.V...‘ ~‘v-1a 11-» -

W111; 1.111;:
‘.1‘;‘§§ P31913115
1:1,»...

..

fiat-1w“
, ‘uQ w... .;.....1......
..

..' . .1 1 “Mu-r...
...“

EL .11:- :iv v5 v
rx-cn-Tn

u ,
‘T’g‘gi: n‘u‘
It I
1.1 a“; '3. Q‘Wu‘ifln": L‘ ‘
{.431 ‘35.
:‘1

m . 1; 1 1 F1 11-“
.. 1 1.. . W..-
V‘.AUL§I «div-'11:. - 1'39...» “.1 -\.~.. .7“
I \1-Iu‘u "flu 5.. §:\
, .4 ' ., - ‘u“"‘ my, “H33" £3113 111:1»... 4.... at}: TV“. »
. 1 '11.“??? «5:12;: “1
Y z)! 1 J < g

1‘. . "131‘ .,1 _

. .3
1.1‘\‘I;i§'RMJ:1 “k ‘ ‘¥. r
41;:

“I -1-
‘__‘..7 ‘4 ..1...\~::ij‘3:n~1;u~1lfl ,"
- fry" ”1:;‘- 51'3". £1 1.. a
«.11 1.5.». ‘w': i... a. 1-1.;
(Ivy-'1 . .L :4!-A-I.A~

:1.-

.2311:-
5 "‘kfik

" ‘1‘1..~;:{:‘3
a“ ,
fir

I Ike
741:1.qu I

' ‘2
135-13;

"5:: .
‘1-
. . h 5
"1319 Mt

 

11$

1
“1.1.
«1- n

A:
1.11;

r
q... w...
r

'1 .(. .7-
. '.' 19‘:

 

.1
.11 rf’
13.91....
.1111
, , ,...-..-..-
..,. n-.-
1-

 

.u.
v; -.
J', “,,,...¢...11~..-
'.11o.1':v~‘1-:1 '
,u
,.,......1.1. m'
.a-.,—-o

aways

Illllllllll'llllll‘lHi1llllllllllllllllllllilllllllllllll

mm377

This is to certify that the
dissertation entitled
Synthetic Approaches to Porphyrin-Fe4S4

Assemblies, FAB-MS Characterization of the

of the Fe S Clusters and Ligand Binding of
4 4 presented by

Ni-oxoporphyrins

Wen-Lian Lee

has been accepted towards fulfillment
of the requirements for

Ph.D. degree“, Chemistry

Chew?

Major professor

 

Date Z- 7 _ f2"

MSU is an Affirmatiw Action/Equal Opportunity Institution 0-12771

 

LIBRARY
Michigan State
"University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

;

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ll

 

 

 

 

 

 

 

 

 

 

 

 

 

l—jl

MSU I. An Affirmative Action/Equal Opportunity Institution
cmmui

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SYNTHETIC APPROACHES TO PORPHYRIN-Fe4S4 ASSEMBLIES,
FAB-MS CHARACTERIZATION OF Fe4S4 CLUSTERS AND
LIGAND BINDING OF Ni-OXOPORPHYRINS

By

Wen-Lian Lee

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Chemistry

1992

ABSTRACT

SYNTHETIC APPROACHES TO PORPHYRIN-Fe4S4 ASSEMBLIES,
FAB-MS CHARACTERIZATION OF Fe4S4 CLUSTERS AND
LIGAND BINDING OF Ni-OXOPORPHYRINS

By
Wen-Lian Lee

The two newly-discovered naturally occurring iron
isobacteriochlorins, siroheme and heme d1, have generated much interest
because both are involved in ecologically significant nitrite reduction
processes. In the first part of this research, nickel porphyrindione and
related porphyrinoids have been examined for their axial ligand affinity.
In general, electron-withdrawing or positively charged Ni(II)
porphyrinoids favor the formation of hexacoordinate complexes. The
ligand binding equilibrium constant is a function of the basicity and
reduction potential of free base porphyrinoids. X-ray crystallographic
studies revealed that the tetracoordinate Ni(AcOOEP-trione) complex has a
characteristic saddle-shaped conformation. However, the hexacoordinate
Ni(AcOOEP-trione)(py)2 and Ni(OEP-dione)(py)2 possess much less ruffled
porphyrin conformations, in accordance with the Ni(II) spin state change
upon axial ligation.

In pursuit of an active site model of the siroheme-dependent nitrite

reductase, we attempted the use of appropriately-tailored porphyrin-thiol

Wen-Lian Lee
ligands to achieve porphyrinyl-Fe4S4 assemblies. A series of nickel-
porphyrin thiol ligands, equipped with tetra-, di-, and mono- thiol function
groups has been successfully synthesized. Three types of nickel
porphyrinyl-Fe484 assemblies have been obtained. The preparation of the
mixed ligand complex [Fe4S4(SPh)2(NiP82)]2' appeared to be the closest
example of a porphyrin-Fe4S4 assembly.

Fast atom bombardment mass spectrometry (FAB-MS) has been used
to analyze a series of iron-sulfur clusters, (A)2Fe4S4X4, where A: R4N or
Ph4P, and X=Cl, Br, SEt, SPh. Clusters with mixed Cl, SPh ligands have
also been studied. The best FAB-MS results for these clusters were obtained
with 3-nitrobenzyl alcohol (NBA) and 2-nitrophenyl octyl ether (NPOE) as
matrices. The most unique feature of the negative-ion FAB mass spectra is
the identification of the intact ionic core, [Fe4S4X4]', preformed anions
[(A)Fe4S4X4]‘, and a series of cluster fragment ions. A mechanism is
proposed to explain the formation of small [FemSn] clusters through
unimolecular reduction processes that involve only +2 and +3 oxidation
states for the Fe atoms. This work demonstrates that FAB-MS can be

employed as a valid method for rapid molecular weight determination as

well as structural elucidation of [Fe4S4] cluster-containing complexes.

To my Family

iv

ACKNOWLEDGEMENTS

I would like to express my sincere gratitude to Professor Chi K.
Chang for his encouragement and support throughout the course of this
research work. I would also like to thank Professor Mercouri Kanatzidis
who taught me the syntheses of iron-sulfur clusters. Special thanks are
extended to Professor John Allison, Dr Douglas Gage, and Dr. Zhi-Heng
Huang who taught me how to apply FAB mass spectrometry to the field of
bioinorganic chemistry. I would also like to thank Professor Shih-Ming
Peng at the Department of Chemistry, National Taiwan University, who

carry out the X-ray structural determination of Ni porphyrinoids.

Financial support from Michigan State University in the form of
teaching assistantships, and National Institute of Health in the form of
research assistantships are gratefully acknowledged.

My deepest appreciation is due to my wife, Dr. I-Chia Chen, and my

family for their love, patience and encouragement.

TABIE OF CONTENTS

Pace
LIST OF TABLES .............................................................................. x
LIST OF FIGURES ........................................................................... xii
LIST OF ABBREVIATIONS .............................................................. xvi
GENERAL INTRODUCTION .............................................................. 1
CHAPTER I AXIAL LIGATION OF NICKEL
PORPHYRINDIONE AND RELATED
PORPHYRINOIDS ....................................................... 6
INTRODUCTION ...................................................................... 7
RESULTS AND DISCUSSION .................................................... 10
1. Ultraviolet-visible Absorption Spectra .............................. 10
2. Equilibrium Constants of Nickel Porphyrinoids and
Basicity of Free Base Porphyrinoids ................................. 15
3. Redox Potentials of free Base and Nickel Porphyrinoids ..... Z)

4. Description of the Structure of Ni(AcOOEP-trione XI),
Ni(AcOOEP-trione) (py)2 (II), and Ni(AcOOEP-

dione)(PY)2(III) .............................................................. 24
CONCLUSION ......................................................................... :9
EXPERIMENTAL SECTION ...................................................... {B

1. Materials ..................................................................... {B
2. Physical Measurements ................................................. 40
3. Cyclic Voltammtry ........................................................ 40
4. Preparation of Ni(II) Porphyrinoids ................................ 40
5. Preparation of Ni(II) (OEP-trione) and Ni(II) (AcOOEP- .....

trione) .......................................................................... 41

6. Preparation of Ni+(PhOEP)ClO4- ..................................... B
7. The Basicity of Porphyrinoids ......................................... 43
8. Spectrometric Titration .................................................. 44
9. Crystallography ............................................................ 44

CHAPTER II SYNTHETIC MODELS FOR ASSIMILATORY

NITRITE REDUCTASE ............................................... 46
INTRODUCTION ..................................................................... 47
RESULTS AND DISCUSSION .................................................... 51

1. Ligand Design .............................................................. 51

2. Synthesis of Appended Nickel Porphyrin Thiol Ligand ...... 53
3. Synthesis of Mono-substituted Porphyrin Thiol Ligand ...... 58
4. Synthesis of Iron-sulfur Cluster ...................................... 6
5. Preparation of Ni-porphyrinyl-Fe4S4 Assemblies ............... 6
6. Characterization of Ni-porphyrinyl-Fe4S4 Assiblies ........... 67
SIMMARY ............................................................................... 72
EXPERIMENTAL SECTION ..................................................... 74
1. Physical Measurements ................................................. 74
2. Materials ..................................................................... 75
3. Synthesis of Appended Porphyrins .................................. 75
4. Synthesis of Mono-substituted Porphyrin Thiol Ligand ...... 81
5. Preparation of Iron-sulfur Cluster .................................. 87
6. Preparation of Ni-porphyrin-Fe484 Assemblies ................. 88

CHAPTER III CHARACTERIZATION OF IRON-SULFUR

CLUSTERS BY FAB MASS SPECTROMETRY ............... 90
INTRODUCTION ..................................................................... 91
EXPERIMENTAL SECTION ...................................................... 93
RESULTS AND DISCUSSION .................................................... 94

1. Mass Spectral Characterization of Iron-sulfur
Clusters ....................................................................... 94

A . Positive- and Negative-ion FAB Mass Spectra of
Halogenated Clusters, (A)2Fe4S4Br4 and

(A)2Fe4S4Cl4....................................... ...................... 95

B. Negative-ion FAB Mass Spectra of Thiolated

Iron-sulfur Cubane Clusters, (A)2Fe4S4(SR)4 ,,,,,,,,,,,,,, 104
C. The FAB-MS Analysis of Mixed Ligand Cubane
Cluster ................................................................... 112

2. Proposed Fragmentation Mechanisms: Correlations
betweenNegative Ion FAB-MS Data and Known

Iron-sulfur Cluster Chemistry in Condensed Phases ....... 115
CONCLUSIONS ...................................................................... 1%
REFERENCES AND NOTES .............................................................. 128

LISTOFTABLES

page
Wavelength of Absorption Maxims (in nm ) of the Nickel
Porphyrinoids in Various Coordinating and Noncoordinating
Solvents ................................................................................... 11
Experimental Ligand Binding Equilibrium Constants for Ni
Porphyrinoids and the Basicity of Related Free Base
Porphyrinoids .......................................................................... 16
Redox Potentials (Volts) of Free Base and Nickel
Porphyrinoids .......................................................................... 21
Crystallographic Data for Ni(AcOOEP-trioneXI),
Ni(OEP-dione)(Py)2(II), and Ni(AcOOEP-trione)(Py)2(III) .............. 27
Comparsion of Selected Bond Distances (A) and Angles
(deg) in Ni(AcOOEP-trioneXI), Ni(AcOOEP-trione)(py)2(II),
and Ni(OEP-dione)(PY)2(III) ....................................................... E
Summary of stereochemical parameters of reduced porphyrin
species .................................................................................... 31
The Various Oxidation States of E. Coli Sulfite/Nitrite
Reductase ................................................................................ 49

10

ll

12

14

10.

ll.

12.

l3.

l4.

15.

1H NMR (8) Assigments for Free Base Porphyrins (Ga-6c)
and Nickel Porphyrins (78-7CX3OOMHZ, CDC13) ............................ 57

1H NMR (8) Assigments of NiPSAC(13), NiPSz(19). (NiPS)2(20)
and NiPSH(21a), NiPSH(21b) (300MHz, CD013) ............................. 64

Selected 1H NMR (5) Assigments for Complexes I-III (300MHz,
CDCla)..........; .......................................................................... 70

Cations in the FAB Mass Spectra of Complexes, (A)2Fe4S4X4
(designated as M), X = Br and Cl, using the matrix NBA .............. 1(1)

Relative Intensities of Anions in the Negative Ion FAB Mass
Spectra of Complexes, (A)2Fe4S4X4, X = Br and Cl, using

the matrix NPOE ..................................................................... 101

Relative Intensities of Peaks in the Negative Ion FAB Mass
Spectra of Complexes, (A)2Fe4S4(SR)4, R=Et and Ph, using

the matrix NPOE ..................................................................... 107

Relative Intensities of Peaks in the Negative FAB Mass
Spectrum of (Ph4P)2Fe4S4(SEt)4, (designed as M), using

the matrix NBA ....................................................................... 111

Relative Intensifies of Peaks in the Negative Ion FAB Mass

Spectrum of the complex, (Ph4P)2Fe4S4(SPh)2012, in DMF/NPOE . . 114

LIST OF FIGURES

POGO
Examples of metalloporphyrinoids and their parent ring 1:-

conjugation ............................................................................... 2

UV-visible spectra of Ni(OEP-trione)and Ni(AcOEP-trione)

in toluene solution .................................................................... 12

Spectrophotometric titration of 5x10‘5 M NiT(F5)PP

in toluene solution with pyrrolidine. The concentrations

of pyrrolidine are as follows : 0, 0.6, 1.2, 2.4, 3.6, 4.8, 6.0, 7.2,

10.2, 30 mM. Inset: plot of log [(Ai-Ao)/(Ac-Ai)] vs. log [B].

B = pyrrolidine ......................................................................... 13

Spectrophotometric titration of 1x10'4 M Ni(AcOOEP-trione) in
toluene with pyridine. The concentration of pyridine are as

follows : 0, 1.24, 2.48, 3.72, 4.96, 6.20, 30 mM. Inset: plot of
log [(Ai-Ao)/(Ac-Ai)] vs. log [Py]. Py = pyridine ............................. 14

Correlation between log Keq of nickel porphyrinoids and
pK3 of free base porphyrinoids (open points: pyrrolidine
binding and solid points: pyridine binding) .................................. 18

Cyclic voltammograms of +2l+1, +110, and 0/-1 processes
of Ni(OEP-trioneXA) and Ni(AcOOEP-trioneXB) recorded at
IOOmVS'1 in CHzClz, 0.1M Bu4N C104 supporting electrolyte .......... 23

(a) Correlation between pK3 and Egg of free base keto-
substituted OEP series: 1. HzOEP, 2. H2(OEP-one),
3. H2(OEP-dione), 4. H2(OEP-trione). (b) Correlation
between Keq and E112 of nickel keto substituted OEP series:
1. NiOEP, 2. Ni(OEP-one), 3. Ni(OEP-dione), 4. Ni(OEP-trione).
5. Ni(AcOOEP-trione) ............................................................... %

Molecular structure and atom names of Ni(AcOOEP-trione)(l).
Hydrogens are omitted for clarity. Themal ellipsoids are drawn

to enclose 50% probability ........................................................... %

Edge-on view of Ni(AcOOEP-trione) that illustrate the
deformation of the macrocyclic skelton. The deviations of the

macrocycle from a plane defined by four nitrogens. ...................... 3)

Molecular structure and atom names of

Ni(AcOOEP-trione)(py)2(II). Hydrogens are omitted for clarity.

Themal ellipsoids are drawn to enclose 50% probablity .................. 33

Molecular structure and atom names of Ni(OEP-dione) (py)2

(III). Hydrogens are omitted for clarity. Themal ellipsoids are
drawn to enclose 50% probablity ................................................. 34

Edge-on view of the skelton of Ni(AcOOEP-trione)(py)2 (II). The

deviations of the macrocycle from a plane defined by

four nitrogens .......................................................................... 35

xiii

13.

14

15

11

Edge-on view of the skelton of Ni(OEP-dioneXpyh (III).
The deviations of the macrocycle from a plane defined

by four nitrogens ...................................................................... $

(a) Schematic description of the bridging cofactors,
Fe4S4 cluster and siroheme of nitrite/sulfite reductases.

(b) The structure of siroheme ..................................................... 48

1H NMR spectrum (300 MHz, CD013) of NiP(SAc)2(7b).
Me4Si peak is not shown ............................................................ 59

1H NMR spectrum (300 MHz, CDC13) of (a) NiPSAc (13).
and NiPSH (218) ....................................................................... 63

Far-infrared spectra of (a) (Ph4P)2[Fe4S4(SPh)2(NiP(Sz))]
and (b) (Ph4P)2[Fe4S4(SPh)2C12]. The vibration mode at
529 cm'1 corresponds to the aromatic C-H bending of Ph4P+ ........... Q

1H NMR spectrum (300 MHz, DMSO-ds) of
[Fe4S4(SPh)2(NiP(82))]2’ (III) ...................................................... 71

1H NMR spectrum (300 MHz, DMSO-ds) of
[Fe4S4(SPh)2012]2’(25) ................................................................ 73

A portion of the Positive-ion FAB mass spectrum of
[(Et4N)2Fe4S4Br4] (M), using the matrix NBA ............................... 97

N egative-ion FAB mass spectrum of [(Et4N)2Fe4S4Br4] using
the matrix NPOE. Matrix ions are designated by an (*) ................. Q

xiv

Comparison of isotope abundance for an experimentally
observed (scribed bar) and theoretically calculated (blank bar)
clusters (a) [(Et4N)3Fe 43 4131.41). and (b) [Fe 43 4131-4} ........................ 102

Negative-ion FAB mass spectrum of (Ph4P)2Fe4S4(SPh)4,
using the matrix NPOE. Matrix ions are designated by an (*) ....... 105

Negative-ion FAB mass spectrum of (Ph4P)2Fe4S4(SEt)4,

using the matrix NPOE. Matrix ions are designated by an (*).
Assignment of cluster peaks: 1. [Fe282(SEt)2]‘, 2. [Fe4S4]', 3.
[Fe4S4(O)]', 4. [Fe4S5]‘, 5. [Fe4S4(SEt)]', 6. [Fe485(SEt)]‘, 7.
[Fe4S4(SEt)2]', 8. [Fe435(SEt)2]', 9. [Fe4S4(SEt)3]‘, 10. [Fe485(SEt)3]‘.

11. [Fe4S4(SEt)4]', 12. (Ph4P)[Fe4S4(SEt)4]‘ .................................... 108

N egative-Ion FAB mass spectrum of (Ph4P)2Fe4S4(SEt)4,
using the matrix NBA. Matrix ions are designated by an (*) ......... 110

353C

AcOO.

DFDP

HF

33:50

acac

AcOOEP-trione

DFDPTMP

DMF

DMSO

iBC

MeOEC

OEP

OEP-one

OEP-dione

OEP-trione

LIST OF ABBREVIATIONS

acetylacetonate

5-acetoxy-2,2,7,7,12,12,17,18-octaethyl-3,8,13-
porphintrione dianion

4.8-diformyl-2,6-di-n-pentyl-1,3,5,7-tetramethylporphyrin
dianion

N ,N-dimethylformamide

dimethyl sulfoxide

isobacteriochlorin

2-hydro-2-methyl-3,3,7,8,12,13,17,18-octaethylporphyrin
dianion

2,3,7,8,12,13,17,18-octaethylporphyrin dianion
3,3,7,8,12,13,17,18-octaethyl-2-porphinone

dianion

3,3,8,8,12,13,17,18-octaethyl-2,7-porphindione

dianion

2,2,7,7,12,12,17,18-octaethy1-3,8,13-
porphintrione dianion

PhO

PE

PhOEP

Py

TBAP

T(F5)PP

TMPyP

TPC

'I‘PiBC

TPP

2,2,7,7,12,12,17,18-octaethyl-2-phenyl-3,8,13-
porphyrin dianion

pyridine
tetra-n-butylammonium perchlorate

5,10,15,20— tetrakis(pentafluorophenyl)porphyrin

dianion
tetrahydrofuran
tetrakis(1-methyl-4-pyridyl)porphine dianion

2,3-dihydro-5,10,15,20-tetraphenylporphyrin

dianion

2,3,7 ,8-tetrahydro-5,10,15,20-tetraphenylporphyrin

dianion

5,10,15,20-tetraphenylporphyrin dianion

dive

ran

it

GENERAL INTRODUCTION

Nature has selected a rich variety of porphyrinoids to take part in a
diversity of fundamental biological functions in all kinds of organisms
ranging from bacteria to plant, and from insect to manl. Examples shown
in Figure 1 are the better known porphyrin family compounds; hemin (iron
porphyrin)2, chlorophyll a (magnesium chlorin)3, bacteriochlorophyll a
(magnesium bacteriochlorin)3, siroheme (iron isobacteriochlorin)3, vitamin
B 12 (cobalt corrin)4 and coenzyme F430 (nickel corphin)5. These
metalloporphyrinoids represent the exceptional examples of nature's fine-
tuning of the macrocych ligand sphere as well as the central coordinated

metal to optimize a particular biological functionl.

This thesis research deals with the isobacteriochlorin class of
compounds. The two naturally occurring iron complexes of
isobacteriochlorin (iBC) whose structures were determined during the last
decade are siroheme3 and heme d15:7; both are involved in ecologically

significant nitrite reduction processes.

As outlined in the Scheme I, nitrite reduction in the biological world
involves many metalloenzymes. Nitrate is first reduced to nitrite by

molybdenum-containing nitrate reductase3v9. Studies have shown that

assimilatory nitrite reductases which are present in plants have Fe4S4

cluster and siroheme cofactors10'12. Siroheme is the active site carrying out

the remarkable 6—e1ectron reduction of N02“ to NH313'14, a process whose

mechanism is essentially unknown.

1

iiiiiiiii l-fi Illa! uuutur‘alutlnll-Ec:u\——-fii=-.-h o‘afi tav~A-~h~&?.rwv~ r~. nth-u‘vv~

 

 

2: 3.53.5

a -m¢kouo3oot3oom

domuauowaoé ME» Ronda £05 and memoaneaoaflofi mo scion—sum .n 0.3.3..—

01

be
on
the

90F

str

 

 

 

 

Scheme I
6 e '
” P1113
AununflauxquunusRcducuuc
2 e- (Fe4S4-sirdreme)
Nitrite \ - -
4 -e 2 e
Rommmmc
auntaouan) "'Pq2() . P‘Z
Ifimfinfihnoqybfiuhc PfiunusChudc
Rcmmnmu: iRammumc
(Cytochrome cd) (Cu min)

In the parallel denitrification route, N02' is reduced to N20 and N2.
Our knowledge on the dissimilatory nitrite reduction is equally inadequate.
The delineation of the general pathway only occured recently. It is now
believed that a common nitrite reductase, cytochrome cd1 present
overwhelmingly in denitrifing bacteria15:16, reduce nitrite to N20 which is
then converted to nitrogen through a separate nitrous oxide reductase (a

copper enzyme)17v13.

Cytochrome cd1 contains an unusual green heme called heme d1, its

structure and properties have recently been studied extensively by Chang

and coworkers6»7:19'21.

 

HOOC COOH
Heme d,

4

Part of our research is to continue the ongoing effort to understand the
intrinsic properties of dioneheme and related metal complexes. We focus
on the study of axial ligation of nickel (II) complexes of porphyrindione and
related porphyrinoids. It is particularly interesting that such a study is

relevant to the chemistry of nickel-containing F430, in which axial

coordination may be the key to its function.

On another front, in order to understand the overall chemical
principles involved in biological nitrite reductase, it is desirable to obtain
chemical models that mimic the active site of siroheme-dependent nitrite

reductase present in plants.

In this regard, the principal objectives of this thesis research

included the following:

 

To determinate the affinity constants between a number of bases and nickel
(II) complexes of porphyrindione and related porphyrinoids.
(II)antheuc_mndels_£Qr_asaimilam:1Jutnthductase To build nickel
porphyrin equipped with appropriately-positioned thiol ligands for
attaching to a Fe4S4 unit in order to achieve a porphyrin-linked iron sulfur
cluster.

(III) u: .‘ctuo 09:. .:.-:,o := 0023:“ Hun.
To characterize synthetic iron-sulfur clusters and biologically relevant

complexes by fast atom bombardment mass spectrometry.

