Synthesis and characterization of diiron dithiolate complexes containing a quinoxaline bridge

Article abstract of DOI:10.1039/c1dt10953g

A potential model complex for the hydrogenase active site, [Fe ₂{(μ-CH₂S)₂R}(CO)₆] (1) (R = quinoxaline), was synthesized by condensation of [(μ-LiS)₂Fe ₂(CO)₆] with 2,3-bis(brom

Full text of DOI:10.1039/c1dt10953g

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2011, 40, 10907
www.rsc.org/dalton
PAPER
Synthesis and characterization of diiron dithiolate complexes containing a
quinoxaline bridge†
Fenfen Xu,a,c Shaowu Du,*a Yuping Liu,b Jacqueline Hassan,b Jie Zhang*b and Alan M. Bond*b
Received 21st May 2011, Accepted 15th July 2011
DOI: 10.1039/c1dt10953g
A potential model complex for the hydrogenase active site, [Fe2{(m-CH2S)2R}(CO)6] (1) (R =
quinoxaline), was synthesized by condensation of [(m-LiS)2Fe2(CO)6] with 2,3-bis(bromomethyl)-
quinoxaline. Reactions of 1 with bis(diphenylphosphino)methane (dppm) under a range of conditions
yielded substituted complexes [Fe2{(m-CH2S)2R}(CO)5(dppm)] (2), [Fe2{(m-CH2S)2R}(CO)4(k2-dppm)]
(3) and [Fe2{(m-CH2S)2R}(CO)4(m-dppm)] (4). X-ray crystallography confirms that in 2, the dppm is
terminally bonded to an iron atom via one phosphorus atom, whereas in 3, it acts as a chelating ligand
to coordinate to an iron center in a dibasal-substituted manner. In 4, the dppm bridges the two iron
atoms in a cis basal/basal fashion with one phosphorus bonded to each iron atom. Treatment of 1 with
various tertiary phosphines at room temperature in acetonitrile (MeCN) generates a range of
mono-substituted products [Fe2{(m-CH2S)2R}(CO)5L] (5, L = PEt3; 6, PMe3; 7, PPh3; 8, Me2PPh). With
ButNC, mono- and di-substituted [Fe2{(m-CH2S)2R}(CO)5(ButNC)] (9) and [Fe2{(m-CH2S)2R}(CO)4-
(ButNC)2] (10) complexes are generated. All the complexes were characterized by elemental analysis,
IR, MS and NMR spectroscopy. IR and NMR spectroscopic studies suggest that addition of excess
HBF4·OEt2 acid to 1–4 led to the protonation of quinoxaline nitrogen atoms. In contrast, 5–10 were not
stable in acidic media. Electrochemistry of 1–4 was investigated in the acetonitrile medium (0.1 M
Bu4NPF6). The electrochemical instability of the reduced ligand, quinoxaline, and the reduced forms of
these complexes revealed from the electrochemical studies suggests that they do not provide ideal
models of the hydrogenase active site.
2Fe2S unit.7–9 In order to mimic the behavior of metal–sulfur sites
Introduction
in hydrogenase, extensive research has been devoted to studies
Hydrogenases are a class of natural enzymes that can catalyse
concerning synthetic models of the [FeFe]-hydrogenases by either
reversible and rapid hydrogen oxidation and evolution in many
microorganisms.1–4,7 Among them, [FeFe]-hydrogenases have been
under intense investigations because of the unusual structures
CO substitution or bridge variation based on the well-known
diiron complex [Fe2(m-pdt)(CO)6] (pdt = propane-1,3-dithiolato)
with low overpotential and high efficiency towards hydrogen
evolution reaction.10–19 In contrast to the CO substitution, the
variations of the S-to-S bridge linkers of the model complexes
have received less attention. In the past few years, much effort
has been focused on the functionalized azadithiolate Fe-only
hydrogenases, [Fe2{m-(SCH2)2NR}(CO)6], where the R groups
range from alkyl,16,20–25 aryl,26–31 or heterocyclic substituents.32–35
The N atoms in these model complexes can be protonated, leading
to the reduction of the electron-donating capacity of the ligand,
which may subsequently affect the electron density at the appended
diiron site. Such effects upon protonation on N atoms may on
one hand lead to positive shift in the reversible potentials of iron
centers, thus lower the overpotential for hydrogen evolution, due
to the introduction of positive charge, and on the other hand,
provide a low energy pathway for molecular hydrogen production
via proton–hydride combination.28,32,36
of their active sites and their extremely high capability for the
production of “clean” and highly efficient hydrogen fuel.4–6 X-
Ray crystallographic studies on [FeFe]-hydrogenases isolated from
Clostridium pasteurianum and Desulfovibrio desulfuricans have
revealed that the active site is composed of a 2Fe2S butterfly struc-
ture (H-cluster), which is linked to an Fe4S4 cubane-like subcluster
by a cysteine-S residue, and a three-atom linker (–CH2XCH2–,
+
X = CH2, NH or NH2 ) bridged between the two S atoms of the
aState Key Laboratory of Structural Chemistry, Fujian Institute of Research
on the Structure of Matter, Chinese Academy of Sciences, Fuzhou, Fujian,
350002, P. R. China. E-mail: swdu@fjirsm.ac.cn; Fax: +86 591 83709470;
Tel: +86 591 83709470
bSchool of Chemistry, Monash University, Clayton, Vic, 3800, Aus-
tralia. E-mail: Jie.zhang@monash.edu, alan.bond@monash.edu; Fax: +61
3 99054597; Tel: +61 3 99056289,+61 3 99051338
cGraduate University of Chinese Academy of Sciences, Beijing, 100039, P.
R. China
The overpotential of hydrogen evolution reaction also can
be reduced by the introduction of unsaturated bridging motifs
with electron-withdrawing character to the model complexes,
† CCDC reference numbers 816112–816119. For crystallographic data in
CIF or other electronic format see DOI: 10.1039/c1dt10953g
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 10907–10917 | 10907
©

View Article Online
[Fe2(m-S2R)(CO)6] (S2R
=
hydroquinone-2,3-dithiolate, 3,6-
followed by addition of 2,3-bis(bromomethy)quinoxaline (Scheme
1). Column chromatography using petroleum ether and ethyl
acetate (v/v: 6/1) as the eluents gave 1 as the major product.
The infrared spectrum of 1 exhibits three absorption bands
between 2079–2003 cm-1 that can be attributed to terminal
carbonyl ligands. The 1H NMR spectrum of 1 shows two
doublets at 3.40 and 4.29 ppm (JHH = 12.8 Hz) that can be
assigned to the resonances of two methylene groups in the
bismethylenequinoxaline linker. The inequivalence of the two
protons on each methylene group could be due to the presence
of bulky quinoxaline moiety which hinders the interconversion
of the FeS(1)C(7)C(9)C(10)C(8)S(2) dithiacycloheptane ring (Fig.
1(a)), as observed in [Fe2{m-(SCH2)2CHCOOH}(CO)6], in which
the dithiacyclohexane ring carries a carboxylic substituent.40
The substitution of 1 with dppm in the presence of Me3NO
afforded different products depending on the reaction conditions
employed (Scheme 2). When the reaction was carried out in
acetonitrile (MeCN) at room temperature, complex 2, where
the ligand is monodentate, was isolated as the only product.
Extension of the reaction time or addition of more CO scavenger
did not give other products. However, when the reaction was
performed in refluxing toluene, in addition to a small amount
of 2, two other products were both obtained in moderate
dichloro-1,2-benzenedithiolate, quinoxaline-2,3-dithiolate and
1,2-closo-C2B10H10-1,2-dithiolate).37–39 Among these model com-
plexes, [Fe2(m-quinoxaline-2,3-dithiolate)(CO)6] briefly reported
by Ott and co-workers38 is of particular interest since the N sites
of quinoxaline may be protonated leading to the advantages asso-
ciated with the [Fe2{m-(SCH2)2NR}(CO)6] complexes mentioned
above.16,20,22–35
Herein, we report the synthesis, and the spectroscopic and
crystallographic characterization of a new diiron complex [Fe2{(m-
CH2S)2R}(CO)6] (1) (R = quinoxaline) and its derivatives which
contain the rigid, conjugated, N atom containing an electron with-
drawing quinoxaline group. We also describe the electrochemical
studies employed to assess whether the new compounds could
mimic the behaviors of hydrogenase.
