Long-distance electronic interaction in a molecular wire consisting of a ferrocenyl-ethynyl unit bridging two [(η5-C5H5)(dppe)M] metal centers

Article abstract of DOI:10.1016/j.jorganchem.2008.12.056

Several multinuclear ferrocenyl-ethynyl complexes of formula [(η⁵-C₅H₅)(dppe)M^II-C{triple bond, long}C-(fc)_n-C{triple bond, long}C-M^II(dppe)(η⁵-C₅H₅)] (fc =

Full text of DOI:10.1016/j.jorganchem.2008.12.056

Journal of Organometallic Chemistry 694 (2009) 1529–1541
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Long-distance electronic interaction in a molecular wire consisting of a
ferrocenyl–ethynyl unit bridging two [(g
5-C5H5)(dppe)M] metal centers
*
Teng-Yuan Dong , Shu-Fan Lin, Chiao-Pei Chen, Shu-Wen Yeh, Hsing-Yin Chen, Yuh-Sheng Wen
Department of Chemistry, National Sun Yat-Sen University, Kaohsiung 804, Taiwan
Faculty of Medicinal and Applied Chemistry, Kaohsiung Medical University, Kaohsiung 804, Taiwan
Institute of Chemistry, Academia Sinica, Nankang, Taipei, Taiwan
a r t i c l e i n f o
a b s t r a c t
5-C5H5)(dppe)MIIꢀC„C–(fc)n–C„C–
Article history:
Several multinuclear ferrocenyl–ethynyl complexes of formula [(
g
MII(dppe)( 5-C5H5)] (fc = ferrocenyl; dppe = Ph2PCH2CH2PPh2; 1: MII = Ru2+, n = 1; 2: MII = Ru2+, n = 2;
g
Received 19 November 2008
Received in revised form 19 December 2008
Accepted 24 December 2008
Available online 3 January 2009
3: MII = Ru2+, n = 3; 4: MII = Fe2+, n = 2; 5: MII = Fe2+, n = 3) were studied. Structural determinations of 2
and 4 confirm the ferrocenyl group directly linked to the ethynyl linkage which is linked to the
pseudo-octahedral [(g
5-C5H5)(dppe)M] metal center. Complexes of 1–5 undergo sequential reversible
oxidation events from 0.0 V to 1.0 V referred to the Ag/AgCl electrode in anhydrous CH2Cl2 solution
and the low-potential waves have been assigned to the end-capped metallic centers. The solid-state
and solution-state electronic configurations in the resulting oxidation products of [1]+ and [2]2+ were
characterized by IR, X-band EPR spectroscopy, and UV–Vis at room temperature and 77 K. In [1]+ and
[2]2+, broad intervalence transition band near 1600 nm is assigned to the intervalence transition involv-
ing photo-induced electron transfer between the Ru3+ and Fe2+ metal centers, indicating the existence of
strong metal-to-metal interaction. Application of Hush’s theoretical analysis of intervalence transition
band to determine the nature and magnitude of the electronic coupling between the metal sites in com-
plexes [1]+ and [2]2+ is also reported. Computational calculations reveal that the ferrocenyl–ethynyl-
Keywords:
Molecular wire
Redox systems
Alkynyl ligands
Ferrocene
Bridging ligands
based orbitals do mix significantly with the (g
5-C5H5)(dppe)Ru metallic orbitals. It clearly appears from
this work that the ferrocenyl–ethynyl spacers strongly contribute in propagating electron delocalization.
Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction
metal center and
important role in determining the magnitude of electronic cou-
pling. The magnitude of electrochemical data E1/2 has been taken
as an indication of the magnitude of electronic coupling between
two redox metal centers, where E1/2 is the difference in redox
potentials associated with the first and second oxidation of the me-
tal centers. The electrochemical response of binuclear transition
metal complexes bridged with polyynes and related ligands de-
p-conjugated organic spacer plays significantly
The insight provided by the rich chemistry of bridged mixed-va-
lence binuclear complexes has promoted a great deal of both
experimental and theoretical studies. Considerable work has been
focused on the use of mixed-valence complexes to probe electronic
coupling between well-separated metal sites [1–3]. Recently, the
study of homo- and hetero-metallic binuclear transition metal
complexes in which the end-capped metal centers are connected
D
D
rived from polyynes (
2,5-thiophene, and 2,5-pyridine) has been comprehensively stud-
ied. The compounds [( -C„C)n (n = 2,
5-C5Me5)Re(NO)(PPh3)]2(
l-C„C–X–C„C, X = C„CC„C, 1,4-C6H4,
by p-conjugated organic linear spacers has been an intriguing area
of research, since such system may provide the possibility to study
the electronic coupling between the redox-active end-capped me-
tal centers, or propose as models for molecular wires. In this con-
text, end-capping of unsaturated organic spacers with various
redox-active groups, such as ferrocene, ruthenium(II) polypyridine
[4–9], [RuL2Cl] (L2 = 2PPh3, Ph2PCH2CH2PPh2 (dppe)) [10,11], and
g
l
3 and 4) [21–24] display reversible electrochemical response, the
cyclic voltammogram (CV) of these compounds being character-
ized by two one-electron redox waves. Furthermore, the decreas-
ing of the
DE1/2 value from 0.53 V in n = 2, 0.38 V in n = 3 to
0.28 V in n = 4 indicates that the magnitude of Re2+–Re2+ interac-
tion is decreased significantly on the number of C„C moiety. Fur-
thermore, complexes with an odd numbered carbon linear or cyclic
bridge has been recently found to promote very efficient electronic
communication between two [RuCl(dppe)2]+ metal centers [25].
Very recently, we have described the electrochemical and
photophysical properties for a series of complexes containing
[M(g
5-C5H5)L2] (M = Fe(II), Ru(II), and Os(II)) metal centers [12–
21], have been most studied where they are intended to promote
the long-range electronic coupling. The nature of the end-capped
* Corresponding author. Tel.: +886 7 5253937; fax: +886 7 525 3908.
E-mail address: dty@mail.nsysu.edu.tw (T.-Y. Dong).
0022-328X/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2008.12.056

1530
T.-Y. Dong et al. / Journal of Organometallic Chemistry 694 (2009) 1529–1541
bis-(2,20:60,200-terpyridyl)polyferrocene redox-active spacers end-
capped with photoactive and redox-active Ru2+-terpyridine
terminals ([(tpy)RuII-(tpy-(fc)n-tpy)-RuII(tpy)]4+ (tpy = terpyridine,
fc = ferrocenyl, n = 1–3) and [(tpy)RuII-(tpy-C„C–(fc)n–C„C-tpy)-
RuII(tpy)]4+ (n = 2–3)) to study the electronic interaction between
the end-capped Ru2+ metal centers [26–28]. In our previous papers,
the CV electrochemical measurements for these binuclear Ru2+
complexes were dominated by the Ru2+/Ru3+ redox couple (E1/2
from 1.35 to 1.38 V), Fe2+/Fe3+ redox couples (E1/2 from 0.4 to
1.0 V) and tpy/tpyꢀ/tpy2ꢀ redox couples (E1/2 from ꢀ1.3 to
ꢀ1.5 V). Note worthily, a single irreversible wave was found for
the Ru2+/Ru3+ redox couple. We suggested that the electronic cou-
pling between the two Ru2+ centers is relatively weak. In attempt-
ing to modulate the electronic coupling between the two terminal
Ru2+ centers by manipulation of the energetic of the end-capped
metal centers and the connecting spacer, we now describe the
electrochemical and photophysical properties of 1–7 (Schemes 1
and 2). The preparation and X-ray crystallographic structure anal-
ysis of 1 were previously reported by Bruce [19]. Quasi-reversible
and irreversible waves were observed in the cyclic voltammetry
(CV) measurements of complexes 1. Our design principle for
wire-like molecules 1–5 is that the redox-active ferrocenyl–ethy-
nyl spacer can enhance the capability of transfer information along
the stepwise molecular axis through stronger
r –
-bonding of M2+
(CH3)3Si
Cl Ru
+
C„C in comparison with tpy-ferrocenyl-tpy spacer. Furthermore,
it has been found that the 18-electron ferrocene complex can be
easily oxidized to form a stable 17-electron ferrocenium complex.
As shown in Scheme 3, rapid intramolecular electron transfer be-
tween the two ruthenium centers in mixed-valence diruthenium
complex could occur through the ferrocenyl spacer. Thus, the ferr-
ocenyl spacer that can provide an effective hetero-nuclear electron
delocalization along main chain can serve as model system for
molecular wire.
Fe
P
P
Si(CH3)3
n
8 n = 1
9 n = 2
10 n = 3
MeOH KF/KPF6
2. Results and discussion
P
2.1. Synthesis and characterization
P
Ru
Binuclear complexes of 1–7 were prepared by reacting ferro-
cenylethynyl spacers (8–12) with stoichiometric amount of
Fe
Ru
(g
5-C5H5)(dppe)MCl. Complexes were purified by column chroma-
n
P
P
tography and recrystallization. All complexes were fully character-
ized by 1D and 2D NMR techniques, IR, MS and elemental analysis.
Structures of complexes 1–5 were characterized by 31P NMR show-
ing a singlet for the four equivalent phosphorus nuclei at ꢁ87 ppm
1 n = 1
2 n = 2
3 n = 3
in the [Ru(g
5-C5H5)L2] end-capped complexes (1–3) and at
ꢁ107 ppm in the [Fe(
g
5-C5H5)L2] end-capped complexes (4–5).
The ethynyl carbons in 1–3 were identified by 13C NMR showing
a singlet at ꢁ107 ppm for the carbon bonded to the ferrocenyl
Scheme 1. Synthesis of the discussed Ru2+ complexes with ferrocenyl–ethynyl
spacers.
H
Fe Cl
Fe
P
P
H
+
n
11 n=2
12 n=3
1. KPF6
MeOH/THF
2. t-BuOK
Fe
P
Fe
P
P
Fe
P
P
Fe
Fe
P
+
H
n
n
6 n=2
7 n=3
4 n=2
5 n=3
Scheme 2. Synthesis of the discussed Fe2+ complexes with ferrocenyl–ethynyl spacers.

