Synthesis and characterization of ruthenium(II)-nitrile complexes with bisamide-tpa ligands (tpa = tris(2-pyridylmethyl)amine)

Article abstract of DOI:10.1246/bcsj.78.2152

Ruthenium(II)-acetonitrile complexes having pentadentate tris(2-pyridylmethyl)amine (tpa) derivatives with coordinated amide CO groups were prepared and characterized by various spectroscopic methods, electrochemical measurements, and DFT calculations. The acetonitrile ligand was indicated to tightly bind to the ruthenium(II) center in an η¹-N fashion. The Mulliken charge distribution, obtained by a calculation, indicated that the ruthenium(II)-nitrile bond is more covalent than other coordination bonds of the tpa moiety. The redox potentials of Ru centers and the chemical shifts of methyl groups of the acetonitrile ligands exhibited a linear relationship, indicating that electron density of the Ru center controls that of the acetonitrile ligand. Those complexes showed fluxional behavior in CD ₃CN solutions to exhibit one mode of thermal motion; it also altered the symmetry of the complexes.

Full text of DOI:10.1246/bcsj.78.2152

2152 Bull. Chem. Soc. Jpn., 78, 2152–2158 (2005)
Ó 2005 The Chemical Society of Japan
Synthesis and Characterization of Ruthenium(II)–Nitrile Complexes
with Bisamide-tpa Ligands (tpa = Tris(2-pyridylmethyl)amine)
y
ꢀ;
Takahiko Kojima, Ken-ichi Hayashi, Yoshihito Shiota,1 Yoshimitsu Tachi,1
Yoshinori Naruta,1 Takuya Suzuki,2 Kazuya Uezu,2 and Kazunari Yoshizawa1
Department of Chemistry, Faculty of Sciences, Kyushu University, Hakozaki, Higashi-ku, Fukuoka 812-8581
1Institute for Materials Chemistry and Engineering, Kyushu University, Hakozaki, Higashi-ku, Fukuoka 812-8581
2Department of Environmental Engineering, The University of Kitakyushu,
1-1 Hibikino, Wakamatsu, Kitakyushu 808-0135
Received June 30, 2005; E-mail: cosyscc@mbox.nc.kyushu-u.ac.jp
Ruthenium(II)–acetonitrile complexes having pentadentate tris(2-pyridylmethyl)amine (tpa) derivatives with coor-
dinated amide CO groups were prepared and characterized by various spectroscopic methods, electrochemical measure-
ments, and DFT calculations. The acetonitrile ligand was indicated to tightly bind to the ruthenium(II) center in an ꢀ1-N
fashion. The Mulliken charge distribution, obtained by a calculation, indicated that the ruthenium(II)–nitrile bond is
more covalent than other coordination bonds of the tpa moiety. The redox potentials of Ru centers and the chemical
shifts of methyl groups of the acetonitrile ligands exhibited a linear relationship, indicating that electron density of
the Ru center controls that of the acetonitrile ligand. Those complexes showed fluxional behavior in CD3CN solutions
to exhibit one mode of thermal motion; it also altered the symmetry of the complexes.
Transition-metal complexes having organonitriles have
been recognized to be reactive intermediates toward nucleo-
philic reactions due to ꢁ-donation of the non-bonding lone pair
to electron-deficient metal centers in an ꢀ1-N monodentate
fashion, accompanied by reduction of the electron density at
the CꢁN moiety.1,2 For example, those ꢀ1-nitrile ligands can
be converted to amides via the hydration to nitrile CꢁN triple
bonds.3 In addition, a cationic [RuCp(acetonitrile)]þ has been
shown to perform catalytic cyclotrimerization of alkynes to
afford benzene derivatives.4 On the contrary, the nitriles coor-
dinated to electron-rich and ꢂ-donating metal complexes are
expected to exhibit reactivity toward electrophiles owing to
their ꢂ-accepting character. Thus, the electron density of the
metal center can regulate the reactivity of those coordinated
nitriles. As ligands, organonitriles have been classified to be
moderate ꢁ-donors and ꢂ-acceptors toward transition metals.
The ordering of ꢂ-acceptor character was examined by using
We have reported on the synthesis and reactivity of rutheni-
um complexes having tris(2-pyridylmethyl)amine (tpa) and its
derivatives as ligands.8 In the course of our research on Ru–tpa
complexes having two naphthoylamide groups at the 6-posi-
tions of two pyridine rings in tpa, we found that intramolecular
hydrogen bonding is indispensable to construct and stabilize
specific coordination environments.9 Masuda and co-workers
have also reported on the synthesis and characterization of sim-
ilar ruthenium complexes, involving their application toward
hydrocarbon oxidations.10
We have focused on the role of the intramolecular hydrogen
bonding to regulate the reactivity and property of the rutheni-
um center. We have proposed that the hydrogen bonding could
enrich the electron density at the ruthenium center. Thus, we
chose acetonitrile as a ligand to estimate the ꢂ-back bonding
from the ruthenium center to the ligand by spectroscopic and
electrochemical methods. In this article, we describe the
synthesis and characterization of ruthenium(II)–acetonitrile
complexes with the use of spectroscopic methods, including
EXAFS analysis, electrochemical measurements, and DFT
calculations.
Mossbauer5 and IR6 spectroscopies to demonstrate that the
¨
ꢂ-acceptor character of the nitriles is less than that of N2
and more than that of pyridine.