Results and discussion
Synthesis and characterization of the model compound and its
derivatives
The parent complex 1 was prepared by treatment of [(m-
S2)Fe2(CO)6] with 2 equiv. of LiBEt3H in THF at –78 ◦C,
Fig. 1 Molecular structures of 1 (a), 2 (b), 3 (c) and 4 (d) with thermal ellipsoid set at 50% probability. The hydrogen atoms are omitted for clarity. The
labels for C atoms of benzene rings are not shown for clarity.
10908 | Dalton Trans., 2011, 40, 10907–10917
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
NMR spectrum. A similar situation was also observed for 4 whose
31P NMR spectrum exhibits two very closely spaced doublets at
d = 52.42 and 54.08 ppm, indicating the existence of two slightly
inequivalent P atoms.
Mono-substitution reactions of 1 with tertiary phosphines PR3
were carried out according to literature procedures (Scheme 3).
These reactions can be controlled at the mono-substituted stage at
room temperature in MeCN with the aid of Me3NO, resulting in
the formation of a series of complexes [Fe2{(m-CH2S)2R}(CO)5L]
(5, L = PEt3; 6, PMe3; 7, PPh3; 8, PMe2Ph). Similar to the dppm
derivatives (complexes 2–4), CO bands of complexes 5–8 in the IR
spectra also shift to lower energy (2051–1926 cm-1) compared to
those of 1, displaying the increase of electron density at the metal
centers which again suggests that complexes 5–8 may be harder
to reduce, but have higher proton affinity. However, variation in
phosphine ligands appears to have a smaller influence on the shift
Scheme 1
1
of the n(CO) bands of 5–8. Their H NMR spectra show the
methylene protons as two doublets or singlets in the region of 3.35
to 4.46 ppm, thus again revealing the inequivalence of the two
protons in each methylene group. As expected, their 31P NMR
spectra feature two resonance signals that are very close to each
other, and are similar to those found in 2–4.
Scheme 2
yields. They were characterized as the dppm-chelated complex
[Fe2{(m-CH2S)2R}(CO)4(k2-dppm)] (3) and dppm-bridged com-
plex [Fe2{(m-CH2S)2R}(CO)4(m-dppm)] (4). Compounds 3 and 4
can also be prepared by heating 2 in toluene. This indicates that 2
may be the intermediate for the formation of 3 and 4. Complexes
2–4 each show three infrared bands in the n(CO) stretching region
(1916–2052 cm-1). These substantial red-shifts of the n(CO) bands
compared to those of parent complexes reflect the increase of
electron density of Fe atoms in complexes 2–4 which implies that
2–4 may be harder to reduce than complex 1, but have a higher
affinity to proton. In the 31P NMR spectrum of 2, a doublet for
the coordinated P is found at 55.83 ppm, and a doublet for the
uncoordinated phosphorus is found at -26.43 ppm. Complex 3 has
diagnostic 31P NMR signals at 8.44 and 12.43 ppm, and is different
from [Fe2(m-pdt)(CO)4(k2-dppm)] reported by Adam et al.41 where
only one 31P NMR signal was found for the chelated dppm ligand
because of the fast flipping of the central methylene at room
temperature which makes the two P atoms equivalent. The slight
inequivalence of the two phosphorus atoms in 3 is most likely due
to the fixed position of quinoxaline unit rather than the existence of
apical-basal and dibasal isomers that normally display a difference
in 31P NMR chemical shifts about 10 ppm to each other.42 This
is in accordance with the observation of two resonances at 3.48
Scheme 3
Crystal structures of complexes 1–5, 7, 9 and 10
The molecular structures of 1–5, 7, 9 and 10 were established by
X-ray crystallography. The molecular structures of 6 and 8 could
not be obtained since attempts to obtain single crystals of these
complexes were not successful. The ORTEP plots of complexes
whose X-Ray structures could be obtained are presented in Fig.
1 and 2 and selected bond lengths and angles are tabulated in
Table 1. As shown in Fig. 1 and 2, all the complexes contain a
butterfly 2Fe2S core and a quinoxaline-2, 3-diyldimethanethiolate
tether, which acts as a bidentate bridging ligand across the metal–
metal bond and donates six electrons to make these complexes
a 34 electron valance saturated bimetallic compound. Each iron
center in all these complexes displays approximately square-
pyramidal coordination geometry. The average Fe–Fe distance
1
and 4.24 ppm associated with the methylene protons in the H
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 10907–10917 | 10909
©

View Article Online
˚
Table 1 Selected bond lengths (A) and angles (degrees) for 1–5, 7, 9 and
Table 1 (Contd.)
10
S(1)–Fe(1)–S(2)
S(1)–Fe(2)–S(2)
C(15)–Fe(1)–S(1)
C(15)–Fe(1)–S(2)
Fe(1)–S(1)–Fe(2)
87.67(4)
87.42(3)
154.60(11)
83.09(11)
66.88(3)
Fe(1)–S(2)–Fe(2)
C(20)–Fe(2)–S(2)
C(20)–Fe(2)–S(1)
C(15)–N(3)–C(16)
C(20)–N(4)–C(21)
66.87(3)
1
100.64(11)
101.91(10)
177.0(3)
Fe(1)–S(1)
Fe(1)–S(2)
Fe(1)–C(1)
Fe(1)–C(2)
Fe(1)–C(3)
S(1)–Fe(1)–S(2)
S(1)–Fe(2)–S(2)
2
2.2587(11)
2.2598(11)
1.804(4)
1.798(4)
1.799(4)
88.02(4)
88.05(4)
Fe(2)–S(1)
Fe(2)–S(2)
Fe(2)–C(4)
Fe(2)–C(5)
Fe(2)–C(6)
Fe(2)–S(1)–Fe(1)
Fe(1)–S(2)–Fe(2)
2.2569(12)
2.2604(12)
1.798(4)
1.804(4)
1.797(4)
67.88(3)
67.80(3)
177.7(4)
˚
of these complexes (2.527 A) is similar to the values reported for
diiron dithiolate model complexes but slightly longer than those
bearing rigid S-to-S linkers, such as [Fe2(m-bdt)(CO)6] (2.480(2)
Fe(1)–S(1)
Fe(1)–S(2)
Fe(2)–S(1)
Fe(1)–C(2)
Fe(2)–C(3)
S(1)–Fe(2)–S(2)
Fe(2)–S(2)–Fe(1)
3
Fe(1)–S(1)
Fe(1)–S(2)
Fe(2)–S(1)
Fe(1)–C(1)
Fe(2)–C(2)
S(1)–Fe(1)–S(2)
S(1)–Fe(2)–S(2)
P(1)–Fe(1)–P(2)
P(1)–Fe(1)–S(1)
P(1)–Fe(1)–S(2)
4
Fe(1)–S(1)
Fe(1)–S(2)
Fe(2)–S(1)
Fe(1)–C(1)
Fe(1)–C(2)
S(1)–Fe(1)–S(2)
S(1)–Fe(2)–S(2)
5
Fe(1)–S(1)
Fe(1)–S(2)
Fe(2)–S(1)
Fe(1)–C(2)
Fe(2)–C(4)
S(1)–Fe(1)–S(2)
S(2)–Fe(2)–S(1)
P–Fe(1)–S(1)
7
Fe(1)–S(1)
Fe(1)–S(2)
Fe(2)–S(1)
Fe(1)–C(2)
Fe(2)–C(4)
S(1)–Fe(1)–S(2)
S(2)–Fe(1)–S(1)
P–Fe(1)–S(1)
9
Fe(1)–S(1)
Fe(1)–S(2)
Fe(2)–S(1)
Fe(1)–C(1)
Fe(1)–C(2)
Fe(2)–C(5)
S(1)–Fe(1)–S(2)
S(1)–Fe(2)–S(2)
Fe(1)–S(1)–Fe(2)
10
2.2668(8)
2.2706(8)
2.2819(8)
1.771(3)
1.786(4)
86.57(3)
67.12(3)
