Observation of a large current on the cyclic voltammetry of acylaminoferrocenes in the solid state: An efficient electron-transfer pathway through continuous NH...O=C hydrogen-bond chains and π-conjugation

Article abstract of DOI:10.1246/bcsj.78.1270

A series of crystal structures of acylaminoferrocenes, [(C ₅H₅)Fe(C₅H₄NHCOR)] (R = CH ₃ (1), CH₂CH₃ (2), t-Bu (3), and CF₃ (4)) revealed an unusually delicate dependence of the alkyl substituent structure on electron transfer through NH...O=C bonding and π-conjugation. The presence of strong intermolecular NH...O=C hydrogen bonds in the crystal was established by X-ray, IR, and CRAMPS. Compounds 1 and 2 are without steric or electrostatic hindrance, and both exhibit continuous and strong hydrogen-bond chains, or an effective π-electrons system. However, the hydrogen-bond chains in 3 and 4 are weak, or broken, by the bulkiness of t-Bu or the electrostatic repulsion of CF₃. Solid-state cyclic voltammograms of 1-4 have revealed the electrochemical processes, which are critically dependent on the array and the strength of the hydrogen bonds. Although both 3 and 4 show typical reversible cyclic voltammograms, 1 and 2 exhibit irreversible redox behaviors and remarkably large current values, which indicate the presence of an effective electron transfer through the continuous hydrogen-bond chains.

Full text of DOI:10.1246/bcsj.78.1270

1270 Bull. Chem. Soc. Jpn., 78, 1270–1278 (2005)
ꢀ 2005 The Chemical Society of Japan
Observation of a Large Current on the Cyclic Voltammetry
of Acylaminoferrocenes in the Solid State: An Efficient
Electron-Transfer Pathway through Continuous
NHÁÁÁO=C Hydrogen-Bond Chains and ꢀ-Conjugation
Taka-aki Okamura, Keiko Sakauye, Mototsugu Doi, Hitoshi Yamamoto,
ꢀ
Norikazu Ueyama, and Akira Nakamura1
Department of Macromolecular Science, Graduate School of Science, Osaka University, Toyonaka, Osaka 560-0043
1OM Research, 7-2-1308 Minami Ogimachi, Kita-ku, Osaka 530-0052
Received December 14, 2004; E-mail: ueyama@chem.sci.osaka-u.ac.jp
A series of crystal structures of acylaminoferrocenes, [(C5H5)Fe(C5H4NHCOR)] (R ¼ CH3 (1), CH2CH3 (2), t-Bu
(3), and CF3 (4)) revealed an unusually delicate dependence of the alkyl substituent structure on electron transfer through
NH O=C bonding and ꢀ-conjugation. The presence of strong intermolecular NH O=C hydrogen bonds in the crystal
ꢁꢁꢁ
ꢁꢁꢁ
was established by X-ray, IR, and CRAMPS. Compounds 1 and 2 are without steric or electrostatic hindrance, and both
exhibit continuous and strong hydrogen-bond chains, or an effective ꢀ-electrons system. However, the hydrogen-bond
chains in 3 and 4 are weak, or broken, by the bulkiness of t-Bu or the electrostatic repulsion of CF3. Solid-state cyclic
voltammograms of 1–4 have revealed the electrochemical processes, which are critically dependent on the array and the
strength of the hydrogen bonds. Although both 3 and 4 show typical reversible cyclic voltammograms, 1 and 2 exhibit
irreversible redox behaviors and remarkably large current values, which indicate the presence of an effective electron
transfer through the continuous hydrogen-bond chains.
Extensively conjugated organic systems or polynuclear met-
al coordination systems are the most promising candidate to
develop novel conductive or semi-conductive materials. Many
inorganic polymers have been investigated for this purpose.
Polynuclear ꢁ- or ꢀ-conjugated ferrocene derivatives connect-
ed by various linkages, especially, are well-known by concern-
ing their ease of chemical modification, good solubility, satis-
factory redox stability, and potential utilities, leading to vari-
ous other physical applications. Poly(ferrocenylsilane)s have
been known to show a semi-conductivity through a conjugated
pathway after I2- or TCNQ-doping. Partially oxidized polynu-
clear ferrocene derivatives are found in a mixed-valence state,
and show intramolecular electron transfer in solution. Howev-
er, the electron transfer is limited within several metal centers
because the conjugation is broken at many points by the rota-
tion around the bridging –SiR2– (R ¼ alkyl) moiety.1 Crystal
engineering using hydrogen bonds presents us with an infinite
conjugation chain, though it is limited by the crystal size.
Various arrays of organic2–7 and inorganic8,9 compounds have
been reported. We present here, the crystal structures of redox-
active acylaminoferrocenes (Fc-NHCOR, Fc = ferrocenyl)
having the intermolecular hydrogen bonds, and we describe
the relationships between the strength and electron-transfer
capability.
conjugation, which results in an increase of the protonic acid-
ity of NH or the capacity for hydrogen bonding.11 Indeed,
[Fe(C H CONHCH ) ] has no NH O=C hydrogen bond in
ꢁꢁꢁ
5
4
3 2
the crystal despite the presence of amide groups.12 However,
applications had been limited in FcCONHR because of the dif-
ficulty of synthesizing aminoferrocene until the recent work,13
and an improved synthetic procedure14 has been known. N-Al-
kyl- or N-aryl-carbamoylferrocenes, [FcCONHR] (R = alkyl
or aryl group), have been reported as substrate receptors that
recognize anions by using intermolecular hydrogen bonds.15–17
Recently, some acylaminoferrocene derivatives have been
reported.13,18
The synthesis of acetylaminoferrocene has been known from
long ago,19 and recently its structure has been reported, but it
shows an unacceptably large R1 factor (15.8%), which must be
because of an incorrect solution. Here, we report on the struc-
tures of a series of acylaminoferrocenes and the formation of
intermolecular hydrogen bonds in the solid state; we estimated
1
the strength of the hydrogen bonds using X-ray, IR, H NMR,
and electrochemical analysis. As we reported previously, the
variation of R in the acylamino (RCONH–) group effectively
and successfully controls the strength of the hydrogen
bond.20–24
Electrochemical measurements, especially cyclic voltam-
metry in the solid state surrounded by aqueous medium, are
useful techniques for the analysis of insoluble materials (e.g.,
Prussian Blue [PB]),25–30 as well as for the analysis of unstable
compounds by dissociation of the ligand or total decomposi-
tion in solution.31 The shape of cyclic voltammograms in the
Acylaminoferrocenes form stronger NH O=C hydrogen
ꢁꢁꢁ
bonds than the corresponding carbamoyl derivatives
(FcCONHR), as mentioned in a previous paper.10 When amide
NH is directly connected to a Cp ring, the electron on the lone
pair of the amide NH is delocalized to the Cp ring through ꢀ-
Published on the web July 1, 2005; DOI 10.1246/bcsj.78.1270

T. Okamura et al.
Bull. Chem. Soc. Jpn., 78, No. 7 (2005) 1271
for C15H19NOFe: C, 63.18; H, 6.72; N, 4.91%.