Since the ruthenium(II) ion has the d6 low-spin configura-
tion, and is classified to be a moderately soft acid based on
the Hard–Soft Acid–Base (HSAB) theory,7 it has an ability to
donate the dꢂ electron to a coordinated ꢂ-acceptor. Thus, the
coordination of nitriles to ruthenium(II) centers should be at-
tributable to not only ꢁ-donation of the lone pair of the nitrile
ligand to the ruthenium(II) center, but ꢂ-back donation from
Experimental
Materials. Precursor complexes, [RuCl(1-Naph2-tpa)]PF6 (1-
Naph2-tpa = N,N-bis(6-(1-naphthoylamide)-2-pyridylmethyl)-N-
(2-pyridylmethyl)amine) (1), [RuCl(2-Naph2-tpa)]PF6 (2-Naph2-
tpa = N,N-bis(6-(2-naphthoylamide)-2-pyridylmethyl)-N-(2-pyri-
dylmethyl)amine) (2), and [RuCl(isob2-tpa)]PF6 (isob2-tpa =
N,N-bis(6-(isobutyrylamide)-2-pyridylmethyl)-N-(2-pyridylmeth-
yl)amine) (3) were prepared as reported.8 CH3CN was dried on
CaH2 and then distilled under N2 and stored on a Molecular
Sieves 3A. All other reagents were of analytical and/or reagent
ꢀ
the dꢂ orbital of ruthenium(II) to the ꢂ orbital of the nitrile.
y Present address: Department of Material and Life Science,
Graduate School of Engineering, Osaka University, 2-1
Yamada-oka, Suita, Osaka 565-0871
Published on the web December 9, 2005; DOI 10.1246/bcsj.78.2152

T. Kojima et al.
Bull. Chem. Soc. Jpn., 78, No. 12 (2005) 2153
grade and used without further purification.
gave satisfactory results. For 5: Anal. Calcd for C42H35N7O2-
P2F12Ru H2O 1/2CH3CN (5 H2O 1/2CH3CN): C, 46.98; H,
3.53; N, 9.56%. Found: C, 47.18; H, 3.42; N, 9.75%. 1H NMR
(CD3CN): ꢄ 2.34 (3H, s, CH3), 4.39 and 4.58 (2H, ABq, J ¼ 18
Hz, CH2), 4.82 and 4.85 (2H, ABq, J ¼ 18 Hz, CH2), 4.91 (2H,
s, CH2), 7.0–8.2 (multiplets, aromatic).
Instrumentation. 1H and 13C NMR spectra were recorded on
JEOL GX-400 and/or EX-270 spectrometers. 1H–13C HMBC
spectra were recorded on a JEOL ꢃ-600 spectrometer. Chemical
shifts were determined relative to residual solvent peaks. UV–
vis absorption spectra were measured in CH3CN on a Jasco Ub-
est-55 UV–vis spectrophotometer at room temperature. Infrared
spectra were recorded as KBr disks in the range of 4000–400
cmꢂ1 on a Jasco IR model 800 infrared spectrophotometer. ESI-
MS spectra were obtained on a Perkin-Elmer Sciex API-300 mass
spectrometer. FAB-MS spectra were measured on a JMS-SX/
SX102A Tandem Mass spectrometer.
ꢃ ꢃ
ꢃ ꢃ
For 6: Anal. Calcd for C28H35N7O2P2F12Ru 1.5H2O (6
ꢃ
ꢃ
1.5H2O): C, 36.57; H, 4.16; N, 10.66%. Found: C, 36.27; H, 3.94;
1
N, 11.03%. H NMR (CD3CN): ꢄ 1.18, 1.28, 1.38, and 1.47 (for
all; 3H, d, J ¼ 7 Hz, CH3 of isob), 2.31 (3H, s, CH3 of CH3CN),
2.80 and 3.13 (1H, 7-tet, J ¼ 7 Hz, CH of isob), 4.13 and 4.58
(2H, ABq, J ¼ 18 Hz, CH2), 4.74 (2H, s, CH2), 4.78 (2H, s, CH2),
7.11–7.27 (multiplets, aromatic). 13C{1H} NMR (CD2Cl2): ꢄ
181.2 and 176.4 (CO of amide), 162.6, 162.0, 157.8, 149.9, 140.2,
139.4, 138.3, 130.6 (CN of CH3CN), 125.1, 121.8, 118.8, 118.6,
116.2, 115.5 (pyr of tpa), 70.5, 70.4, 68.8 (–CH2–), 38.3, 37.5
(CH of isob), 19.83, 19.77, 19.5, 19.1 (CH3 of isob), 4.7 (CH3
of CH3CN).
DFT Calculations. We optimized the geometry of a model
complex, {[Ru(bis(acetoamide)-tpa)(CH3CN)]Cl}þ, which was
constructed based on the crystal structure of 1, using the
B3LYP method,11 which has been reported to provide excellent
descriptions of various reaction profiles, particularly concerning
geometries, heats of reaction, barrier heights, and vibrational anal-
yses. For the Ru atom we used Los Alamos ECP with double-ꢅ
basis,12 and for the H, C, O, and N atoms we used the D95 basis
set,13 a standard double-z basis.
Cyclic voltammograms were recorded on a HECS 312B dc
pulse polarograph (Fuso Electrochemical System) attached to a
HECS 321B potential-sweep unit of the same manufacture, and
an electrochemical analyzer, Model 720 (ALS/chi). A glassy car-
bon (3 mm o.d.) was employed as a working electrode, a platinum
coil as a counter electrode, and a silver/silver nitrate (Ag/AgNO3)
electrode as a reference electrode, respectively. All measurements
were carried out in CH3CN containing 0.1 M [(n-Bu)4N]ClO4 as a
supporting electrolyte under N2 at ambient temperatures. The
redox potentials were determined relative to a ferrocene/ferro-
cenium couple as a reference (0 V).
