Synthesis, characterization, and fiber-optic infrared reflectance spectroelectrochemical studies of some dinitrosyl iron diphosphine complexes Fe(NO)2L2 (L = P(C6H4X)3)

Article abstract of DOI:10.1016/j.jorganchem.2013.12.040

A series of iron dinitrosyl complexes of the form Fe(NO) ₂(P(C₆H₄X)₃)₂ (X = p-OMe (1), p-Me (2), m-Me (3), p-H (4), p-F (5), p-Cl (6), p-CF₃ (7)) has been prepared from the reactions of Fe(NO)₂(CO)₂ and the respective triarylphosphines. Complexes 1-7 have been characterized by IR and ³¹P NMR spectroscopy, and by X-ray crystallography for 1 and 7. In general, the compounds with the more basic phosphines display lower υ_NO stretches in the IR spectra than those with the less basic phosphines, and the trends in υ_NO as a function of Hammett parameter and solvent donor/acceptor number were analyzed. The redox behavior of compounds 1-7 in CH₂Cl₂ were studied by cyclic voltammetry at a Pt electrode. In general, the compounds undergo one-electron oxidations. Infrared spectroelectrochemistry revealed that the oxidations generate the derivatives with υ_NOs that are ～100 cm ^-1 higher in energy indicative of Fe(NO)₂-centered oxidations.

Full text of DOI:10.1016/j.jorganchem.2013.12.040

Accepted Manuscript
Synthesis, Characterization, and Fiber-optic Infrared Reflectance
Spectroelectrochemical Studies of Some Dinitrosyl Iron Diphosphine Complexes
Fe(NO) L (L = P(C H X) )
6 4
2
2
3
Myron W. Jones, Douglas R. Powell, George B. Richter-Addo
PII:
S0022-328X(13)00921-2
10.1016/j.jorganchem.2013.12.040
JOM 18426
DOI:
Reference:
To appear in: Journal of Organometallic Chemistry
Received Date: 17 September 2013
Revised Date: 12 December 2013
Accepted Date: 20 December 2013
Please cite this article as: M.W. Jones, D.R. Powell, G.B. Richter-Addo, Synthesis, Characterization, and
Fiber-optic Infrared Reflectance Spectroelectrochemical Studies of Some Dinitrosyl Iron Diphosphine
Complexes Fe(NO) L (L = P(C H X) ), Journal of Organometallic Chemistry (2014), doi: 10.1016/
6 4
2
2
3
j.jorganchem.2013.12.040.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Graphical Abstract Synopsis:
Dinitrosyl iron complexes (DNICs) are important in biology and have utility in
chemical catalysis. Several new DNICs of the form Fe(NO) (PAr ) (PAr =
2
3 2
3
triarylphosphine) have been prepared and characterized by spectroscopy and X-ray
crystallography. The results of the spectroscopy and electrochemistry reveal subtle
substituent effects upon phosphine variation.

ACCEPTED MANUSCRIPT
Highlights:
1
2
3
4
. New dinitrosyl iron complexes with phosphine ligands have been prepared.
. X-ray crystallography reveals essentially linear FeNO linkages in these compounds.
. Phosphine substitutions have effects on the spectroscopy and electrochemistry.
. Infrared spectroelectrochemistry reveals metal-centered oxidations.

ACCEPTED MANUSCRIPT
REVISED Ms. Submitted to J. Organomet. Chem. 12.12.2013
Synthesis, Characterization, and Fiber-optic Infrared Reflectance
Spectroelectrochemical Studies of Some Dinitrosyl Iron Diphosphine
Complexes Fe(NO) L (L = P(C H X) )
2
2
6
4
3
a,b
a
a
Myron W. Jones, * Douglas R. Powell, and George B. Richter-Addo *
a
Department of Chemistry and Biochemistry, University of Oklahoma, 101 Stephenson Parkway,
Norman, OK 73019. Email: grichteraddo@ou.edu
b
Department of Chemistry, Southern Illinois University Edwardsville, IL 62026. Email:
myrjone@siue.edu.
_
____________________________________________________________________________
ABSTRACT: A series of iron dinitrosyl complexes of the form Fe(NO) (P(C H X) ) (X = p-
2
6
4
3 2
OMe (1), p-Me (2), m-Me (3), p-H (4), p-F (5), p-Cl (6), p-CF (7)) has been prepared from the
3
reactions of Fe(NO) (CO) and the respective triarylphosphines. Complexes 1-7 have been
2
2
3
1
characterized by IR and P NMR spectroscopy, and by X-ray crystallography for 1 and 7. In
general, the compounds with the more basic phosphines display lower υNO stretches in the IR
spectra than those with the less basic phosphines, and the trends in υNO as a function of Hammett
parameter and solvent donor/acceptor number was analyzed. The redox behavior of compounds
1
-7 in CH Cl were studied by cyclic voltammetry at a Pt electrode. In general, the compounds
2 2
undergo one-electron oxidations. Infrared spectroelectrochemistry revealed that the oxidations
-1
generate the derivatives with υ s that are ~100 cm higher in energy indicative of Fe(NO) -
NO
2
centered oxidations.
Keywords: iron, nitrosyl, nitric oxide, X-ray, spectroelectrochemistry