T.-Y. Dong et al. / Journal of Organometallic Chemistry 694 (2009) 1529–1541
1531
M3+
P
Fe2+
P
M 2+
P
P
n
π*
(
)
e-
e-
M3+
M2+
Fe2+
+
π (
)
Vs
P
P
2+
Ru3+
Ru
unsat. org. spacer
P
P
π*
(
)
e-
e-
Ru2+
Ru3+
Scheme 3. Intramolecular electron transfer between the two ruthenium centers in a mixed-valence diruthenium complex through ferrocenyl spacer or through unsaturated
organic spacer.
moiety and a triplet ꢁ106 ppm for the carbon bonded to the Ru2+
2.233(2); Ru1–C14, 2.247(2); Ru1–C15, 2.244(2)) are also
within previously observed ranges. Based on the angles of P2–
Ru1–P1 (83.63(2)°), P1–Ru1–C41 (82.57(5)°), and P2–Ru1–C41
(83.10(5)°), the Ru2+ metal center has pseudo-octahedral
geometry.
The biferrocenyl moiety exists in a trans conformation with the
two iron ions on opposite sides of the fulvalenide ligand. The
molecular structure of 2 can be described as step-like of ferrocenyl
moieties. The refinement of the structure imposed an inversion
center. Thus, the two Ru and Fe centers are crystallographically
equivalent and the two Ru-ethynyl axes are parallel. The dihedral
metal (2JPC = ꢁ25 Hz). In the case of [Fe(
g
5-C5H5)L2] end-capped
complexes, the ethynyl carbons were also identified by 13C NMR
showing a singlet at ꢁ116 ppm for the carbon bonded to the ferr-
ocenyl moiety and a triplet ꢁ114 ppm for the carbon bonded to the
Fe2+ metal (2JPC = ꢁ42 Hz). The IR spectroscopic data of 1–3 also
showed the characteristic of triple bond vibration from 2079 to
2089 cmꢀ1. In the case of 4 and 5, a red-shift of C„C vibration
was observed at 2067 cmꢀ1
. Selected spectroscopic data of
compounds 1–5 are given in Table 1.
0
2.2. Molecular structure of complex 2
angle of Ru–C6–C1–C1 (125.4°) indicates that complex 2 does not
have linear molecular wire geometry as shown in Fig. 1. The two
least-squares-fitting Cp planes in a given ferrocenyl moiety are
nearly parallel, and the dihedral angle is 0.86(8)°. Inspection
of the average distances of Fe–C (2.048(3) Å) and Fe–Cp
(1.6549(2) Å) indicates that the metallocenes are in Fe2+ oxidation
state [30,31]. The bond distances and angles about the Cp rings
vary little and they are close to those reported for analogous ferro-
cenes. Furthermore, the two Cp rings associated with Fe center are
nearly eclipsed, with an average staggering angle of 18.34°. Attach-
ment of ethynyl moiety to the biferrocene has minimal influence
The X-ray crystallographic structure analysis of 2 showed that
its space group was P21/c at 120 K. As shown in Fig. 1, the ORTEP
view confirms the molecular structure with the ferrocenyl group
directly linked to the ethynyl linkage which is linked to the
[(
(dppe)Ru] fragment is similar to those examples reported previ-
ously, such as complex of [(
5-C5H5)(dppe)Ru(C„Cfc)] [29].
g g
5-C5H5)(dppe)Ru] metal center. Geometry of the [( 5-C5H5)-
g
Distances of Ru–P (Ru1–P1, 2.2449(5); Ru1–P2, 2.2611(5)) and
Ru1–Ccp (Ru1–C11, 2.236(2); Ru1–C12, 2.227(2); Ru1–C13,

1532
T.-Y. Dong et al. / Journal of Organometallic Chemistry 694 (2009) 1529–1541
Table 1
on the molecular structure in comparison with analogous
biferrocene.
Selected spectroscopic data and important bond distances and angles of compounds
1–5.
The C40–C41 triple bond distance (1.203(2) Å) in ethynyl moi-
ety is slightly longer than that in [(g g
5-C5H4)Fe( 5-C5H4C„CH)]2
Compound
1
2
3
4
5
a
(C„C stretching, cmꢀ1
2+–C„ (13C NMR)b
)
2089
2080
106.55
107.36
87.53
1.203(2)
2.018(2)
2.2449(5)
2.2611(5)
2.237(2)
2.048(3)
1.6549(2)
0.86
2079
2067
2067
(1.176(5) Å) [32], indicating the existence of
p characteristic be-
m
M
tween the Ru2+ center and the ethynyl moiety. Furthermore, the
105.18
107.93
87.10
107.06
107.44
87.11
113.75
115.82
107.13
1.226(9)
1.883(9)
2.153(2)
2.133(2)
2.083(7)
2.040(7)
1.648(1)
0.67
114.00
115.72
107.28
fc–C„ (13C NMR)b
triple bond distance in 2 is 0.047 Å shorter than that in 1 [19], indi-
31P NMRb
cating an increasing of
p
characteristic between the Ru2+ center
C„C (Å)
1.25(2)c
2.00(1)c
2.259(3)
2.254(3)
2.25(2)
2.05(2)c
1.64c
2+–C„ (Å)
and the ethynyl moiety in 1. The Ru1–C41 (2.018(2) Å) distance
involving the ethynyl moiety in 2 is slightly longer than that in 1
M
M
2+–P1
M2+–P2
(2.00(1) Å). There have been reported that the
Ru–C distances are 2.214(5) Å in [(
5-C5H5)(PPh3)(PMe3)Ru–
CH2(CH(CH3)2)] and 2.186(9) Å in [(
5-C5H5)(PMe3)2Ru-CH2CH2-
IrCl( characteristic was also evidenced
5-C5H5)2] [33,34]. The
r-bonding
M
2+–C(Cp) (av. Å)d
g
Fe–C (av. Å)e
g
Fe–Cp (centroids, av. Å)f
Cp–Cp dihedral angle (°)g
Cp–Cp stagger angle (°)h
0.0c
g
p
by the red-shift of the triple bond IR stretch from 2149 cmꢀ1 in
spacer 9 to 2080 cmꢀ1 in Ru2+ complex 2. The averaged Ru1–
C(Cp) distances in 2, e.g., Ru1–C(11–15) between 2.227(2) and
2.247(2) Å, is 2.237(2) Å which is similar to that in 1 (2.25(2) Å).
As given in Table 1, a direct comparison of 2 with 1 was made.
There is little difference between 1 and 2 in ferrocenyl moiety
and Ru2+-dppe ligand environment.
18.34
7.64
M
2+–C„C (°)
176(1)c
84.0(1)
90.7(4)
80.6(3)
13.48c
178.6(2)
83.63(2)
82.57(5)
83.10(5)
19.07
178.4(7)
87.16(8)
85.3(2)
85.4(2)
18.82
P1–M2+–P2 (°)
P1–M2+–C„
P2–M2+–C„
M
2+/M2+ distance (Å)i
24.64j
24.40j
a
b
c
d
e
f
KBr mull.
In C6D6.
From Ref. [19]. The structure analysis of 1 was carried out at room temperature.
Average M2+–C(Cp) distance.
2.3. Molecular structure of complex 4 ꢂ C6H6
Average Fe–C distance for each ferrocenyl moiety.
Distance from the Fe atom to the center of mass of the Cp ring in each ferrocenyl
The X-ray crystallographic structure analysis of 4 ꢂ C6H6
showed that its space group was P21/n at 150 K. As shown in
Fig. 2, the ORTEP view confirms the molecular structure is similar
to that of 2 described previously. Distances of Fe2–P (Fe2–P1,
2.153(2); Fe2–P2, 2.133(2)) and Fe2–Ccp (Fe2–C11, 2.093(7); Fe2–
C12, 2.085(7); Fe2–C13, 2.075(7); Fe2–C14, 2.077(7); Fe2–C15,
moiety.
g
Dihedral angle between the two least-squares-fitting Cp ring in each ferrocenyl
moiety.
h
Average stagger angle between the two Cp rings in each ferrocenyl moiety.
i
j
The M2+–M2+ metal–metal distances (M2+–C6–Fe2+–C1–C10–Fe2+–C60–M2+).
For complexes 3 and 5, the M2+–M2+ metal–metal distance was estimated from
theoretical geometry generated by optimization.
2.087(7)) are similar to those of [(g
5-C5(CH3)5)(dppe)-
Ru
Ru
C
º
125.4
Fig. 1. ORTEP drawing of 2 with the atom-numbering scheme. The determination was carried out at 120.0(1) K. Selected bond distances (Å) and angles (°): Fe–C1, 2.061(2);
Fe–C2, 2.050(2); Fe–C3, 2.045(2); Fe–C4, 2.042(2); Fe–C5, 2.039(2); Fe–C6, 2.079(2); Fe–C7, 2.047(2); Fe–C8, 2.032(2); Fe–C9, 2.034(2); Fe–C10, 2.049(2); Ru1–C11, 2.236(2);
Ru1–C12, 2.227(2); Ru1–C13, 2.233(2); Ru1–C14, 2.247(2); Ru1–C15, 2.244(2); Ru1–C41, 2.018(2); Ru1–P1, 2.2449(5); Ru1–P2, 2.2611(5); C1–C10, 1.460(4); C6–C40,
1.442(2); C40–C41, 1.203(2); P2–Ru1–P1, 83.63(2); P1–Ru1–C41, 82.57(5); P2–Ru1–C41, 83.10(5); Ru1–C41–C40, 178.6(2); C41–C40–C6, 177.3(2).