Fe(2)–S(2)
Fe(1)–P(1)
Fe(1)–C(1)
Fe(2)–C(4)
Fe(2)–C(5)
S(1)–Fe(1)–S(2)
Fe(1)–S(1)–Fe(2)
2.2619(8)
2.2284(8)
1.779(3)
1.789(4)
1.810(4)
86.72(3)
66.85(2)
43
37
˚
˚
A), [Fe2{m-S2C6H2(OH)2}(CO)6} (2.4815(4) A)
and [Fe2(m-
44
˚
Mebdt)(CO)6] (2.4754(14) A)
(bdt = benzene-1,2-dithiolato;
Mebdt = 3,4-toluenedithiolate). Particularly noteworthy is that
the Fe–Fe distance of 3 (2.5851(17) A) is very close to those
found in the structures of Fe-only hydrogenases (ca. 2.62 A). In
the large number of previously reported 2Fe2S model complexes,
˚
3
˚
2.2373(16)
2.2243(14)
2.2658(16)
1.751(5)
1.780(5)
88.90(6)
87.05(6)
74.57(6)
93.41(6)
165.18(5)
Fe(2)–S(2)
Fe(1)–P(1)
Fe(1)–P(2)
Fe(2)–C(3)
2.2712(18)
2.2142(14)
2.2121(17)
1.801(5)
1.810(5)
152.33(5)
96.87(6)
˚
few have Fe–Fe bond lengths larger than 2.58 A. Examples
reported so far include [Fe2{(m-o-xyldt)}(CO)4(PMe3)2] (2.5825
A)42 and [Fe2{(m-SCH2)2NBut}(CO)5(IMes)] (2.5860 A)45 (o-xyldt =
SCH2C6H4CH2S; IMes = 1,3-bis(mesityl)imidazole-2-ylidene). The
˚
˚
Fe(2)–C(4)
P(2)–Fe(1)–S(1)
P(2)–Fe(1)–S(2)
Fe(1)–S(1)–Fe(2)
Fe(1)–S(2)–Fe(2)
˚
mean bridging Fe–S distances are 2.2243(14)–2.2848(7) A and the
70.07(5)
70.19(5)
average Fe–S–Fe and S–Fe–S angles lie in the narrow range of
66.85(2)–70.19(5)◦ and 81.41(6)–88.90(6)◦, respectively. The small
variations in bond lengths and angles in this series of complexes
indicate that the bonding characteristics of the 2Fe2S skeleton of
parent complex 1 are slightly affected upon substitution of CO
by other donor ligands. The average Fe–C(CO) and C–O bond
2.2675(18)
2.252(2)
2.252(2)
1.787(7)
1.774(6)
86.84(7)
87.49(6)
Fe(2)–S(2)
Fe(1)–P(1)
Fe(2)–P(2)
Fe(2)–C(3)
Fe(2)–C(4)
Fe(1)–S(1)–Fe(2)
Fe(1)–S(2)–Fe(2)
2.2416(18)
2.2025(19)
2.2311(17)
1.764(6)
1.774(7)
67.43(5)
˚
lengths (1.787 and 1.145 A, respectively) and Fe–C–O angles
(178◦) are as expected for these types of complexes. For each
mono-substituted complex, the distances between the substituted
iron atom and its attached CO-carbon atoms are shorter than
those around the unsubstituted iron atom and reveals the increased
strength of back-bonding from Fe to CO caused by coordination
of the strong s-donor ligand. For example, in 2, 5 and 7, the average
Fe–C distances of phosphine-substituted Fe atoms is 1.775, 1.773
67.86(5)
2.2652(15)
2.2426(14)
2.2709(16)
1.781(5)
1.795(5)
87.92(6)
Fe(2)–S(2)
Fe(1)–P
Fe(1)–C(1)
Fe(2)–C(3)
2.2591(16)
2.2436(16)
1.765(5)
1.796(5)
1.800(5)
90.28(6)
68.56(5)
69.15(4)
Fe(2)–C(5)
P–Fe(1)–S(2)
Fe(1)–S(1)–Fe(2)
Fe(1)–S(2)–Fe(2)
87.39(5)
168.08(5)
˚
˚
and 1.775 A, which are shorter by ca. 0.02 A respectively than those
of their corresponding unsubstituted Fe atoms. This Fe–C bond
shortening becomes particularly significant in the dppm chelated
complex 3, where the mean Fe–C bond length of substituted Fe is
2.2805(7)
2.2735(7)
2.2679(7)
1.773(3)
1.785(3)
86.52(2)
86.52(2)
108.47(3)
Fe(2)–S(2)
Fe(1)–P
Fe(1)–C(1)
Fe(2)–C(3)
2.2847(7)
2.2441(7)
1.777(3)
1.786(3)
1.812(3)
107.38(3)
67.13(2)
66.96(2)
˚
0.05 A shorter than that of the unsubstituted one.
Fe(2)–C(5)
In contrast to 2, where one of the phosphorus atoms of dppm
is coordinated to the Fe(1) center and the other is uncoordinated,
in 3, both the phosphorus atoms occupy basal sites and chelate
to the Fe(1) atom. While there are a number of examples of
dppm acting in a chelating manner bridging two iron atoms in
hydrogenase model complexes, this dibasal geometry of dppm
seems uncommon as this adoption requires a significant decrease
in normal dibasal angle for iron center and P–C–P angle for
dppm. The only crystallographically characterized example of this
dibasal type model complex is in [Fe2(m-pdt)(CO)4(k2-dppm)].41
The bite angle P(1)–Fe(1)–P(2) of 3 is 74.57(6)◦, which is very
small but comparable with the value found in [Fe2(m-pdt)(CO)4(k2-
dppm)]. For complex 4, the dppm bridges the two iron atoms with
one P atom bonded to each iron atom and is cis to Fe–Fe bond.
Unlike in 3 where the two Fe–P bond lengths are essentially equal,
P–Fe(1)–S(2)
Fe(1)–S(1)–Fe(2)
Fe(1)–S(2)–Fe(2)
2.2447(13)
2.2632(13)
2.2679(15)
1.793(4)
1.773(4)
1.808(4)
88.24(5)
87.64(4)
68.02(4)
Fe(2)–S(2)
Fe(1)–C(16)
N(3)–C(16)
Fe(2)–C(3)
2.2648(14)
1.873(4)
1.160(5)
1.791(5)
1.790(5)
67.76(4)
155.41(12)
85.60(12)
174.9(4)
Fe(2)–C(4)
Fe(1)–S(2)–Fe(2)
C(16)–Fe(1)–S(1)
C(16)–Fe(1)–S(2)
C(16)–N(3)–C(17)
Fe(1)–S(1)
Fe(1)–S(2)
Fe(2)–S(1)
Fe(2)–S(2)
Fe(1)–C(15)
Fe(1)–C(1)
Fe(1)–C(2)
2.2646(10)
2.2707(10)
2.2756(10)
2.2701(11)
1.867(3)
1.788(4)
1.770(4)
Fe(2)–C(20)
N(3)–C(15)
N(3)–C(16)
N(4)–C(20)
N(4)–C(21)
Fe(2)–C(3)
Fe(2)–C(4)
1.886(4)
1.149(4)
1.451(4)
1.151(4)
1.463(4)
1.786(4)
1.780(4)
˚
the Fe(1)–P(1) distance (2.2025 A) is slightly shorter ◦than that of
˚
Fe(2)–P(2) (2.2311 A) in 4. The P–C–P angle is 113.1 , very close
to that of 2 with a different coordination mode.
10910 | Dalton Trans., 2011, 40, 10907–10917
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig. 2 Molecular structures of 5 (a), 7 (b), 9 (c) and 10 (d) with thermal ellipsoid set at 50% probability. The hydrogen atoms are omitted for clarity. The
labels for C atoms of benzene rings are not shown for clarity.
It has been reported that there are four possible coordination
configurations for double-CO displacement by two identical non-
carbonyl ligands on each iron atom of several model com-
plexes, which are, ap/ap, ap/ba, ba/ba (trans) and ba/ba (cis)
configurations. For the doubly substituted complex 10, one
isocyanide occupies at the apical position while the other at the
basal position (Fig. 2(d)). Such ap/ba arrangement is different
from those in related complexes [Fe2(m-pdt)(CO)4(ButNC)2]12
and [Fe2{(m-SCH2)2N(4-CH3C6H4)}(CO)4(4-IC6H4NC)2],46 which
have the isocyanide ligands in the ap/ap (trans) and ba/ba (cis)
conformations in the solid state structure, respectively. The C–N–
C angles of the t-Butylisocyanide in 10 [177.7(4)] and 177.0(3)◦,
respectively] are relatively linear, whereas that of 9 [174.9(4)◦] is
slightly bent.