ACCEPTED MANUSCRIPT
2
1
. Introduction
Iron is a major target for nitric oxide (NO) in biological systems and in the environment.
The resulting ground-state Fe–N–O moiety may be linear or bent [1]. The dinitrosyl Fe(NO)2
moiety has been suggested as a biologically relevant entity [2-7]. Exogenous and endogenous
NO, upon exposure to and contact with Fe-containing biomolecules, generates a species
exhibiting a strong EPR signal at g = 2.03 attributed to an “Fe(NO) ”-containing species [3].
2
Other ligands (L) coordinate to the Fe(NO) unit to give Fe(NO) L derivatives [1]; these latter
2
2 x
species are collectively referred to as “dinitrosyl iron compounds” (DNICs). DNICs, although
mostly discussed in the literature in the realm of biology, also have significance in catalysis as
described below. Their syntheses and characterization, to a large extent, preceded their
discovery in biology.
Synthetic DNICs display a variety of chemical properties including the ability to transfer
oxygen atoms from molecular oxygen to phosphines or alkenes [8-17]. For example, Postel and
co-workers demonstrated that the dinitrosyl compounds [Fe(NO) X] (X = Cl, I) react with
2
2
oxygen in the presence of PPh or OPPh to produce nitrato complexes such as
3
3
Fe(NO ) X(OPPh ) [8-12]. These latter nitrato complexes transfer oxygen atoms to phosphines
3
2
3 2
or cyclohexene and regenerate the nitrosyl moiety [10][12].
In 1965, Maxfield reported in a patent the use of dinitrosyl iron halides as effective
catalysts for the dimerization of diolefins in the presence of reducing agents [17]. Others later
demonstrated that some synthetic DNICs could be used for the cyclodimerization of diolefins
when reduced electrochemically [15]. Candlin [13] reported the dimerization of butadiene and
isoprene by Fe(NO) (CO) , and Gadd [14] proposed that the photocatalytic dimerization might
2
2
2
4
occur via Fe(NO) (η -C H )(η -C H ) as an intermediate. In 1994, Li [18] reported
2
4
6
4
6

ACCEPTED MANUSCRIPT
3
2
Fe(NO) PPh (η -TCNE) as the first stable compound containing an olefin π-bonded to an iron
2
3
dinitrosyl group.
Returning to the biological significance of DNICs, it was originally proposed that DNICs
formed from the reaction of NO with the active center of iron sulfur proteins [19-21]. Recently,
Ricci and co-workers reported the crystal structure of a stable complex of human glutathione
transferase P1-1 containing a post-translationally modified dinitrosyliron glutathionyl moiety
[
22]. It is now generally accepted that the iron in biological Fe(NO) DNICs comes
2
predominately from “freely chelatable iron storage units” and not necessarily from FeS proteins
23-25]. For example, Lancaster et al. recently published work describing the direct reaction of
[
NO with the intercellular labile iron and found that NO rapidly and quantitatively reacts with the
iron to form stable DNICs detectable by EPR [24].
All DNICs derive their basic functionality from the dinitrosyl Fe(NO) moiety. It is
2
known that the carbonyl ligands of Fe(NO) (CO) can be replaced with σ donor ligands (L, L’)
2
2
to yield dinitrosyl complexes of the form Fe(NO) L or Fe(NO) L’L [1, 26, 27]. Although many
2
2
2
such complexes are known, it is somewhat surprising that only a handful of mono-phosphine
Fe(NO) (CO)(phosphine) [26-29] and diphosphine Fe(NO) (phosphine) [8, 26, 29-35]
2
2
2
compounds have been reported. In order to determine how the spectroscopic and
electrochemical properties of the Fe(NO) group are affected by small changes on the periphery
2
of some diphosphine complexes, we have prepared and characterized a homologous series of
Fe(NO) (PAr ) complexes (Ar = aryl group; allowing for a systematic variation in Ar) by FTIR
2
3 2
3
1
and P NMR spectroscopy, X-ray crystallography, cyclic voltammetry and fiber-optic IR
spectroelectrochemistry.

ACCEPTED MANUSCRIPT
4
2
. Experimental
2
.1 General
All reactions were performed under an atmosphere of pre-purified nitrogen using
standard Schlenk glassware and/or in a Labmaster 100 inert atmosphere glove box (Innovative
Technology, Inc. Newburyport, MA). Unless otherwise indicated, all experiments were carried
out at ambient room temperature.
2
.2 Chemicals
Solvents were distilled under nitrogen from appropriate drying agents (CaH or Na) or
2
collected under nitrogen from a Pure Solv 400-5-MD Solvent Purification System (Innovative
Technology). Fe(NO) (CO) was prepared using a published method developed by Hieber and
2
2
Beutner, as described by King [36]. Tris(p-methoxyphenyl)phosphine (98%), tri-p-
tolylphosphine (98%), tri-m-tolylphosphine (98%), tri(p-chlorophenyl)phosphine (99%), and
tris(p-fluorophenyl)phosphine (99%), were purchased from Strem Chemical Company
(Newburyport, MA). Iron pentacarbonyl, triphenylphosphine, (99%), tris(p-
trifluoromethylphenyl)phosphine (97%) and tetrabutylammonium hexafluorophosphate (98%)
were purchased from Sigma-Aldrich Chemical Company (Milwaukee, WI). Chloroform-d
(99.8%) was purchased from Cambridge Isotope Laboratories (Andover, MA) in single-use
ampoules and used as received or drawn from a stock reagent bottle and subjected to at least
three freeze-pump-thaw cycles and stored over Grade 514 Type 4A molecular sieves.
2
.3 Instrumentation
Infrared spectra were recorded using a Bio-Rad FTS 155 FT-IR spectrometer. P{ H}
NMR spectra were recorded using a Varian Mercury-VX 300 MHz spectrometer equipped with a
3
1
1

ACCEPTED MANUSCRIPT
5
four-nuclei autoswitchable pulsed field gradient probe. All chemical shifts (δ, ppm) are reported
relative to 85% H PO as an external reference standard (δ = 0 ppm).
3
4
Cyclic voltammograms were recorded using a BAS CV-50W Voltammetric Analyzer
(Bioanalytical Systems, West Lafayette, IN) equipped with a three-electrode cell (3 mm Pt disk
working electrode, Pt wire auxiliary electrode and a Ag/AgCl or Ag wire quasi-reference
electrode). Solutions were 1 mM in analyte and 0.1 M in [NBu ][PF ] in CH Cl . Ferrocene
4
6
2
2
5
5
+
(
(η -C H ) Fe; Fc) or decamethylferrocene ((η -C Me ) Fe) (Fc*; E°' = –0.55 V vs. Fc/Fc ) were
5 5 2 5 5 2
+
used as internal reference standards. All potentials (V) are reported relative to the Fc/Fc couple.
IR spectroelectrochemical measurements were recorded using a Bruker Vector 22 FT-IR
spectrometer equipped with a Remspec mid-IR fiber-optic dip probe and a liquid nitrogen cooled
MCT detector (Remspec Corporation, Charlton City, MA). The stainless steel mirror on the
liquid transmission head of the fiber-optic dip probe was replaced with a 3 mm Pt disk working
electrode and equipped with a custom-made electrochemical cell including a Pt wire auxiliary
electrode and a Ag/AgCl or Ag wire quasi-reference electrode as described previously [37].
2
.4 Synthesis
Each of the dinitrosyl iron diphosphine complexes was prepared using a method based on
slightly modified literature procedures [33, 38]. To the best of our knowledge, only
Fe(NO) (PPh ) (compound 4) [34, 35, 38] from this series has been reported prior to our entry
2
3 2
3
1
in this area. The reactions can be readily monitored by IR spectroscopy and/or P NMR
3
1
spectroscopy. For example, the reactions were considered complete when the P NMR spectra
3
1
indicated the generation of a single P NMR peak assigned to the desired products. At the
conclusion of each preparative reaction, the complexes were isolated by solvent removal in
vacuo and purified by dissolution in CH Cl or CHCl followed by filtration through Celite®.
2
2
3