In the following chapters, chapter 1 is devoted to investigating the
axial ligation of a series of Ni(II) porphyrinoids to determine the effects of

5

pyrroline ring saturation and substituents on nickel coordination. Chapter
I also describes the crystal and molecular structures of a novel
tetracoordinate Ni(II)porphyrinoids, as well as two hexacoordinate Ni(II)
complexes, which were characterized by a single crystal x-ray difl‘raction
study. In chapter II, we demonstrate an initial approach to synthesize the
models for assimilatory nitrite reductase. Finally, the characterization of
Fe4S4 clusters by fast atom bombardment mass spectrometry is presented
in Chapter III, a mechanism is proposed to explain the formation of small

[FemSn] chusters through unimolecular reduction processes.

CHAPTERI

AXIAL LIGATION OF NICKEL PORPHYRINDIONE
AND RELATED PORPHYRINOIDS

13

CC

of

(I)
A

1':

['71

.m _

INTRODUCTION

Metallohydroporphyrins have been identified as being essential in a
variety of biological systems including nitrite and sulfite reductases 10'
13:22-23, and S-methyl coenzyme M reductase24:25. Of special interest is
cofactor F430, a nickel containing porphyrinoid, found in the latter enzyme
of methanogenic bacteria, which catalyzes the final reduction step in
methane genesis, where MeSCHzCH2803' is reduced to methane as shown

in eq. 1.

CH33-CH2CH2803' + 2H+ + 2e' ----> CH4 +HS-CHZCHZSO3' (1)

Althrough X-ray structure data are still lacking, the detailed
structure of F430 has been determined5:25'27. Systematic chemical and X-
ray structural studies on related hydroporphyrinoid carried out by
Eschenmoser, Kratky and coworkers have revealed that these low spin
nickel (II) complexes share a general structural characteristic: a saddle
shaped, ruffled conformation of the macrocycle28'30. This saddle
conformation has been interpreted “as resulting from the tendency of the Ni
(II) to achieve saturation of its electrophilicity by pulling the four nitrogen
atoms within the equatorial plane towards the coordination center". Thus,
release of ruffling strain energy on going from 4-coordinate Ni(II) to the
larger 6-coordinate nickel(II) form is said to be responsible for increasing
the residual electrophilicity of nickel in the axial direction, which is why
F430 exhibits a high axial reactivity.

01

re
he:
Nit

8

However, counter-examples have been found. Fabrizzi?’1 et. al.
described macrocyclic nickel complexes that have planar structure but
show high axial reactivity. Recently, Kaplan, Scott and Suslick have
measured the equilibrium constants for the binding of six imidazoles with
ZnTPP, ZnTPC, and ZnTPiBC as well as two pyrrolidines with NiTPP,
NiTPC, and NiTPiBC32. Only very small difference in Keg is obtained as a
function of the macrocycles. These authors suggested that even though
ring reductions result in distortions from planarity they do not cause an

increased ability for axial ligation.

In order to examine the effects of pyrrole ring saturation and
substituent groups of the macrocycle on the axial ligation, the purpose of
this reseach is to determine the affinity constants between nitrogen bases
and a number of nickel porphyrinoids. Particularly, we focus on the study
of axial ligation of nickel porphyrindione and related systems including:
NiOEP(1), Ni(OEP-one)(2), Ni(OEP-dione)(3), Ni(OEP-trioneX4), Ni(AcO-
OEP-trione)(5), Ni(MeOEC)(6), and Ni+(PhOEP)ClO4'(7), as well as three
reference compounds; NiTPP(8), NiT(F5)PP (9), and Ni(DFDMTMP)(10) as

shown in Scheme 1.

The results illustrated that the basicity of the tetra-aza porphyrinoids
appear to be more important that structural effects; the decrease in the ring
basicity would favor the formation of hem-coordinate Ni(II) complexes. We
also reported three molecular structures characterized by single crystal x-
ray diffraction: tetracoordinate Ni(AcOOEP-trione)(I), as well as two
hexacoordinate Ni(II)porphyrinoids, Ni(AcOOEP-trione)(py)2(II), and
Ni(OEP-dioneXpy)2(III).

Sociezegz 3:..sz $5.:sz

  

 

I
O

O

I

ADWO—UENOEVLZ onUmOogz AnXoaoEdmOOo<xz Avxogmbdmoxz
a:

O .. ,,
axoaoedmoxz axocodmozz :Emgz

Hana

  

1. Ultra

'1
the mat
maxim
391 rm:
porphw
(tide i1
absorpt
pattern

Ni(OEF

i
converg
Coordin
that a
solutio
Coordin
Ni(OEi
“Don t]
PETidin

Shifts t}

eleCtrOD

10

RESULTS AND DISCUSSION

1. Ultraviolet-Visible Absorption Spectra

The UV-visible spectra of Ni-porphyrinoids are strongly dependent on
the nature of the peripheral substituents at the ring. For example, the Soret
maximum in dichloromethane occurs at 424 nm for Ni(OEP-trione) and at
391 nm for NiOEP (Table 1). The trend indicates a greater red shift for
porphyrins with more conjugated keto substituents at the pyrrole periphery
(vide infra). Moreover, the electronegative groups not only shift the
absorption bands to longer wavelengths but produce large variations in the
pattern of the visible bands. Figure 2 shows the intensive red shift of
Ni(OEP-trione) and Ni(AcOOEP-trione).

Particularly striking is the red shift for the Soret band upon
conversion of four-coordinate species to six-coordinate complexes in the
coordinating solvents, pyrrolidine and pyridine. Figure 3 demonstrates
that a red shift ca. 20 nm in the Soret region is observed when toluene
solution of NiT(F5 )PP is titrated with pyrrolidine. However, the
coordination-induced shift of visible spectra is variable. In the case of the
Ni(OEP-trione) and Ni(AcOOEP-trione), a dramatic blue shift is observed
upon the formation of six-coordinate complexes. As shown in Figure 4,
pyridine binding to Ni(AcOOEP-trione) to form six coordinate complex
ShifistheQbandfrom706nmto688nm.

The spectral shift upon ligand binding is indicative of changes in the
electron configuration of the Ni(II) complexes, from low spin in a typical

11

 

 

 

Rm .m5 .m3 Sc 6mm .83. Sc 65 .fiv Agfinmemz

H8438". Beams .3. £34.34: $38.22

mum .vmv .va mom «on .oum .mmv 5.3. Em .mam .Ev mgmz

mum .me .8» .5 4% .2w dam .va mac 4mm .43 segmmofixz

8w .mmv .mwv .aov cam 65 .HS .mmv cam .93 .wwv .vov amofizxz

wwm .va .wmm .mev cam 63 .wum .va m2. .wmw $8 .wuv Amootadmooofimz

«.8 4mm 6% .33 >8 {mm .N8 .8“. .NE. o2. £8 dam .vav $85¢m0x2

vac .vwm .Hmv .mmv 6mm 0mm .mwm .bmo .va .mmm Hum .wmm .wmv .wwm Aooomodmcxz

3m .NE .mmv .Hdv .vmm 3m .mum .3 .3... .mmm So .H S .wvm 43. .Num Aonodmozz

3m .mHm .Sm mum .Em .omv 3m .mS .Hmm mmomz
86:8 Shanghai 885.

.3538 gafivuoooccz 98

”533330 got: 5 mvmcoannuom 3:32 23 mo 33 cc «SEE 539823. mo finds—963 .m 0358

43333 2333
E anagmmoosmz nauseousdmoxz as «58% 03335 .a 2.6m...—

 

 

Es a
2.3 8» 8o 8» 8.. 8n
J1; - - - u G
s I I \-
l. a; \ ll \\ .
— _ \ I 1 a
. . x
, a a
.. : .
1 3
l I. 8
.l I. 8

 

 

 

 

8—.

c-OtXW3

wczqfiufiucuu‘

432

t '2:

0.6 - g

in

2
5: 0‘

§

i‘ r
0.4 i- ,

ABSORBANCE

.0
N

13

 

  

 

 

 

. vvvvvv
-l.6 -l.2 ~1.° ~O.I -0..

10: [Bl

 

 

 

340 400 450 500

2. (nm)
Figure 3. Spectrophotometric titration of 5x 10'5 M N iT(F5)PP

in toluene solution with pyrrolidine. The concentrations of

pyrrolidine are as follows : 0, 0.6, 1.2, 2.4, 3.6, 4.8, 6.0, 7.2, 10.2,
30 mM. Inset: plot of log [(Ai-Ao)/(A¢-Ai)] vs. log [B]. B =

pyrrolidine.

0.6

A.”
o

moz<mm°wm<

9m
0

rrtnanu.

14

 

 

 

 

 

 

 

0.6
w 0.4 P
O
z
<
a » ~
0
to
‘2
0.2 i- .
J
l r I ‘
610 650 700 750 780
1 (nm)

Figure 4. Spectrophotometric titration of 1x10'4 M Ni(AcOOEP-
trione) in toluene with pyridine. The concentration of pyridine
are as follows : 0, 1.24, 2.48, 3.72, 4.96, 6.20, 30 mM. Inset: plot
of log [(Ai'Ao)/(A¢-Ai)] vs. log [Py]. Py = pyridine.

tetracoo
are font
ineufl.l
onto th
complex
subsfitt
formati

favorab'.

2. Equii

Porphy;

T
Dirroli

Spectre]
When t
bases,
tetracm
except

GOgKeq
ifitA£()

15

tetracoordinate to high spin in the hexacoordinate complex. As complexes
are formed with additional ligands, electron density builds up in the central
metal. If this electron density may be partially dissipated by delocalizing
onto the porphyrin ring through dn-n“ back bonding, the formation of
complexes with additional ligands will be more favored. If the peripheral
substituent is able to withdraw electron density from the porphyrin, the
formation of complexes with additional ligands should be even more

favorable.

2. Equilibrium Constants of Nickel Porphyrinoids and Basicity of Free Base
Porphyrinoids

The equilibrium constants for the binding of two different bases,

pyrrolidine (6 base) and pyridine (1: base), were carried out by
spectrophotometric titration in toluene. The values are given in Table 2.
When tetracoordinate nickel porphyrinoids are treated with coordinating
bases, hexacoordinate adducts are formed in toluene. However,
tetracoordinate species might be present in the strong coordinating solvent
except for those nickel porphyrinoids with high ligand binding aflinities
(logKml > 4). Figures 3 and 4 show the titration studies of NiT(F5)PP and
Ni(AcOOEP-trione), respectively.

In general, the average values of equilibrium constant Keq obtained
in our studies were determined at four wavelengths in each case, and

reported in Table 2. The logarithmic method was used for the
determination of Keq33, In the present case, eq 2 could be rewritten as eq. 3,

16

.Amm codename:

3.4 3M mac Sand mm 393:8 Esta—Ego “scam: Bacofitonuo SE. 0 .vm monouomom n .ooofioeuaooZom a

 

 

Nd Nd Em AESQMQVMZ
N...“ H6 md mmam'mvfimz
m6. oNA nvfi @952
ad md VOHXAHOAQXZ
Nd. m4 NJ. AOHOezvmz
fin m.m Aoflomhuumnm000<vmz
mgv fin CA Aefic€9-mm0vmz
N.” Wm m4 Aeflogtmflcvmz
Nd 9N mum AenoxmmmOvmz
- md Qm AHHOMZ
SEES 832.5 85388 SE
«-2 .Oomm um Sufi 0&3

 

.aaoafiafiom 82. 8E 33%

as steam 2: as. masseuse...“ 2 .5 3.588 grandam Beam 255 3:08:85 .« 3:3.

17

Keq=NiP(B)2KNiPXB)2 (2)

Keg = [(Ai-A‘\<a)/(1’x¢-Ai)llBl‘2 (3)

where A0 = absorbance of a solution of NiP in the absence of Lewis bases, Ac
= absorbance of solution of the same concentration of NiP in the presence of
a large excess of base, Ai is the absorbance of a solution containing an
intermediate concentration of bases and [B]=[B]o=[B]eq. A plot of log [Ai-
Ao/(Ac-Aifl vs. log [13] is expected to give a slope of 2 and an intercept of log
Keq. Table 2 lists the values of Keq for the reaction of bases with Ni
porphyrinoids. The log-log plots of NiT(F5)PP (Figure 3) and Ni(AcOOEP-

trione) (Figure 4) were linear and all gave slopes close to n=2.

Our results suggest that the keto substituent on the pyrroline ring is
one of the major factor determining the ligand affinity. The tendency of
ligand binding of NiOEPs increases with the number of the keto groups.
The conjugation interaction of the carbonyl group with the u-system of the
porphyrinoid macrocycle reduces the electron density at the macrocycle
nitrogen and should, therefore, lead to an decrease in electron density at

the metal center.

Combining o-electron withdrawing effects (inductive effect) and 1c-
interactions (resonance effect) of substituents either in the pyrroline ring or
meso position, the basicity of free base of porphyrinoids offers a parallel
trend for the axial binding affinity. Examination of Table 2 shows a
remarkable correlation between pK3 and logKeq. Figure 5 depicts the
excellent linear relationship obtained between logKeq and pK3 for the OEPs

18

 

    
 
   
 

pyrrolidine

logKeq

T

 

 

 

9K3

man-95, Correlation between log ch of nickel
porphyrinoids and pK3 of free base porphyrinoids (open
points: pyrrolidine binding and solid points: pyridine

binding).

19

and the porphyronoids that were chosen for the comparison purpose. The
basicity of the ring system can ofi‘er a relatively clear prediction of the axial
ligand aflinity. For example, with almost identical basicity of HzMeOEC
(pK3 4.2) and H2TPP (pKa 4.4)34 one finds the similarity in the axial ligand
affinity of NiMeOEC (logKeq 1.5) and NiTPP(logKeq 1.2) with pyrrolidine.
With different basicity between pyridine (pKa 5.7) and pyrrolidine (PKa 2.9)
this correlation holds true.

The distinct axial ligand affinity of the nickel ion in enzyme F430 may
well be related to its biological function524’25. However, its specific catalytic
task remains to be established. In order to explain the high axial reactivity
of F430, a model complex Ni+(PhOEP)'ClO4‘, which contains a monovalent
anionic tetraaza ring, similar to the corphin(i. e. hexahydroprphyrin) ring
system of F430, has been synthesized for comparison. Complexation of F430
with imidazole in dichloromethane shows a large Keg (logK1 is 2.7 and
logK2 is 2.0)”. However, Ni+(PhOEP)'ClO4' did not have high ligand
affinity for imidazole. The ligand binding equilibrium constant (logKeq 2.3)
of the bis pyrrolidine adducts of Ni+(PhOEP)(ClO4)' is close to Ni(OEP-one)
(logKeq 2.5) and Ni(DFDMTMP) (logKeq 2.2). However, in comparison with
NiOEP or NiTPP, the axial ligand binding affinity of Ni+(PhOEP)(ClO4)' is
relatively high, since the substituent electronic effect from the phenyl group
is not expected to be as large as from peripheral keto or formyl groups. The
unusually high axial ligand afinity of Ni+(PhOEP)(CIO4)' with pyrrolidine
might be due to the positive charge of the Ni metal center accepting axial
ligands in order to build up the electron density in the metal center. This
might be also the reason why [F430M]+ClO4' exhibits a high axial reactivity.

20

Regardless of the nature of the substituents or their position on the
porphyrinoid periphery, the pK3 values of the macrocycles give a
quantitative prediction of the reactivity of Ni porphyrinoids in terms of axial
ligation. In general, basic axial ligand and acidic porphyrinoids favor the
formation of hexacoordinate complexes. The tendency toward axial ligation
in Ni(II) tetrapyrroles does not necessarily correlate with the extent of ring
ruffling. The ligand binding equilibrium constants (Keg) for NiTPP
(2.3x101), NiTPC (4.0x101), and NiTPiBC (1.9x101) in pyrrolidine solution
demonstrate that the ruffled macrocycles do not increase the propensity to
bind axial ligands32. Thus, the notion that porphyrin ring saturation as
occurred in hydroporphyrinoids in which the macrocycle appears to have
greater flexibility for the ligand binding deserves a closer scrutinization. It
seems clear that substituent electronic effects can have a greater influence
on the afinity for axial ligands, regardless of the saturation level of the ring
as postulated for the series of porphyrin, chlorin, isobacteriochlorin,

bacteriochlorin, and pyrocorphin27.

3. Redox Potentials of Free Base and Nickel Porphyrinoids

The influence of the keto, formyl, phenyl and fluorophenyl group on
porphyrinoid properties can also be examined by redox properties. Redox
potentials for Ni porphyrinoids and its free base in 0.1M TBAP/methylene
chloride solution are listed in Table 3.

The redox potentials of free base porphyrinoids examined agree well
with the published data36°33. For example, B-keto porphyrinoids exhibit

ring oxidation potentials very similar to those of the parent porphyrin and

21

Table 3. Redox Potentials (Volts) of Free Base and Nickel Porphyrinoids (vs.

 

 

SCE) 1!.

Ni-porphyrinoids 9x4. 93.2 [ESL]. 12951.2
H20EP 0.83 1.30 -1.45b

Ni(OEP) 0.76 1.23 -1.46
H2(OEP-one) 0.84 1.33 -1.36
Ni(OEP-one) 0.73 1.13 -1.30
H2(OEP-dione) 0.82 1.28c -1.28
Ni(OEP-dione) 0.71 1.03 -1.13
H2(OEP-trione) 0.73 1.30d -1.29 -1.60°
Ni(OEP-trione) 0.631 0.81 -0.90c
Ni(AcOOEP-trione) 0.70d 0.84 -0.91

H2MeOEC 0.83 1.11c -1.58

Ni(MeOEC) 0.48° 1.05c -1.53b
Ni(PhOEP)C104c 0.441 0.75 0.900

H2TPP 0.70b 0.92 -1.20 -1.58
NiTPP 1.20d 1.29f -1.40

H2T(F5)PP 1.18b 1.47‘3 0.90 -1.30
NiT(F5)PP 1.30 -0.95 -1.52b
H2(DFDPTMP) 1.06 1.48 -0.94 -1.15
Ni(DFDPTMP) 0.98 1.38 -1.06

 

‘3 0.1 M Tetrabutyl ammonium chloride (TBAP) in methylene chloride with

100 mV/s scan speed. b Quasi-reversible. c Irreversible. d Redox potential
for Ni(II)/NI(III). e Redox potential for Ni(II)/Ni(I). f Ni(III)TPP"‘-.

22

the ring reduction is made easier by increasing the number of B-keto group

as shown in Table 3.

The first and second oxidation process of Ni porphyrinoids are
characteristic of diffusion-controlled one-electron process, except that
Ni(OEP-trione) and Ni(AcOOEP-trione) show an overlapping but reversible
cyclic voltammogram (exemplified in Figure 6). As seen in Table 3, redox
potentials for the first oxidation process in methylene chloride solution
exhibit a considerably small difference: 0.76 V (NiOEP), 0.73 V (NiOEP-one),
0.71 V (NiOEP-dione), and 0.70 V(NiOEP-trione), and 0.63 V (NiAcOOEP-
trione). The first one-electron on’dation of Ni porphyrinoids is tentatively
assigned for the cation radical species. The second electron oxidation is
possibly removing electron from Ni center to form Ni(III). However, recent
evidence39 from EPR data demonstrate that Ni(III)(OEP-trione) could be
produced in methylene chloride at liquid nitrogen temperature. The data
in Table 3 indicate that within the OEP series, the second oxidation product
Ni(III) is stabilized by increasing the number of B-keto groups. With an
acetoxy subsituent at meso position, Ni(AcOOEP-trione) appears to behave
like Ni(OEP-trione).

The B-keto NiOEP derivatives demonstrate that ring reduction
becomes easier with increasing number of B-keto groups as shown in Table
3. The most dramatic difference of the reduction potential could be ca. 560
mV between Ni(OEP) (-1.46 V) and Ni(OEP-trione) (-0.90 V). In these one-
electron reduction process, typically, the formation of anionic radicals is

assumed. However, an irreversible wave was observed for

Ni(PhOEP)(ClO4') at -0.90V. This could be the reduction of Ni(II) to Ni(I)

(A)

 

511A I

 

(B) '

 

 

4.80

 

L ' L 1 1 l
1 .0 0.5 0 -0.5 -1.0

vous (vs. scs)

Figure 6. Cyclic voltammograms of +2/+l, +1/0, and 0/-1
processes of Ni(OEP-trione)(A) and Ni(AcOOEP-trione)(B)
recorded at 100mVs'1 in CH2C12, 0.1M Bu4NClO4 supporting
electrolyte.

24

analogous to the observation that a model complex of F430 has a reduction
potential at -0.86 V in CH2C12 with Bu4N C104 as electrolyte”.

Reduction potentials of free base and nickel keto-substituted OEPs are

markedly dependent on their basicity“). In most cases the porphyrinoids
that are more difficult to reduce also have larger pK3. A correlation of the
reduction potential EH2 values of Ni porphyrinoids and free bases versus
their basicity was studied. Figure 7A shows the linear relationship
obtained between pK3 and E1/2. This is especially true for H20EP and its
keto substituted compounds. An excellent linear relationship between the
equilibrium constant Keq of NiOEPs and E112 could also be obtained (Figure
7B).

The studies of reduction potentials of free base porphyrinoids provide
a quantitative measurement of porphyrin electron density on the central
porphyrin nitrogen atom influenced by peripheral substituents. The trend
of reduction potentials (especially for OEP series compounds) reflects the
difference between B-ketoOEP and hydroporphyrin.

4. Description of the Structure of Ni(AcOOEP-trione)(I), Ni(AcOOEP-
trione) (py)2 (II), and Ni (OEP-dione)(py)2 (III)

A. Ni(AcOOEP-trione) (I).
The crystal structure of Ni(AcOOEP-trione)(l) is shown in Figure 8.

The crystallographic data for the compound are given in Table 4. The
selected bond distances and angles are listed in Table 5. The molecule

 

9K3

 

 

 

 

log ch

 

 

 

 

U V U r

4.5 . -1.4 4.3 V 1.2 . -1.1 -1.0 -o.9
E1/2

Figure 7. (a) Correlation between pK3 and E1/2 of free base
keto-substituted OEP series: 1. H20EP, 2. H2(OEP-one), 3.
H2(OEP-dione), 4. H2(OEP-trione). (b) Correlation between K9,,
and Egg of nickel keto substituted OEP series: 1. NiOEP, 2.
Ni(OEP-one), 3. Ni(OEP-dione), 4. Ni(OEP-trione), 5.
Ni(AcOOEP—trione).

330:
o
. 8 and 38
2qu A5000
2.5 98
. .9...” 383
<sz . Be
no means SE. .3
833 was tame you w :36
0.583% ”was changes.“ 83
300 some
32 . him
a 9:5
#—

88 838
‘h
we moan.

.5

  

13
mm 8.4 8

EU 0.,
flex, mmu

 

 

 

8.3.8 38... 8.8... .8 ES. .88

.8 85.3383 83

8H a «.5 8.88 .8.— 8838 we 62

8... 8m 8.. 858...? 8 .82
down—03.83

8.8.8.8 8.8.8.8 8.8.8.8 .38 was .53 .32
88 88 2.8 2885888883 35

88 88 88 838:8 33 8 .oz

8.3. 8.3. 8.8 88 3383..

8... .8... .8... 85 .8... .8... 8... .88 .8... as .38 .980
3. 3. «.8 18 .s

8.2 82 38.. «so.» . 8.89
Eng Tm : m 959% madam
.988 .e 88.38 .N .8888 .N a. .> .u
8.8 .8888 8.8 8.. .m
.838 .3888 88.8 8.. ...
8.8 3.3.8.. 8.8 8.. .8
388.8 888.8 888.8 4. .8
688.8 888.8 883...: 4 ...
88...». .888...S .88... .8. .8
8.88 8.8 8.88 .38
28828880 2822.880 280328830 8.888...
a H— a $589.30
9.8.888...

.8892 was .88....388588885.2 Assassinmooosfi 3.. 38: 838.83.88.80 .3 338a.

28

Table 5. Comparsion of Selected Bond Distances (A) and Angles (deg) in
Ni(AcOOEP-trioneXD, Ni(AcOOEP-trione)(py)2(lI), and Ni(OEP-
dione)(py)2(III).