A comparison of Fe–Fe bond lengths of these complexes shows
that the metal–metal distance in the diiron compounds is slightly
affected by the substitution position and the nature of the terminal
ligands with a better electron-donating ability. For the mono-
substituted complexes, substitution by an electron-donor ligand at
the basal position leads to the elongation of Fe–Fe bond distance,
whereas at the apical position it causes the decrease of the Fe–Fe
bond lengths. For example, in comparison to the parent complex
˚
1, the Fe–Fe bond length of 5 is apparently longer by ca. 0.034 A,
˚
while that of 2 shows slight decrease of 0.016 A upon coordination
of dppm to Fe(1). Complex 7 with PPh3 lying in the apical position
only shows a little decrease in Fe–Fe distance, which is in contrast
to [Fe2(m-Mebdt)(CO)5(PPh3)], where the axial PPh3 causes the
elongation of Fe–Fe bond.44 In comparison with 1, the Fe–Fe
˚
bond length in 3 shows significant lengthening by 0.064 A, whereas
˚
in 4, it shortens by 0.013 A. This could be due to their different
coordination manners of dppm ligands.
In the mono-substituted complexes 5 and 9, the quinoxaline
rings are oriented toward the sites occupied by basal PEt3 or t-
butylisocyanide, a geometry that seems to be sterically disfavored.
However, when the substitution happens at the apical positions,
as in 2 and 7, the quinoxaline rings are pushed away from the
substituted Fe center to avoid the strong steric repulsion between
the quinoxalin rings and phosphine ligands. It is perhaps due to the
spatial requirements of the quinoxaline rings, further displacement
of CO by tertiary phosphines on the unsubstituted Fe atoms
is unfavorable and the second substitution only occurs for the
less bulky t-Butylisocyanide resulting in the formation of 10. The
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 10907–10917 | 10911
©

View Article Online
less spatial demand of t-Butylisocyanide is also evident from the
small distortion away from an eclipsed state of the molecular
structures of 9 and 10, where the average torsional angles between
the three non-sulfur substituents on each iron atom are 1.1 and
1.5◦ respectively. For the more bulky substituents such as PPh3,
the twisting away from the eclipsed state becomes a little more
appreciableas theaverage torsionalangle of 7 increases to7.9◦ with
a maximum of 16.6◦ associated with P–Fe(1)–Fe(2)–C(5) (Fig.
2(b)). This is especially so in 3 where the average torsional angle is
14.7◦, which includes angles of 23.7 and 18.4◦ between C1–C4 and
P2–C2 respectively (Fig. 1(c)). It could be for this reason that 3 is
somewhat unstable in solution and readily decomposes during the
crystallization process. Despite this asymmetry, the substitution of
donor ligands does not change the unrotated nature of the parent
complex.
The protonation of complexes 5–10 has also been investigated.
The results suggest that these complexes were not stable under
acidic conditions.
Electrochemistry of complexes 1–4
Since complexes 5–10 were not stable under acidic conditions,
they are not suitable for use as catalysts for hydrogen evolution.
Therefore, only the electrochemical properties of complexes 1–
4 were investigated with respect to their potential application
as catalysts that mimic hydrogenase. The hydrogen evolution
reaction catalyzed by the diiron based model hydrogenases is
complicated. It involves both electron transfer and protonation
reactions. The ability to achieve catalysis is determined by the
acid–base properties of the diiron complexes in different redox
states. One of the possible reaction mechanisms is summarized in
Scheme 4.
Spectroscopic studies of the protonation of complexes 1–4 by FTIR
and 1H NMR
Since protonation of N atoms in the model complexes affects the
electrochemical properties of these complexes and their catalytic
activity towards hydrogen evolution,16,20–35 protonation of our
complexes in MeCN with HBF4 were investigated spectroscopi-
cally using both FTIR and 1H NMR.
Scheme 4 A possible mechanism for hydrogen evolution catalyzed by a
hydrogenase type model diiron complex.
The FTIR spectra of the protonation of 1–4 in MeCN solution
with HBF4 showed similar behavior. Upon addition of one equiv-
alent of acid, the infrared absorption bands of these complexes
have no clear shifts. However, after five equivalent acid was added,
the carbonyl frequencies in the IR spectra shifted towards higher
energy by about 5, 5, 10, and 7 cm-1 for 1, 2, 3 and 4, respectively,
compared to their corresponding non-protonated forms. These
blue shifts clearly indicated that the protonation occurred on the
N atoms of quinoxaline rings as they are of the same magnitude
as those observed for the protonated model complexes with
pyridine or bridgehead N atoms.14,23,28,47 While 1 requires at least
5 equiv. acid for complete protonation, addition of 3 equiv. acid
leads to complete protonation on N atoms for complexes 2–4.
The electron-donating dppm improves the proton binding ability
for 2–4. The protonation could be completely reversed by the
addition of 5 equiv. of triethylamine to regenerate the parent
complexes.
Ideally, if the diiron model complex is to be the catalyst, it should
be fully stable during all stages of catalytic reaction. This requires
the stability in all three FeIFeI, FeIFe0 and Fe◦Fe0 redox states both
in the presence and absence of acid. From the electrochemical
point of view, two one-electron reduction processes involving
diiron centers should be chemically reversible. Cyclic voltammetric
measurements were therefore, carried out to investigate whether
this criterion is satisfied in the case of our diiron complexes.
A cyclic voltammogram of complex 1 obtained at a scan rate of
0.1 V s-1 (Fig. 3) is rather complex, exhibiting three reduction and
With regard to the protonated complexes, their 1H NMR spectra
show no peak at d < 0 ppm even at temperatures lower than -70 ◦C,
indicating again that the protonation occurred on the quinoxaline
N atoms rather than at the Fe–Fe site to form m-H complexes.
The peaks for the methylene protons of 1H+ appear at 4.72 and
3.65 ppm, shifted to higher magnetic field by 0.49 and 0.25 ppm,
respectively, than its parent complex 1. A similar trend has been
1
observed in the H NMR spectra for the protonated species of
2–4. In addition, a new broad peak at around d 14.00 emerges in
1
all the H-NMR spectra of 1–4 after protonation. These newly-
formed signals are tentatively assigned to the NH of protonated
quinoxalines. As expected, only slight shifts for the 31P signals of
2–4 were observed when the protonation on N atoms occurred.
For example, after protonation, the two doublets in the 31P NMR
of 4 shift high-field from 52.4 and 54.1 ppm to 50.7 and 52.5 ppm.
Attempts to get single crystals of the protonated species of 1–4
were not successful.
Fig. 3 Cyclic voltammograms of 1 mM 1 in MeCN (0.1 M Bu4NPF6)
obtained at a 1 mm diameter glassy carbon electrode with a scan rate of
0.1 V s-1. The black curve was obtained by initially scanning the potential
in the negative direction. The red curve was obtained by initially scanning
the potential in the positive direction. The cyclic voltammogram for 1 mM
ligand, 2,3-bis(bromomethyl)quinoxaline (blue curve), was obtained under
the same experimental conditions.
10912 | Dalton Trans., 2011, 40, 10907–10917
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
two oxidation processes within the solvent/electrolyte potential
window. As shown in Fig. 4, the first reduction process has a
reversible potential of -1.439 V vs. Fc/Fc+ as calculated from
the average of the reduction and oxidation peak potentials (a
small shoulder observed at low scan rate is attributed to the
protonation of the one-electron reduced form due to the presence
of trace water). At moderate scan rates, this is an electrochemically
first reduction process is assigned to the one-electron reduction
of FeIFeI to FeIFe0 as in eqn (2),
FeIFeI + e- ꢀ FeIFe0
(2)
All other processes are chemically irreversible with peak
potentials of -1.612 and -1.920 V vs. Fc/Fc+ for the two
remaining reduction processes, and 0.362, 0.891 and 1.124 V vs.