ACCEPTED MANUSCRIPT
6
Small amounts of pentane or hexane were added to the filtrates, and the products were allowed to
recrystallize by slow solvent evaporation under nitrogen.
Fe(NO) (P(C H -p-OCH ) ) (1): Dark red Fe(NO) (CO) (20 µL, 0.18 mmol) was
2
6
4
3 3 2
2
2
added by syringe to a toluene solution (5 mL) of P(C H -p-OCH ) (0.129 g, 0.37 mmol) in a
6
4
3 3
Schlenk tube. The light red solution was stirred, heated and allowed to reflux under nitrogen
over a period of 3 h. The solution changed from light red to black/dark brown within the first 20
min. The reaction was monitored by IR spectroscopy and stopped when the IR spectrum
indicated the absence of characteristic carbonyl stretching frequencies for Fe(NO) (CO) (ν =
CO
2
2
-1
-1
-1
2
090 cm and 2040 cm ) and for the expected mono-carbonyl species (ν = 2002 cm ).
CO
-1
Isolated yield: 27%. IR (toluene, cm ): ν = 1711 s and 1667 s; also 1306 w, 1287 s, 1255 s,
NO
3
1
1
1
184 w, 1097 w, 826 w, 798 w. P{ H} NMR (CDCl ): δ 56.4 (s).
3
The other complexes were prepared similarly.
Fe(NO) (P(C H -p-CH ) ) (2): Reaction time = 3 h. Isolated yield: 31%. IR (toluene,
2
6
4
3 3 2
-1
31
1
cm ): ν = 1714 s and 1670 s; also 1197 w, 1189 w, 1116 w, 1095 w, 806 s. P{ H} NMR
NO
(
CDCl ): δ 58.4 (s).
3
Fe(NO) (P(C H -m-CH ) ) (3): Reaction time = 5 h. Isolated yield: 23%. IR (toluene,
2
6
4
3 3 2
-1
31
1
cm ): ν = 1715 s and 1671 s; also 779 m, 588 w, 549 w. P{ H} NMR (CDCl ): δ 60.9 (s).
NO
3
Fe(NO) (PPh ) (4): The known Fe(NO) (PPh ) was prepared using a modified
2
3 2
2
3 2
literature method [38]. A toluene solution ( 5 mL) of P(C H ) (94 mg, 0.36 mmol) was treated
6
5 3
with Fe(NO) (CO) ( 20 µL, 0.18 mmol) under nitrogen. The mixture was heated and allowed to
2
2
reflux over a period of ~3 h. The reaction was monitored by IR spectroscopy and stopped once
-1
-1
Fe(NO) (CO) (ν = 2090 cm and 2040 cm ) or the known Fe(NO) (PPh )(CO) (ν = 2005
2
2
CO
2
3
CO

ACCEPTED MANUSCRIPT
7
-1
-1
cm ) were no longer detected. Isolated yield: 22%. IR (toluene, cm ): ν = 1719 s and 1678 s;
NO
31
1
also 1203 m, 1119 m, 543 m. P{ H} NMR (CDCl ): δ 60.9 (s).
3
-
Fe(NO) (P(C H -p-F) ) (5): Reaction time = 3 h. Isolated yield: 23%. IR (toluene, cm
2
6
4
3 2
1
3
)
: νNO = 1720 s and 1682 s; also 1589 w, 1300 w, 1234 s, 1161 m, 1094 w, 1013 w, 828 s.
1
1
P{ H} NMR (CDCl ): δ 59.3 (s).
3
Fe(NO) (P(C H -p-Cl) ) (6): Reaction time = 3.5 h. Isolated yield: 33%. IR (toluene,
2
6
4
3 2
-1
31
1
cm ): ν = 1722 s and 1682 s; also 1099 w, 1012 m 818 m. P{ H} NMR (CDCl ): δ 60.9 (s).
NO
3
Fe(NO) (P(C H -p-CF ) ) (7): Reaction time = 3.5 h. Isolated yield: 33%. IR
2
6
4
3 3 2
-1
(
toluene, cm ): ν = 1728 m and 1687 s; also 1397 w, 1321 m, 1281 w, 1185 w, 1169 m, 1129
NO
31
1
m, 1062 w, 1015 w, 832 m. P{ H} NMR (CDCl ): δ 63.8 (s).
3
2
.5 X-ray Crystallography
Suitable crystals for the structural analyses of compounds 1 and 7 were grown by slow
evaporation of solvent under nitrogen at ambient room temperature. All samples were mounted
on the end of a plastic loop using an inert oil (Paratone N). The samples were cooled to 100(2)
K and maintained at this temperature throughout the duration of the data collection. Intensity
data for each compound were collected using a diffractometer equipped with a Bruker APEX ccd
area detector and graphite-monochromated Mo Kα radiation (λ = 0.71073 Å). The structures
were solved by direct methods using the SHELXTL software package (Version 6.10) and refined
2
by full-matrix least-squares methods on F . All non-hydrogen atoms were refined
anisotropically and all hydrogen atoms were included using idealized parameters.
Crystallographic collection and refinement parameters for the Fe(NO) (PAr ) crystals are given
2
3 2
in Table 1.