 

 

I I I II I

Bond distance8(A)

Ni-N( 1) 1.949(5) 2.0458(25) 2.090(5)
Ni-N(2) 1.918(5) 2.079(3) 2.097(5)
Ni-N(3) 1.923(5) 2.0859(25) 2.030(5)
Ni-N (4) 1.957(4) 2.075(3) 2.052(5)
Ni-N(5) 2.189(3) 2.216(7)
Ni-N(6) 2.190(3) 2.186(6)
Bond Angles(deg)

N(1)-Ni-N(2) 88,7909) 88.4200) 90.58(18)
N (1)-Ni-N(3) 179.43(21) 176.3400) 178.01(25)
N(1)-Ni-N(4) 90.2409) 92.0200) 90.1309)
N(1)-Ni-N(5) 91.09(20) 87 .43(10) 86.56(22)
N(1)-Ni-N(6) 178.9409) 92.89(10) 91.06(22)
N(2)-Ni-N(3) 89.8909) 90.79(10) 88.3908)
N(2)-Ni-N(4) 179.41(10) 178.91(22)
N(2)-Ni-N(5) 90.50(10) 89.41(22)
N (2)-Ni-N(6) 89.7600) 9092(22)
N(3)-Ni-N(4) 88.75(10) 90.9309)
N(3)-Ni-N(5) 89.01(10) 91.72(24)
N(3)-Ni-N(6) 90.6800) 90.66(24)
N(4)-Ni-N(5) 89.11(10) 91.45(22)
N(4)—Ni-N(6) 90.63(10) 88.25(22)
N(5)-Ni-N(6) 179.5900) 177.60(m)

29

structure demonstrates a pronounced saddle- shape ruflled conformation of
the macrocyclic ligand system as shown in Figure 9. The molecule is
significantly distorted in the conventional fashion; the dihedral mirror
planes contain the S4 axis and opposite pairs of methine carbon atoms. The
methine carbons (mesa-carbons) have the largest displacements of any
macrocycle atom from the mean plane (NiN4 core is caplanar), alternating
0.66 A above and below. The average absolute value of the deviation of four
meso-carbons (dm) from the plan is 0.64 A. However, the Ni(P-dione)“1
exihibits a slight deformation with a dm parameter of 0.32 as shown in
Table 6.

In the structure of Ni(AcOOEP-trioneXI), the average equatorial Ni-
Nred bond distances are 1.933 A which agrees well with the Ni-Nred bond
distance in other nickel porphyrinoids. A summary of selected
stereochemical parameters of metallohydroporphyrinoids is given in Table
642. If comparing the Ni-Nred bond distance with two related nickel B-
oxoOEP complexes, Ni(oxoOEP) and Ni(P-dione) in Table 6, both Ni-Nred

and Ni-Np distance of Ni(AcOOEP-trione) is shorter.

In general, the ring ruining feature of Ni(AcOOEP-trione) is very
similar to Ni (II) hydroporphyrinoid327'30 regardless of the keto and acetoxy
substituents. As noted in the introduction, the saddle conformation has
been interpreted to result from contraction of the coordination core in nickel
hydroprophyrinoids. According to the principle that the porphyrin core is
enlarged upon pyrrolic ring saturation, the Ni-N bonds would become
unfavorably long for Ni(II) ion in Ni(AcOOEP-trione). In order to adjust the
Ni-N bond lengths to stablize the tetracoordinate nickel(II) ion, the four

.m:mm.5m: .5...“ .3 “5.5% 283 a 8on

£93....an one no 2358306 25. dofioxm $388988 23 mo nomaaauomov

2... 32:3: 3.: 325-8852 8 33> 8-8.x... .8 2.6....

 

88.322: “saga—ovum 25,—. :5 8.85888» voEE.vm 9::— 838 =< .8538 sup—am:
.58 85 .3 8583 S35 85 89: 8838 £5.88 -32: 85 no 85838388 33838 owns: 8.: 8m Eu 3 ”4 «can .3 8:03:33 88.5 85
:33 95.838 aunt woo—62 85 :o 8965 85 go .339 5 6853518 958—8.. 85 838.88% 8539:6938 25. 3V “4 E 82...; =< A3

 

«.33 828 8 88.8 88888.8 8888838 38388388882828
8.83 828 8 88.8 8:38 8888.8 383888.deng
8.83 £5 8 88.8 8883 88888.8 8538888382
88 8 88.8 @883 8888.8 8888-582
3. 8 88.8 8538.8 8383 88038585
88 H 88.8 8888.8 3883
88 H 88.8 8:883 E883 88808582
88 8 88.8 388.8 $883 838 889582
88 88 8 88.8 8:3 8:3 :3 880582
B 8 88.8 88883 8883 § 6888082
88 8 88.8 388.8 8:3 88 69.8082
88 8 88.8 8588.8 8883 E 693082
8 8 88.8 8:838 8:883 833 £88.85
88 8 38.8 8:883 6883 888 2838888
88 8 88.8 8:883 8:883 383 $8832
88 8 8.8 8:83 8883 88. $8832
88 8 3.8 8888.8 8883 888 $8832
Juan 838
«am .82 can 8952-2 92-2 8539:5980 883880

883838 :Efi—uog @8383 we 38888.89 3838:8338 mo 585m .c 038,—.

32

nitrogen atoms are pulled toward the metal center, preserving the square
planar of coordination geometry; thus the four mesa-carbon atoms are
situated alternatively above and below this coordination plane. The
resulting distortion of the ligand framework leads to the saddle
conformation as shown in Figure 8. However, the deformation also builds

up conformation strain.

Finally, it should be noted that the expected increase in the Cp-Cp and
Ca'Cfi bond distances (not shown in Table 5) of the reduced pyrrole ring are
observed. It is also expected and generally observed that pyrrole ring is no

longer planar”.

B. Ni(AcOOEP-trione)(py)2(ll) and Ni(OEP-dioneXpy)2(III)

We have succeeded in isolating single crystals of hexa-coordinate

Ni(AcOOEP-trione)(py)2(II) from a saturated pyridine solution of
Ni(AcOOEP-trione). Similarly, single crystals of hexacoordinate Ni(OEP-
dione)(py)2(III) were also obtained in the same fashion from a saturated

pyridine solution of Ni(OEP-dioneX4). The molecular structure and atom
names for Ni(AcOOEP-trioneXpyh (II) are presented in Figures 10 and 11.

Selected bond distances and angles are listed in Table 5.

Figures 12 and 13 are the edge-on view of the molecules II and III
including the deviation from NiN4 core plane. Both molecules are six
coordinate as expected, with the pyridine rings characteristically tilted with
respect to the normal of the mean plane. The pyridine plane is oriented to

C31 C31

0!

C).

 

a

Figure 10. Molecular structure and atom names of Ni(AcOOEP-
trione)(py)2(II). Hydrogens are omitted for clarity. Themal ellipsoids are

drawn to enclose 50% probablity.

G

C3
,. 9
CH
C‘ CO
01

Cl

I CZ

C1

3 c3:

 

Figure 11. Molecular structure and atom names of Ni(OEP-dione)
(PY)2(III). Hydrogens are omitted for clarity. Themal ellipsoids are drawn

to enclose 50% probablity.

damage? .38 .3 @232. 283 a Spa ofiohuobaa 23 no 33333
25. .Ae «En—Xenotadmocoflmz no usage—m on... we 33> noomvm .Nu 9:5?—

 

m~.N+

6:0»..qu 3cm .3 626% 233 a Spa 398°wa 05 mo 33333

23. :8 xmgxmcaomedmoxz co 5:3: 2: .3 32> 8&5 .2 9:53

ad.

8d-
0 do.

26+ 86+

 

C . +
bad... «cc

 
 

37

minimize the steric interaction between porphyrin nitrogen atoms and

pyridine hydrogens43 .

The nonplanarity of the macrocycle of II is not as significant as
structure I. The mesa-carbon atom has displacements from the mean
plane, alternating 0.35 A above and 0.28 A below as shown in Figure 12.
However, the maximum Cp atoms of adjacent pyrrole rings are
alternatively displaced above and below the macrocycle plane by more than
0.6 A. The most dramatic difference between complex III and II is that
complex III has an almost flat ring gemometry as shown in Figure 13. The
average and maximum absolute value of the deviations from the plane are
0.08 A and 0.21 A. The methine carbons have small displacements from the
mean plane, alternating 0.10 A and 0.09 A above and below, respectively.

The electronic configuration is expected to be manifested in the bond
lengths of Ni-N44. The spin state of Ni(II) change upon axial ligation. The
nickel d32-y2 orbital is empty in the tetracoordinate complex, however, the
addition of two axial ligands to form hexacoordinate complexes II forces
one of the dz2 electrons into the dx2.y2 orbital. The presence of this electron
should lengthen the Ni-N bond compared to the tetracoordinate complexes.
Similarly, the axial Ni-N pyridine bonds are expected to be long, because the
unparied electron in the dz2 orbital.

The diameter of the coordination core, 4.14 A for structure II and 4.13
A for structure III (average of the two transannular N, N distances) is
distinctly larger than the core size of 3.8711, for the structure I and 3.91
A.for Ni(P-dione)41. The average lengths of the four equatorial Ni-N bonds

38
is 2.07A for II and 2.08 A for III. It closely matches the corresponding
values (2.11A) found in unstrained Ni(II)N4 octahedral complexes”, and
2.04 A of bis(imidazole) NiTMPyP43.

The axial nickel-nitrogen bond is free of steric constraints of the
porphyrinato core, and is extented somewhat further to 2.190 A, and 2.20
A, respectively. These are comparable to Ni-N bonds in other hexa-
coordinate nickel porphyrinoid complexes, for example, the distance of Ni-
N pyrrole bond in bis(imidazole)NiTMPyP43 is 2.160 A. The Ni-N bond
length of the pyridine ligands also agrees well with the the bond length of
2.115 A in trans-Ni(acac)2(py)244.

As noted in structure I, the deformation allows the ligand system to
adjust its coordination sphere to the specific requirements of the Ni(II) ion,
but at the same time it builds up the conformation strain. Ni(P-dione), with
two reduced pyrrole ring compared with structure I, should have less
conformation strain. We believed that a process leading to lengthen Ni-N
bonds is due to the presence of an electron in the nickel dxz-y2 orbital, and at
the same time the strain energy associated with the saddle deformation
could be released upon the axial ligation. As consequence of the longer
equatorial Ni-N bonds, the hexacoodinate complex III exhibits an almost
flat ring geometry compared with tetracoodinate complex Ni(P-dione).
However, the hexacoodinate complex II does show a residue ruffled

geometry possibly due to the steric crowd at the methine bridge.

CONCLUSION -

In this study, we illustrated the fact that the tendency toward axial
ligation in Ni(II) tetrapyrroles does not necessarily correlate with the extent
of ring rufiling. The presence of electron-withdrawing substituents is one
of the major factors contributing to the increase of axial ligand amnity of
nickel porphyrinoids. We also demonstrated that basicity and reduction
potentials of free base porphyrinoids provide a quantitative prediction of the
ligand binding equilibrium constants. In general, electron-withdrawing or
positively charged Ni(II) porphyrinoids favor the formation of

hexacoordinate complexes.

A novel tetracoordinate Ni(AcOOEP-trione) complex has been
isolated and characterized by single crystal X-ray studies. The structural
analysis reveals that Ni(AcOOEP-trione) has a characteristic saddle shaped
conformation of most nickel hydroporphyrinoids. Two most significant
hexacoordinate nickel(II) porphyrinoids, which were grown in neat

pyridine, have been successfully isolated and chracterized. X-ray structure
of Ni(AcOOEP-dione)(py)2 and Ni(OEP-dione)(py)2 illustrates Ni(OEP-

dione)(py)2 has an almost perfect planar structure, however, Ni(AcOOEP-
trione)(py)2 shows a residue rufiled geometry possibly due to the steric

crowd at methine bridge.

EXPERIMENTAL SECTION

1. Materials

00m
Pm
met
Silit
(34(

ana

2P.

spe
Inc
FA]

mm

3.(

Sys

of:

V011

ran

H

40

All solvents and regents were of reagent grade, purchased
commercially, and used without further purification except mentioned.
Pyridine and pyrrolidine were distilled from barium oxide. Toluene and
methylene chloride were distilled from calcium hydride shortly before use.
Silica gel for column chromatography (60-200 mesh) was from J .T. Baker
(3405). Preparative silica gel plates were from Analtech, Inc. For
analytical TLC, Eastman 13181 chromatography sheets were used.

2.Physical Measurements

1H NMR spectra was recorded on a Varian Germini-300
spectrometer, in “100%” chloroform-d (min. 99.96 atom% D, from Isotec
Inc.) with the residual CHCl3 as the internal standard set at 7.26 ppm.
FAB mass spectra were obtained on a JEOL HX-110 HF double focusing

mass spectrometer operating in the positive ion detection mode.

3. Cyclic voltammetry

All cyclic voltammograms were measured using a Bioanalytical
System CV -1A unit in a specieally constructed glass cell. The cell consists
of two platinum bead electrodes sealed through the cell wall and a total
volume of 0.2 ml. All measurements were carried out in 0.1M tetra-n-
butylammonium perchlorate-dichloromethane solution under argon, scan

rate 100mVs’1.

4. Preparation of Ni(II)Porphyrinoids

41 ‘

H20EP was synthesized and oxidized with H202 in concentrated
H2804 according to literature procedures“. The crude reaction products
were separated to afford H20EP, H2(OEP-one), H2(OEP-dione), and H2(OEP-
trione). H2(MeOEC) was prepared from H2(OEP-one) by reacting with
freshly prepared methyl lithium in THF and working up according to
literature procedures45. H2TPP and H2T(F5)PP were synthesized by the
method of Alder et. a1“. H2(DFDPTMP) was prepared as described“.

Nickel(II) porphyrinoids were generally prepared in the following
method: The free base porphyrin (50mg) was dissolved in a saturated
solution of Ni(OAc)2 in 20 ml of DMF containing 2mg NaOAc. The mixture
was refluxed under nitrogen and monitored spectrophotometrically until
there was no residual Soret band from the starting free base (ca. 30-60 min).
The DMF solution was then poured into an ice bath to give microcrystalline
precipitates. After filtration the precipitate was redissoved in
dichloromethane (20ml) and washed with brine (2x50ml). The organic
solution was dried over Na2804 and the solvent evaporated under reduced
pressure. The collected material is allowed to crystallize by slow
evaporation of the solvent or recrystallize in a minimum of CH2C19/CH3OH
to yield shining crystals of microcrystalline material. The yield were 85-
95%. All the nickel porphyrinoids were further characterized by 1H-N MR
and high resolution mass spectrometry. The material is stable in the solid
form and is stable in solution when protected from light and stored under

inert gas.

5. Preparation of Ni( OEP-trione) (4) and Ni(AcOOEP-trione) ( 5)

42

2,2,7,7,12,12,17,18-octaethyl-3,8,13-porphinetrione (58.2mg, 0.1mmol)
and Ni(OAC)2-4H20 (100mg, 0.4mmol) were dissolved in glacial acetic acid
(30ml). A trace amount of sodium acetate (5mg) was added and the solution
was refluxed for 2h. Dichloromethane (40ml) and water (80ml) were added
when the reaction vessel reached room temperature. The nickel complex
was extracted into dichloromethane. The mixture was further washed with
water (30ml), sat’d. NaHCO3 solution (30ml), and brine (2x30ml). The
organic solution was then dried over Na2804 and the solvent removed
under reduced pressure. The crude product was purified by
chromatography (30% hexane/CHzClz) on a TLC plate and gave the
nickel(II)-2,2,7,7,12,12,17,18-octaethyl-3,8,13-porphinetrione (21mg, 33%)
followed by the nickel(II)-15-acetoxy-2,2,7,7,12,12,17,18-octaethyl-3,8,13-
porphinetrione (36mg, 52%).
Ni(OEP-trione)(4): UV-visible 1mm: (toluene): nm (8M ), 424 (42 000), 626 (11
500), 668 (15 000), 716 (61 200). 1H-NMR (CDC13), 300MHz): d 0.41 (12H, t,
CH20H3), 0.52 (6H, t, CH20H3) 1.46, 1,56 (3H each, t, CH20H3), 2.28-2.36
(12H, m, CH 2CH3), 3.36, 3.39 (2H each, q, sat. CH zCH3), 7.90 (1H, s, 10-H),
7.95 (1H, s, 5-H), 7.97 (1H, s, 15-H), 8.75 (1H, s, 20-H). HRMS found: m/z
638.2763 for (M)+, C35H4403N4Ni requires m/z 638.2767.
Ni(AcOOEP-trioneX5): UV-visible km” (toluene): nm (8M ), 426 (38 200), 617
(9 600), 658 (11 000), 706 (41 500). 1H-NMR (CD013), 300MHz): 0.30, 0.32, 0.33,
0.43, 0.52, 0.57, 1.41, and 1.49 (3H each, t, CH2CH3), 2.0-2.25 (6H, m,
CH2CH3), 2.35-2.50 (6H, m, CH20H3), 2.52 (3H, s CH3CO2-), 2.70-2.82 (m, 1H,
CH gCH3), 2.98-3.10 (m, 1H, CHzCHg), 3.20-3.42 (m, 2H, CH2CH3), 7.70 (1H,
s, 10-H), 7.81 (1H, s, 5-H), 8.61 (1H, s, 20-H). HRMS: m/z 696.2856 for (M)+,
C33H4605N4Ni requires 696.2822.

43
6. Preparation of .PhOEPNi+ ( 0104‘)( 6)“8

To a THF solution of H2(OEP-one) (22mg, .04 mmol), phenyl lithium

reagent (0.5ml of 1.8M in cyclohexane/diethyl ether) was added at room
temperature until the solution turned green. The reaction mixture was
quenched and washed with water; the organic layer was seperated and
evaporated. The crude product was further purified on a silica gel column
with 40% hexane/0H2012 as eluent to afford 2-hydroxy-2-
phenyloctaethylporphyrin (22mg, 81%). After nickel insertion by the
methodology discussed above, the Ni complex was further washed by H0104
(70%) to yield Ni+(PhOEP)ClO4' quantitatively.
Ni+(PhOEP)(CLO4')(7): UV-visible km” (toluene): nm (8M ). 406 (27 000), 584
(5 400), 628 (8 600). 1H-NMR (CD013), 300MHz): d 0.22 (6H, t, CH2CH3), 0.85,
0.88 (3H each, t, CH20H3) 1.35 (3H, t, CH20H3), 1.39-1.50 (9H, m, 0H20H3),
2.28-2.36 (2H, m, CH20H3), 2.51-2.60 (2H, m, CH20H3), 2.89-3.00 (4H, m,
CH20H3), 3.02-3.25(8H, m, CH20H3). 7.85 (1H, s, 10-H), 7.87 (1H, s, 5-H), 7.91
(1H, s, 15-H), 7.96 (1H, s, 20-H). HRMS found: m/z 667.3311 for (My,
C44H49N4Ni requires m/z 676.3330.

7. The basicity of Porphyrinoids

To the anionic SDS detergent solution (2.5% sodium dodecyl sulfate
with fixed pH value), a small quantity of porphyrinoids in CH2012 was
added. The mixture was stirred vigorously until the colored porphyrinoids
solution became homogenerous (if not soluble, the mixture was adjusted to

a lower pH value to enhance solublility in micelles). pK3 values were

measured on a Cary 219 spectrophotometer by monitoring the absorbance of

44

titration curves at different pH. The pK3 was calculated with the average
pH at five different wavelengths. However, if the pK3 was lower than 1.5 the
pH value of the micelle solution was hard to control precisely. Thus, a
series of organic acids were chosen as acidic benchmarks to approximate
the pK3 of the porphyrinoid. The acids used were CHgCOOH (PKa 4.75),
HCOOH (pKa 3.75), 010H20OOH (pKa 2.86), 0120HCOOH (PKa 1.26),
0013000H (pKa 0.64), and CF3000H (PKa .23).

8. Spectrophotometric Titrations

Typical ligand binding titrations were performed in toluene. The
concentration of Ni(II) porphyrinoids were approximately 10'5 M and
ligand concentrations were varied over as a wide range as posssible. Data
were collected typically at six to eight different concentrations on Cary 219
spectrophotometer. The equilibrium constants Keq for the formation of
pyridine or pyrrolidine adduct(s) of Ni(II)porphyrinoids have been
determined by the method of Walker and coworkers33f; the data analysis
was verified by the method of Bent and French”.

9. Crystallography

Crystals of Ni(AcOOEP-trione) (I) suitable for x-ray study were grown
by the slow diffusion of hexanes into a methylene chloride solution. Crystals
of Ni(AcOOEP-trione)-05H14 were obtained. Ni(AcOOEP-dioneXpy)2 (II) and
Ni(OEP-dione)(Py)2 (III) were crystallized from pyridine to afford
(NiC46H54N602)-C5H5N and (NiC43H55N605)-C5H5N, respectively.
Although the crystals appears air-stable, they were covered with thin coats

45

of epoxy resin to prevent losing hexane or pyridine solvent molecules before
mounting on a glass fiber for x-ray single-crystal structure analysis.

The X-ray structural determinations were carried out at the
Department of Chemistry, National Taiwan University, in collaboration
with Professor Shih-Ming Peng. The data were collected on a Nonius CAD-
4 diffractometer with graphite-monochromated Mo Ka radiation in the
scane range. Accurate unit cell parameters for all the compounds were
obtained from the least-squares refinement on the 24) , o), w and 0 values of
25 machine-centered reflections. The stability of the experimental setup
and crystal integrity were monitered by measuring three standard
reflections periodically during the course of data collection. No crystal
decay was detected. the raw data were reduced to net intensifies. Empirical
absorption convertions (1t scan) were applied and the equivalent reflections
were averaged. In all compounds the hydrogen atom positions were
calculated and were included in the structure factor calculations but were

not refined.

CHAPTERII

SYNTHETIC MODELS FOR ASSIIVIILATORY
NITRITE RESUCTASE.

INTRODUCTION

On a global scale, it is estimated that more than 10 billion tons of
nitrogen are incorporated into plants annually“. Apart from a relatively
small number of species that have a symbiotic association with nitrogen-
fixing bacteria15v15:51, the bulk of the plant nitrogen arises from the
reduction of nitrate taken from soil. Studies have shown that assimilatory

nitrite reductase has two metal cofactors, a Fe4S4 cluster and an iron
isobacteriochlorin named siroheme (Figure 14). Siroheme is the active site
where NOz' interacts and receives electrons derived photosynthetically via
an electron transport chain consisting of chlorophyll-->NADPH-->FAD--
>FMN-->siroheme1°'12. The ability of this enzyme to mediate the 6-electron
reduction of N02'--> NH4+ without releasing any intermediate is
remarkable. Siroheme is also found in assimilatory/dissimilatory sulfite
reductase52. That these 2 enzymes can often reduce both substrates“):11

suggests that they have common structural and mechanistic features.

Owing to the now classic investigations of Siege] and coworkers, E.
Coli. sulfite reductase (EcSiR) (which also reduce NOz') is the best

understood of these enzymes. The enzyme is a complex hemoprotein (Mr =
685,000) with an (1304 subunit composition”. Spectroscopic
investigation822»53 of the catalytically-active [3 subunit of this enzyme
indicated that siroheme and an Fe4S4 cluster are exchange-coupled. This
coupling is maintained at all levels of oxidation states (shown in Table 7),
implicating the presence of a common bridging ligand. The active site
structure from X-ray analysis54 is depicted schematically in Figure 14, with

a 4.4 A separation between the siroheme iron atom and the nearest iron

47

.0838? me «.5633 25 as £333.88 353333 me 0838?
e8 .2336 63m .3338 «users 2: .3 8358.. seesaw 3 .3 83E

9.5225 n

 

 

322.com 25555.2

 

 

Asa ..zu .08 cu m 5:0". 8.18.13?»
2." .2 a: .3 .2 u a

 

8a .3." .2." u a 33...:
N a _ u m .53 aims”: $58 3262
.0
60 .28 o u m a. u a
n 8 .u m 5:»... $120": 358 333.38
.0
nzu .snoe S u m 3 u a
8228. 3 S u m 5:5 $.53": :59 322.8
082.87. .233 28%

 

83.3% snazEuEm =8 .93 32% 5325 332$ as. s 3.3.

50

atom in the cluster. The bridging ligand is most likely a cysteinyl sulfur
atom; however, crystallography at 3.0 A resolution is not suficient to reveal
the identity of the bridge. This view has recently been challenged on the
basis of a study of a dissimilatory sulfite reductase from Desulfovibrio
vulgaris (Du SiR)55. The ligand bridging the [F4841-siroheme cluster is
apparently exchanged by 358'2 in both oxidized and reduced enzyme. No X-
ray structural data are available for these enzymes.