solid state has been interpreted as a diffusion-controlled redox
process, just as is true for ordinary ones in solution. Ferrocene
in the solid state using a carbon-paste electrode typically
shows a reversible redox FeIII/FeII couple, just as in solution,
although the current value is very small because of the limited
movement of molecules, or its low solubility in water.31 On the
other hand, PB shows a larger current value that is enabled by
the diffusion of potassium ions in the PB lattice. The current
values and the redox process are considered to represent the
mobility of electrons and ions or the electronic interaction in
the solid state. If electrons move through conjugated hydro-
gen-bond chains in the solid state, the electric current should
be accelerated, compared with a diffusion-controlled process.
In this paper, we discuss the relationship between the strength
of the hydrogen bond and the ease of electron transfer in the
solid state.
Trifluoroacetylaminoferrocene (4). Aminoferrocene (0.50 g,
2.5 mmol) was dissolved in CH2Cl2 (10 mL), and trifluoroacetic
anhydride (0.8 mL, 5.7 mmol) was added to the solution. After
stirring overnight at room temperature, crushed ice was added to
a solution cooling in an ice bath, and the solvent was evaporated
under reduced pressure. The residue was dissolved in ethyl acetate
and the solution was washed with sat. NaCl aq, dried over
Na2SO4, and concentrated. The crude product was purified by col-
umn chromatography using silica gel and ethyl acetate as the elu-
ent. Recrystallization from ethyl acetate gave orange plates. Yield
0.64 g (87%). 1H NMR (chloroform-d1): ꢂ 7.32 (s, 1H), 4.59 (t,
2H), 4.13 (s, 5H), 4.04 (t, 2H). Anal. Found: C, 47.99; H, 3.68;
N, 4.39%. Calcd for C12H10NOF3Fe: C, 48.52; H, 3.39; N, 4.72%.
Physical Measurements. 1H NMR spectra in solution were
taken on a Jeol GSX-400 spectrometer in CDCl3 or CD2Cl2. Sol-
1
id-state H NMR spectra using the CRAMPS (combined rotation
and multipulse spectroscopy) technique were taken on a Chem-
magnetics CMX-300. MREV-8 pulse sequence and 3 kHz spin-
ning were used to detect amide NH protons. Tetrakis(trimethyl-
silyl)silane (TTMS) was used as an internal standard at 0.2 ppm
from TMS. IR spectra were recorded on a Jasco FT/IR 8300 spec-
trometer. Samples were prepared as 10 mM CH2Cl2 solutions or
as KBr pellets. Cyclic voltammograms were obtained by BAS
100B/W in solution and in the solid state using a three-electrode
method. The solution samples were prepared as a 2.5 mM CH2Cl2
solution, including 0.2 M tetrabutylammonium perchlorate as a
supporting electrolyte. The solid samples were prepared as fol-
lows. The electrode body and carbon paste were purchased from
Bioanalytical Systems, Inc. (BASi). The carbon paste was mixed
well with 5 wt % of ferrocene derivatives and filled into an elec-
trode body with metallic rod, which was used as a working elec-
trode. The measurements were performed in an aqueous 0.1 M
NaClO4 solution. The scan rate was 100 mV/s for the solution
and 50 mV/s for the solid state. The E1=2 value was determined
using the SCE electrode as a reference at room temperature.
Ab initio Calculations. The bond order of the hydrogen-bond-
ed amide groups was calculated using a dimeric model. Two
Experimental
Materials. All procedures, except for column chromatogra-
phy, were performed under an argon atmosphere. The carbon
paste and electrode for electrochemical measurements were pur-
chased from BAS Inc. All solvents were dried and distilled under
argon before use. The other reagents were commercially available.
Acetylaminoferrocene (1). This compound was prepared by a
literature method.19 Single crystals for X-ray analysis were pre-
pared by recrystallization from ethyl acetate. 1H NMR (chloro-
form-d1): ꢂ 6.51 (s, 1H), 4.56 (t, 2H), 4.16 (s, 5H), 3.99 (t, 2H),
2.06 (s, 3H). Anal. Found: C, 59.29; H, 5.32; N, 5.86%. Calcd
for C12H13NOFe: C, 59.29; H, 5.39; N, 5.76%.
Preparation of Aminoferrocene. Acetylaminoferrocene (100
mg, 0.ꢂ41 mmol) was suspended in 6 M HCl aq (20 mL) and heated
at 80 C for 1 h. Insoluble materials were removed by filtration.
The obtained clear filtrate was washed with diethyl ether twice.
The acidic solution was neutralized by NaHCO3 and diethyl ether
was added. The ethereal layer was washed with a NaCl-saturated
aqueous solution, dried over Na2SO4, and concentrated. The crude
product was purified by column chromatography using silica gel
and ethyl acetate as the eluent. Yield 46 mg (56%). 1H NMR
(chloroform-d1): ꢂ 4.10 (s, 5H), 4.00 (t, 2H), 3.85 (t, 2H), 2.56
(s, 2H). Anal. Found: C, 59.78; H, 5.36; N, 6.73%. Calcd for
C10H11NFe: C, 59.74; H, 5.51; N, 6.97%.
Propionylaminoferrocene (2). Aminoferrocene (0.31 g, 1.5
mmol) and triethylamine (0.2 mL, 1.7 mmol) were dissolved in
CH2Cl2 (10 mL). Propionyl chloride (0.2 mL, 1.7 mmol) was add-
ed to the solution cooling in an ice bath. After stirring overnight,
water (30 mL) and diethyl ether (20 mL) were added to the reac-
tion mixture. The product was extracted with ethyl acetate. The
combined organic layer was washed successively with 2% HCl
aq, sat. NaCl aq, 4% NaHCO3 aq, and sat. NaCl aq, dried over
Na2SO4, and concentrated to dryness in vacuo. The crude product
was recrystallized from hot ethyl acetate to give orange needles.
Single crystals for X-ray analysis were prepared by recrystalliza-
tion from propionitrile. Yield 0.36 g (92%). 1H NMR (chloroform-
d1): ꢂ 6.41 (s, 1H), 4.58 (t, 2H), 4.15 (s, 5H), 3.99 (t, 2H), 2.27 (q,
2H), 1.22 (t, 3H). Anal. Found: C, 60.23; H, 5.93; N, 5.22%. Calcd
for C13H15NOFe: C, 60.73; H, 5.88; N, 5.45%.
neighboring molecules forming an intermolecular NH O=C hy-
ꢁꢁꢁ
drogen bond with each other were positioned for the dimeric mod-
el according to the atomic coordinates, determined by X-ray anal-
ysis. For 1 and 3, which have two distinct molecules in the asym-
metric unit, two kinds of dimeric models were used. Ab initio cal-
culations (single point energy) were performed by Wavefunction
Inc. SPARTAN (ver. 4.1.2) using 3-21GðꢀÞ basis set.