Elemental analysis data for all compounds were obtained at the
Service Center of the Elemental Analysis of Organic Compounds,
Department of Chemistry, Kyushu University.
EXAFS Measurements. Ru K-edge X-ray absorption ex-
periments were carried out on a BL0B1 beam-line of SPring-8
(Hyogo, Japan). Spectra were collected in the transmission mode.
Solid samples (0.1 g) were mixed and grinded with boron nitride
(0.5 g) and pressed to disks of 0.5 cmf. Curve-fitting analysis of
the EXAFS spectra was performed on the inversely Fourier trans-
formed radial distribution function, and was made by using the
FEFFIT program.
Results and Discussion
Synthesis of Nitrile Complexes. We established the crys-
tal structures of the precursors [RuCl(1-Naph2-tpa)](PF6) (1),9
[RuCl(2-Naph2-tpa)](PF6) (2),9 and [RuCl(isob2-tpa)]PF6 (3)9
to reveal one amide oxygen coordinated to the ruthenium cen-
ter. It bound to the uncoordinated amide NH via an intramolec-
ular hydrogen bonding.
Acetonitrile complexes [Ru(1-Naph2-tpa)(CH3CN)](PF6)2
(4), [Ru(2-Naph2-tpa)(CH3CN)](PF6)2 (5), and [Ru(isob2-
tpa)(CH3CN)](PF6)2 (6) were synthesized in good yields by
treating 1–3 with AgNO3 in CH3CN to remove the chloride li-
gands (Scheme 1).14 In the course of the reaction, the red color
of the chloride complexes turned to be brown, and the LMCT
band from Clꢂ to ruthenium(II) around 430 nm disappeared.
The products were characterized by means of elemental analy-
sis, ESI-MS, IR, NMR, and EXAFS spectroscopies, and also
electrochemical measurements.
Synthesis of [Ru(1-Naph -tpa)(CH CN)](PF ) 1/2CH CN
2
2Á
3
6
3
(4). AgNO3 (47.6 mg, 0.280 mmol) was added to a red solution
of [RuCl(1-Naph2-tpa)]PF6 (1) (200 mg, 0.215 mmol) in CH3CN
(40 mL). The mixture was refluxed for 4 h and the color turned
to be orange; a white precipitate of AgCl was observed. After re-
moving the precipitate by filtration, the filtrate was dried up under
reduced pressure. The residue was dissolved into a small volume
of acetone/EtOH, and then NH4PF6 (40 mg, 0.239 mmol) was
added as solids. The mixture was concentrated to a small volume
to obtain a brown precipitate. The precipitate was filtered and
washed with a small volume of EtOH, followed by Et2O, and then
dried in vacuo to obtain the brown powder of 1 in 84% yield
Concerning the ESI-MS spectrum of 4, peaks assigned to
(178 mg). Anal. Calcd for C42H35N7O2RuP2F12 1/2CH3CN:
ꢃ
C, 47.76; H, 3.40; N, 9.72%. Found: C, 47.80; H, 3.39; N,
9.64%. ESI-MS: 770.2, {[Ru(1-Naph2-tpa)(CH3CN)]-H}þ; 916.3,
{[Ru(1-Naph2-tpa)(CH3CN)]PF6}þ. 1H NMR (CD3CN): ꢄ 2.39
(3H, s, CH3), 4.39 and 4.58 (2H, ABq, J ¼ 18 Hz, CH2), 4.81 and
4.85 (2H, ABq, J ¼ 16 Hz, CH2), 4.91 (2H, pseudo-s, CH2), 6.42
(1H, dd, J ¼ 8 and 7 Hz, 1-Naph-H6), 6.9–8.0 (multiplets, aromat-
ic), 8.25 (1H, d, J ¼ 5 Hz, py-H6), 8.54 (1H, d, J ¼ 8 Hz, 1-Naph-
H2), 9.84 (1H, brs, amide-NH, uncoordinated), 12.04 (ca. 1H, brs,
amide-NH). 13C NMR (CD2Cl2): ꢄ 131.3 (CN).
N
N
N
N
N
AgNO3
N
Ru
H
N
Ru
N
H
N
N
O
CH3CN, ∆
N
O
R
H
Cl
R
N
R
H
R
N
C
C
CH3
Synthesis
1/2CH3CN
of
[Ru(2-Naph -tpa)(CH CN)](PF ) H O
and
2Á
(CH CN)](PF ) 1.5H O (6 1.5H O). These complexes were
Á
O
C
2
3
6
2
O
(5 H O 1/2CH CN)
Á
[Ru(isob2-tpa)-
Á
2
3
R = 1-naphthyl; 1
R = 2-naphthyl; 2
R = isopropyl; 3
R = 1-naphthyl; 4
R = 2-naphthyl; 5
R = isopropyl; 6
2Á
Á
3
6
2
2
prepared by the same procedure as described for 4, except
[RuCl(2-Naph2-tpa)]PF6 (2) and [RuCl(isob2-tpa)]PF6 (3) were
used as the starting materials, respectively. Elemental analysis
Scheme 1. Synthesis of acetonitrile complexes.