A simple and attractive picture of substrate reduction by Ec SiR
involves (i) substrate binding at the vacant axial site of siroheme, (ii)
coupled steps of electron transfer from cluster to substrate, and (iii)
protonation until sulfide is formed. The electon-transfer pathway may
involve the bridge, or possibly pass directly to siroheme. From electron

density maps, the Fe4S4 cluster and periphery of the siroheme ring appear

at or near van der Waals contact“.

Synthetic models of protein-active sites have provided valuable
information concerning catalytic principles and structure-function
relationships of many metalloproteins. As the structural features of the
active site of sulfite/nitrite reductases are now available, it becomes timely
to construct models for these redox centers. Only recently has an initial
approach to the synthesis of biologically bridged assemblies been reported by
Holm et 0155.

Here we describe our efforts in persuit of an active site model of the

siroheme-dependent nitrite reductase. The present systems are not

necessarily intended as true structural models of the native [Fe4S4]-

51
siroheme assembly. Rather, we demonstate the design to synthesize
appropriately tailored porphyrin-based ligands to achieve particular Fe4S4-
nickel porphyrin configurations.

RESULTS AND DISCUSSION

1. Ligand Design

Our aim was to synthesize appropriately tailored porphyrin
macrocycles equipped with thiol ligands capable of attaching to Fe4S4 units.
Specifically modified porphyrins with appended thiols such as NiP(S4) are
designed to achieve Fe4S4-heme linkages (Scheme I). Nickel porphyrin
tetrathiol ligand 7a was thought to be well-suited in achieving our goal
because (i) they are synthetically practical57; (ii) the appended thiols are
placed in proper arrangement to span Fe--Fe distances in the Fe4S4 cube
and hold it in close proximity to the macrocych metal center; (iii) extrusion
of Fe4S4 core with similar thiols, i.e. m-xylyl dithiol has been demonstrated
by Holm et 0153. The model compound provides a full complement of
thiolato ligands and is expected to hold the Fe4S4 center with maximum
stability, as shown in the structure [Fe4S4(NiP(S4))]2' of scheme I. In
connection to this approach, we also prepared some bidentate thiol ligands,
such as NiP(SAc)2 (7b) and NiP(SAc)2 (7c), as well as two meso-substituted
porphyrin nickel complexes (21a and 21b), which provide a monodentate
thiol function. These thiol ligands would be able to link themselves to the
Fe4S4 cluster in various configurations as depicted by the structures I, II,
and III in Scheme I.

52

 

3,7

cis

isc

Si]

is:

CE

d.

53

2. Synthesis of appended nickel porphyrin-thiol ligands

The key ligand precursor bis(o-aminophenyl)-2,8,12,18-tetraethyl-
3,7,13,17-tetramethy1 prophyrin(NH2)2DPE(l) was obtained as a mixture of
cis and trans isomers via the procedure shown in Scheme II, published by
Young and Chang“. After seperation of the mixture of cis and trans
isomers, thermal atropisomerization of the trans isomer in a manner
similar to that reported by Lindsey59 allowed isolation of more pure cis

isomer (1 ).

The strategy for the preparation of the appended porphyrin-thiol
ligands is depicted in Scheme III. Attachment of sulfur-containing
appendages to the o-amino group was accomplished with acid chloride
derivatives (5a-5c) bearing thiol-acetates. As shown in Scheme IV, the acid
chlorides were freshly prepared by treating 80012 with the corresponding
acid: 3,5-bis(thioacetoxymethyl}benzoic acid (4a), 3-(thioacetoxymethyl)
benzoic acid (4b), and 3-thioacetoxypropionic acid (4c), respectively. Thus,
the appended porphyrin thiol derivatives 6a-6c were synthesized by treating
cis-(NH2)2OPE(1) with the corresponding acid chloride in the presence of
N (Et)3. The desired compounds 6a-6c can be obtained in excellent yields:
72%, 58%, and 67% respectively. The IR spectrum exhibits an amide
carbonyl stretching at 1670-1680 cm'l. The 1H NMR spectra readily
demonstrate the presence of ligation to the o-aminophenyl substituents.
Detailed assignments of proton resonances of complexes 6a-6c are listed in
Table 8. The upfield shift of the terminal CH3COS indicated that the
appended thioacetoxy functional group is within the shielding porphyrin
ring current. A particularly useful information in the 1H NMR spectra of

: oeofim

 

GENE.» a

£24104zoé n _

 

owxuzoa : z
.52: .NNOO Z Z ONO.U

2:
Z:

:9:

 

 

oxol© /\ /\i :9. /\ /\
N

02

_= 2:055

 

 

 

>_ oEonom

0v 0m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

, 0.3 0.8
l 2200:5320...
_ 500:“.2052 l 20.20208 2.036.203. 20.20.38.
.3
A... an
an 08 2.8 2.8 260
0.3 0.8 0.2.
t‘ .5 ‘lllll
20.20200... 20.20.28. 200.82
. u a
23520 33.20 a 20
2.
260
3%20\M.U/0.2mq<
mm

260

0.3..
‘
. £882.82
0.2...

 

 

5.20

57

 

 

 

$0 000. .00 00. 000 0N... 2-88.
.0 000 .0 .80 0 .80
0 .80 .0 000 .0 .80 .0 .80 .0 .80 .0 .80
0 .30 0 .80 0 0...... 0 0.0... 0 000 0 05
0 0.00 0 000 0 0...... 0 00... 0 000 0 .80
0 000 0 000 m 0.0 m .80 .0 .80 0 0.0 00....
.0 .80 .0 000
a 000.00 .0 0.0 .0 .80 .0 00.0
.0 0.0 s 0.00.00 .0 0.0 .0 ...0 .00
m 000 0 000 0.0.
0 0.... .0 00.. m .20 m 000 .0 0.00 m1.0.0 mamo
.0 000 .0 000 m .80 0 .80 .0 000 0 000 .3002.
.0 0.0 .0 .8... .0 0.0... .0 0.0... .0 000 .0 .8... .03....
0 00.. m 0.... m 00.. .0 0.... .0 00.. .0 0... .03 .3...
0 00.. 0 0.... .0 0.... m 00.. m 0.... .0 00.. .0000
m .80- m 000- m .80- m...
.z 0m .z 0m .z 0m
0.0.00.0). 0.0.00.0. 5400...). 80.8..
.3000

002800.355 8.0390. .20.... .23 .8000 8.5.0.3.. 80.. 8.... .80 00580.02 .00 ms... 2. 0 .030...

58

these types of compounds is the meso-proton of porphyrins which is
typically well-seperated from other signals. The chemical shift of the meso
proton can be diagnostic of the type of complex; while the number of signals
in this region gives an indication of the purity of the sample. In the free-
base ligand, such as 6a, the mesa proton appears at 10.27 ppm. The UV-
visible spectrum exhibits a Soret band at 408 nm and four visible bands at
508, 542, 577, 627 nm.

To prepare nickel porphyrins, free base porphyrin (Ba-6c) in CHC13
was treated with freshly prepared methanolic solutions of Ni(II) ions to give
nickel complexes NiP(SAc)4(7 a), NiP(SAc)2(7 b), and NiP(SAc)2(7c),
respectively. After 4-6 h of heating at reflux, the UV-visible spectra of the
reaction mixtures showed that the four bands of the free base collapsed to a
two band pattern, which signals the completion of metal insertion. FAB
mass spectra of the complexes exhibit the expected molecular ion cluster
peaks for nickel complexes. Further evidence for the insertion is derived
from 1H NMR spectrum, for example, NiP(SAc)2(7b) in Figure 15. The
proton resonances of internal pyrrole NH are no longer present in the NMR
spectra. Interestingly, the purity of the nickel complexes could be readily
ascertained by a dramatic high-field shift of the single mesa-proton

resonance from 10.27 to 9.45 ppm.
3. Synthesis of mono-subsituted Porphyrin Thiol Ligand
As shown in Scheme V, the key ligand precursor, biladiene-ac

dihydrobromide (8) was prepared by condensation of 3,3’-diethyl-5,5’-
diformyl-4,4’-dimethyldiprromethane and 3,4-diethylprrole via the

go... 0.0.. a. 0.8.. .05.. 80.00.0002... 0.0 3.000 020,. 8.00 6300.0... .8... 2. 0. 8.6....

tan N m w

m m N m m o.

kph-pp-errPPbbb—PbbrnppbppPlppprrppbbbPrppbprhp—bfbph.L.>—>»..pupp.—>P>pp»-p—~b.pbprhbpp

 

 

 

1147!...

 

 

59

 

 

 

 

 

 

 

 

ON
£020.20 [Ollml ' g

 

m."

> 0.202%

61

Q."

. , n _> 0.202%
a 0.... 0 ..

 

 

 

 

 

 

 

 

 

 

:0.
90 0.0 0.0 90
0.0.0.. 0.00.0020
0 l 0 .1 0 - 0
20.20 2.... 0200.200 «0.0-...
20 0000200 0 0.). m
o E \ #0
H H o H m
m H
00 0.0 70 70
0.... 0.0.0 .02.. 0. E
x. 10020 .352 :2 3 +7355 :00 .mmZ
0.0.20 5.20 00.2 .20

 

 

62

procedures developed by Johnson and Kay“. The strategy for the
preparation of mono-substituted porphyrin thiol ligand is based on the
method of condensing linear tetrapyrrole and aldehydes bearing a thiol
function“. p-Thioacetoxymethylbenzaldehyde( 12), obtained in three steps
from the a—bromo-p-tolunitrile (9) (Scheme VI)”, was allowed to react with
ac-biladiene (8) to afford 5-(p-thioacetoxyphenyl)-etioporphyrin. Similarly,
4,4’-diformydiphenyl disulfide (17) obtained in four steps from 4-
(methylthio)benzaldehyde (14)“, was reacted with ac-biladiene (8) to afford
the disulfides. All of these porphyrins were converted to nickel (II)
complexes to give NiPSAc (13), (NiPS)2 (l9), and NiPSg (20) which were
formed in high yield from the corresponding free base porphyrin by the
reaction of methanolic nickel acetate. The UV-visible spectra of 13, 19, and
20 exhibit a Soret band at 403 nm and two visible bands at 526, 561 nm. The
1H NMR spectrum of nickel porphyrin(13) (Figure 16a) demonstrate the
presence of the thioacetoxymethyl phenyl group. For example, proton
resonance of CH3COS centered at 2.5 ppm and CH 2$Ac at 4.43 ppm. Again,
the chemical shift of the meso proton can be diagnostic of these porphyrins.
The chemical shift and their integration of two different mesa-protons gives
an indication of the purity of the compound 13., (e.g. 15-H at 9.51 ppm, and
10H, 20H at 9.58 ppm., respectively). Detailed assignments of proton
resonances of nickel porphyrins 13, 19, and 20 are listed in Table 9. FAB-
MS data also provided the expected molecular ion peaks for nickel

complexes.

The removal of the acetyl group from Ni(PSAc)(l3) was readily
achieved to afford NiPSH(21a) by treating with 3N HCl in methanol/CHCI3,

1H NMR spectroscopy (Figure 16b) demonstated the proton resonance of SH

.300 20...... .08 .8: 20...... .00... 3000 .0022. 8... 8.508.. 22.... 2. 0. 9.qu

:0... m n v m m H m m

pb-P-bLL—prPP+bphp—~nn-P~bb-_-pbb-p~.phh;pbpPpr—Lpppphb-n—bh-bb-uphP-ppbnupkp—nbpp_-L.P—L.ppplph»~

i

 

 

 

 

20.20

 

A...

 

 

rad . m m w m m H m m

m pbbbPP—ppppnpp-n—pprr-P-bpp-ppbprpthp-pb-ppP—pup-bpprpb-pptpb—pptpp-ph—-»rP_-»bb—-.thLb

1 l 1‘ 4 1‘ I: 1I.‘Ilu.lll1

 

 

 

 

 

 

 

 

20.20 .
30

 

 

 

 

 

5 555 5 .55 5 555 5 555 5 .55 m8 5 5:
5 555 5 .35 5 55.5 5 .35 5 .35 8-355-858
5 .555 5 55

5 555 5 55.5 5 555 5 .85 5 555 5 53. 5 555 5 555 5 555 5.555 .2
5 55.5. 5 .35 «$20va
$me
5 53. .52 3.3.8
5 555 5 555 5 .555 5555 a 555 a 555 a $55.55 55.8me
a 55.5 a 555 5 555 £550va
5 55.5 5 55.5 5 555 5 55.5 5 .35 556-0
5 555 5 .85 5 555 a .555. 5 555 £055
5 555 5 55.5 mm
5 55.5 .52 8me
5 55 ”5 55 5 55 ”5 5S 5 .35 ”5 55.5 5 55 55 55.5 5 55 55 55 £0me
5 .555 5 555 5 555 5 555 5 .85 £0me

35 mmnmz 3 8 mmfiz 8%?sz acmmaz a: 2352
.380 55,858

Annuvmmmmz .Aafimvmmmmz can vauammmzv .Awfivmmmmz .SCX‘mmmz mo 95083355 A3 $42 $3 .a 033—.

65

at 1.98 ppm, as well as the chemical shift of methylene at 4.05 ppm. In
order to prepare NiPSH(2 lb) from disulfide complexes 19 and 20, reduction
was performed with NaBH4. Again, 1H NMR spectroscopy revealed the

proton resonance of SH at 3.5 ppm. FAB-MS data provided the expected
molecular ion peaks for both compounds.

4. Synthesis of Iron-Sulfur Clusters

The synthesis of all iron-sulfur cluster complexes were carried out
under strictly anaerobic and moisture-free conditions. The synthesis of

[Fe4S4Cl412' clusters has been reported previously by a simple spontaneous

self-assembly process as shown in equation 364.

2FeClz + 3KSPh + (BUNCI or (Ph4P)Cl + SS --> [Fe4S4Cl412' (3)

The synthesis of the [Fe4S4(SPh)2012]2' cluster is readily accomplished by
the stoichiometic reaction between [Fe4S4CI412' and PhS' ligands shown in

equation 455.

[Fe4S4Cl412' + 2KSPh --> [Fe4S4(SPh)2012]2- + ZKCl (4)

The successful isolation of these clusters from equilibrium mixtures of
[Fe4S4(SPh)nCl4-n]2- is attributed mainly to the crystallization
characteristics of the (Ph4P)+ salts that are relatively insoluble in the

solvent used.

5. Preparation of the Ni-Porphyrinyl-Fe4S4 assemblies

To connect an Fe4S4 cluster with nickel-porphyrin thiol ligands,
various types of complexes can be conceived. During the course of two
years, we attempted ligand exchange reactions using the deprotected
NiP(SAc)n 7 a-7c, (n=4, or 2), the monothiol NiPSH (21a-21b) and the Fe4S4
clusters with either halide ligands(22-24) or mixed ligand [Fe4S4(SPh)2012]2'
(25) as starting materials. Unfortunately, despite repeated efforts, this
approach to the most desired assembly Fe4S4-NiP(S4) was unsuccessful.
The results described here focus on the preparation and the

characterization of assembly complexes I-III (shown in Scheme I).

(i) Deprotection (Deacetylation). The removal of the acetyl group of

NiP(SAc)2(7b or 7c) was readily achieved under basic conditions.

Generally, a slightly excess amount of freshly prepared NaOMe was added
into a NiP(SAc)2-DMF solution (eq 5),

NiP(SAc)2 + 2 NaOMe ---> (N a+)2[NiP(S)2] + ZACOMe (5)

the mixture was kept at room temperature for ca. 3h with stirring. The
resultant thiol groups are susceptible to oxidation to form disulfide, thus,
all manipulations were carried out under nitrogen. The sodium salts were
not isolated for characterization and used freshly for the ligand exchange

reactidn.

(ii) Ligand exchange reaction. In order to synthesize the [Fe4S4(X)4-
2n(NiP(Sg)n]2- clusters, the deprotected NiP(Sg) was first prepared by a
simple metathetical reaction which was performed in DMF (eq 6).

aHInaJ
DMFs:
drovai
Precipi‘
produc
Two c

Schenm

Conse<
succes

[(Na)2[

[F9484

6. Che

major
ray di

becau

67

[Fe4S4Cl4JZ' + n((Na)2[NiP(S)2] —->
[Fe4S4CI4—2n(NiP(S)2)nJ + ZnNaX (6)
n = 1, 2

all manipulations were carried out under nitrogen. The freshly prepared
DMF solution of (Na)2[NiP(S)2] is added into a Fe4S4 clusters-DMF solution
dropwise keeping the mixture at 50-60°C for 6h. The product was then
precipitated by the addition of a THF/ diethyl ether mixture. In general, the
product was further purified by re-precipitation from DMF/ether twice.
Two classes of nickel-porphynyl-Fe4S4 clusters (I and II)(as shown in
Scheme 1) were obtained in relatively low yield (< 30%) as reddish powder.
Consequently, an assembly complex [Fe4S4(SPh)2(NiP(Sz))]2‘ (III) was
successfully prepared by treating [Fe4S4(SPh)2Clz)]2' with 1 eq.
[(Na)2[NiP(S)2] as shown in eq 7.

[Fe4S4(SPh)2C12)]2' + [(Na)2[NiP(S)2] —->
[Fe4S4(SPh)2(NiP(S)2)]2‘ + 2NaX (7)

6. Characterization of the Ni-Porphyrinyl-Fe4S4 Assembly

In the course of our work, both IR and 1H NMR spectra provided the
major evidence of the formation of nickel-porphyrinyl-Fe4S4 assemblies. X-

ray diffraction quality crystals of the products have not been obtained,
because the possibility of polymer formation cannot be completely ruled out .

alnfi'aredSpectmscopy

The most useful diagnostic features that differentiate iron sulfur
clusters and the assembly complexes are found in the far-infrared spectra,
exemplified in Figure 17. The skeletal vibrations of the clusters result in
absorption bands with distinct energies and profiles that make it possible to
distinguish III with weak vibrations at 349 and 362 cm‘1 from
[Fe4S4(SPh)2Clz]2' at 339, 356, 371, and 383 ch. Especially interesting is
the disappearance of 356 cm'1 corresponding to an Fe-Cl vibration band. In
the mid-IR region, features corresponding to (Ph4P)+ cation ion at 529, and
534 cm'1 further give evidence of the existence of complexes (III) in the solid
state. The feature of NHCO at 1670 cm"1 and the other vibrational modes

corresponding to nickel porphyrin were not shown in Figure 17.

b. 1H NMRspectmscopy.

Selected assignments of proton resonances of complexes I-III are
listed in Table 10. The isotropic shifts of the methylene protons of S-CHz
group centered ca. 1315 ppm, can be diagnostic of the formation of nickel-
porphriynyl-Fe4S4 assembly complexes. The integrations of the methylene
protons, however, demonstrate that the equilibria intervene in the solution

behavior; various components could present in the solution such as

[NiP(Sz)2-Fe4S4]2', and [NiP(Sz)-Fe4S4012]2' (according to reaction 6).

The 1H NMR spectrum of complex (III), Figure 18, deserves more
discussion. In general, chemical shifts corresponding to the two portions
of the assembly, the cluster and nickel porphyrin, are satisfactionly

identified.

(i) Clusters. The phenyl ring proton resonances of the mixed terminal

ligand “cubane” clusters, [Fe4S4(SPh)2NiP(Sz)]2' show isotropic shifts

 

 

 

 

 

 

 

 

 

A
362 349
B
383 339
355
529
553 ' 453 Y 323

Figure 17 . Far-infrared spectra of (a) (Ph4P)2[Fe4S4(SPh)2(NiP(Sz))l
and (b) (Ph4P)2[Fe4S4(SPh)ZClz]. The vibration mode at 529 cm-1
corresponds to the aromatic C-H bending of Ph4P+.

 

 

555 55 555 55 85 55 555 55
8.5.5-55 55.555 .8555 5555.3. .8355 3.5555 53558
8.55 £5
$3 «N5 5%
55.38
.5555
3.5835 85 85 S5 :5 555 oomz
m 555 85 555 85 555 55.5 55-858
555 S5 mamoég
5555
.855 55.2 55.: 5555 55.55 55.55 5555 855 .5555 mums
2.5.3.5
55:: 5.535 5.535 at: at: mass
:5 .E a: 3 a .5— 85558.5

 

.35an £35,899 mafia 58838.50 .8.“ magmammg A8 g2 ma uSoflom .3 935,—.

71

.85 Esmfizvfimmavmsa .s 3.82: am: 89 3:558... 522 E 55 835..—

omé OYN

   

3

 

 

 

«105:0

 

«.20

820

 

 

 

 

 

 

 

«55 «a.» ms.» 85 3.5

. J. .5
55

1:...

 

 

 

 

 

 

auto

Em...EKanVJQENEEaV

72

(Figure 18) similar to those of [Fe4S4(SPh)2C12]2' (Figure 19). The o-H and p.
H signals remain essentially unchanged at 5.22 and 5.75 ppm respectively,
while the m-H signal appears at 7.7-8.4 ppm overlapping with the signals
corresponding to NHCO, and aromatic proton of cation(Ph4P+).

(ii) Nickel porphyrin thiol ligand. Taking the spectra of NiP(SAc)2 (7b) as
references (Figure 15), the isotropic shift of methylene proton centered at
13.36 ppm, as shown in Figure 18, is characteristic of the binding of the
thiol ligand to the Fe4S4 core. The particularly useful signal of the mesa-H
of nickel porphyrin remains unchanged at 9.6 ppm. Signals corresponding
to ethyl and methyl functional groups of etioporphyrin are similar to the
NiP(SAc)2 (7b) in the region 1.8 (CHgCHz) 3.8 (CH 2CH3) and 2.4 ppm (CH3)
respectively. Most of the phenyl protons corresponding to nickel porphyrin
remain undisturbed in the region of 7 .6-8.8 ppm. The only exception is the
upfield shift of the o-H signal of the tailored phenyl ring centered at 5.80

ppm.

An extensive 1H NMR investigation of the synthetic assemblies (III)
convincingly demonstrates that the Fe4S4 cluster and the nickel porphyrin

are coupled in a 1:1 fashion, rather than as an equilibrium mixture in the

solution.

SUMMARY

The following are the principal findings and conclusions of this
investigation.

(1) A series of nickel-porphyrin thiol ligands (7a-7c) has been
successfully synthesized. Attaching of ligating appendages to the amino

73

.Guvunugnmmvvmeoa mo 36.09.55 .umz 89 35.53% 522 mu .3 Pun-urn

 

‘

OmEO

 

“.20

 

 

 

0...:

 

sausamvemefiuiza.

Q

 

 

 

 

mud Nam;

 

 

 

 

74
substituents was accomplished by adding acid chlorides bearing thiol
functions(5a-5c) to the key precursor cis-(NH2)DPE(1). The purity of free
base porphyrins (Ba-6c) and their nickel complexes (7a-7c) has been
characterized using various spectroscpoic means.

(2) Using biladiene dihydrobromide (8) condensation with aldehydes
l2 and 17, which bear mono-thiol function, we could successfully synthesize
two meso-substituted porphyrin thiol ligand 21a and 21b.

(3) [Fe4S4] clusters 22-24 reacted with nickel porphyrins 7b and 70 to
give two novel classes of nickel porphyrinyl-Fe4S4 assemblies I and II.
However, 1H NMR demonstrate that disproportionation components could
be present in solution.

(4) The assembly complex [Fe4S4(SPh)2(NiP(S)2)]2'(III) was prepared
by treating [Fe4S4(SPh)2C12]2°(25) with 1 equivalent of (Na+)2[NiP(S)2]. Both
IR and 1H NMR provided evidence to support the existence III. However,
satisfactory FAB-MS and elemental analysis has not been obtained.

This work provides the initial experimental protocol for the

construction of [Fe4S4]-siroheme assemblies. Unfortunately, a deliberate
approach to synthesize the complex [Fe4S4(NiP(S4)]2' was not successful.
The preparation of the mixed ligand complex III appeared to be closest
example of nickel-porphyrinyl-Fe4S4 assembly, even there the polymer
formation could not be completely ruled out.