T.-Y. Dong et al. / Journal of Organometallic Chemistry 694 (2009) 1529–1541
1533
Fe
º
80.2
Fe
Fig. 2. ORTEP drawing of 4 with the atom-numbering scheme. The determination was carried out at 150.0(1) K. Selected bond distances (Å) and angles (°): Fe1–C1, 2.053(7);
Fe1–C2, 2.026(7); Fe1–C3, 2.027(6); Fe1–C4, 2.048(7); Fe1–C5, 2.037(7); Fe1–C6, 2.072(7); Fe1–C7, 2.035(6); Fe1–C8, 2.033(6); Fe1–C9, 2.026(7); Fe1–C10, 2.041(7); Fe2–
C11, 2.093(7); Fe2–C12, 2.085(7); Fe2–C13, 2.075(7); Fe2–C14, 2.077(7); Fe2–C15, 2.087(7); Fe2–C17, 1.883(9); Fe2–P1, 2.153(2); Fe2–P2, 2.133(2); C1–C10, 1.47(1); C6–C16,
1.441(10); C16–C17, 1.226(9); P2–Fe2–P1, 87.16(8); P1–Fe2–C17, 85.3(2); P2–Fe2–C17, 85.4(2); Fe2–C17–C16, 178.4(7); C17–C16–C6, 175.2(7).
Fe(C„CfcC„C)Fe(dppe)(
g
5-C5(CH3)5)] reported previously [19].
2) and reversible ferrocene-based Fe2+/Fe3+ redox couples (E1/2 at
0.86 V in 1; 0.73 and 0.87 V in 2) vs. Ag/AgCl. For complex 3,
one reversible Ru2+/Ru3+ redox couple (E1/2 at 0.14 V) and three
Based on the angles of P2–Fe2–P1 (87.16(8)°), P1–Fe2–C17
(85.3(2)°), and P2–Fe2–C17 (85.4(2)°), the end-capped Fe2+ metal
center has pseudo-octahedral geometry.
Table 2
DPV data of 1–10 with scan rate of 20 mV sꢀ1 a
.
The biferrocenyl moiety also exists in a trans conformation. The
refinement of the structure imposed an inversion center. Thus, the
Compound M2+/3+
Fe2+/3+
two end-capped [Fe(g
5-C5H5)L2] centers are crystallographically
E1/2 (V)
D
E1/2 (V) Kc
E1/2 (V)
DE1/2 (V) Kc
equivalent and the two Fe-ethynyl axes are parallel. The dihedral
angle of Fe–C6–C1–C01 (80.20°) indicates that complex 4 does not
have linear molecular wire geometry as shown in Fig. 2. The two
least-squares-fitting Cp planes in a given ferrocenyl moiety are
nearly parallel, and the dihedral angle is 0.67°. Inspection of the
average distances of Fe1–C (2.040(7) Å) and Fe1–Cp (1.648(1) Å)
indicates that the metallocenes are also in Fe2+ oxidation state. Fur-
thermore, the two Cp rings associated with Fe center are nearly
eclipsed, with an average staggering angle of 7.64°.
1
2
3
0.05
0.57
0.52
0.17
0.00
6.51 ꢃ 108 0.86
0.10
0.27
7.61 ꢃ 102 0.73
0.14
2.36 ꢃ 102
0.87
0.14
ꢁ4b
0.62
0.83
0.96
0.21
0.13
3.63 ꢃ 103
1.60 ꢃ 102
4
5
0.04
0.04
0.00
0.00
ꢁ4b
ꢁ4b
0.60
0.89
0.29
8.23 ꢃ 104
The C16–C17 triple bond distance (1.226(9) Å) in ethynyl moi-
0.47
0.75
1.02
0.28
0.27
5.57 ꢃ 104
3.77 ꢃ 104
ety is slightly longer than that in 2 (1.203(2) Å), indicating the exis-
tence of more
metallic center and the ethynyl moiety. The more
p
characteristic between the [Fe(
characteristic
g
5-C5H5)L2]
p
6
7
ꢀ0.03
ꢀ0.04
0.54
0.86
0.32
2.65 ꢃ 105
was also evidenced by the red-shift of the triple bond IR stretch
from ꢁ2085 cmꢀ1 in 1–3 to 2067 cmꢀ1 in 4–5. As given in Table
1, a direct comparison of 2 with 4 was also made.
0.38
0.72
0.98
0.34
0.26
5.79 ꢃ 105
2.55 ꢃ 104
2.4. Electrochemical measurements of complexes 1–7
8
9
0.81
0.60
0.97
0.37
1.87 ꢃ 106
One of the interesting attributes of 1–5 is the magnitude of the
interaction between the two terminal metallic sites. Electrochem-
ical voltammetry affords a simple and effective way for estimating
this interaction. In this study, DPV measurements of 1–5 were car-
ried out in anhydrous CH2Cl2 solution. The electrochemical results
for 1–5 and the spacers are given in Table 2. As shown in Fig. 3, the
10
0.42
0.78
1.07
0.36
0.29
1.26 ꢃ 106
8.23 ꢃ 104
a
E1/2: all half-wave potentials are referred to the Ag/AgCl electrode in anhydrous
CH2Cl2 solution. E1/2: the difference of E1/2 between two redox waves of metal
centers. Kc: disproportionation constant Kc = 10
D
redox behavior in 1 and 2 is dominated by the two reversible Ru2+
Ru3+ redox couples (E1/2 at 0.05 and 0.57 V in 1; 0.10 and 0.27 V in
/
D
E(1/2)/0.059
.
b
In the statistical limit.

1534
T.-Y. Dong et al. / Journal of Organometallic Chemistry 694 (2009) 1529–1541
In the CV measurement of complex 1, two reversible waves at
ꢀ0.03 and 0.50 V and one irreversible wave at 0.81 V vs. Ag wire
pseudo-reference electrode were observed by Bruce and his
coworkers and the assignment of the waves to the redox reaction
of a particular metal center could not be definitely accomplished.
We suggest that the low-potential wave(s) in 1–5 could be as-
signed to the end-capped metallic centers. In our study, the 2:1:1
integrated area-ratio of the redox couples at half-potential (E1/2
)
5μA
of 0.04, 0.60 and 0.89 V in the DPV of 4 and the 2:1:1:1 integrated
area-ratio of the redox couples at E1/2 of 0.04, 0.47, 0.75, and 1.02 V
in the DPV of 5 suggest that the low-potential wave could be as-
signed to the end-capped metallic centers. Making a comparison
with monocapped complexes of 6 and 7, this assignment of the
1
Fe2+/Fe3+ redox couples for the end-capped [( 5-C5H5)(dppe)Fe]
g
moieties at low-potential wave in 4–5 can be further supported.
As shown in Fig. 4, the integrated area-ratio of the redox couples
at half-potential (E1/2) of ꢀ0.03, 0.54 and 0.86 V in the DPV of
monocapped complex of 6 was varied from 1:1:1 in 6 to 2:1:1 in
4. The integrated area-ratio was varied from 2:1:1:1 in 5 to
1:1:1:1 in 7. Furthermore, making a comparison of the redox
potentials with the corresponding the free spacer also suggest that
the low-potential wave could be assigned to the end-capped
metallic centers. This assignment is not in agreement with the
solid-state Mössbauer observation for one-electron-oxidized
2
complex [(
g
5-C5H5)(dppe)Ru–C„C–fc]+. The DPV measurements
3
1.2
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
Potential (V vs Ag/AgCl)
Fig. 3. Electrochemical measurements of 1–3 in CH2Cl2 at room temperature vs.
4μΑ
4
Ag+/AgCl.
reversible ferrocene-based Fe2+/Fe3+ redox couples (E1/2 at 0.62,
0.83, 0.96 V) vs. Ag/AgCl were observed.
Quasi-reversible and irreversible waves were observed in the
CV measurements of complexes 1 [20], [(
C„C–Ph] [35], [(
5-C5H5)(PPh3)2Ru–C„C–fc] (fc = ferrocenyl)
[30], and [(
5-C5H5)(dppe)Ru–C„C–fc]. Therefore, differential
g
5-C5H5)(PPh3)2Ru–
g
g
5
6
7
pulse voltammetry was employed to obtain better-resolved poten-
tial information in our study, because the Fe2+/Fe3+ redox process
for 2 and 3 was quasi-reversible in the CV experiment. In our CV
voltammograms of 4–7, reversible Ru2+/Ru3+ and Fe2+/Fe3+ redox
couples were observed in these complexes. In the case of
[(g =
5-C5H5)(dppe)Ru–C„C–fc], a quasi-reversible wave at E1/2
ꢀ0.16 V and a irreversible wave at Epa = 0.48 V vs. Ag/AgCl were
observed in Sato0s previous report [29]. An additional cathodic
wave was observed at Epc = 0.35 V, but this wave was small or van-
ishing when the scanning turned back. Furthermore, it was unam-
biguously assigned from the solid-state Mössbauer result that the
low-potential wave was assigned to the Fe2+/Fe3+ redox couple of
the ferrocenyl moiety. The solid-state Mössbauer spectrum of
one-electron-oxidized complex [(g
5-C5H5)(P(Ph)3)2Ru–C„C–fc]+
showed one doublet with quadrupole splitting value of
0.77 mm sꢀ1 and isomer shift value of 0.54 mm sꢀ1 at 78 K. The
smaller quadrupole splitting value in comparison with neutral
[(
was assigned to the iron metal center in ferrocenyl moiety.
1.2
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
Potential (V, vsAg / AgCl)
g
5-C5H5)(dppe)Ru–C„C–fc] suggested that the oxidized site
Fig. 4. Electrochemical measurements of 4–5 in CH2Cl2 at room temperature vs.
Ag+/AgCl.