2154 Bull. Chem. Soc. Jpn., 78, No. 12 (2005)
Table 1. Spectroscopic Data for Acetonitrile Complexes
Ru–tpa Nitrile Complexes
Complex
ꢆ(CN)aÞ/cmꢂ1
ꢄ(13C/CN)bÞ
ꢄ(1H/CH3)cÞ
4
2268
(2220)dÞ
—
2251
2266
131.3
2.39
eÞ
fÞ
5
—
130.6
2.34
2.31
2.45hÞ
6
[Ru(TPA)(bpy)(CH3CN)](PF6)2
gÞ
fÞ
—
a) Data obtained by KBr method. b) In CD2Cl2. c) In CD3CN. d) Synthesized with use of
CH313CN. e) Too weak to be observed. f) Not available. g) See Ref. 4g. h) In CD3OD.
{[Ru(1-Naph2-tpa)(CH3CN)]-H}þ at 770.2 and those of {[Ru-
(1-Naph2-tpa)(CH3CN)](PF6)}þ at 916.3 in addition to those
of {[Ru(1-Naph2-tpa)(CH3CN)]}2þ at 385.5 were observed
and their isotopic patterns were consistent with their computer
simulations. Complex 5 also showed the same spectral pattern
as 4. These complexes have a tetradentate tpa ligand and con-
comitant coordination of one of the amide oxygen, as dis-
cussed below.
Spectroscopic Characterization. Representative spectro-
scopic data are summarized in Table 1. In the IR spectra of 4
and 6 (KBr method), very weak absorption due to ꢆ(CN) were
observed at 2268 and 2251 cmꢂ1, respectively, which were
slightly higher than those of free nitriles (2200–2240 cmꢂ1).
This difference between 4 and 6 may be derived from electron-
ic effects of the substituents; the isopropyl groups are more
electron-donating than the 1-naphthyl groups as reflected on
the redox potentials of 1 and 3. This more electron-donating
character renders the ruthenium center to be more electron-rich
to facilitate ꢂ-back bonding toward the CꢁN moiety. The
ꢆ(CN) for 5 was not clearly determined due to the weakness
of the peak. An isotope shift was observed for 4 by using
CH313CN, and ꢆ(CN) was shifted to 2220 cmꢂ1 (ꢀꢆ ¼ 48
cmꢂ1), which exhibited excellent agreement with the calculat-
ed value. It is known that the ꢀ1-coordination of nitriles is usu-
ally accompanied by an increase in ꢆ(CN),15 although in sev-
eral cases it has been found to remain virtually unchanged, or
at lower frequencies than those of free nitriles, due to signifi-
cant back-bonding from the dꢂ orbitals of electron-rich metal
Fig. 1. Relationship between redox potentials of 1–3 and
the chemical shifts of the methyl groups of the acetonitrile
ligands of 4–6.
trile (2.1 ppm) and upfield shifts relative to those of ꢀ2-coordi-
nation (>3 ppm).18 In addition, the upfield shift in the chemi-
cal shifts of the methyl group of the acetonitrile ligand in 3
compared to that of 1 corresponds to the lower energy shift
of ꢆ(CN) in 3, suggesting that the electron density at the ruthe-
nium center can be transferred to the acetonitrile ligand.
In the 13C NMR spectrum of 4, a peak due to CꢁN was
observed at 131.3 ppm in CD2Cl2. As for that of 6, a peak as-
signed to the methyl group of CH3CN ligand was observed at
4.72 ppm, and that of CꢁN was observed at 130.6 ppm, which
was assigned by 1H–13C HMBC spectroscopy (Figure S1 in
Supporting Information). Based on IR and 13C NMR spectro-
scopic data, the coordination mode of the acetonitrile ligand
in 4–6 is a usual ꢀ1-N fashion.
ꢀ
centers to the ꢂ orbitals of the nitriles.16 In our cases, the
ruthenium(II) centers undergo sufficient electron donation
from the pentadentate tpa ligands to be electron-rich and ꢂ-
back bonding from the metal center can be facilitated toward
the acetonitrile ligands, which are the strongest ꢂ-acceptor
in those complexes, to give a relatively small increase of
ꢆ(CN).
In CD3CN, resonances due to the methylene groups of the
tpa ligand in 4 were observed as an AB quartet at 4.39 and
4.58 ppm (JAB ¼ 18 Hz), pseudo-AB quartet at 4.81 and 4.85
ppm (JAB ¼ 16 Hz), and a pseudo-singlet at 4.91 ppm at room
temperature. This multiplicity indicates that one amide oxygen
atom still binds to the ruthenium(II) center to hold an asym-
metric coordination environment, as found in the precursor.9
These signals showed a temperature dependency of multiplic-
ity and other peaks also showed a temperature-dependent
change of those chemical shifts in variable-temperature
1H NMR spectroscopy in CD3CN in the range from ꢂ40 to
50 ꢄC, as shown in Fig. 2.
1
In the H NMR spectra of acetonitrile complexes, singlets
ascribed to the methyl groups of the coordinated acetonitriles
were observed at 2.39 ppm for 4, 2.34 ppm for 5, and
2.31 ppm for 6 in CD3CN at room temperature without any re-
markable exchange with the deuterated solvent. This observa-
tion suggests that the acetonitrile ligand binds tightly to the
ruthenium(II) center.17 Those chemical shifts exhibited a good
linear relationship with their redox potentials of RuII/RuIII
couples of the corresponding precursor complexes 1–3, as
shown in Fig. 1, indicating the more electron-rich is the ruthe-
nium center, the more does the ꢂ-back donation occur toward
ꢀ
ꢂ
orbital of the nitrile. The chemical shifts showed slight
As for 6, its 1H NMR spectrum in CD3CN exhibited two ap-
parent singlets and one AB quartet for the methylene moieties
downfield shifts compared to those of ꢀ1-coordinated acetoni-

T. Kojima et al.
Bull. Chem. Soc. Jpn., 78, No. 12 (2005) 2155
Fig. 3. Temperature-dependence of chemical shifts of the
methyl groups in 4–6 in CD3CN.