X-ray Data Collection and Structural Determination. Each
of suitable crystals of [(C5H5)Fe(C5H4NHCOR)] (R ¼ CH3 (1),
CH2CH3 (2), t-Bu (3), and CF3 (4)) was sealed in a glass capillary
under an argon atmosphere. X-ray measurements were made at 23
ꢂC on a Rigaku AFC5R (1 and 4) or AFC7R (2 and 3) diffractom-
ꢀ
eter with graphite monochromated Mo Kꢃ radiation (0.71069 A)
using the !-2ꢄ scan up to 2ꢄmax ¼ 55ꢂ. The unit cell dimensions
were refined with 25 reflections. The basic crystallographic
parameters are listed in Table 1. Three standard reflections were
chosen and monitored with every 150 reflections, and did not
show any significant change. An absorption correction based on
azimuthal scans of several reflections was applied for 2 and 4.
The structures were solved by a direct method (SHELXS 86)32 us-
ing the teXsan crystallographic software package of the Molecular
Structure Corp. The positions of one t-Bu group in 3 were disor-
dered. The multiplicity for the major position (C107–C109) was
converged to 0.511. All non-hydrogen atoms were refined
Pivaloylaminoferrocene (3). The compound was synthesized
by a similar method as described for 2. Recrystallization from hot
ethyl acetate gave orange needles in 74% yield. 1H NMR (chloro-
form-d1): ꢂ 6.62 (s, 1H), 4.61 (t, 2H), 4.13 (s, 5H), 3.98 (t, 2H),
1.27 (s, 9H). Anal. Found: C, 62.98; H, 6.52; N, 4.97%. Calcd

1272 Bull. Chem. Soc. Jpn., 78, No. 7 (2005)
CV of Acylaminoferrocenes in the Solid State
Table 1. Crystallographic Data for [(C5H4NHCOR)Fe(C5H5)] (1–4)
R
CH3 (1)
CH2CH3 (2)
t-Bu (3)
CF3 (4)
Empirical formula
Formula weight
Crystal system
C12H13NOFe
243.09
monoclinic
5.9845(4)
7.520(6)
23.58(1)
92.93(4)
1060(1)
P21
C13H15ONFe
257.11
monoclinic
9.033(3)
13.112(4)
10.158(5)
108.29(3)
1142.4(8)
P21=c
C15H19ONFe
285.17
orthorhombic
14.939(3)
18.097(3)
10.158(2)
90
2746.1(9)
P212121
8
C12H10ONFeF3
297.06
monoclinic
15.344(1)
7.640(2)
10.202(1)
109.043(7)
1130.5(3)
P21=c
ꢀ
a/A
ꢀ
b/A
ꢀ
c/A
ꢅ/deg
ꢀ 3
V/A
Space group
Z
4
1.523
1.390
2621
0.036
0.106
1.07
4
1.495
1.295
2625
0.046
0.143
1.04
4
1.745
1.358
2588
0.040
0.110
0.99
Dcalc/g cmꢃ3
ꢆ(Mo Kꢃ)/mmꢃ1
No. of reflections
1.379
1.085
3542
0.036
0.106
1.00
2
2
R1½Fo > 2ꢁðFo Þꢄ
wR2
GOF
O
C
R
N
H
Fe
R = CH3 (1), Et (2), t-Bu (3), CF3 (4)
Chart 1.
anisotropically on F2 by the full-matrix least-squares method
(SHELXL97).33 All hydrogen atoms were placed at the calculated
positions. The final difference Fourier map showed no significant
features. Atom scattering factors and dispersion corrections were
taken from the International Table.34 Crystallographic data have
been deposited with Cambridge Crystallographic Data Centre:
Deposition numbers CCDC 257497–257500 for 1–4. Copies of
the data can be obtained free of charge via http://www.ccdc.cam.
ac.uk/conts/retrieving.html (or from the Cambridge Crystallo-
graphic Data Centre, 12, Union Road, Cambridge, CB2 1EZ,
UK; Fax: +44 1223 336033; e-mail: deposit@ccdc.cam.ac.uk).
Fig. 1. One-dimensional NH O=C hydrogen bond chain
ꢁꢁꢁ
in [(C5H5)Fe(C5H4NHCOR)] where R indicates a) CH3
(1), b) CH2CH3 (2), c) t-Bu (3), and d) CF3 (4).
angles are listed in Table 2. Conventional values of the bond
lengths and the angles were found. The amide planes of 1
and 4 show high coplanarity (dihedral angles between two
planes are 7.2ꢂ for 1 and 34.0ꢂ for 4), and their hydrogen-bond-
ing motifs exhibit the typical glide for 1 and translation for 4,
respectively.36 On the other hand, the amide planes of 2 (85.2ꢂ)
and 3 (86.8ꢂ) are mutually perpendicular. Semiempirical and
Results and Discussion
Synthesis. A series of acylaminoferrocenes, [(C5H5)Fe-
(C5H4NHCOR)] (R ¼ CH3 (1), CH2CH3 (2), t-Bu (3), and
CF3 (4)) (Chart 1), was systematically synthesized to reveal
the electrostatic and steric contributions to the hydrogen-bond
formation. The smallest substituent, formyl group (R ¼ H),
was not used. The derivative exists as an equilibrium mixture
of monomers and oligomers in solution, and does not give
single crystals due to cis amide hydrogen-bonding bridge.35
Intermolecular NHÁÁÁO=C Hydrogen Bonds in the
Crystal. Compounds 1–4 were readily crystallized to form
ab initio calculations showed that the strength of the NH O=
ꢁꢁꢁ
C hydrogen bond, or the interaction energy between two
hydrogen-bonding amide planes, is independent of the torsion
angle(O)C–N–C(O)–N between the two amide planes.37 Thus,
we disregard the coplanarity of the two amide planes in any
discussion about the strength of the hydrogen bond.