EXPERIMENTAL SECTION

1. Physical Measurements

75

1H NMR spectra and 130 NMR spectra were recorded on a Bruker
WM-250 or Varian Germini-300 spectrometer, in “100%” chloroform-d or
dimethyl sulfoxide-d6 (min. 99.96 atom% D, both fi'om Isotec Inc.) with the
residual CHCl3 or DMSO as the internal standard set at 7.26 or 2.49 ppm,
respectively. IR spectra were recorded on a N icolet MIR-42 spectrometer,
the samples were prepared as a thin film on NaCl plates or as a KBr pellet.
Melting points were obtained on an Electrothermal melting point apparatus
uncorrected. Mass spectra were obtained with a Finnignan 4000 GC/MS
system using the direct inlet mode at 70eV ionization energy. FAB mass
spectra were obtained on a J EOL HX-110 HF double focusing mass
spectrometer operating in the positive ion detection mode.

2. Materials

All solvents and reagents were of reagent grade quality, purchased
commercially, and used without further purification except where
mentioned. Methylene chloride, triethylamine and acetonitrile were
distilled from calcium hydride; THF was distilled from LiA1H4; methanol
and ethanol were distilled from sodium. DMF was stored in 4A molecule
sieves for 24h, then distilled at ca. 30°C under reduced pressure. Thionyl
chloride was distilled from triethylphosphite. Silica gel for column
chromatography (60-200 mesh) was from J .T. Baker (3405). Preparative
silica gel plates were from Analtech, Inc. For analytical TLC, Eastman
13181 chromatography sheets were used.

3. Syntheses of appended porphyrins

76

O .0 O . . O
“-.:0:IIIOOO‘O o 0"W -.'¢=H‘~.I'0JI°UIJ|

The cis and trans mixture of bis(o-aminophenyl)-2,8,12,18-tetraethyl-
3,7 ,13,17-tetramethylporphyrin, (NH2)2DPE, was prepared by the well
developed procedure57. These isomer mixtures could be easily seperated by
silica gel column chromatography using 2% CH3OH/CH2012 as eluents.
The trans isomer could be further converted to the cis/ trans isomer mixture
by (atropimerization), using the following procedure.59 A mechanically
stirred mixture of toluene (1.5 1, dried over molecular sieves) and 100g of
silica gel was heated at reflux for 1h. under nitrogen. Trans-(NH2)20PE
(3g) was added and heating at reflux was continued for 24h. After cooling
to room temperature, the slurry was poured into a 15 cm diameter coarse
glass frit and rinsed with toluene to remove the remaining trans isomer
first, until the washings were clear. Cis-(NH2)2DPE was eluted from the
silica gel with 2% CH30H/CH2012, and isolated by removing the solvent
under reduced pressure to yield 1.35 g (45%) of purple solid. The purity of
the cis and trans isomers was checked by analytical TLC (Si02, CHzClz-
CH30H; 99:1) and 1H NMR spectroscopy. The 1H NMR spectra agree well

with literature values57.

iii-B'fl I] ll] . .“31

A stirred mixture of 3,5-dimethylbenzoic acid(2a)(12g, 0.08mol), N-
bromosuccinimide (31.33g, 176 mmol), benzoyl peroxide (200mg), and
methyl formate (200 ml) was refluxed under illumination (270W Sun
Lamp) for 12h. The resulting red solution was pumped to dryness in vacuo.
After extracting with ether and washing two times with H20 (300ml x 2),

the organic fraction was dried over N a2804 and the solvent was evaporated

77

in vacuo . The residue was recrystallized several times from
CH2012/hexane. The white product was intense by lachrymatory, and yield
12.5g (51%). mp: 139-14000. 1H NMR (0D013): 5 4.58 (s, 4H, CH2), 7.65 (s,
1H, Ar), 8.04 (s, 2H, Ar); 130 NMR (CD013): 8 31.52 (CHgBr), 130.71, 130.83,
134.83, 139.33, 170.41 (00211). MS (RD: 305/307/309 W"; 13/25/12).

35-3“]. | I] ll] . .1“)
A mixture of 3,5-bis(bromomethyl) benzoic acid(3a) (2.0g, 6.5 mmol)

and potassium thioacetate (3.7g, 32.4 mmol) in methanol (50ml) was heated
to 50°C for 30 min. The solution was cooled to room temperature and
potassium bromide precipitates were filtered out. The solvent was removed
in vacuo and the residue was redissolved in ether (50ml) and water (50ml).
After the pH value was adjusted to 6 using conc. HCl, the aqueous phase
was extracted with ether (50ml x 3). The combined organic fractions were
washed with brine and dried over Na2804. The solvent was removed in
vacuo and the residue was recrystallized from CH2019/hexane. Yield: 1.71g
(88%). mp: 147-149 00. 1H NMR (CD013): 5 2.38 (s, 6H,00-CH3), 4.17 (s, 4H,
Ar-CH2), 7.49 (t, 1H, Ar), 7.9 (d, 2H, Ar); 130 NMR (CD013): 5 30.15, 32.66,
129.62, 130.19, 134.80, 139.03, 171.20. MS (RD: 298 (M"', 19).

C .0 O O O . .
”:c --:0 -o: 90:,‘ogun'o 0‘... :_uoooo'¢ o .-

To a 20 ml 0H2012 solution of 3,5-bis(thioacetoxymethyl)benzoic
acid(4a) (200mg, 0.66 mmol), 0.2 ml freshly distilled thionyl chloride was
added dropwise. The mixture was heated to reflux for 3h and the excess
80012 and 0H2012 was then removed in vacuo. The crude acid chloride (5a)
was redissolved in dry CH2012 (20ml) and added dropwise to a 50 ml CH2012

78

solution of cis-bis(o-aminophenyl) porphyrin (150mg, 0.23 mole) and
triethylamine (lml). The resulting mixture was refluxed under argon and

the progress of the coupling reaction was monitored by TLC. After 6h, the
solution was washed with 5% H01 (2 x 100ml), H20 (100ml), NaHCO3(sat.

aq. 100ml), and H20 (100ml), then evaporated to dryness. The reaction
mixture was purified on a silica gel plate, eluting with 0H2012 containing
1% methanol. Recrystallization of the major band from 0H2012/0H30H
produced 200mg (72%). 1H NMR: 8 -2.30 (br s, 2H, Nprnole), 1.53 (s, 12H,
COCH3), 1.75 (t, 12H, CH30H2), 2.58 (s, 12H, CH3), 2.70 (s, 8H, CHgs-COMe).
4.01 (m, 8H, CH20H3), 6.11 (d, 4H, Ar’), 6.46 (s, 2H, Ar’), 7.60 (t, 2H, Ar),
7.77 (s, 2H, NHCO), 7.92 (t, 2H, Ar), 8.01 (d, 2H, Ar), 8.91 (d, 2H, Ar), 10.26
(s, 2H, mesa-H). FAB-MS: m/z 1221 (M+H)+. UV-visible (kmax): 408, 508,
542, 577 , 627 nm

3-B ll 1] . 113!)

A stirred mixture of m-toluic acid(2b) (11.7g, 0.1mol), N-
bromosuccinimide (19.58g, 0.11mol), and benzoyl peroxide (200mg) in 0014
(200 ml) was refluxed vigorously. The progress of the reaction was

monitored by 1H NMR. Afier 3h the mixture was cooled and the solid was
filtered and washed with 0014, the filtrate was evaporated to dryness to

yield a yellow solid which was recrystallized in CHZClz-hexanes to give an
off white power (13.5g, 73%); mp 150-15200; 1H NMR (CD013-acetone-d5): 8
4.53 (s, 2H, CH2), 7.47 (t, 1H, J =7 .8Hz, Ar), 7.66 (d, 1H, J=7.8Hz, Ar), 8.06 (d,
1H, J=7.8Hz, Ar), 8.14 (s, 1H, Ar). MS(RI): 214/216 (M+, 12/11), 116 ((M-Br)+,
100).

3-1’1' I ll 11 . 'lllll

79
This benzoic acid was prepared as described for the acid 4a. Yield:
90%, mp. 90-9100, 1H NMR (0D013): 5 2.37 (s, 3H, 00-0H3), 4.18 (s, 2H, Ar-
CH2), 7.40 (t, 1H, J=7.8Hz, Ar), 7.55 (d, 1H, J=7.8Hz, Ar), 7.99 (d, 1H,
J=7.8Hz, Ar), 8.01 (s, 1H, J=7.8Hz, Ar); 130 NMR (0D013): 5 30.17, 32.84,
129.07, 129.33, 130.00, 130.69, 134.52, 138.61, 172.05. MS (RD: 210 (M+, 16),
193 ((MOH)‘. 16), 150 ((M-OH-Me00)+, 100).

J, _--_ _,,;_.,f.m., ,., , ,,,,,,,., _ ; _
WW6!»

The appended porphyrin 6b was prepared from the 3-
thioacetoxymethyl benzoic acid (5b) as described for the appended porphyrin
6a. Yield: 58%. IR: 1670 cm'1 and 1710 cm'l. 1H NMR (0D013): 5 -2.40 (s,
2H, Nprn-ole), 1.56 (s, 6H, CH3CO), 1.80 (t, 12H, 0H20H3), 2.61 (s, 12H,
CH3), 2.86 (s, 4H, CH 2Ar’), 4.03 (q, 8H, CH20H3 ), 6.29-6.38 (m, 4H, Ar’), 6.43
(s, 2H, Ar’), 6.66 (d, 2H, Ar’), 7.56 (t, 2H, Ar), 7.80 (m, 4H, Ar), 7.93 (s, 2H,
NHCO), 9.01 (d, 2H, Ar); 10.27 (s, 2H, mesa-H). FAB-MS : 1045 (M+H)+.

B-Il . | . . . 1 [4c)
This benzoic acid was prepared from 3-bromopropionic acid(3c) as
described for the acid 4b. Yield: 80%. mp 51-5200. 1H NMR (0D013): d :

2.32 (s, 3H, CH3CO), 2.68. (t, 2H, CH gsCOMe), 3.09 (t, 2H, CH3002H), 10.65
(s, 1H, 00211). MS(RI): 148 (M43 16), 131((M-OH)+, 14), 88 ((M-OH-MeCO)+,
100).

o .o o 0 ' o o
q:- --‘o -‘gro:_ 'Ao,‘oooog :Luooou'g - o o-A' :_‘| -

80

The appended porphyrin (6c) was made from the 3-
thioacetoxypropionic acid (5c) in an analogous fashion as porphyrin 2c
Yield: 67%. 1H NMR (0D013): 5: -2.51 (s, 2H, Nprmle), 1.43 (s, 6H,
CHaCO), 1.58 (t, 4H, -CH2SCOMe), 1.76 (t, 12H, 0H20H3), 2.52 (s, 12H, CH3),
2.59 (s, 2H, -0H20H20O), 4.01 (q, 8H, CH20H3 ), 6.85 (s, 2H, NHCO), 7.52 (t,
2H, Ar), 7.84 (m, 4H, Ar), 8.75 (d, 2H, Ar); 10.27 (s, 2H, mesa-H). FAB-MS :
921(M+H)"’.

H' l l E l . C 1

To prepare a nickel complex, NiP(SAc)4(7a), for example, a stirred
solution of appended porphyrin P(SAc)4(5a) (0.20 g, 0.25mmol) in 10 ml
0H013/MeOH (9:1) was added a solution of nickel acetate (0.25 mmol) in 5
ml MeOH with stirring. The resulting reaction mixture was heated at
reflux for 4h. After evaporating the solvent to dryness, 200 ml water and 20

m1 1N H01 were added to the flask and the suspension was stirred for 20
min before being extracted with 0H2012. The 0H2012 solution was dried

over Na2SO4, filtered and concentrated to 2m]. After recrystallization from
0H2012/MeOH, the precipiated solid was filtered, washed with n-hexane
and dried in vacuo at ambient temperature. Yield: 80% of red solid. UV-
visible (1mg): 408 (soret), 530, 566 nm. NiP(SAc)2(7b) and NiP(SAc)2(7c)
were prepared in an analogous fashion as described above.

NiP(SAc)4(7a) : 1H NMR (0D013): 5: 1.57 (t, 12H, 0H20H3), 1.78 (s, 6H,
CH 200), 2.30 (s, 12H, CH3), 3.06 (s, 2H, ArCHz-S), 3.67 (m, 8H, CH20H3 ),
6.31 (d, 4H, Ar), 7.57 (d, 2H, Ar), 7.82 (t, 2H, Ar), 8.04 (s, 2H, NHCO), 8.90
(d, 2H, Ar), 9.55 (s, 2H, mesa-H) FAB-MS: m/z 1276.2 (M)+

NiP(SAc)2(7b) : 1H NMR (0D013): 5: 1.58 (t, 12H, 0H20H3), 1.75 (s, 6H, CH2-
00), 2.31 (s, 12H, CH3), 3.11 (s, 4H, ArCH2«S), 3.67 (q, 8H, CH20H3 ), 6.45 (s,

81

2H, Ar'), 6.53-6.56 (m, 4H, Ar’), 6.84-6.88 (m, 2H, Ar’). 7.42 (t, 2H, Ar), 7.49
(d, 2H, Ar), 7.81 (t, 2H, Ar), 8.14 (s, 2H, NHCO), 8.94 (d, 2H, Ar), 9.54 (s, 2H,
mesa-H) FAB-MS: m/z 1100.7 (M)+

NiP(SAc)2(7c) : 1H NMR (0D013): 5: 1.56 (s, 6H, CH3-00), 1.65 (t, 12H,
0H20H3), 1.96 (t, 4H, CH2CH2800), 2.28 (s, 12H, CH3), 2.69 (t 4H, 00-
CHgCHz). 3.73 (q, 8H, CH20H3 ), 7.22 (s, 2H, NHCO), 7.37 (t, 2H, Ar), 7.44 (d,
2H, Ar), 7.74 (t, 2H, Ar), 8.68 (d, 2H, Ar), 9.49 (s, 2H, mesa-H). FAB-MS:
m/z 976.3 (M)+

4. Syntheses of mono-substituted porphyrin thiol ligand(21a-2Ib)

,_._').a.“ (Ht. _ : ::-.-:n.: “ulna...“
W60

3,3'-Diethyl-5,5'-diformyl-4,4'-dimethyl-2,2’-dipyrrylmethane (286
mg, 1 mole) and 3,4-diethylpyrrole (250 mg, 2.01 mmol) in methanol (20
ml) were treated with hydrobromic acid (2 ml, 48%), the solution was
heated on a steam bath for 5 min, and kept at room temperature for ca. 2 h.
The product was separated, washed with methanol (containing a little
hydrobromic acid), then with ether, and dried in air to give red-brown
prisms with a green luster (491 mg, 81%). mp > 300 00. 1H NMR(0D013): 5:
0.68 (t, 6H, 0H20H3), 1.22 (t, 12H, CH20H3), 2.27 (s, 6H, CH3), 2.50, 2.57 (q,
4H each, CH20H3), 2.72 (q, 4H, CH20H3), 7.23 (s, 1H x 2, 15-H and 20-H),
7.69(d, 2H , 10-H), 13.32, 13.54 (s, 2H each, NH).

WW9)
A stirred mixture of p-tolunitrile(9) (11.7 g, 0.1 mmol), N-
bromosuccinimde (19.58 g, 0.11mmol), and benzoyl peroxide (200mg) in 200

ml of methyl formate was refluxed under illumination (270 Watt Sun
Lamp) for 2 h. The volatile component were removed by evaporation, the
residue was taken into ether (100 ml) and washed with water (100 ml x 2).
After drying over N a2SO4, the solvent was evaporated to dryness.
Recrystallization twice from 0H2012-hexane afi‘orded pure bromide 13.5 g
(73%). mp 115-116 00. 1H NMR(0D013): 5 4.45 (s, 2H, CH gBr). 7.47 (d, 2H,
J=8Hz, Ar), 7.62 (d, 2H, J=8Hz, Ar). 130 NMR(0D013): 5 31.43, 112.13,
118.31, 129.66, 132.52, 142.78. MS(RI): 195/197 (Mt 12/11), 116 (NI-Br)"; 100.

W162

A 1M solution of diisobutylaluminum hydride (DIBAL) in hexane
(36.8 ml) was added in a dropwise manner over a period of 20 min to a
solution of a-bromo-p-tolunitrile(10) (6.00 g, 30.6 mmol) in 60 ml of
chlorobenzene at 0 00. The resulting mixture was stirred at 0 00 for 1 h and
then diluted with 100 ml of 0H013. This solution was shaken with 10% aq.
H01 for ca. 10 min. The layer was seperated and the aqueous layer
extracted with 0H013 (20 ml x 2). The organic solution were combined ,
dried over Na2SO4, evaporated to dryness in vacuo. The residue was
recrystallized by dissolving in a minimum amount of 0H2012 (ca. 10 ml)
and layering with a small amount of ice-cold hexane to yield 4.7g (77%) of
aldehyde. m p 97-98 00. 1H NMR(0D013): 5 4.49 (s, 2H, CHzBr), 7.53 (d, 2H,
J=8Hz, Ar), 7.84 (d, 2H, J=8Hz, Ar), 9.98 (s, 1H, CHO). 130 NMR(0D013): 5
31.96, 129.64, 130.13, 136.06, 144.20, 191.49. MS(RI): m/z 198/200 (M+, 20/19).
119 [(M-Br)+, 100], 91 [(M-Br-0H0)+, 82].

Wand“

83

A mixture of p-bromomethylbenzaldehyde(1l) (3.95g, 0.02 mmol) and
potassium thioacetate (3.42g, 0.03 mmol) in 50 ml 0H3OH was heated to 50
00. Potassium bromide was. gradually precipitated out after 10 min. The
reaction was kept stirring for 20 min. After filtration, the filtrate was

stripped to dryness. The residue was redissoved in ether, and washed with
H20(50 ml x 2) and brine (50ml). The organic fraction was dried over

Na2804, and then concentrated to give a yellow oil. The crude product was
further purified on silica gel column chromatography using 30% hexane-
0H2012 for elution. The main fraction was collected and the solvent
removed in vacuo to give colorless oily product 3.3g (85%). 1H NMR(0D013):
5 2.35 (s, 3H, CH3COS), 4.14 (s, 2H, CHgsCOMe), 7.44 (d, 2H, J=8Hz, Ar).
7 .80 (d, 2H, J=8Hz, Ar), 9.99 (s, 1H CH0). 130 NMR(0D013): 5 30.29, 33.13,
129.46, 130.05, 135.37, 144.84, 191.69. MS(RI): m/z 194(M+, 35), 152 [(M+H-
0H300)+, 100].

‘.0‘10_I‘_1110_1‘ -05....” . -2; . .3 , 61

The mixture of 1,19-dideoxy1-7,13,-dimethyl-2,3,8,12,17,18-hexaethyl-
biladiene-ac dihydrobromide (102 mg, 0.15 mmol), p -
thioacetoxymethylbenzaldehyde(12) (250 mg, 1.29 mmol), and 4 drops of
acetic acid saturated with hydrobromide in methanol (25ml) was refluxed
for 24 h. The resulting solution was cooled and treated with an excess of
solid sodium bicarbonate. After filtration, the crude product was washed
with water, dried in the air, and redissolved in chloroform-acetic acid
(25:1). A saturated methanolic solution (2 ml) of nickel acetate was then
added, and the mixture was heated under reflux for 3 h. The solution was

cooled, and washed with water (50 ml x 2) and then sat. N aH003 solution
(50 ml). The solvent was removed from the dried chloroform extract, and

84

the residue redissolved in 0H2012. The crude product was
chromatographed on silica gel column using 50% hexane-0H2012 for
elution. The main fraction was collected and the solvent removed in vacuo.
Recrystallization from 0H2012-hexane give the nickel derivative(l3) as
purple crystals. 1H NMR(0D013): 5 0.91 (t, 6H, 0H20H3), 1.70, 1.76 (t, 6H
each, 0H20H3), 2.49 (s, 3H, CH3COS), 2.63 (d, 4H, CH 20H3), 3.41 (s, 6H,
CH3), 3.74-3.87 (m, 12H, CH20H3), 4.44 (s, 2H, CHZSAc), 7.51 (d, 2H, Ar),
7 .93 (d, 2H, Ar), 9.51 (s, 1H, 15—H), 9.57 (s, 2H, 10-H and 20-H). FAB-MS: m/z
726 (M+).

WWW

To a stirred solution of 4-(methylthio)benzaldehyde(l4) (0.76 g, 5
male) in dry dichloromethane (100 ml) at 0°C, 85% m-chloroperbenzoic
acid (1.0 g, 5.9 mmole) in dichloromethane (50 ml) was added dropwise.
The mixture was stirred at 000 for lb. and then treated with 0a(OH)2 (0.55
g, 7.5 mmole) for 15 min. After filtration, the solvent was evaporated in
vacuo. The crude product was recrystallized from hexane-CH2012 to give
colorless plates. mp: 85-8600. 1H NMR(0D013): 5 2.80 (s, 3H, CH3SO), 7.83
(d, 2H, Ar), 8.06 (d, 2H, Ar), 10.10 (s, 1H, CHO). 13C NMR(0D013): 5 43.54,
124.07, 130.26, 137.98, 152.15, 191.04. MS(RI): m/z 168 (M+, 100); 139 [(M-
0H0)+, 91].

I |’-D°fl ll. 1 l l. 151 [”163

The mixture of the sulfoxide(l5) (1.68 g, 0.01 mole) with
trifluoroacetic anhydride (10ml) was heated under reflux for 1h. The
volatile compoments were removed by evaporation and the crude

trifluroacetoxymethylthiobenzaldehyde(16) was directly hydrolyzed in

85

methanol-triethylamine (100 m1, 1:1) without isolation. After stirring at
room temperature for 30 min, the volatile components were removed in
vacuo. The residue was redissolved in chloroform (50 ml), washed with 1N
H01 (50 ml x 1), and then with sat. NH401 (50 ml x 1). After drying over
Na2804, the solvent was evaporated to dryness to provide 4-
mercaptobenzaldehyde(l7). The crude thiol was converted to the
corresponding disulfide(18) by shaking a 0H013 solution with excess
aquous K13. After washing with saturated. Na2SO3 solution, drying over
NazSO4, the solvent was evaporated to dryness. Recrystallization of the
residue from EtOAc-hexane gave yellow crystals 0.97g (71%). mp: 105-
10600. 1H NMR(0D013): 5 7.61 (d, 2H, Ar), 7.80 (d, 2H, Ar), 9.94 (s, 1H,
CHO). 130 NMR(0D013): 5 126.23, 130. 31, 135.07, 143.74, 190.92. MS (RD:
m/z 274 (M+, 46), 137 (M/2+, 27), 109 (M/2++H-00H, 100).

The disulfide was prepared from the diformyldiphenyl disulfide(l8)

and biladiene-ac dihydrobromide(8) as described for the nickel porphyrin
(13). The crude compounds were separated on silica gel TLC using 70%
hexane/0H2012 for elution. The desired disulfide(l9) was shown to be the
minor product which eluted faster, and the major component was the
dimethylacetal of the formyl compound(20). Recrystallization from
methylene chloride-hexane give small red needles.

Disulfide(l9). 1H NMR(0D013): 5 0.97 (t, 12 H, 0H20H3), 1.66, 1.75 (t, 12H
each, 0H20H3), 2.71 (q, 8H, CH20H3), 3.40 (s, 12H, CH3), 3.76, 3.86 (q, 8H
each, CH20H3), 7.90 (d, 4H, Ar), 8.08 (d, 4H, Ar), 9.51(s, 2H, 15-H), 9.57 (s,
4H, 10-H and 20-H). FAB-MS: m/z 1338 (M+).

86
Disulfide(20). 1H NMR(0D013): 5 0.86 (t, 6 H, 0H2CH3), 1.68, 1.74 (t, 6H each,
0H20H3), 2.62 (q, 4H, CH20H3), 3.36 (s, 6H, CH3), 3.39 (s, 6H, 0H(OCH3)2),
3.77, 3.84 (q, 8H each, CH 20H3), 5.44 (s, 1H, CH(OMe)2), 7.49 (d, 2H, Ar),
7.65(d, 2H, Ar), 7.76(d, 2H, Ar), 7.94 (d, 2H, Ar), 9.51(s, 1H, 15-H), 9.57 (s,
2H, 10-H and 20—H). FAB-MS: m/z 852 (M+).

