S., 100, 5338
(1978).

All nmr data are in units of 6 (delta).

   

   

MIC

WM)[[filflfifllflfl'fljiflflflflfijfyfi[I

RIES

H