intermolecular NH O=C hydrogen bonds. Hydrogen-bond
ꢁꢁꢁ
chains in the crystal are shown as broken lines in Fig. 1. Both
compounds 1 and 3 exhibit two crystallographically independ-
ent molecules in the asymmetric unit. Recently, the structure
of 1 was reported, but the space group is P21=n and the R1 fac-
tor is unacceptably large (15.8%),18 a finding that must be the
result of an incorrect solution. The selected bond length and
Taylor et al. have reported a geometrical analysis of the
NH O=C hydrogen bonds and a statistically significant
ꢁꢁꢁ
tendency for hydrogen-bond formation.38,39 We evaluated
Taylor’s geometrical parameters for the hydrogen bonds in
1–4, and summarize the results in Table 3. The definitions of
the parameters are represented in Fig. 2. All N–H bond dis-

T. Okamura et al.
Bull. Chem. Soc. Jpn., 78, No. 7 (2005) 1273
ꢀ
Table 2. Selected Bond Length (A) and Angles (deg) in [(C5H5)Fe(C5H4NHCOR)]
R
CH3 (1)
molecule
CH2CH3 (2)
t-Bu (3)
molecule
CF3 (4)
molecule
B
molecule
B
A
A
Fe–C mean
Fe–Cp mean
2.024
1.641
2.042
1.656
2.047
1.655
2.040
1.650
2.028
1.649
2.033
1.650
C11–N1–C16–O1
(C21–N2–O26–O2)aÞ
C16–N1–C11–C12
(C26–N2–C21–C22)aÞ
ꢃ1:5ð9Þ
1.6(11)
1.0(6)
ꢃ1:3ð9Þ
3.2(9)
ꢃ1:8ð7Þ
ꢃ17:6ð10Þ
ꢃ159:3ð6Þ
ꢃ3:3ð6Þ
ꢃ172:2ð6Þ
179(6)
29.5(6)
Cp–Cp0bÞ
Cp–amide planecÞ
1.2
18.9
0.5
23.7
1.0
2.1
2.5
6.4
1.5
5.1
1.2
26.2
a) Parenthesis indicates the numbering for molecule B. b) Dihedral angle between two Cp rings in a molecule.
c) Dihedral angle between an amide plane and the adjacent Cp ring.
Table 3. The Distances and Angles of [(C5H5)Fe(C5H4NHCOR)]aÞ
R
CH3 (1)
CH2CH3 (2)
t-Bu (3)
CF3 (4)
reported valuebÞ
1.812
1.866
2.840(7)
2.878(8)
175.7
166.5
83.6
86.3
56.5
55.0
1.960
1.990
2.918(6)
2.972(6)
153.5
158.3
88.4
86.2
69.8
75.5
ꢀ
NH O (A)
2.059
2.088
1:869 ꢅ 0:108
ꢁꢁꢁ
ꢀ
N O (A)
ꢁꢁꢁ
3.038(4)
157.8
87.4
2.950(5)
139.8
81.9
2:850 ꢅ 0:089
161:3 ꢅ 11:6
NH O (deg)
ꢁꢁꢁ
ꢄ
ꢇ
50.1
63.0
ꢀ
a) NH bond distances are normalized to 1.030 A which is the standard value determined by neutron diffrac-
tion.38,39 b) Ref. 38.
oretical calculations.37 The wider angle (175.7ꢂ) of 1 reflects
z
ꢀ
the corresponding shorter N O contact (2.840(9) A). The lin-
ꢁꢁꢁ
θ
earity of NH O=C suggests that the order of the strength of
hydrogen bond is 1 > 2 > 3 > 4.
ꢁꢁꢁ
y
C
O
In an ideal NH O=C hydrogen bond, the ꢄ and ꢇ values
ꢁꢁꢁ
H
N
x
equal 90ꢂ and 30ꢂ, respectively; that is, the hydrogen atom is
placed on the lobe of the sp2 lone-pair of the carbonyl group.39
Actually, statistics have indicated a significant tendency for
hydrogen bonds to form near the idealized sp2 lone-pair direc-
tion (i.e., 77ꢂ < ꢄ ꢆ 90ꢂ, 20ꢂ < ꢇ < 30ꢂ).38,39 Theoretical cal-
culations have revealed the optimal angles to be ꢄ = ca. 50ꢂ
and ꢇ ¼ 90ꢂ.37 If the discussion is limited to the contribution
of ꢄ and ꢇ, the observed trend for the hydrogen bonds is
2 > 1 > 4 > 3. Because many competitive geometrical factors
contribute to the strength of the hydrogen bonds, we were un-
able to determine the exact order of the strength. At the very
least, we can conclude here that the hydrogen bonds in 1 are
significantly and remarkably strong.
φ
Fig. 2. Illustration for the angles defined by the orientation
of amide NH toward the lone pair on carbonyl group.38
ꢀ
tances were normalized to 1.030 A, which is the reported
mean value determined by neutron diffraction. The statistical
mean values of 57 observations have been reported, (i.e.,
ꢀ
ꢀ
NH O=C ¼ 1:869 ꢅ 0:108 A, N O ¼ 2:850 ꢅ 0:089 A,
ꢁꢁꢁ ꢁꢁꢁ
and N{H O ¼ 161:3 ꢅ 11:6ꢂ).38 The two NH O=C dis-
ꢁꢁꢁ
ꢁꢁꢁ
ꢀ
tances (1.812 and 1.866 A) of 1 are significantly shorter than
the reported mean value. The NH O=C distances of 2
ꢁꢁꢁ
ꢀ
(2.059 A) and 3 (1.960, 1.990 A) are slightly longer. The cal-
ꢀ
culated optimal NH O=C distances have been reported to be
1.94 A in the gas phase and 2.20 A in aqueous solution.37 Be-
cause the distances for 2 and 3 are close to these values, they
IR Spectra. The IR spectral data of 1–4 in the amide re-
gion are summarized in Table 4. In a dilute CH2Cl2 solution,
only non-hydrogen bonding, (i.e. the ‘‘free’’ amide group)
was found. The differences in the wavenumber of the absorp-
tion maxima are ascribed to an inductive effect of the substitu-
ent, R. In the solid state, the presence of two sets of distinct
ꢈ(NH) and ꢈ(C=O) bands of 1 with the 1:1 ratio of the inten-
sity is attributed to the two unequivalent molecules in the crys-
tal. The ꢈ(NH) at the lowest wavenumber (3211 cmꢃ1) is as-
ꢁꢁꢁ
ꢀ
ꢀ
are considered to form effective NH O=C hydrogen bonds.