ACCEPTED MANUSCRIPT
8
2
.5.1 Fe(NO) (P(C H -p-OCH ) ) (1)
2 6 4 3 3 2
A red plate-shaped crystal of dimensions 0.50 x 0.31 x 0.08 mm was selected for
structural analysis. Cell parameters were determined from a non-linear least squares fit of 6550
peaks in the range 2.34 < θ < 27.91°. A total of 29988 reflections were measured in the range
2
.02 < θ < 26.00° using ω oscillation frames. The data were corrected for absorption by a semi-
empirical method giving minimum and maximum transmission factors of 0.774 and 0.962. The
data were merged to form a set of 7618 independent data with R(int) = 0.0353 and a coverage of
9.6 %. The monoclinic space group P2 /c was determined by systematic absences and
9
1
statistical tests and verified by subsequent refinement.
One of the p-methoxyphenyl groups of one of the phosphines was disordered; restraints
on the positional parameters of the disordered atoms were required. The occupancies of these
disordered groups refined to 0.521(3) and 0.479(3).
2
.5.2 Fe(NO) (P(C H -p-CF ) ) (7)
2 6 4 3 3 2
A red prism-shaped crystal of dimensions 0.33 x 0.25 x 0.15 mm was selected for
structural analysis. Cell parameters were determined from a non-linear least squares fit of 9530
peaks in the range 2.29 < θ < 28.26°. A total of 22582 reflections were measured in the range
1
.61 < θ < 26.00° using ω oscillation frames. The data were corrected for absorption by a semi-
empirical method giving minimum and maximum transmission factors of 0.843 and 0.928. The
data were merged to form a set of 8855 independent data with R(int) = 0.0183 and a coverage of
9
9.1 %. The triclinic space group P 1 was determined by statistical tests and verified by
subsequent refinement. The structure included one n-pentane solvent molecule (not shown).
The n-pentane was best modeled using the SQUEEZE program [39].

ACCEPTED MANUSCRIPT
9
CCDC 959031 (1) and CCDC 959032 (7) contain the supplementary crystallographic
data. These data can be obtained free of charge from The Cambridge Crystallographic Data
Center via http://www.ccdc.cam.ac.uk/conts/ retrieving.html. The crystal structures of 2 [40], 5
[
41], and 6 [42] have been published separately.
Table 1. Crystallographic collection and refinement parameters for the Fe(NO) (PAr ) products.
2
3 2
Complex
1
7•(C H )
5 12
Empirical formula
Formula weight
Crystal system
Space group
C H FeN O P
C H F FeN O P
47 36 18 2 2 2
42
42
2
8
2
820.57
1120.57
Monoclinic
Triclinic
P2 /c
P 1
1
Unit cell dimensions
a (Å)
b (Å)
c (Å)
14.467(4)
10.936(4)
24.848(8)
10.933(2)
13.024(2)
16.851(3)
α
β
γ
(°)
(°)
(°)
90
98.498(8)
90
103.280(5)
101.957(5)
92.653(6)
3
Volume (Å )
Z, Z'
3888(2)
4, 1
2273.5(7)
2, 1
3
Density (calculated) Mg/m
Wavelength (Å)
F(000)
1.402
0.71073
1712
1.637
0.71073
1132
-1
Absorption coefficient (mm )
Max. and min. transmission
Theta range for data collection (°)
Reflections collected
0.527
0.519
0.962 and 0.774
2.02 to 26.00
29988
0.928 and 0.843
1.61 to 26.00
22582
Independent reflections
7618 [R(int) = 0.0353]
7618/141/569
wR2 = 0.1136
R1 = 0.0420
1.038
8855 [R(int) = 0.0183]
8855/0/604
Data/restraints/parameters
2
a
wR(F all data)
wR2 = 0.1031
R1 = 0.0380
1.004
b
R(F obsd data)
2
Goodness-of-fit on F
Observed data [I > 2σ(I)]
6330
8196
Largest and mean shift / s.u.
0.000 and 0.000
0.494 and –0.348
0.001 and 0.000
1.116 and –0.434
3
Largest diff. peak and hole (e/Å )
a
2
2 2
2 2
1/2
b
wR2 = { Σ [w(F – F ) ] / Σ [w(F ) ] }
R1 = Σ ||F | – |F || / Σ |F |
o
c
o
o
c
o

ACCEPTED MANUSCRIPT
1
0
3
. Results
3
.1 Synthesis
The Fe(NO) (PAr ) compounds were prepared in refluxing toluene using a modified
2
3 2
literature procedure as depicted in Figure 1 [33, 38].
1
OCH3
H
2
CH3
H
3
H
CH3
4
H
H
5
F
H
6
Cl
H
7
CF3
H
X1
X2
Figure 1. Reaction scheme depicting the formation of the Fe(NO) (PAr ) products.
2
3 2
The compounds were isolated in 22–27% non-optimized yields as brown (1–3 and 6)
black (4–5), or red-brown (7) solids. Of the compounds in this series, only the parent compound
, and a brief mention of 2 [43], had been reported previously [29, 34, 35, 38], although the
related Fe(NO) (PR ) (PR = phosphine or phosphite) [32, 33, 35] and Fe(NO) (P–P) (P–P =
4
2
3 2
3
2
chelating/bidentate phosphine) [8, 26, 30, 31, 33, 34] compounds are known. The solids appear
to be stable under nitrogen in for more than twelve months as determined by IR spectroscopy
and/or cyclic voltammetry.
3
.2 IR Spectroscopy
The progress of the reactions to prepare the Fe(NO) (PAr ) compounds were monitored
2
3 2
by IR spectroscopy. The solution IR spectrum of the starting compound Fe(NO) (CO) is shown
2
2