Fig. 2. VT 1H NMR spectra of 4 (methylene reagion) in
CD3CN.
Fig. 4. Temperature-dependence of the chemical shifts of
CH313CN in 13C NMR spectra of 4 in CD2Cl2.
of isob2-tpa. In acetone-d6, however, the spectrum showed
three AB quartets at room temperature: one was observed at
4.66 and 4.90 ppm (JAB ¼ 18 Hz), the other two overlapped
at 5.07–5.22 ppm with JAB ¼ 15 and 16 Hz. Also, broad sin-
glets due to amide N–H protons were observed at 9.28 (br)
and 11.4 ppm (very broad); the former peak should be ascribed
to the uncoordinated amide moiety with hydrogen bonding,
and the latter should be assigned to the coordinated amide
group. These results were almost identical to those of 4 (see
Experimental section). The higher positive charge of the ruthe-
nium center renders the coordinated amide N–H more acidic
than that in 3. These observations suggest that the complexes
have a similar geometry to those of 1–3, possessing coordinat-
ed amide moieties without any symmetric plane.
Interestingly, the singlet assigned to the methyl group of the
coordinated acetonitrile showed a temperature dependence in
its chemical shift. The singlet exhibited upfield shifts upon
cooling. In all cases of the three complexes, the chemical shifts
of the methyl groups and that of the CN carbon (for 4) of co-
ordinated acetonitrile showed a nice linear relationship with
1=T, as depicted in Figs. 3 and 4, respectively. These observa-
tions indicate that only one mode of fluxional behavior is con-
sidered to alter the symmetry of the complexes. However, we
cannot describe the detailed picture of the motion at this point,
unfortunately. Also, the coordination environment of the ace-
tonitrile ligands of those complexes should be similar, as
judged from those chemical shifts and the fluxional behavior.
As for the thermal motion of acetonitrile ligands, it could be
derived from the thermal vibration of the ligand in the coordi-
nation spheres of those ruthenium(II) complexes. At lower
temperatures, the vibration could be retarded and ꢂ-back do-
nation from dꢂ of the ruthenium(II) center to pꢂ of the ace-
tonitrile ligand would be facilitated to make it electron-rich
and cause the observed upfield shifts.
Concerning the existence of the intramolecular hydrogen
bonding between the coordinated amide oxygen and the unco-
ordinated N–H group, we measured the temperature depen-
dence of the chemical shifts of one broad singlet observed
for 4 in CD3CN, which could be assigned to the uncoordinated
N–H group. As shown in Fig. S3 (Supporting Information), the
gradient of the linear relationship was determined to be
1.4 ppb/K, which was comparable to that (1 ppb/K)5 of 1. This
clearly indicates that intramolecular hydrogen bonding is oper-
ating in the acetonitrile complexes discussed here.
ꢀ
EXAFS Analysis of the Nitrile Complexes. In order to
accumulate structural information, we also examined EXAFS
spectroscopy to argue the coordination mode of the acetonitrile
ligand of 4–6.19 Curve fittings of EXAFS data used a model

2156 Bull. Chem. Soc. Jpn., 78, No. 12 (2005)
Ru–tpa Nitrile Complexes
Table 3. Redox Potentials of Complexes 1–6 in CH3CN
(a)
6
Complex
E1=2 (V)aÞ
obs(ab)
obs(im)
fit(ab)
[RuCl(1-Naph2-tpa)]PF6 (1)
[RuCl(2-Naph2-tpa)]PF6 (2)
[RuCl(isob2-tpa)]PF6 (3)
[Ru(CH3CN)(1-Naph2-tpa)](PF6)2 (4)
[Ru(CH3CN)(2-Naph2-tpa)](PF6)2 (5)
[Ru(CH3CN)(isob2-tpa)](PF6)2 (6)
[Ru(3-Me3-tpa)2](PF6)2
0.34bÞ
0.30bÞ
0.27bÞ
0.86
0.82
0.83
0.67cÞ
4
2
fit(im)
0
-2
-4
-6
a) Potentials were determined relative to that of Fc/Fcþ redox
couple as a reference (0 V). b) See Ref. 5. c) T. Kojima,
unpublished result. See Ref. 8e.
0
1
2
3
4
5
6
R/A
(b)
10
8
ꢁ
ꢁ
ꢁ
of 2.125–1.992 A (average 2.053 A) for 1, 2.113–2.005 A
obs(ab)
obs(im)
fit(ab)
ꢁ
(average 2.067 A) for 2, and 2.126–2.011 A (average 2.068
ꢁ
A) for 3. These values are in good agreement with those
ꢁ
6
fit(im)
4
estimated by EXAFS, as listed in Table 2. Those observations
could lend credence to the distances around the ruthenium cen-
ters of 4–6, to be discussed. In the CH3CN complexes, the
bond distances could be almost the same as in the precursor
complexes, except for the loss of those of the Ru–Cl bonds.