ꢁꢁꢁ
When the H O distance is short, the NH O=C angle has a
ꢁꢁꢁ
ꢁꢁꢁ
tendency to increase, thereby avoiding an N O non-bonded
ꢁꢁꢁ
interaction. Actually, the mean of the distribution of the report-
ed NH O=C angles is 161.2(3)ꢂ.38 The extended conforma-
ꢁꢁꢁ
tion (i.e. NH O=C ¼ 180ꢂ) is favored and supported by the-
ꢁꢁꢁ

1274 Bull. Chem. Soc. Jpn., 78, No. 7 (2005)
CV of Acylaminoferrocenes in the Solid State
Table 4. IR Spectral Data (cmꢃ1) of [(C5H5)Fe(C5H4NHCOR)]
in CH2Cl2
ꢈ(C=O)
in the Solid state
R
ꢈ(NH)
ꢈ(NH)
ꢈ(C=O)
ꢁꢈ(NH)aÞ
ꢁꢈ(C=O)bÞ
CH3
CH2CH3
t-Bu
3432
3435
3455
3417
1687
1688
1678
1738
3259, 3211
3295, 3210
3301
1654
1665
1653
1718
173, 221
140, 225
154
33
23
25
20
CF3
3328
89
a) ꢁꢈðNHÞ ¼ ꢈðNHÞsolid ꢃ ꢈðNHÞsolution. b) ꢁꢈðC=OÞ ¼ ꢈðC=OÞsolid ꢃ ꢈðC=OÞsolution
.
signable to the amide NH with the shortest NH O=C contact
ꢁꢁꢁ
Table 5. Chemical Shift (ꢂ) of the Amide NH Proton of
[(C5H5)Fe(C5H4NHCOR)]
ꢀ
(1.812 A). Complex 2 also showed two distinct ꢈ(NH) at 3295
(major) and 3210 cmꢃ1 (minor) with an integral ratio of 3:1.
However, we detected no distinct molecules in the crystal by
X-ray diffraction analysis. This finding suggests the presence
of two or more distinct amide groups on the time scale of
IR, which are averaged and could not be distinguished by
X-ray diffraction analysis. We evaluated the strength of the
hydrogen bonds in the solid state from ꢁꢈ(NH). The values
of ꢈ(NH) and ꢁꢈ(NH) suggest the strength of the hydrogen
bonds to be in the order of 1 > 2 ꢇ 3 > 4.
R
in CD2Cl2
in the Solid state
Differences
CH3
CH2CH3
t-Bu
6.6
6.4
6.6
7.3
10.6
9.8
10.0
9.4
8.6
8.8
4.0
3.4
3.4
2.8
2.2
2.2
CF3
9.2
1.9
established by the X-ray analysis. However, 2 crystallograph-
ically exhibits only one kind of molecule. As described in the
IR spectra, 2 shows two distinct ꢈ(NH) bands. The split peaks
1H NMR Spectra. The chemical shifts of the amide NH
signals in dichloromethane-d2 and in the solid state are sum-
1
in H NMR also suggest the presence of two kinds of amide
1
marized in Table 5. H NMR spectra in the solid state (i.e.,
groups that are indistinguishable from each other by X-ray
analysis.
The peak at the lowest field (10.6 ppm) corresponds to the
CRAMPS) are shown in Fig. 3. In dichloromethane-d2 solu-
tion, amide NH protons of 1–4 were observed at 6.4–7.3
ppm. These amides are free under this condition, and the dif-
ferences of the chemical shifts are because of the inductive ef-
fect of the substituents, such as those found in the IR spectra.
In the solid state, two distinct signals were found in 1, 2, and 3.
For 1 and 3, two independent molecules exist in the crystal, as
strongest NH O=C hydrogen bond in 1. The small difference
(1.9 ppm) of the chemical shift for 4 between the solid state
and in solution agrees with the small ꢁꢈ(NH) value (89
ꢁꢁꢁ
cmꢃ1). In the H NMR spectra, a trend of the strength of the
1
hydrogen bond, 1 > 2 ¼ 3 > 4, was found. The difference be-
ꢀ
Fig. 3. 1H NMR spectra (CRAMPS) of 1–4 in the solid state. denotes the peaks of TTMS (tetrakis(trimethylsilyl)silane as
standard).

T. Okamura et al.
Bull. Chem. Soc. Jpn., 78, No. 7 (2005) 1275
tween 2 and 3, as found in the IR spectra, was not discriminat-
1
caused by intermolecular electronic interactions in the crystal.
In general, the redox behavior is influenced by several fac-
tors: (i) solubility of the sample in an aqueous medium, (ii)
identity of the electrolyte, (iii) concentration of the sample
in carbon paste, (iv) temperature, (v) scan rate, (vi) mobility
of ions in the solid state, etc.42,43 When ferrocene is oxidized
to the ferrocenium cation, it becomes soluble in aqueous me-
dia. Previously, the diffusion-controlled redox behavior was
believed to be induced by diffusion of the ferrocenium cation
from carbon paste to an aqueous medium;42 but this hypothesis
has been disproved. Even when a completely water-insoluble
solid was used during the redox process, a diffusion-controlled
voltammetric process was observed. From this result, mass or
ion transport has been considered to take place across the
solid–solution interface at a sufficiently fast rate to observe
a measurable current, and ions must diffuse within the solid
to achieve its oxidation.28,29 Especially, at a low scan rate, a
diffusion-controlled process occurs in the thick layer of the
solid. Thus, electron and ion transports within the solid are
necessary.29,42
ed by H NMR.
Recently, CRAMPS has come to one of the most powerful
techniques used in detecting hydrogen bonds. In some cases,
however, disagreements between the CRAMPS and the IR data
can occur. These discordant results have been attributed to the
dependency of IR on the transition dipole moment and the sig-
nificant differential time scale between the two methods.40
Cyclic Voltammograms. Each compound 1–4 exhibits a
reversible redox couple in dichloromethane. The E1=2 values
are found at 0.34 V (vs SCE) for 1–3, and 0.49 V for 4. Under
the same condition, ferrocene shows a redox couple at 0.48 V.
The redox potentials of the ferrocene derivatives have been re-
ported to be controlled by the electronic effects of the substitu-
ents The values show a good correlation with the Hammett
substituent parameters, except for NH2 and NHCOMe, as
reported previously by Lever et al.41 The unexpected negative
shifts for 1–3, of which the Hammett constants nearly equal 0,
are considered to be caused by delocalization of the ꢀ-electron
on the amide N to the adjacent Cp ring. The delocalization
increases the acidity of the amide proton and its ability to form
In the case of 1 and 2, the observed large current values sug-
gest the presence of a certain electron-transport mechanism
that is different from those of 3 and 4. Here, we propose a
probable electron-transport process that occurs via intermolec-
intermolecular NH O=C hydrogen bonds.
ꢁꢁꢁ
Cyclic voltammograms of 1–4 in the solid state are shown
in Fig. 4. Ferrocene shows a reversible redox couple at 0.26
V vs SCE under the same condition. Compounds 1 and 2
show an irreversible redox process and a large current value,
although both 3 and 4 show a reversible couple with a small
current value, as was found for ferrocene. The feature of the
cyclic voltammograms was not significantly dependent on
the method of the preparation (particle size, the extent of mix-
ing, concentration of the sample, etc.). The difference must be
ular NH O=C hydrogen bonds. In the case of Prussian blue
(PB), the large current (ꢈ1 mA) is caused by the diffusion
ꢁꢁꢁ
of potassium ions through the PB lattice.25
In our case, electron transfer is considered to occur in the
solid-carbon paste interface, as illustrated in Fig. 5. The redox
reaction probably proceeds in the thick layer of the electrode.