T.-Y. Dong et al. / Journal of Organometallic Chemistry 694 (2009) 1529–1541
1535
of 1–5 demonstrate that the
r
and
p
bonding characters of the
our [2]+ system (Ru2+–Ru2+ distance = 19.07 Å) would be with
C14-alkynyl systems [M–(C„C)7ꢀM]. The synthesis and molecular
end-capped metal centers play significantly important role in
determining the magnitude of electronic coupling. By replacing
structure of C14-alkynyl with electron rich (g
5-C5Me5)(dppe)RuII
the (g
5-C5H5)(dppe)Ru center to (
g
5-C5H5)(dppe)Fe center, a sin-
metal center have been reported. However, the electrochemical
measurement has not been reported in this C14-alkynyl complex
with Ru2+–Ru2+ distance of 20.560(5) Å [38]. Thus, ferrocenyl–
ethynyl spacers appear to be promising spacers which can ensure
stronger coupling between two Ru2+ metal centers. Furthermore,
gle reversible wave was found for the Fe2+/Fe3+ redox couple of
end-capped metal centers in 4 and 5, indicating relatively weak
electronic coupling between the two end-capped metallic centers.
Finally, the assignment of the Ru2+/Ru3+ redox couples in 1–3 at
high-potential waves can be further excluded, because the metal-
to-metal interaction between the two Ru centers should be inver-
there is a correlation of the
DE1/2 values with the C„C stretching
values or with the 13C NMR chemical shifts of Ru2+–C„ in 1–5.
sely proportional to the distance between them. Hence, the
D
E1/2
From Table 1, the DE1/2 value decreases as the C„C stretching va-
value, giving an indication of the interaction between the two me-
tal sites, should decrease fairly with distance. The assignment of
the wave to a particular metal center in complexes 1–3 could also
be accomplished by comparison with the redox potentials of the
lue decreases and the 13C NMR chemical shift of Ru2+–C„
increases.
2þ
½Ru—CBC—ðfcÞn—CBC—Ruꢄ þ ½Ru—CBC—ðfcÞn—CBC—Ruꢄ
g
5-C5H5)(PPh3)2Ru–C„C–Ph]
Kc
ꢀ 2½Ru—CBC—ðfcÞn—CBC—Ruꢄþ
ð1Þ
analogous complexes. Complex [(
afforded one redox wave in a similar potential region (E1/2
0.535 V vs. SCE). Furthermore, complex [(
5-C5H5)(dppe)Ru–
=
g
C„C–Ph] afforded one quasi-reversible redox wave in a similar
Eð1=2Þ=0:059
Kc ¼ 10D
:
potential region (E1/2 = 0.50 V in CH2Cl2 vs. SCE). In the case of
g
5-C5H5)(dppe)RuII–C„C–
Another interesting attributes of 1–5 is the magnitude of the
electronic interaction between the Fe sites. As shown in Figs. 3
and 4, complexes 1–5 show reversible oxidation processes on
sweeping at anodic potentials between ꢁ0.6 V and ꢁ1.0 V, corre-
sponding to the oxidation of the ferrocenyl moieties. When the re-
dox behavior occurred at the ferrocenyl site, the ruthenium metal
center should be in the Ru3+ oxidation state. We found that there
was appreciable variation detected in the potential associated with
each Fe2+/Fe3+ redox couple. Furthermore, the variation of the
unsaturated organic spacers, complex (
C„C–RuII(dppe)( 5-C5H5) shows two reversible Ru2+/Ru3+ redox
g
couples at E1/2 of ꢀ0.24 and 0.35 V vs. SCE [36].
Noticeably, pronounced Ru2+–Ru2+ metal-to-metal interaction
is evidenced directly from the observation of two reversible Ru2+
/
Ru3+ redox couples in 1 and 2. Making a comparison of the
DE1/2
values of 1–2 with those of [(tpy)RuII–(tpy-C„C–(fc)n–C„C–
tpy)–RuII(tpy)]4+ indicates that the interactions between the two
Ru2+ sites in 1–2 are all pronouncedly larger. Furthermore, the
D
E1/2 value for the Fe2+ 3+
/ redox couples (0.37 V in the free spacer
decreasing of the
DE1/2 value from 0.52 V in 1, 0.17 V in 2 to
9, 0.14 V in its Ru2+ complex 2 and 0.29 V in its Fe2+ complex 4;
0.36 and 0.29 V in the free spacer 10, 0.21 and 0.13 V in its Ru2+
complex 3 and 0.28 and 0.27 V in its Fe2+ complex 5) strongly sug-
gest that there is an interaction between the iron and the ruthe-
nium metal centers. The decreasing of Thus, ferrocenyl–ethynyl
spacers appear to be promising spacers which can ensure stronger
coupling between two Ru2+ metal centers. Furthermore, there is a
0.0 V in 3 indicates that the magnitude of Ru2+–Ru2+ interaction
is changed significantly on the number of ferrocenyl moieties.
The smaller value of
Ru2+ interaction. As shown in Table 1, we have found that there
is a correlation between the
E1/2 values and the Ru2+–Ru2+
metal–metal distances (13.48 Å in 1, 19.07 Å in 2, and 24.64 Å in
3). Metal-to-metal electronic interaction decreases as the Ru2+
D –
E1/2 gives an indication of smaller Ru2+
D
–
Ru2+ distance increases. The Ru2+–Ru2+ metal–metal distances
(Ru2+–Ccp–Fe2+–Ccp–Ru2+) in 1 and 2 could be directly calculated
from the molecular structure determination. For complex 3, the
Ru2+–Ru2+ metal–metal distance of 24.64 Å was calculated from
the theoretical geometry of 3 generated from the X-ray molecular
structure of 2 by setting an inversion center at the Fe center for the
correlation of the
with the 13C NMR chemical shifts of Ru2+–C„ in 1–5. From Table
1, the E1/2 value decreases as the C„C stretching value decreases
and the 13C NMR chemical shift of Ru2+–C„ increases value for the
Fe
2+ 3+ redox couples in 1–5 indicates that the magnitude of Fe–Fe
interaction is changed pronouncedly on the coordination of Ru3+
ion. The smaller value of E1/2 gives an indication of smaller Fe–
DE1/2 values with the C„C stretching values or
D
/
fragment of (g g
5-C5H5)(dppe)RuII–C„C–fc–( 5-C5H4)Fe.
D
As given in Table 2, the comproportionation constants Kc de-
rived from Eq. (1) for the Ru2+/Ru3+ redox couples are calculated
to be 6.51 ꢃ 108 in the mixed-valence [1]+, 7.61 ꢃ 102 in [2]+,
and very small value (ꢁ4) for [3]+. With increasing Ru–Ru metal
distance on going from [1]+ to [3]+ the thermal stability of the
mixed-valence complex decreases sharply to ꢁ4 for [3]+. Better
Fe interaction. Due to the electronic coupling between the iron
and the ruthenium metal centers, the hetero-nuclear electron
transfer process between the Fe2+ and Ru3+ nuclei is possible. As
an example, on the oxidation to the monocation [2]+, more than
one oxidation isomer can exist. In monocation [2]+, an equilibrium
can exist between the two energetically nonequivalent isomers
3+
2+
Rua3+–Fea2+–Feb2+–Rub2+ M Rua2+–Fea2+–Feb2+–Rub
and Rua –
electronic coupling was observed in C4-alkynyl (g
5-C5H5)(dppe)R-
Fea3+–Feb2+–Rub2+ M Rua2+–Fea2+–Feb3+–Rub2+. The distribution be-
tween the two isomers depends on the zero-point energy differ-
ence between them. Energetically, the energy surface for the two
uII–(C„C)2–RuII(dppe)(
g
5-C5H5) (E1/2 at ꢀ0.24 and 0.35 V vs. SCE;
Kc = 1.0 ꢃ 1010) with Ru2+–Ru2+ distance of 7.74 Å [36]. Replacing
the (g -
5-C5H5)(dppe)RuII metal center by more electron rich (g5
2+
2+
equivalent isomers Rua3+–Fea2+–Feb2+–Rub
and Rua2+–Fea
–
C5Me5)(dppe)FeII center exhibits higher Kc value (1.6 ꢃ 1012). The
electron rich C5Me5 ligand enhanced markedly the stability of
the mixed-valence state. An adequate comparison of our [1]+
system would be with C8-alkynyl systems [M–(C„C)4–M]
(metal–metal distance ꢁ13 Å) [37]. System of C8-alkynyl with
Feb2+–Rub is more stable than that for the 2 equiv. isomers
3+
Rua2+–Fea3+–Feb2+–Rub and Rua2+–Fea2+–Feb3+–Rub2+. In dication
2+
3+
2+
[2]2+
, , –
oxidation isomers Rua3+–Fea2+–Feb2+–Rub Rua3+–Fea
Feb3+–Rub
,
Rua3+–Fea3+–Feb2+–Rub
,
Rua2+–Fea3+–Feb2+–Rub
,
2+
2+
3+
Rua2+–Fea2+–Feb3+–Rub3+, and Rua2+–Fea3+–Feb3+–Rub could be in
2+
(g
5-C5H5)(dppe)RuII metal center has not been reported. Both
C4- and C8-alkynyl systems have been studied for M = [(g
5-C5-
a state of equilibrium.
Me5)Re(NO)(PPh3)]. The study has revealed that Kc value decreases
2.5. IR, EPR, and UV spectra of oxidation complexes [1]+ and [2]2+
dramatically from C4
(D
E1/2: 0.53 V; Kc: 9.62 ꢃ 108) to C8
(
D
E1/2
:
0.28 V; Kc: 5.57 ꢃ 104). In the case of electron rich (
g
5-C5Me5)-
(dppe)FeII metal center, Kc value also decreases dramatically from
C4 (1.6 ꢃ 1012) to C8 (2.0 ꢃ 107). Another adequate comparison of
Thermodynamic stability of the oxidized forms with respect
to the disproportionation constants was matched and chemical