Fc/Fc+ for the three oxidation processes. A cyclic voltammogram
obtained from the ligand, 2,3-bis(bromomethyl)quinoxaline, is
also included in Fig. 3. Via comparison of metal complex and
ligand data, the processes found at -1.612 V and 0.362 V vs. Fc/Fc+
are assigned as being ligand based, while the process at -1.920 V
most probably represents the reaction scheme, on the basis of the
previous studies on the similar compounds.32,49,50
reversible process with a peak-to-peak separation (DEp) of 58
2
mV which is very close to the theoretical value of 57 mV◦predicted
for a reversible one-electron transfer reaction at 25 C.48 The
fact that this DEp only increases slightly (mostly likely due to
uncompensated resistance) and the peak current increases linearly
with the square root of the scan rate over the range of 0.05
to 0.5 V s-1 also suggests that this process is chemically and
electrochemically reversible and diffusion controlled. A diffusion
coefficient, D, of 1.23 ¥ 10-5 cm2 s-1 for 1 was calculated based on
the Randles–Sevcik relationship,48
FeIFe0 + e- ꢀ Fe◦Fe0 → product
(3)
Other small peaks (Fig. 3) that follow the irreversible oxida-
tion or reduction processes are attributed to the decomposition
products arising from the major processes. A related voltammetric
response of complex 1 was obtained in dichloromethane (0.1 M
Bu4NPF6), but with a reversible potential of -1.610 V vs. Fc/Fc+
for the first one-electron reduction process. Thus, this is a
significant thermodynamic effect in changing from acetonitrile
where the reversible potential was -1.439 V vs. Fc/Fc+.
Steady state voltammetric experiments were also conducted
using a 10 mm diameter gold microdisc electrode in order to
provide further understanding of the electrochemical processes.
As shown in Fig. 5, two major oxidation processes with half
wave potentials of 0.418 and 1.106 V vs. Fc/Fc+ and two major
reduction processes with half wave potentials of -1.440 and
-1.978 V vs. Fc/Fc+ are observed. A diffusion coefficient of 1.20 ¥
10-5 cm2 s-1 was calculated from the steady-state diffusion limiting
current for the first reduction process from the relationship,
1/ 2
⎛
⎜
⎜
⎞
nFDv
RT
c*
⎟
(1)
i = 0.4463nFA
⎟
p
⎟
⎟
⎜
⎝
⎠
where ip is the peak current; n is the number of electron transferred
which is assumed to be 1 in this case; F is Faraday’s constant; A is
the electrode area; v is the scan rate; c* is the bulk concentration
of the electroactive species; R is gas constant and T is temperature
(◦K). This diffusion coefficient value in MeCN is typical for
species having a size similar to complex 1. Consequently, the
iss = 4nFDrc*
(4)
where r is the radius of the electrode and the number of electrons
transfer is assumed to be one. The fact that this value of
diffusion coefficient obtained agrees well with the one calculated
from transient voltammetry (1.23 ¥ 10-5 cm2 s-1) supports the
assumption of a one-electron transfer reaction. The steady-state
Fig. 4 (a) Cyclic voltmmograms of the first reduction process obtained
at scan rates of 0.05, 0.1, 0.2 and 0.5 V s-1 with a 1 mm diameter glassy
carbon electrode and (b) dependence of the reduction peak current on the
square root of scan rate.
Fig. 5 Steady state voltammograms of 1 mM 1 (black) and ligand,
2,3-bis(bromomethyl) quinoxaline (red) obtained at a 10 mm diameter Au
microelectrode. A scan rate of 0.05 V s-1 was used.
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 10907–10917 | 10913
©

View Article Online
diffusion controlled limiting current associated with the second
reduction process (half wave potential = -1.978 V vs. Fc/Fc+) is
comparable to that for the first and is assigned to the reduction
scheme given in eqn (3) since the half wave potential of this
process is close to the peak potential of -1.92 V obtained from
cyclic voltammetry (Fig. 3). The steady-state diffusion controlled
limiting current of the first oxidation process (half wave potential =
0.418 V vs. Fc/Fc+) is about half that of the first one-electron
reduction process. The response may therefore be associated with
an irreversible oxidative dimerization involving complex 1 reacting
with the one-electron oxidized form, which is unstable (eqn (5)
and 6).
Conclusion
A new complex 1 that contains a large conjugated quinoxaline
moiety has been prepared as a potential model for hydrogenase
active site. Reaction of 1 with dppm yields complexes 2–4 that
differ in their dppm coordination modes. In 2, the dppm is
terminally bonded to an iron atom with one phosphorus atom,
whereas in 3 or 4, it attaches to a Fe atom in a dibasal-substituted
manner or bridges two Fe atoms in a cis basal/basal fashion.
CO-substituted derivatives of tertiary phosphines with various
electron-donor abilities and t-butylisocyanide derivatives also have
been prepared. Electronic effects of the substituted ligands on
the structure parameters of the parent complex as well as the
electron properties of the diiron site have been studied. In addition,
protonation of 1–4 is shown to occur on the quinoxaline N atoms
by shifts in the nCO bands in IR spectra. Electrochemical studies
reveal that only 1 can be reduced reversibly via a one-electron
transfer reaction. However, the electrochemical ease of reduction
of the ligand, quinoxaline, and the instability of the two-electron
reduced Fe0–Fe0 state of complex 1 suggest that this compound
does not provide an ideal model of the hydrogenase active site. The
even greater instabilities of reduced forms of 2–4 also rule these
materials from being efficient hydrogen catalysts.
FeIFeI ꢀ FeIIFeI + e-
(5)
(6)
FeIIFeI + FeIIFeI → FeIIFeI-FeIIFeI → product
The steady-state diffusion controlled limiting current of the
second oxidation process (half wave potential = 1.106 V vs.
Fc/Fc+) is approximately twice that of the first reduction pro-
cess. Therefore, this oxidation process could be assigned to
a two-electron oxidation process of the product of the first
oxidation process. A steady-state voltammogram for the lig-
and, 2,3-bis(bromomethyl)quinoxaline, reduction process is also
included in Fig. 5. Comparison of the steady-state diffusion
limiting currents of the ligand and complex 1, after taking
into account the differences in their sizes suggests that the
ligand undergoes a two-electron reduction process at a potential
similar to the reduction of FeIFeI. This finding is consistent with
those proposed in other electrochemical studies of quinoxaline
derivatives.51
4. Experimental
Reagents and instruments
All synthetic reactions and operations were carried out under
nitrogen atmosphere with standard Schlenk techniques. All sol-
vents were dried and distilled prior to use according to standard
methods. Tetrahydrofuran was purified by distillation under
N2 from sodium/benzophenone. MeCN and dichloromethane
were distilled from CaH2 and P2O5, respectively, under N2. The
following chemicals were commercially available and used as
received: 2,3-bis(bromomethyl)quinoxaline, phosphines, ButNC,
LiEt3BH and HBF4·OEt2. The compounds [(m–S)2Fe2(CO)6] and
[(m-LiS)2Fe2(CO)6] were synthesized according to the literature
procedure.52,53 Infrared spectra were recorded with a Spectrum-
All of the above findings on the reduction processes are not
conducive to anticipating efficient catalysis to form hydrogen and
this was not found on the addition of acidic acid and examination
of reduction in the acidified medium.
The electrochemistry of other dppm derivatives was also
examined in CH3CN (0.1 M Bu4NPF6). The results reveal that
even the first reduction of 2–4 is not reversible as was found
for 1. The first reduction peak potentials associated with the
FeIFeI to unstable Fe◦FeI and then to final product process of
complexes 1–4 are -1.468, -1.670, -1.916 and -1.944 V vs. Fc/Fc+,
respectively. Thus the first irreversible reductive peaks for 2, 3
and 4 occur at more negative potentials by 202, 448 and 476 mV,
respectively, than that of the parent complex 1. The electron donor
character of dppm would therefore appear to render the reduction
of the iron core even more difficult, with di-substitution of dppm
exerting additional influence on the redox potential of 1 than
the mono-substitution. However, it should be noted that it is not
meaningful to directly compare a reduction peak potential for
an irreversible process (2–4) with that derived from a reversible
process (as found for 1). However, chemical irreversibility of
2–4 is predicted to lead to a positive shift of the reduction
peak potential from the reversible case. Therefore, the conclu-
sions on substituent effect reached are probably qualitatively
correct.