ACCEPTED MANUSCRIPT
1
1
in Figure 2a, and exhibits two carbonyl and two nitrosyl signals; each pair of bands correspond
to the symmetric and antisymmetric stretches of the dicarbonyl and dinitrosyl moieties [1].
-1
-1
The carbonyl stretching vibrations (νCO = 2087 cm and 2036 cm ) appear at higher energy than
-1
-1
the nitrosyl stretches (νNO = 1807 cm and 1760 cm ) and the symmetric stretch of each pair
appears at higher energy than the antisymmetric stretch.
Addition of the respective phosphine to a stirred toluene solution of Fe(NO) (CO) at
2
2
room temperature under nitrogen resulted in the replacement of at least one carbonyl group as
indicated by the disappearance of the associated IR bands of the starting compound. For
example, the IR spectrum of the reaction mixture containing Fe(NO) (CO) and P(C H -p-F)
2
2
6
4
3
before heating to full reflux is shown in Figure 2b. The spectrum indicates that a mixture of the
mono and disubstituted products forms prior to reflux. In addition to the bands assigned to the
-1
-1
dinitrosyl group of the product Fe(NO) (P(C H -p-F) ) at 1720 cm and 1682 cm , the band at
2
6
4
3 2
-1
2
009 cm is assigned to the carbonyl group of Fe(NO) (P(C H -p-F) )(CO) intermediate, and the
2 6 4 3
-1
-1
bands at 1763 cm and 1720 cm are assigned to the dinitrosyl group of this intermediate. All
the bands are shifted to lower frequencies than those present in Fe(NO) (CO) , as shown for
2
2
other monosubstitution reactions with other phosphines [27, 28].
The remaining carbonyl group in Fe(NO) (P(C H -p-F) )(CO) was completely replaced
2
6
4
3
upon the application of heat; the IR spectrum for the product, Fe(NO) (P(C H -p-F) ) is shown
2
6
4
3 2
in Figure 2c. As with the replacement of one carbonyl from Fe(NO) (CO) , replacement of the
2
2
carbonyl from Fe(NO) (P(C H -p-F) )(CO) was accompanied with a decrease in the IR
2
6
4
3
stretching frequencies for the dinitrosyl group. The dinitrosyl stretching frequencies, ν s, in
NO
Fe(NO) (P(C H -p-F) ) are lower relative to the ν s present in both Fe(NO) (CO) and
2
6
4
3 2
NO
2
2

ACCEPTED MANUSCRIPT
1
2
Fe(NO) (P(C H -p-F) )(CO). The IR spectral shifts for the reactions between Fe(NO) (CO) and
2
6
4
3
2
2
the other phosphines at room temperature were similar.
-
1
-1
Figure 2. IR spectra of (A) Fe(NO) (CO) in CH Cl ; ν (sym) = 2087 cm , ν (asym) = 2036 cm ; ν (sym) =
1
2
2
2
2
CO
CO
NO
-
1
-1
807 cm , ν (asym) 1760 cm . (B) the reaction mixture during the preparation of Fe(NO) (P(C H -p-F) ) in
N
O
2
6
4
3
2
toluene prior to full reflux, showing the presence of a mixture of the mono-substituted intermediate
-
1
-1
-1
Fe(NO) (P(C H -p-F) )(CO) (ν = 2009 cm , νNO = 1763 cm and 1720 cm ) and the desired product
CO
2
6
4
3
-
1
-1
Fe(NO) (P(C H -p-F) ) (ν = 1720 cm and 1682 cm ), and (C) the reaction mixture during the preparation of
2
6
4
3 2
NO
Fe(NO) (P(C H -p-F) ) in toluene after several hours of reflux.
2
6
4
3 2
In all cases, it was necessary to carry out the reaction at refluxing toluene temperature
~80 °C) to affect complete replacement of the second carbonyl. This observation is consistent
with previous studies on other dinitrosyl iron complexes [18, 27, 38].
(

ACCEPTED MANUSCRIPT
1
3
The IR data for each complex in three different solvents are summarized in Table 2. The
νNOs for compounds containing electron-donating substituents were lower than those for
compounds containing electron-withdrawing substituents. A solvent effect was observed on the
ν s. In general, the ν s obtained in CH Cl were between those determined in toluene and in
NO
NO
2
2
acetonitrile. These results will be discussed later.
3
1
3
.3 P NMR Spectroscopy
3
1
Proton decoupled P NMR spectra were obtained at room temperature in CDCl . The
3
chemical shifts for each compound along with the chemical shifts for the corresponding free
ligands are shown in Table 2. Each compound exhibited a single resonance consistent with two
chemically equivalent phosphorus atoms in the 56–64 ppm range (vs. 85% H PO as an external
3
4
reference standard), and the complexed P signals were shifted ~70 ppm from those of the free
ligands [26, 30, 31]. In general, the compounds with electron-withdrawing substituents had
higher chemical shifts (δ) appearing further upfield from the parent compound (p-H) while, in
contrast, compounds with electron-donating substituents had lower chemical shifts appearing
farther downfield from the parent compound. There seems to be some inconsistencies in this
trend, however, and this will be addressed later.

ACCEPTED MANUSCRIPT
14
3
1
1
a
Table 2. IR and P{ H} NMR spectroscopic data for the Fe(NO) (PAr ) products.
2
3 2
-1
c
νNO (cm )
δ (ppm)
a
b
Complex Substituent
σ
pKa
∆δ
CH CN
CH Cl
toluene
free ligand
–9.6
complex
56.4
3
2
2
1
1
703
660
1704
1659
1711
1667
1
2
3
4
5
6
7
p-OCH
3
–0.27
–0.17
–0.07
0.00
4.57
3.84
3.30
2.73
1.97
1.03
–1.55
66.0
65.8
65.6
65.8
67.9
68.9
1
1
708
665
1708
1663
1714
1670
p-CH
–7.4
–4.7
–4.9
–8.6
–8.0
–5.5
58.4
60.9
60.9
59.3
60.9
63.8
3
1
1
710
667
1711
1664
1715
1671
m-CH
p-H
3
1
1
712
670
1715
1670
1719
1678
1
1
715
673
1719
1674
1720
1682
p-F
0.06
1
1
720
677
1723
1678
1722
1682
p-Cl
0.23
1
1
728
686
1730
1685
1728
1687
p-CF
0.54
69.3
3
a
b
+
+
σ = the Hammett σ or σ substituent parameter. The cited pK corresponds to the acid dissociation process: HPAr3 ꢀ H + PAr . The
p
m
a
3
c
pK data was obtained from the literature [44, 45]. Chemical shifts, δ (ppm), are referenced to 85% H PO set as δ = 0 ppm.
a
3
4