1536
T.-Y. Dong et al. / Journal of Organometallic Chemistry 694 (2009) 1529–1541
oxidation with ferrocenium PF6 allowed isolation of monooxidized
As shown in Fig. 6, an axial-type EPR spectrum (g = 2.92; g\ =
k
ꢀ
complex of [1]+ and dioxidized complex of [2]2+ as PF6 salts in
1.95) at 77 K was found for [1]+ which did not show any hyperfine
splitting. EPR data of (
5-C5H5)(dppe)RuII–(C„C)n–RuII(dppe)(g5
essentially quantitative yield. Due to the smaller
D
E1/2 and Kc val-
g
-
ues, pure monooxidized complex of [2]+ could not be synthesized
successfully by chemical oxidation. To understand which metal
site was oxidized, the solid-state and solution-state electronic
configurations in multinuclear complexes of [1]+ and [2]2+ were
characterized by IR, X-band EPR spectroscopy, and UV–Vis at room
temperature and 77 K.
C5H5) or comparable ruthenium complexes with same geometry
have not been reported so far. For a mononuclear ferrocenium tri-
iodide with E2g electronic configuration, an axial-type spectrum
was observed with g = 4.35; g\ = 1.26 at 20 K but cannot be seen
k
at room temperature [40]. In the case of a binuclear mixed-valence
biferrocenium cation, the value of g tensor anisotropy (
D
g = g ꢀ
k
IR spectroscopy has proven to be useful to tell which metal site
is oxidized. When Fe2+ metallocene is oxidized to Fe3+ metallocene,
there is a dramatic change in the IR spectrum. The perpendicular
C–H bending band of the ferrocenyl moiety is the best diagnosis
of the oxidation state. In solid-state, this band is seen at
815 cmꢀ1 for ferrocene and at 851 cmꢀ1 for ferrocenium triiodide
g\) is considerably reduced and this is a reflection of admixture
of the S = 0 Fe2+ A1g electronic configuration into the ground state.
It is, therefore, probably reasonable that the oxidized site in [1]+ in
solid-state at 77 K could be assigned to the ferrocenium Fe3+ metal
center. Here, we would suggest that reduced
Dg value results from
the extensive mixing with the Ru2+ metal orbital into the Fe3+
ground state, as found in electrochemical measurement and
molecular calculation. Under this circumstance the iron metal cen-
ter loses some degree of their Fe3+ character. In other words, the
admixture of the S = 0 Fe2+ electronic configuration has to be in-
cluded. In the case of solid-state [2]2+, an axial-type EPR spectrum
[39]. Infrared spectra were run for KBr pellets for [1]+ and [2]2+
For the perpendicular C–H bending region there are relatively
strong bands appeared at 838 and 840 cmꢀ1 for [1]+ and [2]2+
.
,
respectively. This observation suggests that the Fe metallic center
in solid-state complexes of [1]+ and [2]2+ is in Fe3+ oxidation state.
Upon oxꢀid1ation of 1 to [1]+, the stretching vibration of C„C at
(g = 2.89; g\ = 2.04) was also found at 77 K, indicating electron
k
2089 cm
in
1
disappears and
a
strong band appears at
localization at the Fe3+ metal center on the EPR timescale
(<ꢁ1010 sꢀ1). As temperature was increased to 298 K, solid-state
complexes of [1]+ and [2]2+ are EPR silent, indicating a faster elec-
tron relaxation process for the E2g electronic configuration in ferro-
cenium center and perhaps a rapid electron delocalization over the
Ru metal centers and the Fe metal center(s) in ferrocenyl–ethynyl
spacers.
1998 cmꢀ1 in [1]+. This band is intermediate between the stretch-
ing vibration of C„C (ꢁ2070 cmꢀ1
) and that of @C@C@
(ꢁ1926 cmꢀ1). Furthermore, the IR spectra of [2]2+ showed a strong
absorption of C„C at 1988 cmꢀ1. The oxidized complexes of [1]+
and [2]2+ are considerably stable in solid-state. In Bruce’s study
[19], spectroelectrochemically generated IR spectrum of oxidized
complex of [1]+ resulted in a lower energy shift of the stretching
vibration of C„C at 1986 and 2030(sh) cmꢀ1. The IR data suggests
Complexes of [1]+ and [2]2+ in CH2Cl2 solution at 77 K and 298 K
are EPR silent. At 298 K in solution-state, we suggest that the un-
paired electron is mainly localized on the Ru3+ metal centers.
a substantial electron delocalization over the (
metal center and ferrocenyl center.
g
5-C5H5)(dppe)Ru
As shown in Fig. 5, the appearance of the absorption bands at
600 and 607 nm in the UV spectra of [1]+ = 6340 Mꢀ1 cmꢀ1
and [2]2+ = 3933 Mꢀ1 cmꢀ1) in CH2Cl2 solution at 298 K would
(e
)
(e
Δg = 0.97
g = 2.92
further support that there is a substantial electron delocalization
over the metal centers. This absorption band, which is not present
in the neutral complexes of 1 and 2, was assigned to the ligand-to-
g = 1.95
metal (Cp-to-Fe3+
complex of [(
identified by solid-state Mössbauer technique, an absorption band
at 605 nm (
= 8800 Mꢀ1 cmꢀ1) was observed in CH2Cl2 solution at
)
2E2g ? 2E1u transition. In the mono-ruthenium
g
5-C5H5)(PPh3)2Ru–(C„C)–fc]+ with Fe3+ character
e
room temperature. In our case, the appearance of this lower inten-
sity absorption in comparison with the mono-ruthenium complex
of [(g
5-C5H5)(PPh3)2Ru–(C„C)–fc]+ indicates that the iron metal
center has some degree of Fe3+ character.
g = 2.89
g = 0.85
Δ
60
50
40
30
20
10
0
[1]
[1]+
[2]
g = 2.04
[2]2+
250 300 350 400 450 500 550 600 650 700 750 800 850 900
500
1500
2500
3500
4500
5500
Wavelength (nm)
Gauss
Fig. 5. UV spectra of 1, 2, [1]+ and [2]2+ in CH2Cl2 at room temperature.
Fig. 6. Solid-state EPR spectra of [1]+ (top) and [2]2+ (bottom) at 77 K.

T.-Y. Dong et al. / Journal of Organometallic Chemistry 694 (2009) 1529–1541
1537
Due to the rapid electron delocalization over the metal centers, the
Rua2+-Fe3+-Rub
2+
electron relaxation process is faster than the timescale of EPR
technique. Such EPR silence at 298 K in fluid is frequent for Ru3+
complexes, such as monocation mixed-valence [Cl(dppe)Ru–
C„C–(C6H4–C„C)n–Ru(dppe)Cl]+ (n = 1, 2, 3). In this arylethynyl
system with n = 1, a rhombic EPR spectrum (g = 2.15, 2.05, and
1.99) was only observed in the frozen CH2Cl2 solution at 3.4 K
[41]. Related 298 K EPR spectrum of mixed-valence C4-alkynyl iron
E
complex of [(g g
5-C5Me5)(dppe)Fe–(C„C)2–Fe(dppe)( 5-C5Me5)]+
has been reported with gavg = 2.126.
2.6. Intervalence transition
EIT
As shown in Fig. 7, complexes [1]+ and [2]2+ showed broad
intervalence transition (IT) band in the NIR region in CH2Cl2 solu-
tion. Application of Hush’s theoretical analysis [42] of IT band to
determine the nature and magnitude of the electronic coupling be-
tween the metal sites in complexes [1]+ and [2]2+ is of some inter-
est. Broad intervalence transition (IT) band near 1600 nm
(6250 cmꢀ1) could be assigned to the intervalence transition
involving photo-induced electron transfer between the two vib-
E0
Rua3+-Fe2+-Rub
Rua2+-Fe2+-Rub
2+
3+
Rua2+-Fea3+-Feb3+-Rub
2+
Rua3+-Fea3+-Feb2+-Rub
2+
Rua2+-Fea2+-Feb3+-Rub
3+
2+
ronic states of Rua3+–Fe2+–Rub and Rua2+–Fe3+–Rub2+. In other
E
words, this is an assignment to the intervalence transition between
the Ru3+ and Fe2+ metal centers, indicating the existence of strong
metal-to-metal interaction. In the mono-ruthenium complex [(g5
-
C5H5)(PPh3)2Ru–(C„C)–fc]+, a broad absorption band at 1550 nm
= 4080 Mꢀ1 cmꢀ1) was observed in CH2Cl2 solution at room tem-
perature. Sato and coworkers suggested that this absorption band,
which is not present in the neutral complex [(
5-C5H5)(PPh3)2Ru–
(e
g
(C„C)–fc], could be assigned to the intervalence transition be-
tween the Ru2+ and Fe3+ metal centers. They suggested that the
iron center was in Fe3+ oxidation state based on the solid-state
Mössbauer technique.
EIT
E0
In our case, the simple two vibronic states model proposed by
Hush can not explain the origin of the IT band. As shown in
Fig. 8, three and five vibronic states potential energy diagrams
are proposed to explain the electronic structures of complexes
[1]+ and [2]2+. In [1]+, the three potential energy surfaces are re-
2+
3+
Rua3+-Fea2+-Feb3+-Rub
Rua2+-Fea3+-Feb2+-Rub
Fig. 8. Potential energy surfaces of [1]+ and [2]2+
.
2+
2+
lated to Rua3+–Fe2+–Rub
,
Rua2+–Fe3+–Rub
and Rua2+–Fe2+
–
the two vibronic states could be estimated by the difference in
the redox potentials of the two metal centers in the molecule. Fur-
thermore, the magnitude of the delocalization can be obtained by a
calculation of the delocalization parameter a2 and electronic cou-
pling Vab. In the case of complexes [1]+ and [2]2+, the values of a2
and Vab were calculated from the following equations and collected
in Table 3. In these equations, mmax is the energy of intervalence
Rub3+. Applying the Franck-Condon principle, the vertical process
indicated by EIT corresponds to the intervalence transition from
2+
2+
the ground state Rua3+–Fe2+–Rub to the excited state Rua
–
Fe3+–Rub2+. Therefore, photoexcitation provides a mechanism for
charge transfer across the ferrocenium bis-ethynyl spacer. The
upper-limit value of zero-point energy difference (
D
E0) between
transition in cmꢀ1
,
emax is the extinction coefficient,
D
m1/2 is the
band width at half-height observed in cmꢀ1
, the constant
5000
4500
4000
3500
3000
2500
2000
1500
1000
500
(4.24 ꢃ 10ꢀ4) is in the unit of M cmꢀ1, and d is the donor–acceptor
distance in Å. These cations are examples of class II mixed-valence
compounds [43].
[1]+
[2]2+
ꢀ4
a2 ¼ f4:24 ꢃ 10
ꢀ
max
ð
D
m1=2Þg=f
m
maxd2g;
Table 3
Intervalence transition band energies and related data.
eb
D
m1/2
a2 (ꢃ103)d
Vab
a
c
e
Compound
mmax
[1]+
6666
6250
4454
1753
4120
2804
25.6
7.49
1068
527
[2]2+
0
a
900
1100
1300
1500
1700
1900
2100
m
max: the energy of intervalence transition in cmꢀ1
.
Extinction coefficient in Mꢀ1 cmꢀ1
.
b
c
Wavelength (nm)
D
a
m .
1/2: the band width at half-height observed in cmꢀ1
2: delocalization parameter.
d
e
Fig. 7. Intervalence transition band in the NIR region of complexes [1]+ and [2]2+ in
CH2Cl2 solution.
Vab: electronic coupling in cmꢀ1
.

1538
T.-Y. Dong et al. / Journal of Organometallic Chemistry 694 (2009) 1529–1541
Vab
¼
mmaxa:
As shown in Fig. 9, examining the characteristics of the five
HOMOs (HOMO, HOMO-1, HOMO-2, HOMO-3, and HOMO-4), we
found that they commonly correspond to the orbitals of the ferr-
p
2.7. Theoretical calculations
ocenyl–ethynyl spacers mixed extensively with the dx2–y2 orbital of
the Ru2+ metal center and dx2ꢀy2 and dxy orbitals of the Fe2+ metal
center. In the case of 2, it is clear that the metallic contribution to
the HOMO is mainly from Fe2+ centers corresponding to the dx2–y2
(10.6%) and dxy (26.4%) orbitals mixed extensively with the Pz orbi-
tal (10.6%) of Cp–C„ and Pz orbital (10.2%) of Ru–C„, defining the
Cp–Fe–Cp axis as the z-axis of the Cartesian. The molecular orbital
HOMO-1 in 2 is related to the dxy orbital (31.8%) of the Fe2+ centers
and the Pz orbital (14.6%) of Cp–C„ and Pz orbital (12.6%) of Ru-
C„. Orbital HOMO-2 is related to the dx2ꢀy2 orbital (14.4%) of
the Fe2+ centers and the dx2ꢀy2 orbital (23.8%) of the Ru2+ centers.
To study the electronic structures of synthesized complexes (1–
5), we have carried out molecular orbital calculations at the B3LYP
level of density functional theory to see how the HOMO–LUMO gap
correlated with the
DE1/2. In particular, it is desirable to under-
stand the origin of the Ru2+–Ru2+ and Ru2+–Fe2+ metal-to-metal
interactions from the computational chemistry. For complex 3,
the Ru2+–Ru2+ metal–metal distance of 24.64 Å was calculated
from the theoretical geometry of 3 generated from the X-ray
molecular structure of 2.
Cp
Cp
Ph
z
Ph
Ru
P
Ru
Cp
Ph
z
Ph
Ph
Ph
Ph
P
Ph
P
z
P
Ru
P
Fe
y
Fe
y
Ph
Ph
Ph
P
Ph
y
x
Fe
Fe
x
x
1
3
2
Fig. 9. Representative frontier HOMOs of 1, 2, and 3.