One FT-IR spectrophotometer. H and 31P NMR spectra were
1
collected on a BRUKER AVANCE III 400 NMR spectrometer.
Mass spectra were recorded on a DECAX-30000 LCQ Deca XP
instrument. Elemental analyses for C, H, and N were determined
using an Elementar Vario EL III elemental analyzer.
Synthesis of complexes 1–10
[Fe2{(l-CH2S)2R}(CO)6]
(1).
A
mixture
of
2,3-
bis(bromomethyl)quinoxaline (316.01 mg, 1 mmol) and [(m-
LiS)2Fe2(CO)6] (357.76 mg, 1 mmol) in THF (20 ml) was stirred
0.5 h under a temperature of -78 ◦C. After further reaction for 2 h
under room temperature, the solvent was removed under vacuum,
and the residue was purified by chromatography on silica gel with
petroleum ether/ethyl acetate (6 : 1 v/v) as eluent to give 1 as
1
the major product (123 mg, 25%). H NMR (CDCl3): d 3.40 (d,
J = 12.8 Hz, 2H,CCH2S), 4.29 (d, J = 12.8 Hz, 2H, CCH2S),
7.26–7.98 (m, 4H, C6H4) ppm; IR (CH3CN): n(CO) 2079, 2042,
2003 cm-1; MS (ESI): m/z: 499.1 [M–H]-; Anal. Calcd (%) for
C16H8Fe2N2O6S2: C 38.43, H 1.61, N 5.60. Found: C 38.16, H
1.76, N, 5.58.
Again with 2–4, ideal characteristics for mimicking the behav-
iors in Scheme 4 or those used by hydrogenase are not available,
so prospects for hydrogen catalysis are poor.
10914 | Dalton Trans., 2011, 40, 10907–10917
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
2051, 1993, 1939 cm-1; MS (ESI): m/z: 734.5. Anal. Calcd (%) for
C33H23Fe2N2O5PS2: C 53.91, H 3.13, N 3.81. Found: C 54.13, H
3.26, N 3.72. For 8 (38 mg, 63%): 1H NMR (CDCl3): d 0.88–2.01
(m, 6H, 2CH3), 3.35 (d, J = 10.8 Hz, 2H, CCH2S), 4.20 (s, 2H,
CCH2S), 7.26–7.94 (m, 9H, Ph, C6H4) ppm;31P NMR (CDCl3): d
33.65(s), 30.19(s) ppm; IR (CH2Cl2): n(CO) 2046, 1989, 1968, 1931
cm-1; MS (ESI): m/z: 610.6. Calcd (%) for C23H19Fe2N2O5PS2: C
45.27, H 3.14, N 4.59. Found: C 45.39, H 3.10, N 4.49.
[Fe2{(l-CH2S)2R}(CO)5(dppm)] (2). A solution of 1 (50 mg,
0.1 mmol) and Me3NO (7.5 mg, 1 mmol) in CH3CN (20 ml) was
stirred at room temperature for 15 min. Then, dppm (38.4 mg,
0.1 mmol) was added. After stirring for 2 h, the solvent was re-
moved under reduced pressure and the crude product was purified
by chromatography on silica gel with petroleum ether/ethyl acetate
1
(4 : 1 v/v) as eluent to afford 2 as a red solid (180 mg, 77%). H
NMR (CDCl3): d 1.26 (s, 2H, PCH2P), 3.44 (s, 2H, CCH2S), 3.64
(s, 2H, CCH2S), 7.15–8.00 (m, 24H, 4Ph, C6H4) ppm;31P NMR
(CDCl3): d 55.83 (d, J = 220 Hz), –26.43 (d, J = 216 Hz) ppm;
IR (CH3CN): n(CO) 2052, 1992, 1943 cm-1; MS (ESI): m/z 856.5;
Anal. Calcd (%) for C40H30Fe2N2O5P2S2: C 56.10, H 3.53, N 3.27.
Found: C 56.49, H 3.60, N 3.27.
[Fe2{(l-CH2S)2R}(CO)5(ButNC)] (9) and [Fe2{(l-CH2S)2R}-
(CO)4(ButNC)2] (10). Complexes 9 and 10 were synthesized in
a similar manner to 5 except that one or two equiv of ButNC, were
1
employed. For 9 (38 mg, 63%): H NMR (CDCl3): d 1.34, 1.37
(2 s, 9H, 3CH3), 3.31 (m, 2H, CCH2S), 4.13 (m, 2H, CCH2S),
7.26–7.96 (m, 4H, C6H4) ppm; IR (CH3CN): n(CO) 2048, 2007,
1978, 1958 cm-1; MS (ESI): m/z: 556.1 [M+H]+; Anal. Calcd (%)
for C20H17Fe2N3O5S2: C 43.24, H 3.06, N 7.57. Found: C 43.47,
H 3.43, N, 7.82. For 10 (43 mg, 70%): 1H NMR (CDCl3): d 1.27–
1.61 (m, 18H, 6CH3), 3.23(d, J = 9.6 Hz, 2H, CCH2S), 4.05 (d, J =
11.6 Hz, CCH2S), 7.26–7.95 (m, 4H, C6H4) ppm; IR (CH3CN):
n(CO) 2005, 1977, 1946 cm-1; MS (ESI): m/z: 610.7; Anal. Calcd
(%) for C24H26Fe2N4O4S2: C 47.23, H 4.29, N 9.18. Found: C 47.55,
H 4.03, N 9.24.
[Fe2{(l-CH2S)2R}(CO)4(k2–dppm)] (3) and [Fe2{(l-CH2S)2R}-
(CO)4(l-dppm)] (4). A solution of 1 (100 mg, 0.2 mmol), dppm
(76.8 mg, 0.2 mmol), and Me3NO (30 mg, 0.4 mmol) in toluene
(25 ml) was refluxed for 4 h. The solvent was then removed
under reduced pressure, and the crude product was purified by
chromatography on silica gel with petroleum ether/ethyl acetate
(4 : 1 v/v) as eluent. Complexes 3 (35 mg, 21%) and 4 (27 mg,
1
17%) were isolated. For 3: H NMR (CDCl3): d 2.07, 2.38 (2 s,
2H, PCH2P), 3.48 (m, 2H, CCH2S), 4.24 (m, 2H, CCH2S), 7.19–
7.93 (m, 24H, 4Ph, C6H4) ppm; 31P NMR (CDCl3): d 8.44 (s), 12.43
(s) ppm; IR (CH3CN): n(CO) 2023, 1956, 1916 cm-1; MS (ESI):
m/z: 827.2 [M - H]-; Anal. Calcd (%) for C39H30Fe2N2O4P2S2:
C 56.49, H 3.62, N 3.38. Found: C 56.69, H 3.54, N 3.48. For
4: 1H NMR (CDCl3): d 2.04 (s, 2H, PCH2P), 3.22–3.74 (m,
2H, CCH2S), 4.11–4.71 (m, 2H, CCH2S), 7.12–8.00 (m, 24H,
4Ph, C6H4) ppm; 31P NMR (CDCl3): d 52.42 (d, J = 132 Hz),
54.08(d, J = 132 Hz) ppm; IR (CH3CN): n(CO) 1993, 1963,
1929 cm-1; MS (ESI): m/z: 827.0 [M - H]-; Anal. Calcd (%) for
C39H30Fe2N2O4P2S2: C 56.49, H 3.62, N 3.38. Found: C 56.80, H
3.40, N 3.16.
Crystal structure determination of complexes 1–5, 7, 9 and 10
X-ray diffraction data were collected on a Rigaku diffractometer
˚
with a Mercury CCD area detector (Mo-Ka; l = 0.71073 A) at
293(2) K. Empirical absorption corrections were applied to the
data using the CrystalClear program.54 The structures were solved
by the direct method and refined by the full-matrix least-squares
on F2 using the SHELXTL-97 program.55 All the non-hydrogen
atoms were refined anisotropically. The hydrogen atoms were all
treated by geometrical positions. Crystallographic data and other
pertinent information for 1–5, 7, 9 and 10 are summarized in
Tables 2 and 3 .