ACCEPTED MANUSCRIPT
1
5
3
.4 X-ray crystallography
Despite the intense research efforts by many research groups into the chemistry and
relevance of DNICs, it is surprising that very little structural data was available on the dinitrosyl
bisphosphine Fe(NO) (PAr ) compounds prior to our work in this area [8, 29, 30]. Suitable
2
3 2
crystals for X-ray diffraction studies were grown by slow evaporation of solvent under nitrogen
at ambient room temperature. Molecular structures for compounds 1 and 7 are shown in Figures
and 4, respectively. Selected bond distances (Å) and angles (°) for 1 and 7 are presented in
Table 3, and compared with our recently published data for compounds 2 [40], 5 [41], and 6
42].
3
[

ACCEPTED MANUSCRIPT
16
Table 3. Selected bond lengths (Å) and angles (°) for the Fe(NO) (PAr ) products.
2
3 2
Complex
Fe–N
N–O
Fe–P
∠ P–Fe–P
∠ N–Fe–N
∠ O••Fe••O
∠ Fe–N–O
1
77.65(19)
177.5(2)
1
1
.655(2)
.654(2)
1.200(3)
1.184(3)
2.254(1)
2.247(1)
1
105.97(3)
106.00(2)
128.49(11)
129.89(7)
124.29(8)
123.8(4)
127.96
130.31
121.67
122.47
125.62
127.15
118.46
179.31(16)
1
1
.655(2)
.659(2)
1.191(2)
1.195(2)
2.260(1)
2.240(1)
2
[40]
175.05(15)
79.10(16)
173.84(15)
78.2(7)
178.2(7)
78.1(2)
1
1
1
.651(2)
.652(2)
1.190(2)
1.187(2)
2.257(1)
2.245(1)
105.57(2)
1
1.190(10)
4
5
6
[29]
[41]
[42]
7
1.650(7)
2.267(2)
111.9(1)
108.27(4)
106.80(3)
105.25(2)
1.190(10)
1
1
1
.657(2)
.653(2)
1.191(3)
1.197(3)
2.242(1)
2.247(1)
127.02(11)
127.78(7)
122.20(9)
177.0(2)
178.76(15)
1
1
.659(2)
.658(2)
1.188(2)
1.186(2)
2.241(1)
2.231(1)
177.67(14)
75.96(18)
173.50(17)
1
1
1
.655(2)
.657(2)
1.184(2)
1.185(2)
2.248(1)
2.250(1)

ACCEPTED MANUSCRIPT
1
7
C1e
C1e’
Figure 3. Molecular structure of 1, Fe(NO) (P(C H -p-OCH ) ) . Hydrogen atoms have been omitted for
2
6
4
3 3 2
clarity. The figure depicts the disorder of one of the phenyl rings of one of the phosphine ligands. The
displacement ellipsoids are drawn at the 50% probability level.
Figure 4. Molecular structure of 7, Fe(NO) (P(C H -p-CF ) ) . Hydrogen atoms have been omitted for
2
6
4
3 3 2
clarity. The displacement ellipsoids are drawn at the 50% probability level.

ACCEPTED MANUSCRIPT
1
8
3
.5 Electrochemistry
The redox behavior of compounds 1–7 were examined in CH Cl using cyclic
2
2
voltammetry. Each compound undergoes a 1-electron oxidation process within the solvent
potential range examined under the experimental conditions described in the Experimental
section. No reduction processes were observed. The cyclic voltammogram for 5,
Fe(NO) (P(C H -p-F) ) , is shown in Figure 5.
2
6
4
3 2
Figure 5. Cyclic voltammogram for a 1 mM solution of Fe(NO) (P(C H -p-F) ) (5) in CH Cl
2
2
6
4
3 2
2
-1
containing 0.1 M NBu PF at a scan rate of 200 mV s .
4
6
The cyclic voltammograms for compounds 1–3 (p-OCH , p-CH , and m-CH ) and 5–6
3
3
3
(
p-F and p-Cl) were similar to that of 4 (p-H) except that the formal electrode potential, E°’ =
(E + E )/2], was shifted more negative for 1–3 and more positive for 5–6 relative to that of 4.
[
pa
pc
The E for 7, Fe(NO) (P(C H -p-CF ) ) , was the most positive of the seven compounds studied
pa
2
6
4
3 3 2
and its cyclic voltammogram was significantly different from those of the other six compounds.
A representative cyclic voltammogram for 7 is shown in Figure 6.
The electrochemical data for each compound is summarized in Table 4. Compounds
containing electron-donating substituents had lower formal electrode potentials, E°’, while

ACCEPTED MANUSCRIPT
1
9
Figure 6. Cyclic voltammogram for a 1 mM solution of Fe(NO) (P(C H -p-CF ) ) (7) in CH Cl
2
2
6
4
3 3 2
2
-1
containing 0.1 M NBu PF at 200 mV s
4
6
.
+
Table 4. Formal electrode potentials (V vs Fc/Fc ) and related electrochemical data for the
-1 a
Fe(NO) (PAr ) products in CH Cl (scan rate = 0.1 V s ).
2
3 2
2
2
b
c
c
d
e
f
g
Complex Substituent
σ
Epa
Epc
∆Ep
E°’
i /i
pc pa
∆Ep(ref)
1
2
3
4
5
6
7
p-OCH
3
–0.27
–0.17
–0.07
0.00
–0.15 –0.27 0.12 –0.21
–0.12 –0.26 0.14 –0.19
–0.11 –0.21 0.10 –0.16
–0.08 –0.17 0.09 –0.13
1.0
0.97
0.89
1.0
1.0
1.0
--
0.09
0.11
0.08
0.08
0.09
0.06
0.06
p-CH
3
m-CH
p-H
3
p-F
0.06
0.10
0.14
0.00
0.06
--
0.10
0.08
--
0.05
0.10
--
p-Cl
0.23
p-CF
0.54
0.51
3
a
+
b
c
All potentials are referenced to the Fc/Fc couple. σ = the Hammett substituent parameter. Epa and Epc = the
d
e
anodic and cathodic peak potentials, respectively. ∆E = | E – E |. E°’ is the formal electrode potential = (E
+
p
pa
pc
pa
Epc)/2. f i /i represents the ratio of peak currents defined by (i ) /i + (0.485(i ) )/i + 0.086.
g
∆Ep(ref) = ∆Ep
pc pa
pc 0 pa
sp 0 pa
of the internal reference standard;