-I- o_-\_[: 5,0,0": g jog'o ‘ooou‘oa 1.9;, .0 \. .' , .1
Etioporphyrinato Nickel(II) (13) (73 mg, 0.1 mmol) in 0H013 (25 ml)
was treated with 3N H01 in methanol (5 ml) under nitrogen. The mixture

was refluxed for 12 h. The resulting solution was cooled and pumped to
dryness in vacuo. The residue was redissolved in CH013 and washed with
H20 (10 ml) and brine (10 ml). After drying over Na2SO4, the solvent was
evaporized in vacuo. Recrystallization from CH2012-hexane in a glove box
under nitrogen give purple crystals 70 mg (90%).

1H NMR(0D013): 5 0.92 (t, 6H, CH20H3), 1.70, 1.76 (t, 6H each, 0H20H3),
1.98 (t, 1H, SH), 2.65 (d, 4H, CH20H3), 3.40 (s, 6H, CH3), 3.80, 3.8 (q, 6H each,
CH20H3), 4.05 (d, 2H, CH 28H), 7.51 (d, 2H, Ar), 7.93 (d, 2H, Ar), 9.51 (s, 1H,
15-H), 9.57 (s, 2H, 10-H and 20-H). FABMS: m/z 684 (M+).

The diacetal porphyrin dissulfide(20) (170 mg, 0.2 mmol), under a
stream of argon, was dissolved in 100 ml of 100% ethanol and treated with
500 mg of NaBH4. The mixture was stirred at room temperature for 12 h.

An equal volume of H20 was then added to the reddish solution. The

aquous layer was discarded, and the organic layer was further washed

with brine ( 100 ml x 2), dried briefly over Na2804, filtered , and taken to
dryness.

87

The residue was purified by flash chromatography under nitrogen
with 30% hexane-CH2012; the major band was separated from several

trailing bands. The solution was collected, evaporated, and crystallized
inside a nitrogen box from hexane/0H2012. Yield: 33 mg (49). 1H
NMR(0D013): 5 0.91 (t, 6H, 0H20H3), 1.70, 1.76 (t, 6H each, 0H20H3). 2.65
(d, 4H, CH20H3), 3.40 (s, 6H, CH3), 3.50 (s, 1H, SH), 3.80, 3.8 (q, 6H each,
CH20H3), 7.49 (d, 2H, Ar), 7.90 (d, 2H, Ar), 9.50 (s, 1H, 15-H), 9.57 (s, 2H, 10-
H and 20-H). FABMS: m/z 670 (M+).

5. Preparation of Iron-Sulfur clusters

The iron-sulfur cluster complexes were prepared and purified by

published procedure854»65. Typical experimental procedures are as follows.

3."- .‘ ;_.... 00020400.."‘t ;-‘..._‘W‘;- ..
Wflwz)

1.26 g (10 mmol) of anhydrous Fe012, 2.23 g (15 mole) of KSPh, 1.87
g (5 mmole) of Ph4PCl, and 0.4 g (12.5 mmol) of elemental sulfur were
placed in a 125 ml Erlenmeyer flask. To that mixture, CH30N (40 ml) was
added with continuous stirring. The resulting reaction mixture was
stirred at room temperature for ca. 45 min. During that time, the reaction
mixture became brown-green. The solution was filtered to remove K01 and
unreacted elemental sulfur. To the filtrate, 100 mL of diethyl ether was
added and the solution was allowed to stand for ca. 12 h at room
temperature. A black crystalline solid was deposited. The product was
isolated by filtration, washed twice with diethyl ether, and dried under

88
vacuum; yield: 2.14 g (73%). The UV-visible spectrum of the crystalline
product was obtained in 0H30N and was found identical with that of the
previously reported (Ph4P)2Fe4S4014.
(Bu4N)2Fe4S4014(23) and (Bu4N)2Fe4S4Br4(24) were prepared in an
analogous fashion as described above.

.0 O O O O O
b,- - :ou-n 010:0... uo:_oo¢n‘no:,oo no: :a: -. so;

Wmfl
To a solution of 1.00 g (0.86 mmol) of (Ph4P)2[Fe4S4014] in 50 mL of

dimethylformamide (DMF) was added 0.20 g(1.72 mmol) of solidKSPh.
After stirring for 20 min, the brown-green solution was filtered and to the

filtrate was added 200 mL of anhydrous ether. The oil that formed was
crystallized from 0H30N following the addition of anhydrous ether. A 0.95

g sample of analytically pure crystals was obtained; yield: 86%. The X-ray
powder pattern, visble spectra, and isotropically shifted NMR spectra of this

compound are identical with those of an “authentic” sample65
6. Preparation of the Ni-Porphyrinyl-Fe4S4 assemblies

Typical experimental procedures to prepare the nickel porphyrin-

Fe4S4 assemblies are as follows.

Complexes (III). To a 10 m1 DMF solution of NiP(SAc)2 (7b) (66 mg, 0.06

mmol) was added one portion of freshly prepared NaOMe (11 mg, 0,20
mmol) in 5 ml DMF/0H3OH (4:1) solution. The resulting reaction mixture

was stirred at room temperature for ca. 3h. To a 10 ml DMF solution of
(Ph4P)2Fe4S4(SPh)2012(25) (66 mg, 0.05 mmol), the above freshly prepared
solution was added dropwise, and the mixture was heated to 50-60 00 and

89

kept stirring for 6h. The resulting solution was filtered, and to the filtrate
20 ml of ether/THF (1: 1) was added. Afiaer the solution was allowed to stand
for ca. 12 h, the reddish precipitates were collected by filtration and washed
with ether. Recrystallization from DMF/ether afforded 56 mg (25%) of
reddish powder. UV-visible (1mg): 408, 527, 564 nm. 1H NMR (DMSO-d5): 5
1.80 (br s, 12H, CH20H3), 2.40 (br s, 12H, CH3), 3.82 (br s, CH20H3), 5.22 (s,
2H, p-H of SPh), 5.75 (Br 8, 4H, o-H of SPh), 5.80 (s, 2H, Ar’), 6.35-6.48 (m,
4H, Ar’), 7.6-8.4 (m, 2H,40H,4H,2H, NHCO, Ar of Ph4P, m-H of SPh, Ar),
8.65 (s, 2H, Ar), 9.62(s, 2H, mesa-H), 13.36 (br s, 4H, CHZS).

CHAPTERIII

CHARACTERIZATION OF THE IRON-SULFUR CLUSTERS
BY FAB MASS SPECTROMETRY

INTRODUCTION

Proteins containing iron-sulfur clusters frequently serve as redox
enzymes, and participate in electron transfer reactions associated with
processes such as photosynthesis, nitrite reduction and nitrogen
fixation55-57. At present, four distinct types of Fe—S cluster cores, in various
oxidation states, have been identified in such enzymes: FeS4, Fezsz, Fe3S4,
and Fe4S4. The iron centers in these clusters form bridges in proteins,
usually by bonding to sulfur atoms of cysteine residues. Most of these
clusters participate in one-electron redox processes“. They frequently
contain iron atoms in one or more oxidation states, usually Fe3+ and Fe2+.
A variety of iron/sulfur core oxidation states have been established for the

various Fe/S clusters present in proteins. Those identified to date include:
[F923211+v2+. $638413“. and [Fe4s4ll+:2+'3+.

In view of the diversity of structures and biological functions, these
complexes are difficult to characterize by direct studies of the proteins
themselves. Fortunately, synthetic analogs of the mono-, bi- and tetra-iron
centers have been developed to provide insights into their intrinsic
properties in the absence of protein-imposed contraints. Of the structurally
characterized synthetic models for the various [FemSn] clusters now
available, the cubane-type, [Fe4S4], core geometry appears to be the most
commonly encountered, and has been the focus of an intensive body of
structural, spectroscopic and magnetic studies for the last two decades69v70.
A wide variety of model complexes of the type [Fe4S4X4]2' have been made in

which the anionic components, X', are a variety of thiolates (SR')64, halides

91

(01', Br, and I')64:65, and alkoxides (OR')71, as well as combinations of
these ligands“. Recently, the 'subsite-difl‘erentiated' analogs of biological
[Fe4S4]2+ clusters“, as well as synthetic peptide model complexes72'74 have
also been explored. Generally, the characterization of such compounds has
always relied on UV-visible spectroscopy, elemental analysis, NMR
spectroscopy (where applicable for complexes in solution), and single-

crystal x-ray crystallography.

Although conventional electron impact mass spectrometry is a
standard spectroscopic method for the characterization of inorganic
compounds, it is not readily applicable for the analysis of ionic, nonvolatile
compounds such as those that contain the [Fe4S4X4]2' core. Recently, fast
atom bombardment (FAB) ionization75 has been used for the mass
spectrometric analysis of a variety of inorganic compounds including
classical inorganic salts, organometallic compounds, coordination
complexes, and bioinorganic systems“. The main advantage of FAB-M877
and the related techniques of liquid secondary ion mass spectrometry
(LSIMS)78 is the facility with which ions can be generated from
nonvolatile/thermally-labile inorganic compounds. Successfiil completion
of the experiment frequently depends on an appropriate choice of the

viscous liquid matrix to assist in the desorption and ionization processes.

This is the first report documenting the utility of FAB-MS for the
characterization of [Fe4S4X4]2’ clusters. Results of both positive and

negative ion FAB-MS are presented here in the characterization of a series

of salts of anionic iron-sulfur clusters which include:

93
(a) (A)2Fe4S4Br4, A: Bu4N, PuN, Et4N; (b) (A)2Fe4S4014, A=(Ph3P=)2N
(PPN), Ph4P, Bu4N, Me4N; (c) (A)2Fe4S4(SPh)4, A=Ph4P, Bu4N; (d)
(A)2Fe4S4(SEt)4, A=Ph4P, Ph4As; and (e) (Ph4P)2Fe4S4(SPh)2012. The
results reported herein demonstrate that useful mass spectra can be
obtained by choosing proper matrices. When negative ion FAB is employed,
identification of the intact core, as the univalent anion, [Fe4S4X4]', is
straightforward. Thus, FAB-MS analysis can be employed as a valid
method for rapid molecular weight determination. A variety of fragment
ions have also been observed; mechanisms for their formation are

proposed.

The mass spectral results show that, in the gas phase, the iron-
sulfur cubane core undergoes unimolecular dissociation following
desorption/ionization. The fragment ions, which contain 4, 3 and 2 iron
atoms, have interesting parallels with known species that have been
synthesized/identified in condensed phase studies. These parallels, and
their implications, will be discussed.

EXPERIMENTAL SECTION

Mass spectrometric analyses were performed on a JEOL HX-110 HF
double-focusing mass spectrometer, operated in either positive ion or
negative ion mode. Ions were produced by fast atom bombardment (FAB)
With a beam of 6 keV Xe atoms or by LSIMS using a 20 keV cesium ion
beam, The mass spectrometer was operated using an accelerating voltage
0f 10 kV and a resolution of at least 3000. The instrument was scanned at a

94

rate of 2 minutes over the range of 1-6000 Daltons. Data reported represent

mass spectra obtained in a single scan.

A variety of liquid matrices including glycerol, thioglycerol, 3-
nitrobenzyl alcohol (NBA) and 2-nitrophenyl octyl ether (NPOE) were
evaluated. Glycerol and thioglycerol both lead to significant complicated
additions to the ionic clusters formed. The terminal ligand-matrix
exchange reactions such as dehalogenation79 or ligand subsititution have
been observed”. Both NPOE and NBA were found to be most appropriate
for analyzing complexes that contain halides ligands81, because reactions
between analytes and matrices do not occur. Iron-sulfur clusters with
thiolate ligands also show sensitivity to the matrix NBA, and only NPOE
was found to be an suitable matrix in those cases. Samples to be analyzed
were dissolved in DMF, and 1 ml of the solution (ca. 10 mM) was mixed
with 2 m1 of the matrix (NBA or NPOE) on the FAB probe tip.

The iron-sulfur cluster complexes were prepared and purified by
published procedure854r71. The purity of the complexes was generally
established by UV-visible spectroscopy, 1H-NMR spectroscopy, and x-ray
powder pattern analysis. To avoid oxidation, all of the samples were
handled under a flow of pure nitrogen during their preparation and

introduction into the mass spectrometer.

RESULTS AND DISCUSSION

1. Mass Spectral Characten'zation of Iron-Sulfur Clusters

%

Presented here are experimental results and observations in the
FAB-MS analysis of various salts containing the iron-sulfur cubane core.
As will be shown, positive FAB-MS gives molecular weight and structural
information for the analytes (A)2Fe4S4Br4 and (A)2Fe4S4014. However, the
positive ion FAB-MS studies of complexes containing thiolate ligands,
(A)2Fe4S4(SPh)4. and (A)2Fe4S4(SEt)4, as well as the mixed-ligand complex,
(Ph4P)2Fe4S4(SPh)2012, gave limited information related to the molecular
formula. In contrast, negative-ion mass spectra of all iron-sulfur cubane

clusters display peaks representing the unique features of the intact

[Fe4S4X4] core, as well as a variety of fi-agment ions.

Since most of the ions observed contain atoms which have a number
of isotOpic forms, isotopic clusters of peaks are observed to represent a
single elemental formula. In the following discussion, the nominal m/z
value is used to represent an ion of a given composition, calculated by using
the lowest mass isotope of each element present (i.e., 56 u for Fe, 35 u for 01,
79 u for Br and 32 u for S). In the data presented, reported relative

intensifies represent the most abundant isotopic peak of an isotopic cluster.

(A). Positive- and negative-ion FAB mass spectra of halogenated clusters,

(A)2Fe4S4Br4, and (A)2Fe4S4Cl4:

Both NBA and NPOE are suitable matrices for positive- and negative-
ion FAB-MS studies of (A)2Fe4S4X4 compounds, where X: Br, and 01. In
general, when FAB is used in the analysis of ionic analytes of the form
[A+][B‘], the intact cation [A]+, and possibly its fragment ions, will

dominate the positive ion mass spectrum, while, [B]' and charged

96

fragments thereof will be present in the negative ion spectrum. In the
FAB-MS analysis of these ionic complexes, which can be written as
(A+)2[Fe4S4]2+(X-)4, the cation (A)+ is seen as the base peak in the positive-
ion spectra. However, the positive-ion FAB spectra are disappointing in
that they give limited structural information in the mass range below m/z
800. Most of the peaks in this range are due to the ions formed from
interactions between the intact cation, A+, and matrix molecules, to yield
clusters of the type [NBAn+A]+, or molecular fragments that do not contain
either iron or sulfur, such as [2(A)+X]+. However, the most significant
spectral features in the positive ion spectra are three clusters of peaks in
the higher mass range. In the positive-ion spectrum of (Et4N)2Fe4S4Br4 as
shown in Figure. 20, the significant high mass ions represent the ionic
species [(Et4N)2Fe4S4Br3]+ (m/z 849); [(Et4N)2Fe4S4Br4,]+ (m/z 928); and
[(Et4,N)3Fe4S4,Br4]+ (m/z 1058), respectively. These can also be written as
variants of the neutral salt molecule, M, as [M-Br]+, [M]+, and [M+A]+
respectively. The gas phase [M+A]+ adduct represents the peak at the
highest m/z value in the spectrum, and is presumably formed by

complexation of the (Et4N)+ cation with the desorbed neutral
(Et4N)2Fe4S4Br4 molecule. The molecular ion peak, [(Et4N)2Fe4S4Br4]+,

which provides direct molecular weight information, is also observed.

In contrast, the negative-ion FAB mass spectrum of (Et4N)2Fe4S4Br4
(Figure 21), obtained using the matrix NPOE, is much richer than the
cation spectrum. Dominant high mass ions include the [M-AJ' complex,

formed by loss of a tetraethylammonium cation, at m/z 798. The intact iron-
Sulfur cubane cluster, as a -1 ion, [Fe4S4Br4]- (m/z 668) is also observed.

The molecular anion [(Et4N)2Fe4S4Br4l' (m/z 928) is present in the negative

me 8:53: mama m<h nomififimem on» we .85qu <

oom— co—_

\
+=zsmrE

 

 

 

 

——‘

$2 .588 85 was. .Ge Hetamvomazsm:

£8

occ—

L2.

Emvmvommazsmv u 2

.cu ousumm

 

.3......

can

v v f fi‘ 7 T

r

 

.co-

 

Ausuaiul aAnelaa

.3 5 .3 838888 8.3 83 x58: .momz ~58.
2: was: fltfimvomxzsa. .8 8589 .88 mg 83.532 .3 23E

 

 

 

 

~\d=
sec. :88 can can coo can oov can

Emacs—22:8 . .3
.n- mom- mom- H
II' .
m - _ w W .

1' mom- Y. o v
m- n
1'7 f

.n. roe
25538238 W

.8»
Vamvmxpm h

.ao~

 

 

 

Ausuaiul 98.112133

ion spectrum, but the peak is of very low intensity. The positive and
negative ion FAB mass spectra of the halogenated complexes such as
(A)2Fe4S4Br4 and (A)2Fe4S4014 give similar spectra to that shown for
(Et4N)2Fe4S4Br4, and these spectra are tabulated in Tables 11 and 12. All of
the elemental formulae listed for the ions observed were confirmed based on
the close agreement of observed/calculated isotopic distributions for the
molecular formulae listed. Thus, the formula weight of the complex could
be deduced from these clusters of mass spectral peaks without ambiguity.
As an example, Figures 22a and 22b show the comparison between
calculated and observed isotopic abundances of [(Et4,N)3Fe4$4Br4]+ and
[Fe4S4Br4]-, respectively. Combining the spectra fi'om both positive- and
negative-ion FAB-MS studies, a rapid molecular weight determination of

iron-sulfur cluster complexes has been successfully achieved.

In addition to peaks that provide molecular weight information,
Figure 21 exhibits several other peaks formed by fragmentation of the iron-
sulfur cubane core. Fragment ions represented by the quartet at m/z 589,
591, 593, 595 and the triplet peak cluster at m/z 510, 512, 514 correspond to
the anionic species [Fe4S4Br3]' and [Fe4S4Br2J' respectively. The relatively
low intensity quartet at m/z 719, 721, 723, 725 corresponds to the cluster
[(Et4N)Fe4S4Br3]'. Surprisingly, we observed a number of abundant anionic
fragments of the [Fe4S4Br4] core structure. For example, the fragment
[Fe4S3Br4]', appears at m/z 636-644; [Fe4S3Br3]' is formed, and is seen at
m/z 557-563; [Fe3szBr3]', is present at m/z 469-475; [Fe383Br2]', is seen at
m/z 422-426; and [FezszBr2J' is seen at m/z 334-338. In each of these cluster
peaks, the bromine content dominates the isotopic "fingerprint" pattern,

which allows a facile identification of these various fragment ions. The

yamxwmvermasn u +035 Ea .Ldnemeomaé u +35
.Ldnvmemmnsn u +21% 8 .25.."an "2mm 5 .838: :3 £35 85288 :88
23 .8.“ c3 3 9533 98 .mmmoaumumn am v33 £55335 .msom “commune“: 23 mo

833.835 @3328 55 EB 33> a): 3:38: on... 3 5%: 033 2: 5 $38.92 .a

 

85 G: as 8: 35 39 8Q
83 mm: :5 m8 2: Sm mew 535
Ag 83 Aug $8 85 33 88
83 8: mg 38 mm: 9.2 mg Ea
m so: so: so: so: 88 so: so:
1 .88 83 £4 E «man can 82 +H<+§

 

 

22mm ASE 2.5m 232 23m 23m 23m u<

 

:Ovmvom£<v Emvmeermxfi

 

.8<mz arses 2: use: .6 38 cm ..... x .3 8
8388883 Jamsmas 888388 as 883m 382 9E 2: a 88:8 .2 838...

101

.33 Shana 3385a ewes 25 .8.“ c3 3 333$ 98 833:8?“ a

 

Ev
2

Na

3

Na
5%

-§%8£R““8$8

8a

Em5¢£883

Nm
8“

33$3°S$88

o3

 

NH mm
8 mm
mm cm
9. mm
mm hm
5 mm.
av an
8H 8a
«A 3
mm mm
.. b

3 .3383
mm -Rxmmmaa
om -Enammoa

- .zxvmvoa

- -Exmmaa
2 3335
2.. -3635
8 Anyway...
mm -_%mxa
8H -Hgmvoa
S .Hameomfiz
«w -Eqfimmfiz
: -quflomxé

 

Zfiumeanc gm 23m 232
:mevmhfifi

2.5m

23m 23m u< c2530 .88
Eameohflfi

 

«mags 952mm

 

.momz x53 2: man:

.8 can .5 u x .Jdmvomfls 698388 go 38am 932 3...... as cfiamez 2: a 332 «a 82833 252% .fi 039—.

102

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100
h) - Obaarvad

R [:1 Calculated
? 80
a
l
I
v 60
e
I
n
g 40
o
n
l
g 20
y _[LILI_.1

o 1056 105a 10001002 1004 1000 1068 1070

' M/Z
100
lb)
- Obaervcd

R
f 80 [:1 Calculated
a
l
I
v 60
a
l
l'l
l 40
c
n
l
g 20
y l

0 |

000 008 070 072 074 070 678
M12

Figure 22. Comparison of isotope abundance for an experimentally
observed (scribed bar) and theoretically calculated (blank bar) clusters (a)
[(E"-4N)::ll"¢4S4Br4]+ and (b) [Fe4S4Br4]'.

103

rich fragmentation observed in negative ion FAB-MS is also important in
establishing the presence of the intact [Fe4S4] cluster core. Additional
features of the negative ion spectrum that should be noted are the low-
intensity signals appearing 16u and/or 32u above each of the major
fragment peaks. These are most likely due to oxygen atom transfer
reactions involving matrix molecules such as NBA or NPOE. Oxygen
additions have been observed before in the FAB-MS of organometallic
compounds”. The chemistry may well take place in the gas phase,
between iron-sulfur-containing anions and individual matrix molecules.
In support of this possibility, we note that McElvany and Allison33 have
reported the gas phase, bimolecular chemistry of transition-metal
containing anions such as Fe(CO)3' with n-nitroalkanes, using ion
cyclotron resonance (ICR) mass spectrometry. In these systems, products
such as [Fe(CO)3O]' are formed due to oxo-transfer from the N02 group of
the organic molecule.

The negative ion FAB mass spectra of these iron-sulfur cubanes are
dominated by ions with a net charge of -1. Of particular interest in these
experiments would be the detection of the doubly-charged core, [Fe4S4Br412'.
This ion has the same nominal mass as the singly-charged fragment
[FezsgBrz]: The two species can be easily distinghuished, since the -1 ion
will exhibit isotopic peaks due to the presence of Br, which will be separated
by 2 m/z units. If the doubly-charged anion peak is present, it will have the
same nominal mass, but a much richer isotopic pattern, since it contains
four bromine atoms. In addition, for the -2 ion, the major isotopic peaks
will be separated by 1 m/z unit. Doubly-charged ions have been previously .

reported in the FAB analysis of transition metal-containing compounds.

104

For example, abundant doubly-charged cations, such as [Ru(bpy)3]2+, from
ruthenium(II) complexes, have been generated in FAB mass
spectrometry“. The spectrum of (Et4N)2Fe4S4Br4 (Figure 12), reveals an
isotopic cluster of peaks with a nominal m/z value of 334, however careful
analysis of the isotopic pattern indicates that no peak due to [Fe4S4Br412'
species is present; the m/z 334 cluster only represents the anion [FegszBrzh
In this case, the doubly-charged anion may not be a stable species in the gas
phase. While the dianion certainly exists in solution, there are a number of
reasons why it may not be observed in the FAB experiment. In the
desorption process, it may transfer an electron to the nitroaromatic matrix,
following a lower energy pathway. Even if the dianion is desorbed intact, it
may either eject an electron spontaneously, or lose an electron in a

subsequent collision with a desorbed matrix molecule.

(B). Negative-ion FAB mass spectra of thiolated iron-sulfur cubane
clusters, (A)2Fe4S4(SR)4, R = Ph and Et.