Even when the surface on the electrode was covered with pure
carbon paste, an effective redox process was observed. In the
ordinary case (i.e., for 3 and 4), a molecule near the electrode
(part of metallic rod) is electrochemically oxidized in the first
step. The neighboring molecule then diffuses toward the elec-
trode, and is then oxidized (step 2). Because the diffusion of
a molecule is slower than that of the redox reaction, all of
the redox-active species near the electrode are finally oxidized,
resulting in a decrease of the current.
In the case of 1 and 2, the situation is very different (Fig. 5c).
These compounds form strong NH O=C intermolecular hy-
ꢁꢁꢁ
drogen bonds, as established by X-ray diffraction analysis
and spectral measurements in the crystal. Once a molecule
near the electrode is oxidized (step 1), an electron on the
neighboring molecule is then transferred to the oxidized mole-
cule through the hydrogen bond (step 2). These molecules con-
stitute a dimeric mixed-valence structure, FeII–FeIII ꢀ FeIII–
FeII, connected by a stronger NH O=C hydrogen bond than
ꢁꢁꢁ
the FeII–FeII state, because the acidity of NH is increased by
the inductive effect of the electron-deficient Fe(III) ion.
The observed electron transfer through the NH O=C hydro-
ꢁꢁꢁ
gen bond is thus explained as fast electron transfer without dif-
fusion control. In the first step of the oxidation process, mole-
cules near the metallic rod are oxidized at an oxidation poten-
tial similar to that of 3 and 4. The anodic current increases with-
out any diffusion control, as shown in Fig. 4. In the oxidation
process, the oxidized molecules are connected successively
Fig. 4. Cyclic voltammograms of [(C5H5)Fe(C5H4-
NHCOR)] in the solid state, where R indicates a) CH3
(1), b) CH2CH3 (2), c) t-Bu (3), and d) CF3 (4). The scans
for 1 and 2 were started at ꢃ0:4 V and swept to the neg-
ative side (2 cycles).
by very strong intermolecular NH O=C hydrogen bonds to
produce a cationic mixed-valence polymer (Fig. 5c, Eq. 1).
ꢁꢁꢁ

1276 Bull. Chem. Soc. Jpn., 78, No. 7 (2005)
CV of Acylaminoferrocenes in the Solid State
ꢁ ꢁ ꢁ {FeII{FeIII{FeIII{FeII{FeIII{FeII{ ꢁ ꢁ ꢁ
eꢃ
ꢃꢃꢃ! ꢁ ꢁ ꢁ {FeII{FeIII{FeII þ FeII{FeIII{FeII{ ꢁ ꢁ ꢁ :
ð3Þ
FeII-rich oligomers
During the cathodic sweep, potentially reducible FeIII spe-
cies remain in the oligomer at the reduction potential of mono-
meric FeIII. Despite hydrogen bonds being weakened by the re-
duction, oligomeric structures probably remain at the reduction
potential of a monomeric acylaminoferrocenium cation. Thus,
the cathodic current slowly increases at the exceeded potential
for the reduction of monomeric compounds. Finally, the par-
tially oxidized oligomers are reduced all at once; that is, a mul-
ti-electron transfer occurs at a sufficiently negative potential
(i.e., at approximately ꢃ0:8 V, as shown in Fig. 4 and Eq. 4).
At ca. ꢃ0:80 V,
meꢃ
(FeIII)m{(FeII)n
ꢃꢃꢃ! (n þ m) FeII: ð4Þ
monomers
mixed valent oligomer (n > m)
Competitive Intermolecular NHÁÁÁS and NHÁÁÁO=C Hy-
drogen Bond. To confirm the relationship between the pres-
ence of a strong intermolecular NH O=C hydrogen bond
ꢁꢁꢁ
chain and a large current value, we performed a cleavage of
the hydrogen bond chain by addition of the ethanethiolate
anion. Thus, the ethanethiolate anion was added to acetylami-
noferrocene (1) in CH Cl to form the intermolecular NH S
2
ꢁꢁꢁ
2
hydrogen bond, the existence of which was established by
IR measurements; the mixture was then concentrated to
dryness. The IR spectra of the residue were measured. As
shown in Fig. 6, 1 exhibits two kinds of strong intermolecular
Fig. 5. Schematic representation of a) a carbon paste elec-
trode including acylaminoferrocenes and the proposed ox-
idation process within the electrode for the cyclic voltam-
mograms of b) 3 and 4 and c) 1 and 2 in the solid state.
NH O=C hydrogen bonds. The addition of 1 equiv (n-Bu) -
4
ꢁꢁꢁ
NSEt broke half of the NH O=C bonds, and formed intermo-
ꢁꢁꢁ
lecular NH S hydrogen bonds. The resulting free ꢈ(C=O) was
ꢁꢁꢁ
FeII ꢁ ꢁ ꢁ FeII ꢁ ꢁ ꢁ FeII ꢁ ꢁ ꢁ FeII ꢁ ꢁ ꢁ FeII ꢁ ꢁ ꢁ FeII ꢁ ꢁ ꢁ
observed at 1669 cmꢃ1. The corresponding cyclic votammo-
gram showed a quasi-reversible redox process at E1=2 ¼ 0:16
V vs SCE, and a slightly smaller current value (100 mA). After
the addition of 10 equiv of thiolate, all amide groups were
ꢃ2eꢃ
ꢃꢃꢃ! FeIII{FeIII{FeII ꢁ ꢁ ꢁ FeII ꢁ ꢁ ꢁ FeII ꢁ ꢁ ꢁ FeII ꢁ ꢁ ꢁ : ð1Þ
During a cathodic scan, the current increases continuously
and decreases sharply down to 0 A at approximately ꢃ0:8
V. This irreversible reduction process is explained as follows.
In the reduction process, the fully oxidized hydrogen-bonded
polymer is reduced near the metallic rod to form a mixed-va-
lence polymer (Eq. 2). The reduction potential of the polymer
must be more negative than that of the monomer when the ma-
jority is in the FeII state; for example, the associated mixed va-
lence, FeII–FeIII–FeII, is more difficult to be reduced than
monomeric FeIII. The stepwise reduction increases the popula-
tion of FeII in the polymer. As mentioned above, hydrogen
bonds in the FeII state are weaker than those in the FeIII state.