As for the ruthenium(II)–acetonitrile complexes reported so
far, the bond distances of Ru–N(nitrile) have been found to
2
0
-2
-4
-6
-8
-10
0
ꢁ
be, for example, 2.032(4) A for [Ru(2,2 -bipyridine)2(PhCN)]-
0
1
2
3
4
5
6
22
(PF6)2,21 2.041(5) A for [Ru(terpy)(bpy)(CH3CN)](PF6)2,
ꢁ
R/A
23
ꢁ
2.050(2) and 2.052(2) A for [Ru(Cp)(PCy3)(CH3CN)2]PF6,
Fig. 5. Curve-fittings of EXAFS analysis for 1 (a) and 4 (b).
and 2.165(18) A for [RuCl2(cycloocta-1,5-diene)(CO)].24
ꢁ
ꢁ
Thus, the bond distance of 2.06 A in 4–6 is within the range
of the Ru–N(NCCH3) bond lengths.
Table 2. Curve-Fitting Results of EXAFS of Ru Complexes
Electrochemical Analysis. Electrochemical measurements
were made by cyclic voltammetry (CV) to determine the redox
potentials of 4–6 in CH3CN at room temperature in the pres-
ence of 0.1 M [(n-butyl)4N]ClO4 as an electrolyte. The results
are summarized in Table 3. Reversible one-electron redox
waves due to RuII/RuIII couples were observed at 0.86 V (4),
0.82 V (5), and 0.83 V (6) relative to ferrocene/ferrocenium re-
dox couple as 0 V, respectively. Those potentials were ca. 0.5
V higher than those of 1 (0.34 V), 2 (0.30 V), and 3 (0.27 V). A
Ru complexes
Ru–O/N
Ru–Cl
R/%
ꢁ
r/A
ꢁ
r/A
N
N
1
2
3
4
5
6
2.08
2.06
2.06
2.06
2.06
2.06
6.0
5.2
4.8
5.9
5.3
5.9
2.40
2.45
2.45
1.3
0.8
1.2
0.59
1.49
0.45
2.02
3.64
1.55
bis-chelate ruthenium(II) complex of tris(3-methyl-2-pyridyl-
methyl)amine [Ru(3-Me3-tpa)2]2þ
8e
with DFT-generated geometry, which was based on the crystal
structure of 1. The radial distribution functions (RDF) for 1
and 4 are presented in Fig. 5, in which the red lines are the
observed spectra and the blue lines are simulated curves.20
Curve-fitting results for six complexes are summarized in
Table 2.
,
which has four pyridine
groups as ꢂ-acceptors and two tertiary amino groups as ꢁ-
donors, shows a corresponding RuII/RuIII couple at 0.67 V,25
which is lower than those of 4–6. Thus, the higher redox
potentials of the acetonitrile complexes can be derived from
the ꢂ-acceptor character of the acetonitrile ligand.
The EXAFS spectra of 1–3 gave the distances of 2.40–2.45
A for Ru–Cl distances, showing good agreement with crystal-
DFT Calculations. In order to access the structural fea-
tures and electronic properties of the Ru–tpa acetonitrile com-
plexes, DFT calculations were performed on a model com-
plex, {[Ru{N,N-bis(6-acetoamide-2-pyridylmethyl)-N-(2-pyr-
idylmethyl)amine}(CH3CN)](Clꢂ)}þ, which was constructed
on the basis of the crystal structure of 1. In Fig. 6a, we show
an optimized structure of the ruthenium(II)–nitrile complex, in
which the nitrogen atom of acetonitrile directly coordinates to
the ruthenium atom in an ꢀ1-N form. In this complex, the four
Ru–N(tpa) bond distances are 2.081, 2.085, 2.051, and 2.132
ꢁ
ꢁ
ꢁ
lographic data (2.446(2) A for 1, 2.4351(9) A for 2, and
ꢁ
2.437(1) A for 3). As for the CH3CN complexes 4–6, the cor-
responding Ru–Cl absorptions disappeared. The pre-edge fea-
tures of 4–6 were almost identical to those of 1–3, indicating
that the oxidation state of the ruthenium centers in CH3CN
complexes was intact to be divalent (þ2). The bond distances
ꢁ
around the ruthenium center were estimated 2.055 A in a single
shell. Curve fittings for 5 and 6 gave similar bond distances of
ꢁ
2.057 and 2.061 A, respectively.
X-ray crystallography on 1–3 has revealed that the bond
lengths around the ruthenium centers deviate in the range
ꢁ
ꢁ
A, and the Ru–O(amide) bond distance is 2.171 A. These geo-
metrical parameters are fully consistent with the values esti-
mated from an EXAFS analysis. The Ru–N(nitrile) bond dis-

T. Kojima et al.
Bull. Chem. Soc. Jpn., 78, No. 12 (2005) 2157
1-Naph
O
R1
(a)
1-Naph
O
H
N
H N
H3C
1-Naph
O
1-Naph
R2
C
β-diketone/2,6-lutidine
(CH2OH)2/ 100 °C
N
N
N
O
N
C
O
H
N
O
NH
Ru
N
Ru
N
N
N
N
R1 = R2 = Ph (dbm−)
1 = R2 = Me (acac−)
R1 = Me, R2 = Ph (bac−)
4
R
Scheme 2. Formation of ꢇ-diketonato complexes.
Nþ–Mꢂ $ –Cꢂ=Nþ=M.27 This strong covalent character
may stem from a combination of the ꢁ-donation of the lone
pair and the ꢂ-back bonding from the ruthenium(II) center
ꢀ
to the ꢂ orbitals of the CꢁN bond.