ACCEPTED MANUSCRIPT
2
0
those with electron-withdrawing substituents had higher formal electrode potentials. The cyclic
voltammetry results obtained for 1 and 7 will now be described as representative examples and
compared to the other compounds.
The cyclic voltammogram of Fe(NO) (P(C H -p-OCH ) ) (1) in CH Cl exhibits a
2
6
4
3 3 2
2
2
+
-1
reversible one-electron oxidation at E°’ = –0.21 V vs. Fc/Fc at a scan rate (ν) of 100 mV s .
The separation in peak potentials (∆E ) is 120 mV; this ∆E increases slightly with increasing
p
p
-1
-1
scan rate, being in one determination 98 mV at 0.10 V s and 134 mV at 0.50 V s . Under the
-1
same conditions, the ferrocene internal reference standard has a ∆E of 76 mV at 0.10 V s and
p
-1
9
7 mV at 0.50 V s . The anodic to cathodic peak current ratio (i /i ) is 1.0, and remained
pc pa
-1
1/2
constant over the 0.05–1.2 V s range. The i value varies linearly with ν over the entire scan
pa
rate range employed indicative of a diffusion controlled electrooxidation. Lastly, in the case of
Fe(NO) (P(C H -p-OCH ) ) , there was some evidence for a second oxidation when the scan
2
6
4
3 3 2
was extended to the solvent limit; however, this feature could not be fully characterized using
our solvent system.
As mentioned previously the cyclic voltammogram of compound 7 is significantly
different from the other six compounds. The cyclic voltammogram of 7 in CH Cl exhibits a
2
2
-1
chemically irreversible oxidation. At a scan rate of 200 mV s the anodic peak potential occurs
+
at E = 0.51 V vs. Fc/Fc (feature a in Figure 6). A small cathodic peak occurs at E = –0.04 V
pa
pc
+
vs. Fc/Fc . This peak (feature b in Figure 6) appears larger if the scan is taken farther in the
1
/2
anodic direction (i.e. closer to the solvent limit), and the i value varies linearly with ν over
pa
the entire scan rate range employed.

ACCEPTED MANUSCRIPT
2
1
2
.3.6 Infrared Spectroelectrochemistry
The IR spectroelectrochemistry of each compound (1–7) was examined at room
temperature in CH Cl . The applied potential E was typically set at 50–120 mV more positive
2
2
appl
than E , and the data obtained are summarized in Table 5. Each of the electrogenerated
pa
compounds exhibited symmetric and antisymmetric nitrosyl stretching frequencies that were
-1
~
100 cm higher than the ν s observed in the neutral compounds. Not surprisingly, the ν s
NO NO
for compounds containing electron-donating substituents were lower than those for compounds
containing electron-withdrawing substituents.
a
Table 5. IR spectroelectrochemical data obtained for the Fe(NO) (PAr ) products in CH Cl .
2
3 2
2
2
-1
νNO (cm )
Complex
Eappl
Neutral
Oxidized
∆νNO
1
1
704
659
1805
1758
101
99
1
2
3
4
5
6
7
–0.05
–0.06
0.01
0.04
0.17
0.26
0.62
1
1
708
663
1810
1763
102
100
1
1
711
664
1812
1764
101
100
1
1
715
670
1816
1767
101
97
1
1
719
674
1820
1770
101
96
1
1
723
678
1823
1772
100
94
1
1
730
685
1827
1759,1778
97
74
a
+
Eappl is the applied potential (V) relative to the ferrocene-ferrocenium couple, Fc/Fc . Eappl was typically set at
0
.05–0.12 V more positive than E . ∆νNO is the difference between νNO for the singly oxidized compound and the
pa
corresponding νNO for the neutral compound.
A representative IR difference spectrum recorded during controlled potential oxidation of
Fe(NO) (P(C H -p-F) ) is shown at the top of Figure 7. The ν s for the neutral starting
2
6
4
3 2
NO

ACCEPTED MANUSCRIPT
2
2
-1
-1
compound are shown as disappearing at 1674 cm and 1719 cm . The ν s for the oxidized
NO
-1
-1
species appear at 1770 cm and 1820 cm which suggests the formation of a cationic Fe(NO)
2
+
containing species such as [Fe(NO) (P(C H -p-F) ]
2
6
4
2
.
Figure 7. (A) IR difference spectrum recorded during the oxidation of 5, Fe(NO) (P(C H -p-F) ) , in
2
6
4
3 2
+
CH Cl containing 0.1 M [NBu ][PF ] at E = 0.17 V vs. Fc/Fc . ν s for the neutral compound are
shown growing downward at 1674 cm and 1719 cm while ν s for the oxidized species appear
growing upward at 1770 cm and 1820 cm . (B) IR difference spectrum recorded during the oxidation of
2
2
4
6
appl
NO
-
1
-1
NO
-1
-1
+
7
, Fe(NO) (P(C H -p-CF ) ) , in CH Cl containing 0.1 M [NBu ][PF ] at E = 0.62 V vs. Fc/Fc . ν s
2 6 4 3 3 2 2 2 4 6 appl NO
-
1
-1
for the neutral compound are shown growing downward at 1685 cm and 1730 cm while evidence of the
ν s for the oxidized species appear growing upward at 1759 cm , 1778 cm and 1827 cm .
-1
-1
-1
NO
The difference spectra for the other compounds studied were all similar to that obtained
for 5 except for the spectrum recorded during the oxidation of 7 which is depicted at the bottom
of Figure 7. The peaks for the oxidized species in the spectrum for 7 are not as clearly defined as
those for the other compounds, which suggests that the formation of the presumed cationic