Thus, the reduction accompanies a weakening and dissociation
of the hydrogen bonds, resulting in the formation of FeII-rich
oligomers (Eq. 3), in which inducing reduction becomes
increasingly difficult.
engaged in the intermolecular NH S hydrogen bonds. The
ꢁꢁꢁ
reversibility of the cyclic voltammogram increased, and the
current decreased. This phenomenon was not caused by a sim-
ple electrolyte. When 10 equiv Ph4PBr was used instead of (n-
Bu)4NSEt, IR spectra and cyclic voltammograms showed sim-
ilar features to the original ones. These results clearly indicate
the formation of NH S interrupts electron transfer through the
intermolecular NH O=C hydrogen bonds.
ꢁꢁꢁ
ꢁꢁꢁ
Theoretical Calculation of the Strength of NHÁÁÁO=C
Hydrogen Bonds.
The electrochemical measurements
demonstrated that the hydrogen bonds in 1 and 2 are stronger
than those of 3 and 4. This trend was found in the crystal data
and the IR spectra. However, a significant difference of the
strength of the hydrogen bonds was not detected between 2
and 3 based on the 1H NMR and the IR spectra. The measure-
ments evaluated the strength of the N–H bonds, or the elec-
tronic environment around the protons. However, the 1H NMR
ꢁ ꢁ ꢁ {FeIII{FeIII{FeIII{FeIII{FeIII{FeIII{ ꢁ ꢁ ꢁ
fully oxidized polymer
and the IR spectra did not directly detect H O bonds, nor did
these methods estimate the strength of the bond. Because a di-
ꢁꢁꢁ
eꢃ
ꢃꢃꢃ! ꢁ ꢁ ꢁ {FeII{FeIII{FeIII{FeIII{FeIII{FeIII{ ꢁ ꢁ ꢁ ;
partially reduced (mixed valent) polymer
ð2Þ
rect evaluation of H O bonds by spectral measurements is
ꢁꢁꢁ
very difficult, we estimated the strength of the H O bonds
ꢁꢁꢁ

T. Okamura et al.
Bull. Chem. Soc. Jpn., 78, No. 7 (2005) 1277
Fig. 6. IR spectra and cyclic voltammograms of a) acetylaminoferrocene (1) b) 1 + 1 equiv (n-Bu)4NSEt and c) 1 and 10 equiv
(n-Bu)4NSEt in the solid state.
Table 6. Calculated Bond Order of Hydrogen Bonds in
Acylaminoferrocene, [(C5H5)Fe(C5H4NHCOR)], Using
the Dimeric Models
R
H O
ꢁꢁꢁ
1
CH3 longaÞ
CH3 shortaÞ
CH2CH3
t-Bu longaÞ
t-Bu shortaÞ
CF3
0.09556
0.11301
0.07903
0.05922
0.07883
0.05861
2
3
4
a) The dimeric model linked by the longer NH O hydrogen
ꢁꢁꢁ
bonds is expressed by ‘‘long’’ and the other is ‘‘short’’.
by ab initio calculation using the 3-21GðꢀÞ basis set. The calcu-
lated bond orders are summarized in Table 6, where two crys-
tallographically distinct dimeric molecules were used for 1 and
3. The bond orders of H O for 1 are significantly large, a find-
ꢁꢁꢁ
ing that is in agreement with the results of X-ray, spectral and
electrochemical analyses. For 2, the bond order is relatively
large. These results suggest that all of the molecules in 1
Fig. 7. Intermolecular contacts of [(C5H5)Fe(C5H4-
NHCOC2H5)] (2) in the crystal.
and 2 are connected by continuous strong H O bonds, which
means that the hydrogen bonds are neither weak nor broken at
any point. In the crystal of 3, the long and short hydrogen
mechanism has been considered. Figure 7 shows the crystal
packing of 2. Intermolecular short contacts, C14 C15
ꢁꢁꢁ
ꢁꢁꢁ
ꢀ
(3.595(6) A) and C11 C21 (3.529(6) A), are found. The short
ꢀ
ꢁꢁꢁ
ꢀ
Cp Cp contact, where H4 H5 is 2.309 A, probably functions
bonds array alternatively. The bond order of the H O=C hy-
ꢁꢁꢁ
ꢁꢁꢁ
ꢁꢁꢁ
drogen bond in the short one is comparable to that of 2, but the
long one exhibits a small bond order. The presumable hydro-
gen bond chains in 3 are essentially weakened, or broken.
Contribution of Electron Hopping to the Cyclic
Voltammograms. Generally, long electron transfer occurs
by super-exchange (electron tunneling), or electron-hopping
mechanisms.44,45 Recent studies have revealed that hydrogen
bonds in the 310-helix promote an efficient electron transfer,45
and that a suitable array of aromatic rings along the helical
peptide increases the electron hopping.46 In our case, a remark-
able difference of cyclic voltammograms between 2 and 3 was
observed despite an unclear spectral contrast between the hy-
drogen bonds. To elucidate this point, the electron-hopping
as an effective pathway by a hopping mechanism, or ꢀ-conju-
gation. Each molecule is connected by hydrogen bonds to the
adjacent molecules forming a continuous and expanded ꢀ-
electron system. On the other hand, the shortest intermolecular
ꢀ
Cp Cp contact in the crystal of 3 is C22 C33 (3.79(1) A),
ꢁꢁꢁ
ꢁꢁꢁ
where the Cp rings are almost perpendicular to each other,
and too far to interact effectively, resulting in isolated hydro-
gen bonds.
Conclusion
Acylaminoferrocenes, [(C5H5)Fe(C5H4NHCOR)], were
found to form characteristic intermolecular NH O=C hydro-
gen-bond chains for which the bond strength depends quite
ꢁꢁꢁ

1278 Bull. Chem. Soc. Jpn., 78, No. 7 (2005)
CV of Acylaminoferrocenes in the Solid State
20 T. Okamura, S. Takamizawa, N. Ueyama, and A.
Nakamura, Inorg. Chem., 37, 18 (1998).
21 N. Ueyama, N. Nishikawa, Y. Yamada, T. Okamura,
S. Oka, H. Sakurai, and A. Nakamura, Inorg. Chem., 37, 2415
(1998).
22 N. Ueyama, N. Nishikawa, Y. Yamada, T. Okamura, and
A. Nakamura, Inorg. Chim. Acta, 283, 91 (1998).
23 N. Ueyama, Y. Yamada, T. Okamura, S. Kimura, and A.
Nakamura, Inorg. Chem., 35, 6473 (1996).