Our experience with the halide-containing Fe/S compounds
demonstrated that negative ion FAB mass spectrometry is superior for the
analysis of such compounds. Therefore only negative ion spectral results
will be presented for the complexes that contain sulfur-based terminal
ligands. The features of the negative-ion FAB mass spectrum of
(A)2Fe4S4(SPh)4, Figure 23, obtained using the matrix NPOE, are analogous
to those presented for (A)2Fe4S4Br4 and (A)2Fe4S4Cl4. For example, the
negative-ion spectrum of the molecule (Ph4P)2Fe4S4(SPh)4 displays clusters
of peaks representing the [M-A]' fragment, [(Ph4P)Fe4S4(SPh)4]', at m/z =
1127, and the intact [Fe4S4(SPh)4] core, as the -1 ion at m/z 788. Dominant

105

 

.3 5 .3 3838.. as 83 €32 .momz NBS: 2:
mama: .vfimmvvmvommavamv .3 89.58% manna gm Evelyn—~32 .mm 0.33..“

 

 

 

 

 

2:.
8. 2:. 8o 8... 8. . 8." a
a. .07 m. 87 . aé...
Em. _ .
_. a h
.53....wf i
gum. - .av
f
. |§ .
can. 1
/ .3
/ wiflflf m
w
/ fiaumvvmvou- .160
H.555 H
“2:

 

 

Kusueml camp};

1%

fragment ions include [Fe4S3(SPh)4]' (m/z 756), [Fe4S4(SPh)3]' (m/z 679),
[Fe4S3(SPh)3]' (m/z = 647), [Fe4S4(SPh)2]‘ (m/z 570), and [Fe4S4(SPh)]' ( m/z
460 ) (Figure 5). The negative ion spectra, for two different counter cations

(A+), are summarized in Table 13.

In the negative ion FAB analysis of the A2Fe4S4X4 compounds,
certain ions are always seen, and can be thought of as important indicators
for cubane-based analytes. These include: [AFe4S4X4]', [Fe4S4X4]',
[Fe4S4X3]', [Fe4S4X2]', [Fe4S4X]', and [Fe282X2]'. These are observed when
X = SPh and SEt as well as when X = Br and Cl. However, the two thiolato
clusters also reveal important differences in their mass spectra. More
extensive fragmentation is observed when X = SEt, than when X = SPh.
This can be seen in the negative ion spectrum of (Ph4P)2Fe4S4(SEt)4, shown
in Figure 24. A series of unusual fragment ions deriving from C-S cleavage
processes and not found in the spectrum of the SPh analog, are
[Fe4S5(SEt)3]' (m/z 567), [Fe4S5(SEt)2]' (m/z 506), [Fe4S5(SEt)]' (m/z 445), and
[Fe4S5]'. (m/z 384). These observations are consistent with the fact that the
SEt ligand contains a relatively weak bond. For RS radicals, the GS bond of
CH3S is almost 20 kcal/mol weaker than the CS bond in the PhS radica185.
Thus, alkyl C-S bond cleavage is certainly not surprising in the -SEt
containing complexes. Also notable in this spectrum are the peaks
representing [Fe4S4]‘, m/z 352. This is the only example of an iron/sulfur
cubane compound that exhibits an anion which represents the [Fe4S4] core.
The extremely active species [Fe4S4]' can further react with desorbed
matrix molecules, as described earlier, to produce the [Fe4S4(O)]‘ ion (m/z =
368). We note that this is analogous to the [Fe4S5]- ion that is observed as a

107

.33 “.5939“ ”.88 on... he 2: 3 3333 98 836:3": .a

 

 

 

 

 

 

8H m. 8 a. -ammvamsm
- - mm mm amen
- - 2: 8H swam
9 on «H mm -zmmvsmam
- - 3 mm -zmmvmmsm
3. 3 mm a. -ammvemam
- 2 - - -mamvmmsm
- - 8 ma -ammvmmsm
mm 3 a. we sagas...“
- 8 - - -ammvmmxm
- - mm mm -ammvnmsm
8. 82 G 8 -eamvemam
E a. S a -xmmvemeomg
23m Ara
«Emmvgmafi «ammvwfimfia
$5835 guflom

 

.amomz aces 2: as? am as swam

.ammvemamas £3238 .3 £88m as: 92 :3 8%qu 2: a 3st no 83:35 332% .2 2.3.

.fismmvemvoaafie .fi .Sfimvemsa .3 .Efimvsmaa .2 .Efimvvmaa

.m .ausmmvumvoa .m nasmmvemeoa s .Smmvmmsa .m .Emmvvmeoa

.m «33.: a. smegma“: .m saga .N nasmmvmmsa A 3.2 cause

MO aflwammm< 6.0 dd %3 “young—M30“. and mdom 5.552 .HOnHZ NTSGE 038

m5; .«Aaflmvvmvommavamv no 855023 39: arm covouéwmQZ .VN 9:5?—
#8

coo. coa con 2:. cow con cov con
i .1 c

N—

108

 

 

 

 

 

 

 

:

 

 

 

 

 

Aususiul aAnejea

109

fragment ion. The complete data representing the negative ion FAB

analyses of the cubanes containing -SR ligands are Hated in Table13.

The data presented above for clusters with sulfur-containing ligands
were obtained using NPOE as a matrix. When the matrix is nitrobenzyl
alcohol (NBA), the complicated but interesting spectrum shown in Figure
25 is obtained. A rich chemistry occurs between NBA and compounds such
as (Ph4P)2Fe4S4(SEt)4 (M), In the mass range 900-1400, a group of cluster
peaks have been assigned as [M-(Ph4P)]' (m/z 935), [M-(Ph4P)+NBA-HSEt]'
(m/z 1026), [M%Ph4PH2(NBA-HSEt)]' (m/z 1117), W—(Ph4P)+ 3(NBA-HSEt)]'
(m/z 1208), and [M-(Ph4P)+4(NBA-HSEt)]' (m/z 1299). The second group of
cluster peaks in the mass range from 590 to 970 correspond to [M-2(Ph4P)]'
(m/z 596), [M-2(Ph4P)+NBA-HSEt]' (m/z 687), ['M-2(Ph4P)+2(NBA-HSEt)]'
(m/z 778), [M-2(Ph4P)+3(NBA-HSEt)]‘ (m/z 869), and [M-2(Ph4P)+4(NBA-
HSEt)]' (m/z 960). Relative abundances of ion species are listed in Table 14.
The results reveal the occurance of stepwise ligand substitution reactions.
The acid-base reaction (Eq 1) proceeds to the right when R’SH is a stronger
acid than RSH, the conjugate acid of the coordinatated thiolate, and this
methodology has been well exploited to synthesize [Fe4S4] complexes with
virtually any desired ligands or combination of ligands54,65. Averill and

coworkers71 have also applied this method to synthesize phenoxide
complexes such as [Fe4S4(OAr)4]2' (Eq. 2),

[Fe4S4(SR)4]2' + nR'SH -—> [Fe4S4(SR)4-n(SR')n]2' + nRSH (1)
[F84S4(SR)4]2' + nArH —-> [F e4S4(SR)4.n(OAr)n]2‘ + nRSH (2)
n =1-4

Relative Intensity

Relative Intensity

110

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

100‘ 1.
1 15. X 4
301 5—’
1 9. (ammusussouocaum-
M
601 1'1'1'
1 a
1 V
0' 1> Fe 3 ss: oca Ar) -
1 5?, max-1151:: (Punmsuocnzmu-
1 a
201 9:. «1311-1151:: I
‘ ‘ F
; I, , . . . . It
1000 1100 1200 1300 1400
100‘
I Fe4s418603(ocnzm°
1 ss: oca Ar -
751 ' / M41 121 2 12
1 rususszmocuzmr
(
+NBA-HSEt
01
S j / ‘ Fc484(SEt)(OCiizAr)3'
25: ' *NBA-HSEI. Fc484(OCHZAt)4'
1
1 *NBA-HSE! l
o . .
610 700 300 900 1000

01/:

Figure 25. Negative-Ion FAB mass spectrum of (Ph4P)2Fe4S4(SEt)4, using
the matrix NBA. Matrix ions are designated by an (*).

111

Table 14. Relative Intensities of Peaks in the Negative FAB Mass Spectrum
of (Ph4P)2Fe4S4(SEt)4, (designed as M), using the matrix NBA.

 

 

Ions Observed m/z Relative Intensitiesa
[M-(Ph4P)+4(NBA-HSEt)]‘ 1299 2
[M-(Ph4P)+3(NBA-HSEt)]' 1208 6
[M-(Ph4P)+2(NBA-HSEt)]' 1117 11
[M-(Ph4P)+(NBA-HSEt)]' 1026 8
M-(Ph4P)]‘ 935 4
[M-2(Ph4P)+4(NBA-HSEt)]' 960 5
[M-2(Ph4P)+3(NBA-HSEt)]' 869 44
[M-2(Ph4P)+2(NBA-HSEt)]' 778 88
[M-2(Ph4P)+(NBA-HSEt)]' 687 100
[M-2(Ph4P)+NBA—2HSEt]' 626 47
[M-2(Ph4P)]' 596 63

 

a.Intensitiesare relative to 100 for the most abundant analyte ions. Nominal

mass is used for m/z value.

112

Although such ligand substitution reactions with thiols and phenols have
been studied in solution by 1H-NMRMG-7 1, similar studies involving alcohols
were not reported. However, it is intuitively expected that alcohols with
acidic -OH groups will behave similarly, perhaps with smaller equilibrium
constants. Obviously in this work, ligand subsitution reactions are
occuring, involving the -SR ligand on the iron-sulfur complex, and the
matrix molecules (alcohols). There are three possible explanations for this
occurance. First, it could be a simple analyte/matrix reaction that occurs
when the two are mixed. Second, it could be chemistry induced by the fast
atom beam (again, condensed phase chemsitry). Third, it could occur for
the gas phase ions, with desorbed matrix molecules, and not be
representative of solution chemistry at all. Since these peaks are so
dominant, such extensive conversion of reactants to products would be
unlikely if the chemistry occured in the gas phase. On the other hand, we
have used 1H-NMR to study the reaction between [Fe4S4(SEt)4]2' and NBA in
d5-DMSO. Ligand substitution chemistry does occur, to form [Fe4S4(SEt)4-
n(OCH2Ar)n]2‘, although the reaction is exceedingly slow, showing
approximately 15% of the reactant being converted into products after 72 h.
Thus, this may well be an example of fast atom bombardment-induced
chemistry, in which the particle bombardment of the matrix facilitates the

rate of the ligand exchange reactions.
(C).The FAB-MS analysis of mixed-ligand cubane clusters.

The utility of FAB-MS to analyze the mixed ligand cluster,
(Ph4P)2Fe4S4(SPh)2Clg, was evaluated. A simple spectrum representative

of this analyte was not expected, since NMR studie365 have shown that, in

113

solution, this complex disproportionates, and exists in equilibrium with

other mixed ligand clusters represented by equations 3-5

”OMSHIWZIZ——> [Fe4$4(SH1)C1312‘+[F04S4(SH1)3Cl]2' (3)
[Fe4S4(SPh)013]2' + [Fe4S4(SPh)3Cl]2'-->[Fe484(SPh)2Clz]2‘ + [Fe484(SPh)4]2‘ (4)
21PM —.> 11194510412 + WW (5)

The negative ion FAB-MS analysis, using the matrix NPOE, confirms the
existence of these disproportionation species in solution. The ions observed
are listed in Table 15. In the high mass range between 800-1100, there are
four major clusters of peaks corresponding to [(Ph4P)Fe4S4Cl4]‘ (m/z 831),
[(Ph4P)Fe4S4(SPh)Cl3]‘ (m/z 905), [(Ph4P)Fe4S4(SPh)2012]'. (m/z 979), and
[(Ph4P)Fe4S4(SPh)3CI]- (m/z 1053). A second group of peaks in the middle-
mass range are assigned as [Fe4S4Cl4]'/[Fe4S4(SPh)Cl]'/[Fe4S3(SPh)Clz]'
(overlapping cluster peaks with nominal m/z values of 492, 496 and 499,
respectively), [Fe4S4(SPh)Clg]-/[(Fe483(SPh)Cl3]- (overlapping cluster peaks
with nominal m/z values of 531 and 534, respectively), [Fe4S4(SPh)Cl3]' (m/z
566), [Fe4S4(SPh)2Cl]'/[Fe4S3(SPh)2012]' (overlapping cluster peaks with
nominal m/z values of 605 and 608, respectively ), [Fe4S4(SPh)2Clz]', (m/z
640), and [Fe4S4(SPh)3Cl]' (m/z 714). The group of peaks in the low mass
region show fragment ions related to the ligand dissociation or core
decomposition of [Fe4S4Cl4J' as described above. A series of oxidized
fragment ions, appearing 16u and/or 32u above the major fragments, were

also observed as described above.

Depending on the extent of disproportionation shown in equations 3-

5, we might expect to see mass spectral features indicative of

114

Table 15. Relative Intensities of Peaks in the Negative Ion FAB Mass
Spectrum of the complex, (Ph4P)2Fe4S4(SPh)2012, in DMF/NPOE“.

 

 

Fragment Ions m/z Relative Intesity
[(Ph4P)Fe4S4(SPh)3C1]' 1053 7
[(131141) )Fe454(SPh)2012]’ 979 15
[(Ph4P)Fe4S4(SPh)C13]’ 905 22
[(Ph4P)Fe4S4C14]' 831 17
[F8434(SPh)3C1]' 714 20
[Fe4S4(SPh)2C12]' 640 52
[Fe433(SPh)2C12]' 608 18
[Fe4S4(SPh)2C1]' 605 22
[Fe4S4(SPh)C13]‘ 566 63
[F6433(SPh)C13]‘ 534 30
[F e4S4(SPh)C12]' 531 62
[Fe4S3(SPh)C12]‘ 499 25
[Fe4S4(SPh)C1]' 496 45
[Fe4S4Cl4l' 492 47
[Fe483Cl41- 460 35
[Fe4S4Cl3]' 457 85
[Fe4S3C131' 425 56
[Fe4S4C12]‘ 422 100
[Fe4S3C12]' 390 33
[F6484C11' 387 68
[Fe382013l' 337 30
[FesssClzl’ 334 72

 

a. Intensities are relative to 100 for the most abundant analyte ions.
Nominal mass is used for m/z value. b. Peaks with m/z values less

than300 are not listed in this table.

115

(Ph4P)Fe4S4(SPh)4. Representative ions of this species are not seen in the
spectrum of (Ph4P)2Fe4S4(SPh)2012. Most of the fragments can be explained
by the loss and addition of SPh or Cl ligands from the mixed ligand
complex. At present, three clusters of peaks representing [Fe4S4(SPh)C13]',
[Fe4S4(SPh)2012]', and [Fe4S4(SPh)3Cl]' have been identified, but no evidence
was obtained for the existence of [(Ph4P)Fe4S4(SPh)4]‘ and/or [Fe4S4(SPh)4]'.
This maybe due to further equilibria established between [Fe4S4(SPh)4]2-
and [Fe4S4(SPh)4-nCln]2' species. This is may be expected in view of the
known lability of the [Fe4S4X4]2' complexe354°. Our FAB mass spectral
results are in excellent agreement with the aforementioned solution

equilibria.

2. Proposed Fragmentation Mechanisms: Correlations between Negative
Ion FAB-MS Data and Known Iron-Sulfur Cluster Chemistry in Condensed
Phases

Before evaluating the types of ions observed in these experiments,
first, a comment should be made on what might be expected in a mass
spectral study of ions derived from salts containing the Fe4S4 core in which
the iron atoms are in +2 and +3 formal oxidation states, as (Fe2+)2(Fe3+)2(Sz'
)4. When electron impact ionization is used to ionize a compound such as
Fe(CO)5, a dominant ion in the resulting mass spectrum is Fe+, which
represents an oxidation state not typically considered in condensed phases.
Thus, one might expect to generate fragments of the Fe4S4 core in which
unusual oxidation states of the metal are present. On the other hand, it is a

general "rule of thumb" in the mass spectrometry of inorganic and

116

organometallic compounds“, that the dominant ions are formed with a
minimal perturbation to the formal oxidation state of the metal in the
compound undergoing ionization. In the iron pentacarbonyl case, the iron
atom is formally taken from an oxidation state of 0 in the neutral compound
to +1 as the univalent cation (and in all fragment ions such as Fe(CO)3+).
In contrast, the 'l‘i+ peak in the mass spectrum of TiCl4 is a very minor
peak, because its formation would require the conversion of Ti“ to Ti+. In
this case, ions such as TiC13+ dominate, in which no net change in the
oxidation state of the metal occurs. Thus, we might expect, based on the
oxidation states of iron atoms in the Fe4S4 core, that fragment ions will be
formed in which the metal atoms are in their readily accessible +2 and +3

oxidation states.

Ionic complexes such as the (A)2Fe4S4X4 salts that are the subject of
this study, once dissolved in solution, produce a variety of ionic species such
as (A)+, (A)+[Fe4S4X4]2', and [Fe4S4X4]2', as suggested in Scheme I. When
the solution is subjected to fast atom bombardment, desorption of neutral
species and ionic complexes from the analyte/matrix target leads to a
variety of gas phase species. Presumably, the neutral, intact molecule
desorbs to some extent. Also, the ionic components desorb as ions
("preionized species", as they are commonly called in FAB), and from

these, the ions observed in the FAB spectra are derived.

Following desorption, the intact neutral molecule (A)2Fe4S4X4 may be
converted into the molecular ion. [(A)2Fe4S4X4]+, either via a gas phase
charge transfer reaction, or by a subsequent interaction of the desorbed

molecule with a fast atom. The countercation, A+, apparently forms

117

N .m .m .m

N .N.m .m

A

 

 

................................................. 4 III0'0000000000000000000000

 

, -v xvmvo mu<l1
385..
.vx.m.omm< .2
v
uxasag
.3289
.0.
lAmvexvmvomfix l
cousins a
coup—omen
. ............. + ................................................

 

-mxvmssa + +<11h-.x.maa< + 2....” £553 £258 5

m «Begum

118

adducts with gas phase species formed in the FAB process, yielding both
A+-matrix complexes, as well as adducts with the desorbed neutral analyte,
[(A)3Fe4S4X4]+. These are the two possible "primary" cations formed, from
which fragment ions may be generated. The major fragment ion is
[(A)2Fe4S4X3]+, which we propose is formed by the loss of an X radical from
the molecular cation. Consider the reaction of the loss of an X ligand, not
as an anion, but as a radical. When X is lost, the formal oxidation state of
one of the iron atoms must change (reduction). We propose two ways to
consider this simple process (which will become very powerful when
anionic fragments are considered). For the unimolecular dissociation

reaction (eq. 6),

A2F0484X4+ -> A2F94$4X3*+X° (6)

One can keep track of the oxidation states of the iron atoms, or of the [Fe4S4]

core. The formal charges on the four metal atoms change in this
dissociation reaction from {+3, +3, +3, +2} to {+3, +3, +2, +2}. Ifthe [Fe4S4]
core is considered, it changes from [Fe4S4]+3 to [Fe4S4]+2 (i.e., a net

reduction of the core) when a ligand is lost as a radical. Apparently no
fragments evolve from the adduct ion [(A)3Fe4S4X4]+, in which the iron

atoms are also in the oxddation states {+3, +3, +2, +2}.

Returning to Scheme I, the early steps in the chemistry that leads to
the desorption/ionization of aninnig species can be described, by considering
the species that can be desorbed from the analyte/matrix solution. Once

desorbed, the intact neutral molecule can capture an electron to form the
molecular anion. [(A)2Fe4S4X4]'. The "preformed ion", [(A)Fe4S4X4]', can be

119
desorbed directly from the matrix. Presumably, the [Fe4S4X4]2‘ dianion can
be desorbed directly, to some extent, although the spectra suggest that, if
this occurs, it is completely converted into the univalent anion, [Fe4S4X4]'.
Note that, these "primary" cations and anions shown in Scheme I, from
which all fragment ions will be formed, all contain iron atoms that only
involve combinations of the +2 and +3 formal oxidation states. Thus, while
it is certainly possible to generate complexes of iron in mass spectrometry

that contain Fe“, none are present in these ions listed in Scheme I.

Scheme II presents proposed unimolecular fragmentation pathways
to explain the evolution of the rich collection of fragment anions formed by
fast atom bombardment. The pathways proposed here are based on a few
simple assumptions. First, for an ion with a charge of -1 to form a
fragment ion with a charge of -1, either a radical or an uncharged even-
electron fragment must be lost. Obvious candidates for neutral species lost
include fragments such as X and FeS. One may consider loss of the neutral
[FeSX], but not [Fe] alone, because of the bonding environment in the
starting material. That is, we would not expect a fragment ion with 2 or 3
iron atoms, but all four X ligands still present. Next, the loss of neutral “A”
was not considered - that is, while one may expect to lose a chlorine radical,
one would not lose a tetra-alkyl ammonium radical. Third, the loss of “AX”
was considered as an allowable fragmentation. An example of ""AX
elimination is shown in equation 7. If, for example, A+ is Et4N+, and the

cluster contains X=Br, ""AX loss is equivalent to the loss of {Et3N and EtBr}.

[(Et4N)2Fe4S4Br4]'--->[(Et4N)Fe4S4Br3]'+Et3N + EtBr (7)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Scheme 11 . .
ELAm 225.com
W' 3, 3, 3. 2 +3
3, 3, 2, 2 +2
A2333“; 3. 2 2. 2: +1 .....
A,.\ _S
3. 2. 2. 2 +1
-2FcSX
«, 3, 3. 2, 2 +2
I ' , 3, 2, 2, 2 +3
I -X‘ 3, 2. 2. 2 +1
‘ ' 2. 2. 2. 2 +2
M -FcS I -FcS 2. 2. 2. 2 0
2. 2. 2 +2
3, 2, 2 +1
2. 2. 2 __ g. _
3, 2 +1

 

 

 

121

Scheme II lists the 4, 3 and 2-Fe-containing fragment ions that are
observed, with various numbers of sulfur atoms and X ligands
incorporated. The column on the right side of Scheme II lists the formal
oxidation states of the iron atoms associated with each chemical species. It
clearly shows that all of the unimolecular dissociation chemistry occurs
without having to invoke the formation of exotic valence states of any of the
iron atoms within the remaining core. This becomes very important in
understanding and interpreting the mass spectra of such compounds. It
suggests that, in the negative ion spectra of iron sulfur clusters, an anion
representative of the cubane core [Fe4S4]‘ would not be expected, since this
would force one of the iron atoms to be reduced from +2 to +1. In fact, the
fragment ion, [Fe4S4}', is only detected from ionization of [Fe4S4(SEt)4]2'.
Essentially, all of the ions observed are those that can exist in the context of

the one restriction - maintaining the iron atoms as either +2 or +3 metals.

The observation of [Fe4S4]‘, in one case only, deserves comment.
Whatever its structure, the four iron atoms must be in the oxidation states
{+2, +2, +2, +1}. As seen in Table III, this anion is only formed from
compounds that also form the [Fe4S5]' anion. The two may be chemically
linked, with [Fe4S4]' being a fragment of the more abundant [Fe4S5]'
species. We note that this one violation of the oxidation state restriction,
that seems to hold for all other fragment ions, involves a complete cubane-
core, as opposed to some cubane-fragment. Of all of the [FemSn] species
encountered in this work, it would surely be the intact cubane that could
most effectively delocalize an extra electron. Thus it is not surprising that
the only exception to the rule occurs not for a small fragment ion but for a

122

larger species containing four iron atoms, in which the cubic geometry is

presumably intact.

Further insights into the fragmentation mechanisms, which provide
a very interesting link between the ions listed in Scheme H, can be seen by
considering the overall oxidation state of the [FemSn] ms, of the various
ions. For all of the fragmentations suggested, if the oxidation state of the
iron/sulfur core is considered, only reductive eliminations, or
fragmentations in which there is no change in the oxidation state, are
observed. The extent to which a "primary ion" leads to fragment ions is
clearly linked to its [FemSn] core on'dation state. The ion that leads to most
of the fragment anions observed is [Fe4S4X4]', which contains a [Fe4S4]3+
core. Clearly, it leads to fragment ions with [FemSn] core oxidation states of
+3, +2, and +1. When it fragments to form an ion with an FemSn core that
is in a +3 or +2 state, that fragment dissociates further. The most dramatic
example is shown in the pathway in Scheme II that leads from [Fe4S4X4J' to
[Fe3S3X]', in which the [FemSn] cores of the species involved smoothly
change their oxidation state from +3 to +2 to +1 to 0! Thus, we propose here
that, in the FAB analyses of iron-sulfur cluster compounds, the primary
ions formed by FAB fragment through reductive chemistry in which all of
the metal atoms retain +2 and +3 oxidation states, and the [FemSn] core
oxidation states do not increase as the unimolecular dissociations occur.

Both one and two electron reductions are proposed in Scheme II.