ACCEPTED MANUSCRIPT
2
3
Fe(NO) containing species may convert to other species under our spectroelectrochemical
2
conditions.
4
. Discussion
4
.1 Synthesis
The starting compound, Fe(NO) (CO) , is a toxic, air sensitive, deep red liquid at room
2
2
temperature (m.pt. 18 °C) [36]. Previous preparations of Fe(NO) L (L = σ donor ligand)
2
2
compounds have involved carrying out the reaction directly with Fe(NO) (CO) or from
2
2
[
Fe(NO) (µ-X)] (X = Cl, Br, I) under reducing conditions [26, 31, 33, 34, 38, 46, 47]. These
2 2
types of compounds have also been prepared by a reductive substitution reaction involving
_
[
(RS) Fe(NO) ] anions with excess PPh or by reaction of phosphines with Fe(NO) (PR )X [35,
2 2 3 2 3
4
8]. Bitterwolf published a one-pot synthesis of Fe(NO) (PAr ) type compounds based on the
2 3 2
reductive substitution of Na [Fe (NO) (µ-S O ) ] in the presence of phosphines and phosphites
2
2
4
2
3 2
[
32].
The current preparation of Fe(NO) (PAr ) type compounds is based on a modified
2
3 2
literature procedure involving the reaction of Fe(NO) (CO) with substituted triarylphosphines in
2
2
toluene [33, 38]. Toluene was selected because it had a boiling point high enough to allow
displacement of the second carbonyl ligand under refluxing conditions.

ACCEPTED MANUSCRIPT
2
4
4
.2 Infrared Spectroscopy
It is known that the substituents on the phenyl rings of triarylphosphines affect their
basicities [44, 49-51]. The position of the dinitrosyl stretching frequency measured upon
coordination with a phosphine ligand is related to the electron donor capacity or the Lewis
basicity of the ligand and to the donor or acceptor capacity of the solvent in which the
measurement is taken [52-55]. Several methods have been used to quantify this type of
substituent effect. One of the most widely applied and useful tools is the Hammett substituent
constant or Hammett parameter σ [56].
Hammett parameters are linear free energy relationships that can be used as a measure of
the electronic effects of ligands [57, 58]. Hammett constants for each substituent used in this
work and the corresponding pK s of the protonated phosphine ligands are given in Table 2; the
a
Hammett parameter, σ, correlates linearly with the pK s of the phosphines used in this study [44,
a
4
5]. The order of ligand basicity is the same as the order of the Hammett σ and σ parameters
p m
(i.e., Hammett constants for substituents in para and meta positions). We thus set out to
determine if a correlation exists between σ and the spectroscopic properties of the
Fe(NO) (PAr ) products.
2
3 2
A plot showing the IR stretching frequencies for each substituted compound in three
different solvents (toluene, CH Cl and acetonitrile) as a function of Hammett parameter is
2
2
shown in Figure 8. Separate plots of the IR stretching frequencies for the dinitrosyl groups as a
function of Hammett parameter in the same three solvents are shown in Figure 9.
In general, compounds with the more basic phosphines have lower ν s while
NO
compounds with the less basic phosphines have higher ν s. The effect of solvent appears to be
NO
greater on the symmetric ν s of compounds with electron-donating substituents. The ν s of
NO
NO

ACCEPTED MANUSCRIPT
2
5
the compound with the most electron-withdrawing substituent, 7 (p-CF ) appear to not be as
3
affected by solvent differences as those with more electron donating substituents. For most of
the series, frequencies obtained in the non-polar and non-coordinating solvent toluene were
higher than those obtained in CH Cl (a polar non-coordinating solvent) or in acetonitrile (a polar
2
2
coordinating solvent). The ν s obtained in CH Cl were generally in-between those obtained in
NO
2
2
toluene and in acetonitrile although they tended to be closer to those obtained in acetonitrile as
can be seen in Figure 8. The effect of solvent on νNO appears to be greater on the compounds
with the more basic ligands.
Figure 8. Plot depicting the symmetric and antisymmetric ν s in toluene, CH Cl and
NO
2
2
acetonitrile versus the Hammett substituent parameter.
In Figures 8 and 9 it can be observed that electron-withdrawing substituents on the
-1
triarylphosphines, PAr , shifted the ν s to higher frequencies (by 9–16 cm ) relative to the
3
NO
parent compound Fe(NO) (PPh ) (i.e. p-H). In contrast, electron-donating substituents shifted
2
3 2
-1
ν s to lower frequencies. (by 8–11 cm ). The largest shifts in ν were observed in the polar
NO
NO
CH Cl and the polar coordinating solvent acetonitrile.
2
2

ACCEPTED MANUSCRIPT
2
6
The shifts in νNO as a function of Hammett parameter can be rationalized on the basis of
the ligand basicity and the concept of backbonding. The more basic ligands (those with electron-
donating substituents) contribute more electron density to the iron center and thus increase
backbonding from the filled d-orbitals on the metal to the antibonding π* orbitals on the nitrosyls
resulting in relatively lower ν s. Conversely, the less basic ligands (those with electron-
NO
withdrawing substituents) contribute less electron density to the iron center and, as a result, the
metal to ligand backbonding is less as indicated by relatively higher ν s.
NO
While the plots for toluene and CH Cl fit linearly with the Hammett parameter, a slight
2
2
bending can be detected near the point where the substituents change from electron-withdrawing
to electron-donating. It is possible for the data to be fitted to two separate lines at this juncture as
shown in Figures 9A and B. This observation tends to support having different modes of
interaction between the different solvent types and the two classes of Fe(NO) (PAr )
2
3 2
compounds. In addition, the two lines may also be explained by the Fe(NO) moiety in the two
2
classes adopting two slightly different conformations (or extent of attracto distortions) in
different solvent types. Toluene and CH Cl are both non-coordinating solvents and the plots of
2
2
νNO versus σ for both of these solvents possess the two-line feature. Compound 3,
Fe(NO) (P(C H -m-CH ) ) deviates from the lines in CH Cl and in toluene more than any of
2
6
4
3 3 2
2
2
the para substituted compounds. In contrast, the plot of νNO versus σ for acetonitrile is quite
linear and cannot be fitted to two separate lines. This suggests that the coordinating solvent
interacts in a way not present in toluene or CH Cl .
2
2

ACCEPTED MANUSCRIPT
2
7
Figure 9. Plots depicting the symmetric and antisymmetric ν s for Fe(NO) (PAr ) compounds in (A)
NO
2
3 2
toluene, (B) CH Cl and (C) acetonitrile vs. the Hammett substituent parameter.
2
2
If the oxygen atom on the dinitrosyl complexes acts as a donor to the solvent, there
should be a relationship between the IR stretching frequency and the solvent acceptor number