24 N. Ueyama, T. Okamura, and A. Nakamura, J. Am. Chem.
Soc., 114, 8129 (1992).
effectively on the substituent, R. The orientation of the amide
proton with respect to the carbonyl group contributes to the
strength of the hydrogen bonds. For R ¼ CH3 and CH2CH3,
a unique electron transfer was detected critically by cyclic vol-
tammetry in the solid state in terms of a large current and an
irreversible redox process. Ferrocene derivatives having such
strong hydrogen bonds are useful components to construct
and strategically develop supramolecules with electronic
properties. Such a combination of metal complexes with strong
hydrogen-bond chains will widely expand applications, and
will lead to the development of a new type of electronic
devices (e.g., semi-conductor with a regulative path for elec-
tron transfer).
25 N. F. Zakharchuk, B. Meyer, H. Hennig, F. Scholz, A.
Jaworksi, and Z. Stojek, J. Electroanal. Chem., 398, 23 (1995).
26 A. Dostal, B. Meyer, F. Scholz, U. Schroder, A. M. Bond,
¨
F. Marken, and S. J. Shaw, J. Phys. Chem., 99, 2096 (1995).
This work was supported by a Grant-in-Aid from the Min-
istry of Education, Culture, Sports, Science and Technology.
27 A. Dostal, U. Schroder, and F. Scholz, Inorg. Chem., 34,
1711 (1995).
¨
28 A. M. Bond, R. Colton, F. Daniels, D. R. Fernando,
F. Marken, Y. Nagaosa, R. F. V. Steveninck, and J. N. Walter,
J. Am. Chem. Soc., 115, 9556 (1993).
References
1
R. Rulkens, A. J. Lough, I. Manners, S. R. Lovelace, C.
Grant, and W. E. Geiger, J. Am. Chem. Soc., 118, 12683 (1996).
J.-M. Lehn, ‘‘Supramolecular Chemistry—Concepts and
Perspectives,’’ VCH, Tokyo (1995).
J. C. MacDonald and G. M. Whitesides, Chem. Rev., 94,
2383 (1994).
29 A. M. Bond, R. Colton, F. Marken, and J. N. Walter,
Organometallics, 13, 5122 (1994).
2
30 F. Scholz and B. Meyer, Chem. Soc. Rev., 23, 341 (1994).
31 S. Kitagawa, S. Matsuyama, M. Munakata, N. Osawa, and
H. Masuda, J. Chem. Soc., Dalton Trans., 1991, 1717.
32 G. M. Scheldrick, ‘‘Crystallographic Computing 3,’’
Oxford University Press (1985), pp. 175–189.
3
4
(1995).
G. R. Desiraju, Angew. Chem., Int. Ed. Engl., 34, 2311
33 G. M. Scheldrick, ‘‘SHELXL97, Program for the Refine-
5
G. R. Desiraju, Acc. Chem. Res., 29, 441 (1996).
C. B. Aakeroy and K. R. Seddon, Chem. Soc. Rev., 22, 397
¨
ment of Crystal Structures,’’ University of Gottingen, Germany
¨
(1997).
34 D. T. Cromer, ‘‘International Tables for X-ray Crystallog-
raphy,’’ T. K. Press, Birmingham, England (1974).
6
(1993).
7
D. Philp and J. F. Stoddart, Angew. Chem., Int. Ed. Engl.,
ˇ´
´
35, 1154 (1996).
8 A. D. Burrows, C.-W. Chan, M. M. Chowdgry, J. E.
McGray, and D. M. P. Mingos, Chem. Soc. Rev., 24, 329 (1995).
9 D. Braga, F. Grepioni, and G. R. Desiraju, J. Organomet.
Chem., 548, 33 (1997).
10 T. Okamura, K. Sakauye, N. Ueyama, and A. Nakamura,
Inorg. Chem., 37, 6731 (1998).
11 T. Okamura and N. Ueyama, ‘‘Metal-Hydrogen Bond
Conjugated Systems,’’ ed by A. Nakamura, N. Ueyama, and K.
Yamaguchi, Kodansha, Tokyo (2002), pp. 307–330.
12 M. C. Grossel, M. R. Goldspink, J. A. Hriljac, and S. C.
Weston, Organometallics, 10, 851 (1991).
13 K. Kavallieratos, S. Hwang, and R. H. Crabtree, Inorg.
Chem., 38, 5184 (1999).
35 G. R. Knox, P. L. Pauson, D. Willison, E. Solcaniova, and
S. Toma, Organometallics, 9, 301 (1990).
36 L. Leiserowitz and M. Tuval, Acta Crystallogr., B34, 1230
(1978).
37 H. Adalsteinsson, A. H. Maulitz, and T. C. Bruice, J. Am.
Chem. Soc., 118, 7689 (1996).
38 R. Taylor and O. Kennard, Acc. Chem. Res., 17, 320
(1984).
39 R. Taylor and O. Kennard, Acta Crystallogr., B39, 133
(1983).
40 I. Schnell, S. P. Brown, H. Y. Low, H. Ishida, and H. W.
Spiess, J. Am. Chem. Soc., 120, 11784 (1998).
41 S. Lu, V. V. Strelets, M. F. Ryan, W. J. Pietro, and A. B. P.
Lever, Inorg. Chem., 35, 1013 (1996).
14 D. v. Leusen and B. Hessen, Organometallics, 20, 224
(2001).
42 A. M. Bond and F. Marken, J. Electroanal. Chem., 372,
125 (1994).
15 P. D. Beer, F. Szemes, V. Balzani, C. M. Sala, M. G. B.
Drew, S. W. Dent, and M. Maestri, J. Am. Chem. Soc., 119,
11864 (1997).
16 P. D. Beer, A. R. Graydon, A. O. M. Johnson, and D. K.
Smith, Inorg. Chem., 36, 2112 (1997).
43 F. A. Schultz and T. Kuwana, J. Electroanal. Chem., 10, 95
(1965).
44 M. Sisido, S. Hoshino, H. Kusano, M. Kuragaki, M.
Makino, H. Sasaki, T. A. Smith, and K. P. Ghiggino, J. Phys.
Chem. B, 105, 10407 (2001).
17 J. D. Carr, L. Lambert, D. E. Hibbs, M. B. Hursthouse,
K. M. A. Malik, and J. H. R. Tucker, J. Chem. Soc., Chem.
Commun., 1997, 1649.
18 K. Heinze and M. Schlenker, Eur. J. Inorg. Chem., 2004,
2974.
45 F. Polo, S. Antonello, F. Formaggio, C. Toniolo, and
F. Maran, J. Am. Chem. Soc., 127, 492 (2005); S. Antonello, F.
Formaggio, A. Moretto, C. Toniolo, and F. Maran, J. Am. Chem.
Soc., 125, 2874 (2003).
46 K. Yanagisawa, T. Morita, and S. Kimura, J. Am. Chem.
Soc., 126, 12780 (2004).
19 D. W. Hall and J. H. Richards, J. Org. Chem., 28, 1549
(1963).