T.-Y. Dong et al. / Journal of Organometallic Chemistry 694 (2009) 1529–1541
1539
1.0
0.5
dure [44–46]. Preparations of the ferrocenyl–ethynyl spacers (8–
10) were described in our previous paper [26–28]. As shown in
Schemes 1 and 2, complexes of 1–7 could be prepared.
Cp
Ph
Ru
P
z
Ph
Ph
Ph
P
Fe
y
x
0.0
LUMO
HOMO
4.2. Preparation of compound 2
-0.5
A mixture of 9 (50 mg, 0.089 mmol), (g
5-C5H5)(dppe)RuCl
-4.5
-5.0
-5.5
-6.0
-6.5
-7.0
(106 mg, 0.178 mmol) [44], and KF (10 mg, 0.178 mmol) was
heated in refluxing methanol (40 ml) for 16 h [45]. After cooling
to room temperature, the reaction mixture was filtered and
washed with methanol, ether, and recrystallized from dichloro-
methane–hexane to give orange-red solid compound. The yield
was approximately 30%. 1H NMR (C6D6): d 2.13–2.24 (m, 4H,
-PCHCHP-), 2.72–2.83 (m, 4H, -PCHCHP-), 3.84 (s, 8H, fc-Cp), 3.92
(s, 4H, fc-Cp), 4.16 (s, 4H, fc-Cp), 4.75 (s, 10H, Cp), 6.98 (t,
J = 2.0 Hz, 12H, Ph), 7.22–7.25 (m, 12H, Ph), 7.33 (t, J = 7.5 Hz, 8H,
Ph), 8.09 (t, J = 8.5 Hz, 8H, Ph). 13C NMR (C6D6): d 29.15 (t,
Jcp = 23.2 Hz, -PCH2CH2P-), 68.53 (s, fc-Cp), 69.34 (s, fc-Cp), 69.82
(s, fc-Cp), 71.59 (s, fc-Cp), 76.84 (s, fc-Cp), 83.21 (s, Cp), 85.32 (s,
fc-Cp), 106.55 (t, Ru–C„, Jcp = 25.5 Hz), 107.36 (s, fc–C„),
ꢁ128.00 (s, m-Ph, overlap with solvent peak), 129.10, 129.95 (s,
p-Ph), 132.17 (t, J = 5.0 Hz, o-Ph), 135.22 (t, J = 5.0 Hz, o-Ph),
137.85–138.32, 143.77–144.22 (m, ipso-Ph). 31P NMR (C6D6): d
87.53 (s). MS (ESI): M+ at m/z 1546.1. Anal. Calc. for C86H74Fe2-
P4Ru2: C, 66.85; H, 4.83. Found: C, 66.82; H, 4.98%. M.p.: 235–
237 °C.
2Fe
2Ru
0.0
0.2
0.4
0.6
0.8
1.0
Total contribution (2Fe & 2Ru)
Fig. 10. Total contributions of Ru2+ and Fe2+ metal centers at any given energy for 2.
Orbital HOMO-3 is related to the dx2ꢀy2 orbital (35.2%) of the Ru2+
centers and the Px (11.4%) and Py (10.8%) orbitals of Cp–C„. The
molecular orbital HOMO-4 is related to the dx2ꢀy2 orbital (51.6%)
of the Fe2+ centers and the dx2ꢀy2 orbital (13.8%) of the Ru2+ cen-
ters. Molecular calculations indicate a substantial delocalization
of the frontier orbital over the metal centers and the ferrocenyl–
ethynyl spacers. Fig. 10 shows a typical example of total contribu-
tion of Ru2+ and Fe2+ metal centers at any given energy of 2. As
shown in Fig. 9, examining the characteristics of the four HOMOs
in 3, it is interesting to find that the metallic contribution of the
middle ferrocenyl moiety to the HOMO is not found. In other
words, the electron delocalization is not continued through the
middle ferrocenyl moiety in these HOMOs. We would suggest that
this is a possible reflection of the relatively weak electronic cou-
pling between the two Ru2+ centers in 3.
4.3. Preparation of compound 3
A mixture of 10 (50 mg, 0.067 mmol), (g
5-C5H5)(dppe)RuCl
(80.4 mg, 0.134 mmol), and KF (10 mg, 0.178 mmol) was heated
in refluxing methanol/THF (3:1) for 16 h. After cooling to room
temperature, the solvents were removed in vacuo. The crude
product was purified by column chromatography on Al2O3 (act.
III), eluting with hexane/acetone (9:1). The fourth band was the
desired compound. The yield was approximately 10%. 1H NMR
(C6D6):
d 2.17–2.20 (m, 4H, -PCHCHP-), 2.74–2.77 (m, 4H,
-PCHCHP-), 3.73 (s, 8H, fc-Cp), 3.94 (t, 4H, fc-Cp, J = 1.8 Hz), 4.08
(t, 4H, fc-Cp, J = 2.1 Hz), 4.11 (t, 4H, fc-Cp, J = 1.5 Hz), 4.24 (t, 4H,
fc-Cp, J = 1.5 Hz), 4.74 (s, 10H, Cp), 6.98 (t, J = 3.0 Hz, 12H, Ph),
7.20–7.25 (m, 12H, Ph), 7.31 (t, J = 7.8 Hz, 8H, Ph), 8.07 (t,
J = 8.7 Hz, 8H, Ph). 13C NMR (C6D6): d 29.36 (t, -PCH2CH2P-,
Jc-p = 23.1 Hz), 68.70 (s, fc-Cp), 68.82 (s, fc-Cp), 69.87 (s, fc-Cp),
70.21 (s, fc-Cp), 71.95 (s, fc-Cp), 77.25 (s, fc-Cp), 83.42 (s, Cp),
84.23 (s, fc-Cp), 86.17 (s, fc-Cp), 107.06 (t, Ru–C„, Jcp = 25.8 Hz),
107.44 (s, fc–C„), ꢁ128.00 (s, m-Ph, overlap with solvent
peak), 129.36, 130.19 (s, p-Ph), 132.42, 135.42 (s, o-Ph), 138.11–
138.44, 143.99–144.36 (m, ipso-Ph), 31P NMR (C6D6): d 87.11 (s).
MS (MALDI-TOF): M+ at m/z 1730.00. Anal. Calc. for C96H82Fe3-
P4Ru2: C, 66.68; H, 4.78. Found: C, 66.64; H, 4.98%. M.p.:
132–134 °C.
3. Conclusions
Several multinuclear ferrocenyl–ethynyl complexes of formula
[(g g
5-C5H5)(dppe)MII–C„C–(fc)n–C„C–MII(dppe)( 5-C5H5)] were
studied. These complexes undergo sequential reversible oxidation
events from 0.0 V to 1.0 V referred to the Ag/AgCl electrode in
anhydrous CH2Cl2 solution. We suggest that the low-potential
wave(s) in 1–5 could be assigned to the end-capped metallic cen-
ters. The resulting oxidation products are characterized by IR, EPR
and UV–Vis–NIR spectroscopic methods, suggesting a substantial
electron delocalization over the (g
5-C5H5)(dppe)Ru metal center
and ferrocenyl center. A combination of spectroscopic data and
computational studies reveals that the ferrocenyl–ethynyl-based
orbitals do mix significantly with the (g
5-C5H5)(dppe)Ru metallic
orbitals. It clearly appears from this work that the ferrocenyl–ethy-
nyl spacers quite strongly contribute in propagating electron
delocalization.
4.4. Preparation of compound 4
A mixture of 11 (50 mg, 0.12 mmol) [27], (g
5-C5H5)(dppe)FeCl
(200 mg, 0.36 mmol) [46], KPF6 (130 mg, 0.72 mmol) was stirred
in methanol (30 ml) at room temperature. After stirring for 2 h,
potassium tert-butoxide (56 mg, 0.5 mmol) was added to the solu-
tion. The resulting mixture was stirred for a further 15 min. The
solvent was removed under reduced pressure. The crude product
was purified by column chromatography on Al2O3 (act. III), eluting
with hexane/acetone (10:1). The first band was mononuclear Fe2+
metallic complex (6). The yield of 6 was 13%. The second band
was the desired complex (4). The yield was 58%. Complex of 4
could be recrystallized from benzene–ether to give an orange-red
solid compound. The physical properties of 4 are as follows. 1H
4. Experimental
4.1. General information
All manipulations involving air-sensitive materials were carried
out by using standard Schlenk techniques under an atmosphere of
N2. Solvents were dried as follows: THF and ether were distilled
from Na/benzophenone; DMF and CH2Cl2 were distilled from
CaH2; TMEDA was distilled from KOH. Samples of 1 and (g5
-
C5H5)(dppe)MCl were prepared according to the literature proce-