[Fe2{(l-CH2S)2R}(CO)5L] (5: L = PEt3; 6: PMe3; 7: PPh3; 8:
PMe2Ph). In a typical reaction, a solution of 1 (50 mg, 0.1 mmol)
and Me3NO (7.5 mg, 1 mmol) in CH3CN (20 ml) was stirred at
room temperature for 15 min. Then, PMe3 (10.2 ml, 0.1 mmol)
was added. After stirring for 2 h, the solvent was removed
under reduced pressure, and the crude product was purified by
chromatography on silica gel with petroleum ether/ethyl acetate
(4 : 1 v/v) as eluent. Complex 6 was obtained as a red solid (40 mg,
74%). 1H NMR (CDCl3): d 0.87–1.51 (m, 9H, 3CH3), 3.36 (d, J =
12.4 Hz, 2H, CCH2S), 4.23 (d, J = 12.4 Hz, 2H, CCH2S). 7.70–7.95
(m, 4H, C6H4) ppm; 31P NMR (CDCl3): d 23.07 (s), 26.09 (s) ppm;
IR (CH3CN): n(CO) 2044, 1988, 1971, 1930 cm-1; MS (ESI): m/z:
548.9; Anal. Calcd (%) for C18H17Fe2N2O5PS2: C 39.44, H 3.13; N,
5.11. Found: C 39.16, H, 3.11, N 4.85. Complexes 5, 7 and 8 were
synthesized in a similar way as that of 6 by using one equiv of PEt3,
PPh3 and PMe2Ph, respectively, instead of PMe3. For 5 (33 mg,
55%): 1H NMR (CDCl3): d 1.18–1.23 (m, 15H, 3CH2CH3), 3.40
(d, J = 12.4 Hz, 2H, CCH2S), 4.26 (d, J = 12.8 Hz, 2H, CCH2S),
7.26–7.95 (m, 4H, C6H4) ppm;31P NMR (CDCl3): d 56.52 (s), 53.16
(s) ppm; IR (CH3CN): n(CO) 2044, 1986, 1970, 1926 cm-1; MS
(ESI): m/z: 590.5. Anal. Calcd (%) for C21H23Fe2N2O5PS2: C 42.68,
H 3.90, N 4.74. Found: C 42.86, H 3.96, N, 5.00. For 7 (54 mg,
Electrochemistry
Electrochemical measurements were carried out using a CHI 700
Electrochemical Workstation (CH Instruments, Austin, Texas,
USA). Cyclic voltammetric measurements were performed in
MeCN containing 0.1 M Bu4NPF6 as the supporting electrolyte
with a glassy carbon electrode (1 mm or 3 mm diameter) as
the working electrode and platinum wires as both the quasi-
reference and counter electrodes. Steady state voltammograms
were obtained with a gold microelectrode (10 mm diameter) as the
working electrode and platinum wires again as both the quasi-
reference and counter electrodes. The Pt quasi-potential scale
was then calibrated against the Fc/Fc+ couple and potentials
are reported versus this reference system. All electrochemical
experiments were carried out at 25
N2 atmosphere.
2
◦C in a dry box under
Appendix A. Supporting information
CCDC 816112–816119 contains supplementary crystallographic
data for the complexes 1–5, 7, 9 and 10. These data can be obtained
free of charge from the Cambridge Crystallographic Data Center
via http://www.ccdc.cam.ac.uk/data_request/cif.
1
73%): H NMR (CDCl3): d 3.54 (d, 2H, J = 12.4 Hz, CCH2S),
4.46 (s, 1H, CCH2S), 7.26–7.89 (m, 19H, 3Ph, C6H4) ppm;31P
NMR (CDCl3): d 64.21 (s), 64.14 (s) ppm; IR (CH2Cl2): n(CO)
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 10907–10917 | 10915
©

View Article Online
Table 2 Crystal data and structure refinement details for 1–4
Complex
1·0.5CH2Cl2
2
3·CH2Cl2
4·2CH2Cl2
Mol. formula
FW
Cryst. syst.
C16.5H9Cl Fe2N2O6S2
542.53
C40H30Fe2N2 O5P2S2
856.42
C40H32Cl2 Fe2N2O4P2S2
913.34
C41H34Cl4Fe2 N2O4P2S2
998.26
Triclinic
Triclinic
Triclinic
Triclinic
¯
¯
¯
¯
Space group
P1
P1
P1
P1
˚
a/A
6.801(2)
11.254(4)
13.330(5)
89.908(12)
81.324(10)
77.667(10)
984.7(6)
2
9.2471(6)
12.8115(8)
16.3337(10)
98.516(4)
95.659(2)
96.527(3)
1888.1(2)
2
10.199(7)
13.712(9)
14.582(9)
92.981(8)
91.827(7)
100.728(13)
1999(2)
2
10.518(6)
11.149(6)
18.267(11)
86.244(12)
88.721(11)
87.302(13)
2135(2)
2
˚
b/A
˚
c/A
a (◦)
b (◦)
g (◦)
3
˚
V/A
Z
Dc/g cm-3
m/mm-1
F(000)
1.830
1.859
542
7453
4329
290
0.0463
0.1376
1.040
1.506
1.011
876
14834
8510
594
0.0451
0.1075
1.005
1.517
1.087
932
11750
6216
487
0.0500
0.1364
1.026
1.553
1.147
1016
14946
6616
514
0.0748
0.1936
1.087
Rflns collected
Rflns unique
Parameters
R1
wR2
GOF
Table 3 Crystal data and structure refinement details for complexes 5, 7, 9 and 10
Complex
5
7
9·CH2Cl2·H2O
10
Mol. formula
FW
Cryst. syst.
C21H23Fe2N2O5PS2
590.20
C33H23Fe2 N2O5PS2
734.32
C21H21Cl2Fe2 N3O6S2
658.13
C24H26Fe2 N4O4S2
610.31
Triclinic
Triclinic
Triclinic
Triclinic
¯
¯
¯
¯
Space group
P1
P1
P1
P1
˚
a/A
6.934(4)
12.827(7)
13.646(7)
87.864(11)
86.760(12)
81.660(12)
1198.5(11)
2
9.3335(0)
12.73310(10)
14.2960(2)
93.811(11)
105.134(7)
109.359(7)
1525.35(2)
2
9.174(4)
11.761(6)
13.765(7)
89.067(15)
81.459(14)
72.711(10)
1401.7(12)
2
10.390(3)
11.082(3)
13.014(4)
85.726(8)
86.282(7)
68.586(6)
1389.9(6)
2
˚
b/A
˚
c/A
a (◦)
b (◦)
g (◦)
3
˚
V/A
Z
Dc/g cm-3
m/mm-1
F(000)
1.636
1.487
604
9427
5387
298
0.0424
0.1327
1.051
1.599
1.187
748
11926
6846
406
0.0373
0.0913
1.559
1.414
668
10801
6235
325
0.0592
0.1602
1.007
1.458
1.230
628
11001
6271
325
0.0451
0.1064
1.003
Rflns collected
Rflns unique
Parameters
R1
wR2
GOF
1.065
4 D. J. Evans and C. J. Pickett, Chem. Soc. Rev., 2003, 32, 268–
275.
Acknowledgements
5 J. Alper, Science, 2003, 299, 1686–1687.
6 J. L. Dempsey, B. S. Brunschwig, J. R. Winkler and H. B. Gray, Acc.
Chem. Res., 2009, 42, 1995–2004.
7 Y. Nicolet, C. Piras, P. Legrand, C. E. Hatchikian and J. C. Fontecilla-
Camps, Structure, 1999, 7, 13–23.
8 J. W. Peters, W. N. Lanzilotta, B. J. Lemon and L. C. Seefeldt, Science,
1998, 282, 1853–1858.
This work was supported by the grants from the State Key Labo-
ratory of Structural Chemistry, Fujian Institute of Research on the
Structure of Matter, Chinese Academy of Sciences, National Basic
Research Program of China (973 Program, 2007CB815306), and
the National Natural Science Foundation of China (20733003).
JZ and AMB also would like to thank the Australian Research
Council for financial support.
9 Y. Nicolet, B. J. Lemon, J. C. Fontecilla-Camps and J. W. Peters, Trends
Biochem. Sci., 2000, 25, 138–143.
10 E. J. Lyon, I. P. Georgakaki, J. H. Reibenspies and M. Y. Darensbourg,
Angew. Chem., Int. Ed., 1999, 38, 3178–3180.