(b)
To estimate the CꢁN stretching mode of the ruthenium(II)–
nitrile complex, we performed a harmonic-frequency analysis
using a scaling factor21 of 0.98. As mentioned above, the ruthe-
nium(II)–nitrile complex has a high CꢁN stretching mode
relative to that in the free acetonitrile. The calculated CꢁN
stretching modes of the ruthenium(II)–nitrile complex and
the free acetonitrile are 2256 and 2245 cmꢂ1 at the B3LYP/
LANL2DZ level of theory, respectively. This result is fully
ꢁ
consistent with a change of the CꢁN bond distance: 1.176 A
ꢁ
in the ruthenium(II)–nitrile complex and 1.180 A in the free
acetonitirle. The result indicates that the formation of the Ru–
N(nitrile) bond should come from the back bonding of the
ꢀ
metal center to the ꢂ orbital of nitrile. The CꢁN stretching
frequency for the ruthenium(II)–nitrile complex is shifted to
2201 cmꢂ1 with CH313CN substitution, which is consistent
with our experimental observation.
Fig. 6. (a) The optimized structure of the model complex
Reaction of 4 with ꢀ-Diketones. Transition-metal com-
plexes with acetonitrile ligand have been used as starting ma-
terials toward further syntheses of coordination compounds.2
Recently, we reported on the synthesis and characterization
of ꢇ-diketonato complexes with 1-Naph2-tpa as the ligand.28
Those complexes could be obtained from reactions of 1 with
ꢇ-diketones. Based on this, we examined the reaction of 4
with ꢇ-diketones (acetylacetone (Hacac), dibenzoylmethane
(Hdbm), and benzoylacetone (Hbac)) in ethylene glycol in
the presence of 2,6-dimethylpyridine (2,6-lutidine) as a base
at 100 ꢄC. The reactions gave the corresponding ꢇ-diketonato
complexes, [Ru(ꢇ-diketonato)(1-Naph2-tpa)]þ in similar
yields, as shown in Scheme 2. All products were analyzed
ꢁ
with selected bond lengths (A) at the B3LYP/LANL2DZ
level and (b) Mulliken charges of selected atoms.
ꢁ
tance of 2.073 A is in good agreement with the results of an
EXAFS analysis, and those reported in the literature.16–19
The oxygen atom coordinating to the ruthenium ion forms a
ꢁ
hydrogen bond (2.014 A) with a neighboring H atom, which
anchors the uncoordinated amide group. This hydorogen bond
plays an important role in the stability of the Ru–O(amide) co-
ordination bond, since the exchange reaction of the two amide
groups requires hydrogen bond breaking to rotate the uncoor-
dinating amide group. The difference between the coordinating
and uncoordinating amide groups induces the multiplicity of
the methylene groups of the tpa ligand through an asymmetric
coordination environment.
In Fig. 6b, computed Mulliken charges of selected atoms are
described. The charge of the Ru center is þ0:64 and those of
the N and O atoms of the pentadentate tpa ligand are all neg-
ative in the range of ꢂ0:33– ꢂ0:26. In sharp contrast to those,
the acetonitrile ligand has a different charge distribution. The
charge of the coordinating N atom is þ0:07, which derives
from a strong covalent character of Ru–NCCH3 bonding. In
addition, the cyano carbon has a negative charge of ꢂ0:14, al-
though the electronegativity of the nitrogen atom (3.0) is larger
than that of the carbon atom (2.5).26 Previously, Cotton pro-
posed that nitriles bound to zero-valent transition-metal com-
plexes show the following resonance structure hybrid: –Cꢁ
1
by H NMR and UV–vis spectroscopies, and those data were
consistent with those reported previously. The apparent reac-
tion rates were also similar to those from 1, suggesting that
the acetonitrile binding is as strong as that of a negatively
charged chloride anion.
Summary
We prepared and characterized a series of ruthenium(II)–
acetonitrile complexes with bisamide-tpa ligands. A spectro-
scopic study indicated that those complexes should have the
ꢀ1-N acetonitrile ligand and one of amide moieties bound to
the ruthenium(II) center through amide oxygen with intramo-
lecular hydrogen bonding with the uncoordinated counterpart.
A DFT study on a model complex allowed us to access an
optimized structure, which was consistent with other spectro-

2158 Bull. Chem. Soc. Jpn., 78, No. 12 (2005)
Ru–tpa Nitrile Complexes
Sakamoto, Y. Matsuda, K. Ohkubo, and S. Fukuzumi, Angew.
Chem., Int. Ed., 42, 4951 (2003). g) T. Kojima, T. Sakamoto,
and Y. Matsuda, Inorg. Chem., 43, 2243 (2004).
scopic data. More importantly, coordination of the acetonitrile
ligand is suggested to have a strong covalent character com-
pared with that of the tpa ligands.
9
a) T. Kojima, K. Hayashi, and Y. Matsuda, Chem. Lett.,
2000, 1008. b) T. Kojima, K. Hayashi, and Y. Matsuda, Inorg.
Chem., 43, 6793 (2004).
10 K. Jitsukawa, Y. Oka, S. Yamaguchi, and H. Masuda,
Inorg. Chem., 43, 8119 (2004).
11 a) A. D. Becke, J. Chem. Phys., 98, 5648 (1993). b) C. Lee,
W. Yang, and R. G. Parr, Phys. Rev. B, 37, 785 (1988).
12 a) T. H. Dunning and P. J. Hay, ‘‘Modern Theoretical
Chemistry,’’ ed by H. F. Schaefer, III, Plenum, New York
(1976), Vol. 3, p. 1. b) P. J. Hay and W. R. Wadt, J. Chem. Phys.,
82, 270 (1985).