1540
T.-Y. Dong et al. / Journal of Organometallic Chemistry 694 (2009) 1529–1541
NMR (C6D6): d 2.09–2.11 (m, 4H, -PCHCHP-), 2.73–2.76 (m, 4H,
-PCHCHP-), 3.85 (s, 4H, fc-Cp), 3.88 (s, 4H, fc-Cp), 3.99 (s, 4H, fc-
Cp), 4.22 (s, 4H, fc-Cp), 4.28 (s, 10H, Cp), 6.98–7.02 (m, 12H, Ph),
7.21–7.25 (m 12H, Ph), 7.32 (t, J = 7.5 Hz, 8H, Ph), 8.09 (br s, 8H,
Ph). 13C NMR (C6D6): d 29.09 (t, -PCH2CH2P-, J = 22.2 Hz), 68.14
(s, fc-Cp), 68.72 (s, fc-Cp), 69.29 (s, fc-Cp), 71.03 (s, fc-Cp), 76.62
(s, fc-Cp), 79.46 (s, Cp), 84.84 (s, fc-Cp), 113.75 (t, Fe–C„,
Jcp = 42.45 Hz), 115.82 (s, fc-C„), 128.00 (s, m-Ph, overlap solvent
peak), 128.29 (s, m-Ph), 128.64, 129.37 (s, p-Ph), 132.12, 134.51
(s, o-Ph), 138.64–138.94, 143.21–143.38 (m, ipso-Ph). 31P NMR
(C6D6): d 107.13 (s). MS (ESI): (M+1)+ at m/z 1455. Anal. Calc. for
C86H74Fe4P4: C, 71.00; H, 5.13. Found: C, 71.32; H, 5.26%. M.p.:
214–216 °C. The physical properties of 6 are as follows. 1H NMR
(C6D6): d 2.07–2.10 (m, 2H, -PCHCHP-), 2.49 (s, 1H, „–H), 2.68–
2.71 (m, 2H, -PCHCHP-), 3.77 (t, 2H, fc-Cp, J = 1.5 Hz), 3.80 (t, 2H,
fc-Cp, J = 1.8 Hz), 3.85 (t, 2H, fc-Cp, J = 1.5 Hz), 4.00 (t, 2H, fc-Cp,
J = 1.2 Hz), 4.11 (t, 2H, fc-Cp, J = 1.8 Hz), 4.18 (t, 2H, fc-Cp,
J = 1.5 Hz), 4.28 (s, 5H, CpFe), 4.30 (t, 2H, fc-Cp, J = 1.8 Hz), 4.35
(t, 2H, fc-Cp, J = 1.8 Hz), 6.97–7.02 (m, 6H, Ph), 7.18–7.24 (m, 6H,
Ph), 7.31 (t, J = 7.8 Hz, 4H, Ph), 8.07 (t, J = 7.8 Hz, 4H, Ph). 13C
NMR (C6D6): d 29.06 (t, -PCH2CH2P-, J = 22.1 Hz), 68.04 (s, fc-Cp),
68.49 (s, fc-Cp), 68.97 (s, fc-Cp), 69.64 (s, fc-Cp), 70.14 (s,
fc-Cp), 70.60 (s, fc-Cp), 71.16 (s, fc-Cp), 73.25 (s, fc-Cp), 76.90 (s,
fc-Cp), 79.49 (s, Cp), 82.65 (s, fc-Cp), 82.95 (s, fc-Cp), 86.93 (s, fc-
Cp), 114.75 (t, Fe–C„, Jcp = 42.5 Hz), 115.54 (s, fc–C„), 128.00 (s,
m-Ph, overlap solvent peak), 128.29 (s, m-Ph), 128.70, 129.39 (s,
p-Ph), 132.08, 134.49 (s, o-Ph), 138.61–138.92, 143.11–143.29
(m, ipso-Ph). 31P NMR (C6D6): d 107.25 (s). MS (MALDI-TOF): M+
at m/z 936. Anal. Calc. for C55H46Fe3P2: C, 70.54; H, 4.95. Found:
8.06 (br s, 4H, Ph). 13C NMR (C6D6): d 29.05 (t, -PCH2CH2P-,
J = 22.4 Hz), 68.04 (s, fc-Cp), 68.06 (s, fc-Cp), 68.13 (s, fc-Cp),
68.21 (s, fc-Cp), 68.57 (s, fc-Cp), 69.10 (s, fc-Cp), 69.49 (s, fc-Cp),
69.51 (s, fc-Cp), 70.12 (s, fc-Cp), 70.42 (s, fc-Cp), 71.13 (s, fc-Cp),
73.35 (s, fc-Cp), 76.79 (s, fc-Cp), 79.48 (s, Cp), 82.69 (s, fc-Cp),
83.17 (s, fc-Cp), 83.35 (s, fc-Cp), 85.70 (s, fc-Cp), 85.89 (s, fc-Cp),
114.33 (t, Fe–C„, Jcp = 42.5 Hz), 115.64 (s, fc-C„), 128.00 (s,
m-Ph, overlap solvent peak), 128.29 (s, m-Ph), 128.70, 129.36 (s,
p-Ph), 132.11, 134.47 (s, o-Ph), 138.64–138.95, 143.12–143.30
(m, ipso-Ph). 31P NMR (C6D6): d 107.23 (s). MS (ESI): M+ at m/z
1120. Anal. Calc. for C65H54Fe4P2: C, 69.68; H, 4.86. Found: C,
70.14; H, 5.30%. M.p: 70–72 °C. IR (KBr):
m(C„C) 2106 (w) and
2065 (s) cmꢀ1
.
4.6. Preparation of Compounds [1]+ and [2]2+
To a solution of 1 (30 mg, 0.022 mmol) or 2 (30 mg, 0.019
mmol) in dichloromethane (1 mL) and benzene (10 mL) was added
stoichiometric amount of ferrocenium hexafluorophosphate under
N2 an ice bath. The mixture was stirred for 4 h. The resulting blue
precipitates were filtered. The precipitates were recrystallized
from dichloromethane/ether to give dark blue desired compound
(ꢁ80% yield). Anal. Calc. for C76H66F6FeP5Ru2 ([1]+): C, 60.60; H,
4.42. Found: C, 61.19; H, 4.04%. Anal. Calc. for C86H74F12Fe2P6Ru2
([2]2+): C, 56.28; H, 4.06. Found: C, 55.90; H, 4.23%.
4.7. Computational details
Density functional theory calculations at the B3LYP level were
performed to obtain the molecular orbitals of 1–3. The basis set
used for C, H, P and Fe atoms was 6-31g**, while effective core
potentials with a LanL2DZ basis set were employed for Ru. Polari-
zation functions were added for Ru (fd(Ru) = 0.15). All the calcula-
tions were made with the use of GAUSSIAN 03. The molecular orbitals
were plotted with the Gauss View program.
C, 70.17; H, 5.15%. M.p.: 168–170 °C. IR (KBr):
m(C„C) 2108 (w)
and 2067 (s) cmꢀ1
.
4.5. Preparation of compound 5
A mixture of 12 (50 mg, 0.083 mmol) [27], (g
5-C5H5)(dppe)FeCl
(180 mg, 0.33 mmol), KPF6 (100 mg, 0.54 mmol) was stirred in
methanol/THF (5:1 30 ml) at room temperature. After stirring for
2 h, potassium tert-butoxide (20 mg, 0.2 mmol) was added to the
solution. The resulting mixture was stirred for a further 30 min.
The solvent was removed under reduced pressure. The crude prod-
uct was purified by column chromatography on Al2O3 (act. III),
eluting with hexane/acetone (10:1). The first band was mononu-
clear Fe2+ metallic complex (7). The yield of 7 was 4%. The second
band was the desired complex (5). The yield was 54%. Complex of 5
could be recrystallized from dicholoromethane–hexane to give an
orange-red solid compound. The physical properties of 5 are as fol-
lows. 1H NMR (C6D6): d 2.08–2.11 (m, 4H, -PCHCHP-), 2.70–2.73
(m, 4H, -PCHCHP-), 3.74 (s, 4H, fc-Cp), 3.77 (s, 4H, fc-Cp), 4.00 (s,
4H, fc-Cp), 4.07 (s, 4H, fc-Cp), 4.14 (s, 4H, fc-Cp), 4.22 (s, 4H, fc-
Cp), 4.27 (s, 10H, Cp), 6.97–7.02 (m, 12H, Ph), 7.20–7.23 (m 12H,
Ph), 7.30 (t, J = 7.5 Hz, 8H, Ph), 8.07 (br s, 8H, Ph). 13C NMR
(C6D6): d 29.10 (t, -PCH2CH2P-, J = 22.2 Hz), 68.06 (s, fc-Cp), 68.10
(s, fc-Cp), 68.18 (s, fc-Cp), 69.23 (s, fc-Cp), 69.40 (s, fc-Cp), 71.15
(s, fc-Cp), 76.77 (s, fc-Cp), 79.46 (s, Cp), 83.63 (s, fc-Cp), 85.48 (s,
fc-Cp), 114.00 (t, Fe–C„, Jcp = 42.5 Hz), 115.72 (s, fc–C„), 128.00
(s, m-Ph, overlap solvent peak), 128.53 (s, m-Ph), 128.66, 129.38
(s, p-Ph), 132.13, 134.48 (s, o-Ph), 138.62–138.92, 143.18–143.36
(m, ipso-Ph). 31P NMR (C6D6): d 107.28 (s). MS (MALDI-TOF):
(M+1)+ at m/z 1639. Anal. Calc. for C96H82Fe5P4: C, 70.36; H, 5.04.
4.8. Physical methods
1H NMR spectra were run on Varian INOVA-500 MHz spectrom-
eter or INOVA-600 MHz spectrometer. Mass spectra were obtained
with a VG-BLOTECH-QUATTRO 5022 system and ESI-LCQ mass
spectra were obtained with a Thermo Finnigan spectroscopy. Elec-
trochemical measurements were carried out with a CHI 660B sys-
tem. Differential pulse voltammetry (DPV) and cyclic voltammetry
(CV) were performed with a stationary Pt working electrode. These
experiments were carried out with a 1 ꢃ 10ꢀ3 M solution of dried
CH2Cl2 containing 0.1 M of (n-C4H9)4NPF6 as supporting electro-
lyte. The potentials quoted in this work are relative to the Ag/AgCl
electrode at 25 °C. Under these conditions, ferrocene shows a
reversible one-electron redox wave (E1/2 = 0.46 V).
4.9. Structure determination of 2
A yellow-red crystal (0.22 ꢃ 0.14 ꢃ 0.12 mm) was grown when
a layer of hexane was allowed to slowly diffuse into a CH2Cl2 solu-
tion of 2. The single crystal X-ray determination of compound 2
was carried out at 120.0(1) K by using a Bruker X8 APEX CCD dif-
fractometer [k(Mo
Data were collected to a maximum h value of 27.54°. Of the
Ka) = 0.71073 Å], graphite monochromator.
Found: C, 70.13; H, 5.41%. M.p.: 130–132 °C. IR (KBr):
m
(C„C)
35764 reflections collected, there were 7845 independent
2067 (s) cmꢀ1. The physical properties of 7 are as follows. 1H
NMR (C6D6): d 2.08 (br s, 2H, -PCHCHP-), 2.44 (s, 1H, „-H), 2.70
(br s, 2H, -PCHCHP-), 3.74 (br s, 4H, fc-Cp), 3.77 (br s, 2H, fc-Cp),
4.00 (br s, 4H, fc-Cp), 4.07 (br s, 2H, fc-Cp), 4.10 (br s, 4H, fc-Cp),
4.20 (br s, 6H, fc-Cp), 4.24 (br s, 2H, fc-Cp), 4.28 (s, 5H, Cp), 6.95–
7.02 (m, 6H, Ph), 7.20–7.24 (m 6H, Ph), 7.30 (t, J = 6.9 Hz, 4H, Ph),
reflections (Rint = 0.0379) with F2o > 2.0 (F2o). A semi-empirical
r
absorption correction based on azimuthal scans of several
reflections was applied. The structures were solved by an ex-
panded Fourier technique. All non-hydrogen atoms were refined
anisotropically. Hydrogen atoms were included at ideal distance.
Crystal data for 2: C86H74Fe2P4Ru2, M = 1545.17, monoclinic, space