11 F. Gloaguen, J. D. Lawrence, M. Schmidt, S. R. Wilson and T. B.
Rauchfuss, J. Am. Chem. Soc., 2001, 123, 12518–12527.
12 J. L. Nehring and D. M. Heinekey, Inorg. Chem., 2003, 42, 4288–4292.
13 J. W. Tye, J. Lee, H. W. Wang, R. Mejia-Rodriguez, J. H. Reibenspies, B.
Michael and M. Y. Darensbourg, Inorg. Chem., 2005, 44, 5550–5552.
14 P. Li, M. Wang, C. He, G. Li, X. Liu, C. Chen, B. kermark and L. Sun,
Eur. J. Inorg. Chem., 2005, 2506–2513.
References
1 R. Cammack, Nature, 1999, 397, 214–215.
2 H. L. Karunadasa, C. J. Chang and R. J. Long, Nature, 2010, 464,
1329–1333.
3 M. Y. Darensbourg, E. J. Lyon, X. Zhao and I. P. Georgakaki, Proc.
Natl. Acad. Sci. U. S. A., 2003, 100, 3683–3688.
10916 | Dalton Trans., 2011, 40, 10907–10917
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
15 D. Morvan, J. F. Capon, F. Gloaguen, A. Le Goff, M. Marchivie, F.
Michaud, P. Schollhammer, J. Talarmin, J. J. Yaouanc and R. Pichon,
Organometallics, 2007, 26, 2042–2052.
16 W. Gao, J. Ekstrm, J. Liu, C. Chen, L. Eriksson, L. Weng, B. kermark
and L. Sun, Inorg. Chem., 2007, 46, 1981–1991.
17 M. L. Singleton, N. Bhuvanesh, J. H. Reibenspies and M. Y.
Darensbourg, Angew. Chem., Int. Ed., 2008, 47, 9492–
9495.
18 L. C. Song, C. G. Li, J. Gao, B. S. Yin, X. Luo, X. G. Zhang, H. L. Bao
and Q. M. Hu, Inorg. Chem., 2008, 47, 4545–4553.
19 P. Li, M. Wang, L. Chen, J. Liu, Z. Zhao and L. Sun, Dalton Trans.,
2009, 1919–1926.
20 J. D. Lawrence, H. Li, T. B. Rauchfuss, M. Be´nard and M. M. Rohmer,
Angew. Chem., Int. Ed., 2001, 40, 1768–1771.
21 J. D. Lawrence, H. Li and T. B. Rauchfuss, Chem. Commun., 2001,
1482–1493.
22 W. Gao, J. Liu, C. Ma, L. Weng, K. Jin, C. Chen, B. Akermark and L.
Sun, Inorg. Chim. Acta, 2006, 359, 1071–1080.
23 W. Gao, J. Liu, B. Akermark and L. Sun, J. Organomet. Chem., 2007,
692, 1579–1583.
34 F. Xu, C. Tard, X. Wang, S. K. Ibrahim, D. L. Hughes, W. Zhong, X.
Zeng, Q. Luo, X. Liu and C. J. Pickett, Chem. Commun., 2008, 606–608.
35 S. Jiang, J. Liu, Y. Shi, Z. Wang, B. Akermark and L. Sun, Polyhedron,
2007, 26, 1499–1504.
36 G. Eilers, L. Schwartz, M. Stein, G. Zampella, L. de Gioia, S. Ott and
R. Lomoth, Chem.–Eur. J., 2007, 13, 7075–7084.
37 R. D. Adams and S. Miao, Inorg. Chem., 2004, 43, 8414–8426.
38 L. Schwartz, P. S. Singh, L. Eriksson, R. Lomoth and S. Ott, C. R.
Chim., 2008, 11, 875–889.
39 L. Schwartz, L. Eriksson, R. Lomoth, F. Teixidor, C. Vias and S. Ott,
Dalton Trans., 2008, 2379–2381.
40 C. M. Thomas, O. Ru¨diger, T. Liu, C. E. Carson, B. Michael and M.
Y. Darensbourg, Organometallics, 2007, 26, 3976–3984.
41 F. I. Adam, G. Hogarth and I. Richards, J. Organomet. Chem., 2007,
692, 3957–3968.
42 X. Zhao, I. P. Georgakaki, M. L. Miller, R. Mejia-Rodriguez, C. Y.
Chiang and M. Y. Darensbourg, Inorg. Chem., 2002, 41, 3917–3928.
43 J. A. Cabeza, M. A. Mart´ınez-Garc´ıa, V. Riera, D. Ardura and S.
Garcia-Granda, Organometallics, 1998, 17, 1471–1477.
44 M. M. Hasan, M. B. Hursthouse, S. E. Kabir and K. Malik, Polyhedron,
2001, 20, 97–101.
24 J. F. Capon, S. Ezzaher, F. Gloaguen, F. Y. Pe´tillon, P. Schollhammer,
J. Talarmin, T. J. Davin, J. E. McGrady and K. W. Muir, New J. Chem.,
2007, 31, 2052–2064.
45 L. C. Song, X. Luo, Y. Z. Wang, B. Gai and Q. M. Hu, J. Organomet.
Chem., 2009, 694, 103–112.
25 Y. Si, C. Ma, M. Hu, H. Chen, C. Chen and Q. Liu, New J. Chem.,
2007, 31, 1448–1454.
26 S. Ott, M. Kritikos, B. kermark, L. Sun and R. Lomoth, Angew. Chem.,
Int. Ed., 2004, 43, 1006–1009.
27 T. Liu, M. Wang, Z. Shi, H. Cui, W. Dong, J. Chen, B. kermark and L.
Sun, Chem. Eur. J., 2004, 10, 4474–4479.
46 J. Hou, X. Peng, J. Liu, Y. Gao, X. Zhao, S. Gao and K. Han, Eur. J.
Inorg. Chem., 2006, 22, 4679–4686.
47 J. L. Stanley, Z. M. Heiden, T. B. Rauchfuss, S. R. Wilson, L. De Gioia
and G. Zampella, Organometallics, 2007, 26, 1907–1911.
48 A. J. Bard and L. R. Faulkner, Electrochemical methods: fundamentals
and applications, A1bazaar, 2006.
28 L. Schwartz, G. Eilers, L. Eriksson, A. Gogoll, R. Lomoth and S. Ott,
Chem. Commun., 2006, 520–522.
29 L. C. Song, J. H. Ge, X. G. Zhang, Y. Liu and Q. M. Hu, Eur. J. Inorg.
Chem., 2006, 3204–3210.
49 R. Mejia-Rodriguez, D. Chong, J. H. Reibenspies, M. P. Soriaga and
M. Y. Darensbourg, J. Am. Chem. Soc., 2004, 126, 12004–12014.
50 S. J. Borg, T. Behrsing, P. Stephen, M. Razavet, X. Liu and C. J. Pickett,
J. Am. Chem. Soc., 2004, 126, 16988–16999.
30 L. C. Song, H. T. Wang, J. H. Ge, S. Z. Mei, J. Gao, L. X. Wang, B.
Gai, L. Q. Zhao, J. Yan and Y. Z. Wang, Organometallics, 2008, 27,
1409–1416.
51 J. Armand, C. Bellec, L. Boulares, P. Chaquin, D. Masure and J. Pinson,
J. Org. Chem., 1991, 56, 4840–4845.
52 L. E. Bogan, J. Organomet. Chem., 1983, 250, 429–438.
53 D. Seyferth, G. B. Womack, R. S. Henderson, M. Cowie and B. W.
Hames, Organometallics, 1986, 5, 1568–1575.
31 W. G. Wang, H. Y. Wang, G. Si, C. H. Tung and L. Z. Wu, Dalton
Trans., 2009, 2712–2720.
32 P. Li, M. Wang, J. Pan, L. Chen, N. Wang and L. Sun, J. Inorg. Biochem.,
2008, 102, 952–959.
33 S. Jiang, J. Liu, Y. Shi, Z. Wang, B. kermark and L. Sun, Dalton Trans.,
2007, 896–902.
54 G. M. Sheldrick, SADABS, University of Go¨ttingen: Go¨ttingen,
Germany, 1996.
55 G. Sheldrick, SHELXTL. Version 5.1, Siemens Analytical X-ray
Instruments Inc., Madison, Wisconsin, USA, 1994.
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 10907–10917 | 10917
©