The X-ray absorption experiment was carried out under the
approval of Japan Synchrotron Radiation Research Institute
(JASRI) (proposal No. 2002A0397). This work partly support-
ed by Grants-in-Aid (Nos. 11740373 and 16550057 to T. K.)
from the Ministry of Education, Culture, Sports, Science and
Technology of Japan and Japan Society for the Promotion
of Science. We are grateful to Prof. Youichi Ishii (Chuo Uni-
versity) for his helpful discussions. We also appreciate Dr.
Y. Terada (SPring-8, JASRI), Prof. Hisanobu Wakita and
Dr. Tsutomu Kurisaki (Fukuoka University), Dr. Takashi
Yamamoto (Tokyo Institute of Technology) and Prof.
Takafumi Shido (The University of Tokyo) for their generous
help on XANES and EXAFS measurements and analyses.
The HMBC spectrum was measured by JEOL Co., Ltd., which
was appreciated. This work was supported in part by a Grant-
in-Aid from the Ministry of Education, Culture, Sports, Science
and Technology of Japan.
13 See Ref. 11b.
14 Similarly,
a benzonitrile complex, [Ru(2-Naph2-tpa)-
(PhCN)](PF6)2, could be synthesized by the reaction of 2 with
AgNO3 in benzonitrile. FAB-MS: 978.2 (fM ꢂ PF6gþ); 832.3
(fM ꢂ 2PF6 ꢂ Hgþ).
15 a) K. F. Purcell and R. S. Drago, J. Am. Chem. Soc., 88,
919 (1966). b) K. F. Purcell, J. Am. Chem. Soc., 89, 247 (1967).
16 a) Y. Tanabe, H. Seino, Y. Ishii, and M. Hidai, J. Am.
Chem. Soc., 122, 1690 (2000). b) T. Tatsumi, M. Hidai, and Y.
Uchida, Inorg. Chem., 14, 2530 (1975).
Supporting Information
HMBC spectrum (CD3CN) of 6, VT 1H NMR spectra (CD3CN)
for 4, temperature-dependence of the chemical shifts of the NH
signal for 4, and EXAFS spectra (2, 3, 5, and 6) (PDF).
17 See Ref. 8g.
18 S. Thomas and C. G. Young, Organometallics, 17, 182
(1998).
19 EXAFS data were obtained at SPring-8 using BL01B1
beamline. Data analysis was done by using FEFFIT program
package at The University of Tokyo. See also Supporting Infor-
mation.
20 For other complexes, see Supporting Information (Figs. S4–
S6).
21 W. P. Griffith, B. Reddy, A. G. F. Shoair, M. Suriaatmaja,
J. P. White, and D. J. Williams, J. Chem. Soc., Dalton Trans.,
1998, 2819.
22 S. Bonnet, J.-P. Collin, N. Gruber, J.-P. Sauvage, and E. R.
Schofield, Dalton Trans., 2003, 4654.
References
1 R. A. Michelin, M. Mozzon, and R. Bertani, Coord. Chem.
Rev., 147, 299 (1996).
2 B. N. Storhoff and H. C. Lewis, Jr., Coord. Chem. Rev., 23,
1 (1977).
3
a) S.-I. Murahashi, T. Naota, and E. Saito, J. Am. Chem.
Soc., 108, 7846 (1986). b) S.-I. Murahashi, S. Sasao, E. Saito,
and T. Naota, J. Org. Chem., 57, 2521 (1992). c) S.-I. Murahashi
and T. Naota, Bull. Chem. Soc. Jpn., 69, 1805 (1996). d) S.-I.
Murahashi and H. Takaya, Acc. Chem. Res., 33, 225 (2000).
23 E. Ruba, W. Simanko, K. Mauthner, K. M. Soldouzi, C.
¨
4
E. Ruba, R. Schmid, K. Kirchner, and M. J. Calhorda,
Slugovc, K. Mereiter, R. Schmid, and K. Kirchner, Organometal-
lics, 18, 3843 (1999).
24 R. O. Gould, C. L. Jons, D. R. Robertson, and T. A.
Stephenson, J. Chem. Soc., Dalton Trans., 1977, 129.
25 T. Kojima, unpublished result.
26 J. E. Huheey, E. A. Keiter, and R. L. Keiter, ‘‘Inorganic
Chemistry,’’ 4th ed, HarperCollins College Publishers, New York
(1993), p. 182.
¨
J. Organomet. Chem., 682, 204 (2003).
R. A. Prados, C. A. Clausen, III, and M. L. Good, J. Coord.
Chem., 2, 201 (1973).
5
6
7
R. E. Clarke and P. C. Ford, Inorg. Chem., 9, 227 (1970).
F. Basolo and R. G. Pearson, ‘‘Mechanism of Inorganic
Reactions,’’ 2nd ed, John Wiely & Sons, Inc., New York (1967).
a) T. Kojima, Chem. Lett., 1996, 121. b) T. Kojima, T.
8
Amano, Y. Ishii, M. Ohba, Y. Okaue, and Y. Matsuda, Inorg.
Chem., 37, 4076 (1998). c) T. Kojima and Y. Matsuda, Chem.
Lett., 1999, 81. d) T. Kojima, H. Matsuo, and Y. Matsuda, Inorg.
Chim. Acta, 300–302, 661 (2000). e) T. Kojima and Y. Matsuda,
J. Chem. Soc., Dalton Trans., 2001, 958. f) T. Kojima, T.
27 F. A. Cotton, Inorg. Chem., 3, 702 (1964).
28 T. Kojima, S. Miyazaki, K. Hayashi, Y. Shimazaki, F.
Tani, Y. Naruta, and Y. Matsuda, Chem.—Eur. J., 10, 6402
(2004).