These insights make it straightforward to interpret these mass
spectra and will be important guidelines with which spectra of related

compounds can be correlated with structure. For each ion that is a

123

primary ion of FAB, we can predict which fragment ions can be formed and
which can not. An example is shown in Scheme III, beginning with the
ionic species, [Fe4S4X4]', as an example. One might expect that the ion
9.01.1111 lose as possible neutral fragments X, S, FeS, FeX, FeSX, or Fe. We
consider each of these possibilities in Scheme III: pathway a: the
conversion of [Fe4S4X4]' to [Fe4S4X31- suggests the dissociation of a labile
ligand X as a radical. It forms a fragment ion which only incorporates
Fe(II) and Fe(III)’s, thus should be expected and is observed; pathway b:
the same is true for the degradation to [Fe4S3X41‘ by the loss of S, which
produces a {+3, +2, +2, +2] iron core by a two electron reduction; pathway c:
the destruction of the cubane core by the loss of FeS also leads to a fragment
ion in which no Fe atoms need to be in states other than +2 or +3, however,
it is not observed. Presumably, the halogens are bonded to the iron atoms,
and an iron cannot be lost with the ionic fragment retaining four halogens;
pathway d: the fragment from the loss of a relatively large FeSX fragment
is reasonable, yet is not observed (i.e. no [Fe383X3]' is detected). The
stepwise loss of X, and then FeS is a more favorable pathway as shown in
Scheme 11 for other ions; pathway e: Fe atom ejection from the cubane core
would be unexpected for two reasons. As mentionned above, an ion cannot
retain four halogens with only three Fe atoms in the core. Also, if Fe(O)
were lost, the oxidation state of the remaining metals must change to (+4,
+4, +3}, which clearly does not occur in these systems, consistent with the
fact that the product [Fe3S4X41' is never observed; pathway f: the loss of the
neutral FeX would also be unexpected, since the oxidation state of one of the

remaining Fe atom must be changed to Fe“.

 

oqum
\
1...!
tannin
x; x a
calm ..
Lam . .
on. a
51 E
x x
I .
:4.»
an .on..\_
a. p.002:
X) x\
mun—rim mom. ._|
_ ml...
itnlm. Ix m. 3.
R
z
x x
) .
alum
3.1....
On—l
scale. x

:— use-Um

125

The correlation of gas phase [FemSn] anionic species, generated from
the cubane cluster by FAB, with known chemistry of iron-sulfur complexes
is intriguing. In the condense phase (i.e. solution), complexes have been
made in which the [Fe4S4] core is in an oxidation state of +1, +2, or +3537.
All three oxidation states are observed in the negative ion spectra, Scheme
II. However, in the gas phase, the unusual 0 oxidation state of [Fe4S4] core
also exists. The [Fezsgl core is known in solution with +2 and +1 oxidation
states in solution; the +1 state is observed in the gas phase. Thus, the
oxidation states observed in solution to date provide some useful limits as
what to expect in the gas phase. The gas phase data may also suggest
species that mid exist in condensed phases, and could be pursued
synthetically.

With the insights gained from analysis of the negative ion FAB
spectra, correlating possible fragment ions with limitations in available
oxidation states, it becomes obvious why negative ion FAB is the mass
spectral technique of choice, when both molecular weight and structural
information is desired. This also explain why the positive ion spectra are so
simple. In Scheme III, most of the fragment ions trace their origin to
[Fe4S4X4]', where the four iron atoms are in the states {+3, +3, +3, +2}. Why
does the corresponding cation not lead to a variety of fragment cations? In
fact, the corresponding cation is not formed by FAB. To convert this anion
to a cation would require the removal of two electrons, which would result
in the four iron atoms having oxidation states of [+4, +3, +3, +3}. Thus, the
limitations on oxidation states of the iron atoms gives fewer choices for
cationic species. Obviously, there are "oxidation state bottlenecks" that

would be expected for the unimolecular chemistry of cations formed by

1%

FAB; a rich cation chemistry is not expected, that would parallel the

unimolecular fragmentation anion chemistry.

CONCLUSIONS

We have reported herein the utility of FAB-MS for the
characterization of iron/sulfur cubane-containing compounds, and have
discussed aspects of the fragmentation mechanisms that will assist in the
mass spectral interpretation of related compounds. Both 3-nitrobenzyl
alcohol (NBA) and 2-nitrophenyl octyl ether (NPOE) are suitable matrices
for the FAB-MS studies of these series of complexes. It should be
emphasized that, when such compounds are being characterized, it is vital
that the experimentally observed and theoretically calculated isotopic
patterns be compared, to assist in the correct identification of the ions

formed.

Fast atom bombardment of a matrix containing iron/sulfur cubane
clusters has been evaluated here, as a chemical system which generates a
variety of smaller clusters in a variety of oxidation states, through
unimolecular fragmentation processes. The restrictions on oxidation
states that seem to dominate the fragmentation pathways yields interesting

parallels with known clusters in the condensed phase.

In addition, we have successfully introduced FAB-MS analysis as a
new methodology to characterize intermediates of ligand substitution

reaction between analyte and matrix such as [Fe4S4(SEt)4]2' and 3-

127

nitrobenzyl al-chohol, as well as to confirm the existence of
disproportionation species in solution, (i.e., (Ph4P)2Fe4S4(SPh)2C12 in DMF).
Furthermore, both NBA and NPOE participates in an unusual gas phase

oxygen atom transfer reactions with the iron-sulfur complexes.

It will be interesting to compare the results presented here with FAB-
based analysis of small redox enzymes that contain an iron/sulfur cubane
linkage. It will be intriguing to see how the cubane core fragments when
attached to a peptide relative to thiolate, since this may give insight on how
peptides regulate the redox properties of Fe/S clusters. Furthermore, it
should be noted that, using mass spectrometry, ion-molecule reactions of
any of the anions discussed in this work could be studied and used to
determine properties of anions in the gas phase such as electron affinities
and reactivity. Thus, FAB-MS offers not only a facile method for analysis,
but an opportunity for the generation of unusual chemical species, and the

characterization of very rich gas phase chemistry.

REFERENCES AND NOTES

5959‘!“

9°

10.

11.
12.
13.

14.
15.
16.
17.

REFERENCESANDNOTEB

Eschenmoser, A. Angew. Chem. 1988, 100, 5; Angew. chem. Int. Ed.
Engl. 1988,27, 5.

Dolphin, D., Ed., The Porphyrins, Vol. 1-7, Academic Press, New
York, 1978.

(a) Scheer, H. In The Porphyrins, Vol. II, Part B, Dolphin, D., Ed.,
Academic Press, New York, 197 8. Chapter 1. (b) Clayton, R. K.,
Sistrom, W.R. The Photosynthetic Bacteria, Plenum Press, New York,
1978.

Dolphin, D. Ed., 312,, Vol. 1-2, Wiley-interscience, New York, 1982.
Eschenmoser, A. Ann. N. Y. Acad. Sci. 1986, 471, 108.

Timkovich, R.; Cork, M. S.; Taylor, P. V. J. Biol. Chem. 1984, 259, 1577.
(8) Chang, C. K. J. Biol. Chem. 1985, 260, 9520. (b) Chang, C. K;
Timkovich, R.; Wu, W. Biochemistry 1986, 25, 8447.

Losada, M. J. Mol. Catal. 1976, 1, 245.

Forget, P.; DerVartanian, D. V. Biochim. Biophys. Acta. 1972, 256, 600.
Murphy, M. J .; Siegel, L. M.; Tove, S. R.; Kamin, H. Proc. Natl. Acad.
Sci. USA 1974, 71, 612. ‘
Vega, J. M.; Garret, R. H.; Siegel, L. M. J. Biol. Chem. 1975, 250, 7980.
Vega, J. M.; Kamin, H. J. Biol. Chem. 1977, 252, 896.

Scott, A. I.; Irwin, A. J.; Siegel, L. M.; Schoolery, J. N. J. Am. Chem.
Soc. 1978, 100, 7987.

Battersby, A. B.; McDonald, E. Bioorganic Chem. 1978, 7, 161.

Payne, W. J. Denitriflcation 1981, Wiley, New York.

Knowles, R. Microbiol. Rev. 1982, 46, 43.

Iwasaki, H.; Saigo, T.; Matsubara, T. Plant Cell Physiol. 1980, 21; 1573.

128

18.
19.

21.

129

Matsubara, T.; Frunzke, K; Zumfi, W. G. J. Bacteriol. 1982, 149, 816.
Chang, C. K; Wu, W. J. Biol. Chem. 1986, 261, 8593.

Chang, C. K; Barkigia, K M.; Hanson, L. K; Fajer, J. J. Am. Chem.
Soc. 1986, 108,1352.

Wu, W.; Chang, C. K J. Am. Chem. Soc. 1987, 109, 3149.

(a) Janick, P. A.; Siegel, L. M. Biochemistry 1982, 21, 3538; 1983, 22,
504. (b) Christner, J. A.; Munck, E.; Janick, P. A.; Siegel, L. M. J.
Biol. Chem. 1983, 256, 11147. (c) Christner, J. A.; Janick, P. A.; Siegel,
L. M.; Munck, E.; J. Biol. Chem.1983, 256, 11157. (d) Christner, J. A.;
Munck, E.; Kent, T. A.; Janick, P. A.; Salerno, J. C.; Siegel, L. M. J.
Am. Chem. Soc. 1984, 106, 6786. (e) Cline, J. F.; Janick, P. A.; Siegel,
L. M.; Hoffman, B. M. Biochemistry 1986, 25, 4647.

(a) Young, L. J .; Siegel, L. M. Biochemistry 1988, 27, 2790, 5984. (b)
Wilkerson, J.O.; Janick, P. A.; Siegel, L. M. Biochemistry 1988, 22,
5048.

(a)Ellefson, W. L.; Whitman, W. B.; Wolfe, R. S. Proc. Natl. Acad. Sci.
U.S.A. 1982, 79, 3707. (b)Ellefson, W. L.; Wolfe, R. S. J. Biol. Chem.
1980, 255, 8388. (c)Nagle, D. P.; Wolfe, R. S. Proc. Natl. Acad. Sci.
U.S.A. 1983, 80, 2151.

Pfaltz, A. In The Bioinorganic Chemistry of Nickel; Lancaster, J. R.,
Ed.; VCH Publishers: New York, Chapter 12

(a) Pfalz, A.; Jaun, B.; Fassler, A.; Eschenmoser, A.; Jaenchen, R.;
Gilles, H. H.; Diekert, G.; Thauer, R. K Helv. Chim. Acta 1982, 65,
828. (b) Livingston, D. A.; Pfalz, A.; Schreiber, J .; Eschenmoser, A.;
Ankel-Fuchs, D.; Moll, J.; Jaenchen, R.; Thauer, R. K. Helv. Chim.
Acta 1984, 67, 334. (c) Pfalz, A.; Livingston, D. A.; J aun, B.; Diekert,
G.; Thauer, R. K. Eschenmoser, A.; Helv. Chim. Acta 1985, 68, 1338.

27.

30.

31.
32.

33.

3?

37.

13)

(d)Fassler, A.; Kobelt, A.; Pfalz, A.; Eschenmoser, A.; Bladon, C.;
Battersby, A. R;; Thauer, R. K Helv. Chim. Acta 1985, 68, 2287.
Kratky, C.; Waditshatka, R.; Angst, C.; Johnasen, J. E.; Plaquevent, J.
C.; Schreiber, J .; Eschenmoser, A.; Helv. Chim. Acta 1985, 68, 1312.
Kratky, C.; Fassler, A; Pfaltz, A.; Krautler, B.; Jaun, B.;
Eschenmoser, A. J. Chem. Soc., Chem. Commun. 1984, 1368.
Waditshatka, R. Kratky, C.; Jaun, B.; Heinzer, J .; Eschenmoser, A.;
J. Chem. Soc., Chem. Commun. 1985, 1604.

Kratky, C.; Angst, C.; Johnasen, J. E.; Angew. Chem., Int. Ed. Engl.
1981, 20, 211.

Fabrizzi, L. J. Chem. Soc. Dalton. Trans. 1979,1857.

Kaplan, W. A.; Scott, R. A.; Suslick, K S. J. Am. Chem. Soc., 1990,
112, 1283.

(a) Baker, E. W.; Brookhart, M. S.; Corwin, A. H. J. Am. Chem. Soc.
1964, 86, 4587. (b) Mclees, B. D.; Caughey, W. S. Biochemistry 1968, 7,
642. (c) Abraham, R. J .; Swinton, P. F. J. Chem. Soc. B 1969, 903. (d)
Curthoys, G. C.; Magnusson, E. A.; Phillips, J. N. Inorg. Chem. 1972,
11, 1024. (e) Pastemack, R. F.; Spiro, E. G.; Teach, M. J. Inorg. Nucl.
Chem. 1974, 36, 599. (D Walker, F. A.; Hui, E.; Walker, J. M. J. Am.
Chem. soc. 1975, 97, 2390.

Bottomly, L. A.; Olson, L.; Kadish, K M. Inorg. Chem. 1982, 21, 1024
Fassler, A. Dissertation, RTH, Zurich Nr. 7799, 1985.

Davis, D. G. In The Porphyrins; Dolphin, D. Ed., Academic: New
York, 1978; Vol. V, Chapter 4.

(a) Stolzenberg, A. M.; Stershic, M. T.; J. Am. Chem. Soc. 1988, 110,
5397. (b) Stolzenberg, A. M.; Stershic, M. T.; Inorg. Chem. 1987, 26,

39.

41.

42.

47.

131

3082.(c) Stolzenberg, A. M.; Stershic, M. T.; Inorg. Chem. 1988, 27,
1614.

Chang, C. K; Hanson, L. K; Rickardson, P. E; Young, R.; Fajer, J.
Proc. Natl. Acad. Sci. U.S.A. 1981, 78, 2652.

Connick, P.; Macor, K Inorg. Chem. 1991, 30, 4654

Bottomly, L. A.; Olson, L.; Kadish, K. M. In Electrochemical and
Spectrochemical Studies of Biological Redox Components; Kadish, K.
M. Ed.; ACS Sym. Ser. 1982, 201, 279

Montforts, F.-P.; Romanowski, F.; Bates, J. W. Angew. Chem. Int. Ed.
Engl. 1989, 28, 480.

Scheidt. W. R.; Lee, Y. J. Struct. Bonding 1987, 64, 1.

Kirner, J. F.; Garofalo, J .; Scheidt, W. R. Inorg. Nucl. Chem. Lett.
1975, 107.

Lynch, M. W.; Buchanan, R. M.; Pierpont, C. G.; Hendrickson, D. N.
Inorg. Chem. 1981, 20, 1038

Chang, C. K Biochemistry 1980, 19, 1971.

Alder, A. D.; Longo, F. R.; Finarelli, J. O.; Goldmacher, J.; Assour,
J .; Korsakofi‘, L. J. Org. Chem., 1967, 32, 476.

Chang, C. K; Hataela, M. H.; Tuliusky, A. J. Chem. Soc. Perkin
Trans. II 1983, 371.

(a) Inhoffen, H. H.; Jager, P. Tetrahedron Lett. 1964, 1317. (b)
Inhoffen, H. H.; Jager, P.; Mahlhop, R.; Mengler, C. D. J. Liebigs
Ann. Chem. 1967, 704, 188.

Bent, H. E.; French, C. L. J. Am. Chem. Soc. 1941, 63, 568.

Beevers, L.; Hageman, R. H. Annuv. Rev. Plant Physiol., 1969, 20, 495.

51.

52.

57.

59.

61.

132

Ingraham, J. L. In Denitrification, Nitrification and Atmospheric

Nitrous Oxide, Delwiche, C. C. Ed.; John Wiley & Sons, New York,

1981, 45.

Murphy, M. J .; Siegel, L. M.; Kamin, H.; Rosenthal, D. J. Bio. Chem.,
1973, 248, 2801.

Madden, J. F.; Han, S.; Siegel, L. M.; Spiro, T. G. Biochemistry, 1989,
28,5471. '

McRee, E. D.; Richardson, J. S.; Siegel, L. M. J. Biol. Chem. 1986, 261,
10277.

Cowan, J. A.; Sola, M. Inorg. Chem. 1990, 29, 2176.

(a)Holm, R. H.; Ciurli, S; Weigel, J. A. Prog. Inorg. Chem. 1990, 38, 1.
(b) Stack, T. D. P; Weigel, J. A.; Holm, R. H. Inorg. Chem. 1990, 29,
3745. '(c) Ciurli, S.; Carrie~, M.; Weigel, J. A.; Carney, M. J.; Stack, T.
D. P.; Papaefthymiou, G. C.; Holm, R. H. J. Am. Chem. Soc. 1990, 112,
2654. (d) Weigel, J. A.; Holm, R. H.; Srivastava, K K P.; Munck, E. J.
Am Chem. Chem. 1990, 112, 8015. (e) Whitener, M. A., Peng, G.; Holm,
R. H. Inorg. Chem. 1991, 30, 2411. (f) Weigel, J. A.; Holm, R. H. J. Am.
Chem. Soc. 1991, 113, 4184. (g) Liu, H. Y.; Scharbert, B.; Holm, R. H. J.
Am Chem. Soc., 1991, 113, 9529.

Young, R.; Chang, C. K J. Am Chem. Soc. 1985, 107, 898.

Holm, R. H.; Philips, N. D.; Averill, B. A.; Mayerle, J. J .; Herskovitz,

T. J. Am. Chem. Soc., 1974, 96, 2109.

Lindsey, J. J. Org. Chem. 1980, 45, 5215.

Johnson, A. W.; Kay, I. T. J. Chem. Soc., 1965, 1620.

Harris, D.; Johnson, A. W.; Gaete-Holmes, R. Bioorganic Chem. 1980,
9, 63.

62.

133

Sun, Y.; Martel], A. E.; Chen, D.; Macfarlane, R. D.; McNeal, C. J. J.
Heterocycl. Chem. 1986, 23, 1565.

63. Young, R. N.; Ganthier, J. Y.; Coombs, W. Tetrahedron Lett, 1984, 25,

64.

67.

1753.

(a) Wong, G. B.; Bobrik, M. A.; Holm, R. H. Inorg. Chem. 1978, 17, 578.
(b) Coucouvanis, D.; Kanatzidis, M. Simhon, E.; Baenziger, D. C. J.
Am Chem. Soc. 1982, 104, 1874. (c) Kanatzidid, M. G. ; Hagen, W. R.;
Lester, R. K; Coucouvanis, D. J. Am. Chem. Soc. 1985, 107, 953. (d)
Kanatzidis, M. G.; Salifoglou, A.; Coucouvanis, D. Inorg. Chem. 1986,
25, 2460. (e) Coucouvanis, D.; Salifoglou, A.; Kanatzidis, Dunham, W.
R.; Simopoulos, A.; Kostikas, A. Inorg. Chem. 1988, 27, 4066.

(a) Kanatzidis, M. G.; Ryan, M.; Coucouvanis, D.; Simopoulos, A.

Kostikas, A. Inorg. Chem. 1983, 22, 179. (b) Johnson, R. E.;

Papaefthymiou, G. C.; Franke, R. B.; Holm, R. H. J. Am. Chem. Soc.

1983, 105, 7280. (c) Kanatzidis, M. G.; Baenziger, N. C.; Coucouvanis,

D.; Simopoulos, A.; Kostikas, A. J. Am. Chem. Soc. 1984, 106, 4500. (d)
Kanatzidis, M. G.; Coucouvanis, D.; Simopoulos, A.; Kostikas, A.;

Papaefthymiou, V. J. Am. Chem. Soc. 1985, 107, 4925.

Thauer, R. K ; Schoenheit, P. In Iron-Sulfur Proteins; Sprio, T. G.,

Ed.; Wiley Interscience: New York, 1982; 329.

Thompson, A. J. In Metalloproteins; Harrison, P., Ed.; Verlag

Chemie: Weinheim, FRG, 1985; Part I, 79.

(a) Carter, C. W.; Kraut, J. Freer, S. T.; Alden, R. A.; Sieker, L. C.;

Adman, E.; Jensen, L. H. Proc. Natl. Acad. Sci. U. S. A. 1972, 68, 3526.

(b) Herskovitz, T.; Averill, B. A.; Holm, R. H.; Ibers, J. A.; Philips, W.

D.; Weiher, J. F. Proc. Natl. Acad. Sci. U. S. A. 1972, 69, 2437.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

134

(a) Holm, R. H. Acc. Chem. Res. 1977, 10, 427. (b) Berg, J. M.; Holm, R.
H. In Iron-Sulfur Proteins; Spiro, T. G., Ed.; Wiley-Interscience: New
York, 1982; 1.

Averill, B. A. In Metal Clusters in Proteins; Que, L., Jr., Ed.; ACS
SymSer. 1988, 372, 258.

Cleland, W. E. Jr.; Holtman, D. A.; Sabat, M., Ibers, J. A. Defotis, G.
C.; Averill, B. A. J. Am. Chem. Soc. 1983, 105, 6021. (b) Cleland, W. E.,
Jr.; Averill, B. A. Inorg. Chim. Acta 1985, 106, L17.

Nakamura, A.; Ueyama, N. in Metal Clusters in Proteins; Que, L.,
Jr., Ed.;ACS Sym. Ser. , 1988, 372, 292.

Nakamura, A.; Ueyama, N. Advances in Inorganic Chemistry 1989,
33, 39.

(a) Ohno, R.; Ueyama, N.; Nakamura, A. Chem. Lett. 1989, 399. (b)
Ohno, R.; Ueyama, N.; Nakamura, A. Inorg. Chim. Acta, 1990, 253.
Barber, M; Bordoli, R. S.; Sedgwick, R. D.; Tyler, A. N. J. Chem. Soc.
Chem. Commun. 1981, 325.

Miller, J. M. Mass Spectrom. Rev. 1989, 9, 319 and references therein.
Fenselau, C; Cotter, R. J. Chem. Rev. 1987, 87, 501.

Pachuta, S. J .; Cooks, R. G. Chem. Rev. 1987, 87, 647.

Sethi, S. K; Nelson, C. C.; McClosky, J. A.; Anal. Chem. 1984, 56,
1977.

Divisia-Blohorn, B.; Kyriakakov, G.; Ulrich, J. Org. Mass. Spectrom.
1985,22, 463.

Hegetscheveiler, K; Keller, T.; Amrein, W.; Schneider, W. Inorg.
Chem. 1991, 30, 873.

(a) Knowalski, M. H.; Sharp, T. R. Stang, P. I. Org. Mass. Spectrom.
1987, 22, 642. (b) Boyle, P. D.; Johnson, B. J.; Alexander, B. D.;

86.

87.

15

Casalnuovo, J. A.; Gannon, P. R.; Johnson, S. M.; Larka, E. A.;
Mueting, A. M.; Pignolet, L. H. Inorg. Chem. 1987, 26, 1346.
McElvany, S. W.; Allison, J. Organometallics 1986, 5, 1219.

(a) Bojesen, G. Org. Mass Spectrom. 1985, 20, 415. (b) Miller, J. M.
Balasumugan, K; Nye, J.; Deacon, G. B.; Thomas, N. C. Inorg.
Chem. 1987, 26, 560. (c) Liang, X.; Suwanrumpha, S.; Freas, R. B.
Inorg. Chem. 1991, 30, 652.

Unfortunately, thermochemical information on the 02H5S radical is
not available. Heats of formation of gas phase CHgs, CH3 and S
suggest a C-S bond strength of 72.4 kcal/mol, which should be typical of
alkyl-S bond strengths. In contrast, available data suggests a C-S bond
strength of 90.2 kcal/mol in PhS. If the C-S bondstrengths in C2H5SH
and CsH5SH are compared, the alkyl thiol has a bond strength 13

kcal/mol less than the aromatic compound. Complete

thermochemical information on the anions are not available, however

if one considers the species that are isoelectronic with C2H5S' and
CsH5S', which are 02H5Cl and C5H5Cl, again the C-X bond is lower by
greater than 10 kcal/mol for the alkyl group. Thus, it is safe to assume
that alkyl-S bonds are weaker than phenyl-S bonds. All
thermochemical data taken from S.G. Lias, J .E. Bartmess, J .F.
Liebman, J .L. Holmes, R.D. Levin, W.G. Mallard, J. Phys. Chem. Ref.
Data, 1988, 17, Suppl. no. 1.

Charalambous, J. Mass Spectrometry of Metal Compounds,
Butterworth’s, Boston, 197 5.

O’Sullivan, T.; Millar, M. M. J. Am. Chem. Soc. 1985, 107, 4096.

"‘1111111111111111