T.-Y. Dong et al. / Journal of Organometallic Chemistry 694 (2009) 1529–1541
1541
[8] J. Hjelm, R.W. Handel, A. Hagfeldt, E.C. Constable, C.E. Housecroft, R.J. Forster,
Inorg. Chem. 44 (2005) 1073.
[9] S. Flores-Torres, G.R. Hutchison, L.J. Soltzberg, H.D. Abruña, J. Am. Chem. Soc.
128 (2006) 1513.
[10] Q.Y. Hu, W.X. Lu, H.D. Tang, H.H.Y. Sung, T.B. Wen, I.D. Williams, G.K.L. Wong,
Z. Lin, G. Jia, Organometallics 24 (2005) 3966.
group P21/c, a = 9.7846(2) Å, b = 21.8359(5) Å, c = 16.6231(3) Å,
b = 106.2500(10)ꢅ, V = 3409.73(12) Å3, Z = 2, Dcalc = 1.505 g cmꢀ3
,
F(000) = 1580,
R1 = 0.0248 (I > 2
l (Mo Ka
) = 0.993 mmꢀ1, Goodness-of-fit = 0.898,
r
(I)), wR2 = 0.0485.
[11] M.I. Bruce, P.J. Low, K. Costuas, J.F. Halet, S.P. Best, G.A. Heath, J. Am. Chem. Soc.
122 (2000) 1949.
[12] M.I. Bruce, B.G. Ellis, P.J. Low, B.W. Skelton, A.H. White, Organometallics 22
(2003) 3184.
[13] N.J. Long, C.K. Williams, Angew. Chem., Int. Ed. Engl. 42 (2003) 2586.
[14] N.L. Narvor, L. Toupet, C. Lapinte, J. Am. Chem. Soc. 117 (1995) 7129.
[15] M. Sato, Y. Hayashi, S. Kumakura, N. Shimizu, M. Katada, S. Kawata,
Organometallics 15 (1996) 721.
[16] F. Coat, M.A. Guillevic, L. Toupet, F. Paul, C. Lapinte, Organometallics 16 (1997)
5988.
[17] M. Guillemot, L. Toupet, C. Lapinte, Organometallics 17 (1998) 1928.
[18] F. Coat, F. Paul, C. Lapinte, L. Toupet, K. Costuas, J.F. Halet, J. Organomet. Chem.
683 (2003) 368.
[19] M.I. Bruce, P.J. Low, F. Hart, P.A. Humphrey, F. deMontigny, M. Jevric, C. Lapinte,
G.J. Perkins, R.L. Roberts, B.W. Skelton, A.H. White, Organometallics 24 (2005)
5241.
[20] M.I. Bruce, K. Costuas, B.G. Ellis, J.F. Halet, P.J. Low, B. Moubaraki, K.S. Murray,
N. Oudda, G.J. Perkins, B.W. Skelton, A.H. White, Organometallics 26 (2007)
3735.
[21] M.I. Bruce, K. Costuas, T. Davin, J.F. Halet, K.A. Kramarczuk, P.J. Low, B.K.
Nicholson, G.J. Perkins, R.L. Roberts, B.W. Skelton, M.E. Smith, A.H. White,
Dalton Trans. (2007) 5387.
4.10. Structure determination of 4 ꢂ C6H6
A yellow-red crystal (0.13 ꢃ 0.10 ꢃ 0.10 mm) was grown when
a layer of ether was allowed to slowly diffuse into a benzene
solution of 4. The single crystal X-ray determination of compound
4 was carried out at 150.0(1) K by using a Bruker X8 APEX CCD
diffractometer [k(Mo Ka) = 0.71073 Å], graphite monochromator.
Data were collected to a maximum h value of 19.85°. Of the
17075 reflections collected, there were 3323 independent
reflections (Rint = 0.0839) with F2o > 2.0 (F2o). A semi-empirical
r
absorption correction based on azimuthal scans of several
reflections was applied. The structures were solved by an ex-
panded Fourier technique. All non-hydrogen atoms were refined
anisotropically. Hydrogen atoms were included at ideal distance.
Crystal data for 4 ꢂ C6H6: C92H80Fe4P4, M = 1532.84, monoclinic,
space group P21/n, a = 11.6455(6) Å, b = 26.7608(14) Å, c =
11.7964(6) Å, b = 96.732(3)ꢅ, V = 3650.9(3) Å3, Z = 2, Dcalc = 1.394
[22] R. Dembinski, T. Bartik, B. Bartik, M. Jaeger, J.A. Gladysz, J. Am. Chem. Soc. 122
(2000) 810.
[23] T. Bartik, W. Weng, J.A. Ramsden, S. Szafert, S.B. Falloon, A.M. Arif, J.A. Gladysz,
J. Am. Chem. Soc. 120 (1998) 11071.
g cmꢀ3
,
F(000) = 1592,
l
(Mo
K
a
) = 0.915 mmꢀ1
r(I)), wR2 = 0.0957.
,
Goodness-
of-fit = 1.052, R1 = 0.0471 (I > 2
[24] M. Brady, W. Weng, Y. Zhou, J.W. Seyler, A.J. Amoroso, A.M. Arif, M. Bohme, G.
Frenking, J.A. Gladysz, J. Am. Chem. Soc. 119 (1997) 775.
[25] S. Rigaut, C. Olivier, K. Costuas, S. Choua, O.F.J. Massue, P. Turek, J.Y. Saillard,
P.H. Dixneuf, D. Touchard, J. Am. Chem. Soc. 128 (2006) 5859.
[26] T.-Y. Dong, K. Chen, M.C. Lin, L. Lee, Organometallics 24 (2005) 4198.
[27] T.-Y. Dong, S.W. Chang, S.F. Lin, M.C. Lin, Y.S. Wen, L. Lee, Organometallics 25
(2006) 2018.
[28] T.-Y. Dong, M.C. Lin, S.W. Chang, C.C. Ho, S.F. Lin, L. Lee, J. Organomet. Chem.
692 (2007) 2324.
[29] M. Sato, H. Shintate, Y. Kawata, M. Sekino, M. Katada, S. Kawata,
Organometallics 13 (1994) 1956.
Acknowledgements
We would like to acknowledge Charles P. Casey (University of
Wisconsin) for valuable discussions. This work was funded by
the National Science Council (NSC95-2113-M-110-014-MY3),
Taiwan, ROC, and Department of Chemistry and Center for Nano-
science and Nanotechnology at National Sun Yat-Sen University.
[30] P. Seiler, J.D. Dunitz, Acta Crystallogr. Sect. B 35 (1979) 1068.
[31] T.-Y. Dong, C.H. Huang, C.K. Chang, Y.S. Wen, S.L. Lee, J.A. Chen, W.Y. Yeh, A.
Yeh, J. Am. Chem. Soc. 115 (1993) 6357.
Appendix A. Supplementary material
[32] N.J. Long, A.J. Martin, R. Vilar, A.J.P. White, D.J. Williams, M. Younus,
Organometallics 18 (1999) 4261.
[33] H. Lehmkuh, M. Bellenbaum, J. Grundke, H. Mauermann, C. Kruger, Chem. Ber.
121 (1988) 1719.
[34] F.R. Lemke, D.J. Szalda, R.M. Bullock, J. Am. Chem. Soc. 113 (1991) 8466.
[35] C. Bitcon, M.W. Whiteley, J. Organomet. Chem. 336 (1987) 385.
[36] M.I. Bruce, B.G. Ellis, M. Gaudio, C. Lapinte, G. Melino, F. Paul, B.W. Skelton,
M.E. Smith, L. Toupet, A.H. White, J. Chem. Soc., Dalton Trans. 10 (2004) 1601.
[37] M.I. Bruce, B.D. Kelly, B.W. Skelton, A.H. White, J. Organomet. Chem. 640
(2000) 150.
[38] A.B. Antonova, M.I. Bruce, B.G. Ellis, M. Gaudio, P.A. Humphrey, M. Jevric, G.
Melino, B.K. Nicholson, G.J. Perkins, B.W. Skelton, B. Stapleton, A.H. White, N.N.
Zaitseva, Chem. Commun. (2004) 960.
CCDC 673138 and 673142 contain the supplementary crystallo-
graphic data for compounds 2 and 4 ꢂ C6H6. These data can be ob-
tained free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif. Supplementary
data associated with this article can be found, in the online version,
at doi:10.1016/j.jorganchem.2008.12.056.
References
[1] K.D. Demadis, C.M. Hartshorn, T.J. Meyer, Chem. Rev. 101 (2001) 2655.
[2] J.P. Launay, Chem. Soc. Rev. 30 (2001) 386.
[3] B.S. Brunschwig, C. Creutz, N. Sutin, Chem. Soc. Rev. 31 (2002) 168.
[4] A. El-ghayoury, A. Harriman, A. Khatyr, R. Ziessel, Angew. Chem., Int. Ed. Engl.
39 (2000) 185.
[5] A. Harriman, A. Mayeux, A. De Nicola, R. Ziesel, Phys. Chem. Chem. Phys. 4
(2002) 2229.
[6] A. Harriman, A. Khatyr, R. Ziessel, A.C. Benniston, Angew. Chem., Int. Ed. Engl.
39 (2000) 4287.
[39] T.-Y. Dong, C.-C. Schei, T.L. Hsu, S.L. Lee, S.J. Li, Inorg. Chem. 30 (1991) 2457.
[40] R. Prim, F.J. Reinders, J. Am. Chem. Soc. 91 (1969) 4929.
[41] A. Klein, O. Lavastre, J. Fiedler, Organometallics 25 (2006) 635.
[42] N.S. Hush, Inorg. Chem. 8 (1967) 391.
[43] P. Day, Int. Rev. Phys. Chem. 2 (1981) 149.
[44] A.G. Alonso, L.B. Reventós, J. Organomet. Chem. 338 (1988) 249.
[45] M.I. Bruce, B.C. Hall, B.D. Kelly, P.J. Low, B.W. Skelton, A.H. White, J. Chem. Soc.,
Dalton Trans. (1999) 3719.
[7] A. Barbieri, B. Ventura, F. Bariglletti, A. De Nicola, M. Quesada, R. Ziessel, Inorg.
Chem. 43 (2004) 7359.
[46] M.J. Mays, P.L. Sears, J. Chem. Soc., Dalton Trans. (1973) 1873.