Ruthenium and chromium complexes bearing pH-indicators as the η6-arene ligand: Synthesis, characterization, and protonation behavior

Article abstract of DOI:10.1016/j.jorganchem.2006.07.046

A series of ruthenium and chromium complexes bearing pH indicators as the η⁶-arene ligand, (η⁶-X)(ML_n)_y [X = methyl yellow, crystal violet lactone, phenolphthalein; ML_n = RuCp^*+, RuCl₂(L), Cr(CO)₃; y = 1, 2] is prepared and characterized by spectroscopic and crystallographic methods. Of the plural arene rings in the indicators, a specific arene ring can be successfully coordinated to the metal center in a selective manner under appropriate conditions (i.e. use of the precursors of different oxidation states and reaction with the non-protonated and protonated pH indicator). The obtained indicator complexes show halochromic behavior depending on pH as observed for the parent molecules but the transition pH ranges are shifted to the more acidic side because of the attachment of the electron-withdrawing metal fragments, which decrease the basicity of the attached pH indicators.

Full text of DOI:10.1016/j.jorganchem.2006.07.046

Journal of Organometallic Chemistry 692 (2007) 93–110
www.elsevier.com/locate/jorganchem
Ruthenium and chromium complexes bearing pH-indicators as the
g6-arene ligand: Synthesis, characterization, and protonation behavior
*
Miyuki Hirasa, Akiko Inagaki, Munetaka Akita
Chemical Resources Laboratory, Tokyo Institute of Technology, R1-27, 4259 Nagatsuta, Midori-ku, Yokohama 226-8503, Japan
Received 23 May 2006; received in revised form 11 July 2006; accepted 18 July 2006
Available online 3 September 2006
Abstract
A series of ruthenium and chromium complexes bearing pH indicators as the g6-arene ligand, (g6-X)(MLn)y [X = methyl yellow, crys-
tal violet lactone, phenolphthalein; MLn = RuCp*+, RuCl2(L), Cr(CO)3; y = 1,2] is prepared and characterized by spectroscopic and
crystallographic methods. Of the plural arene rings in the indicators, a specific arene ring can be successfully coordinated to the metal
center in a selective manner under appropriate conditions (i.e. use of the precursors of different oxidation states and reaction with the
non-protonated and protonated pH indicator). The obtained indicator complexes show halochromic behavior depending on pH as
observed for the parent molecules but the transition pH ranges are shifted to the more acidic side because of the attachment of the elec-
tron-withdrawing metal fragments, which decrease the basicity of the attached pH indicators.
ꢀ 2006 Elsevier B.V. All rights reserved.
Keywords: pH Indicators; g6-Arene complex; Ruthenium; Chromium
1. Introduction
chosen and subjected to complexation with metal species.
In Scheme 2, the coloring mechanisms of the dyes are
shown.
Organometallic species with an auxiliary, which is sensi-
tive to the change of the environment, could be utilized as a
functional sensor. Chromic compounds are characterized
by their feature that they change their colors upon applica-
tion of a stimulus such as light, heat, and pH change. The
color change is induced by a remarkable change of their
electronic structures, which would also trigger some new
reactivity. Combination of these two systems would lead
to a new chemical system such as a self-curing system
(Scheme 1) [1].
In the present study, halochromic pH-indicators [2] were
chosen as the chromic ligand [3], and we wish to report the
results of synthesis and characterization of the ruthenium
and chromium complexes bearing pH-indicators as the
g6-arene ligand. As typical examples of pH-indicators,
azo dye (methyl yellow (MY; 1)) and phthalein dye (crystal
violet lactone (CVL; 2) and phenolphthalein (PP; 3)) were
Protonation of MY (1) promoted by the NMe2 group
occurs at the nitrogen atom bonded to the C6H5 ring
([1+H]+) to bring about contribution of the colored qui-
noidal form. On the other hand, protonation of the phtha-
lein dye A bearing electron-donating substituents X (e.g.
NMe2 in 2) or deprotonation of the phenol derivatives
(e.g. X = OH (3)) induces heterolysis of the C–O bond of
the lactone moiety. As a result, the p-systems separated
by the central sp3-carbon atom and localized on the three
aromatic rings in the original form is spread over the three
aromatic rings through the central sp2-carbon atom to
cause appearance of absorptions in the visible region, i.e.
coloring (B0, B00, C0, C00, etc.).
Because most of organic chromic compounds consist of
aromatic groups, we choose the RuCp*, RuCl2(L), and
Cr(CO)3 fragments to be attached to them. The organome-
tallic chemistry of the (g6-arene)–ruthenium [4] and –chro-
mium complexes [5,6] has been studied extensively and well
established. While studies on interaction with alkali and
*
Corresponding author.
E-mail address: makita@res.titech.ac.jp (M. Akita).
0022-328X/$ - see front matter ꢀ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2006.07.046

94
M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
2. Results and discussion
stumulus
(e.g. hν, Δ, H+)
2.1. Synthesis and characterization of pH-indicator
complexes
chromic
compounds
color change
new function
M
M
2.1.1. Methyl yellow complexes
Azobenzene dyes constitute an important class of pH-
indicators, and methyl yellow 1 was subjected to complex-
ation with the Ru and Cr fragments.
Scheme 1.
2.1.1.1. Cationic RuCp* complexes, [(g6-MY)RuCp*]PF6
(5+ Æ PF6) (MY = methyl yellow). Reaction of 1 with the
labile RuCp*+ precursor, [Cp*Ru(NCMe)3]PF6 (4 Æ PF6)
[8], in CH2Cl2 gave an isomeric mixture of two mononu-
clear adducts 5a Æ PF6 and 5b Æ PF6 in 1:5 ratio (Scheme
alkali earth metals appeared [7], no organo-transition
metal species has been reported so far. The present study
revealed (1) the capability of the pH indicators working
as the g6-ligands and (2) the modified functions of the
adducts as the pH-sensitive molecules.
N
NMe2
N
NMe2
N
NMe2
+ H
etc.
N
N
H
N
H
[1+H]+
[1+H]'+
1
(methyl orange)
X
X
X
X
X
X
C
Y
C
Y
+ H
X= NMe2 (2) etc.
C
etc.
O
COOH
COOH
X
X
O
H
Y
C
O
B (A+H)
B'
B"
A
O
X
O
X
O
Y
X
O
X=Y= NMe2
C
Y
C
Y
[crystal violtet lactone (2)]
X= OH; Y= H
[phenol phthalein (3)]
+
OH / -H2O
C
O
etc.
COO
COO
X= OH (3) etc.
O
Y
C (C-H)
C'
C"
Other substituenta are
omitted for clarity.
Scheme 2.
MLn
N
NMe2
β
N
N
NMe2
N
α
N
MLn= RuCp*+ (5a·PF6), Cr(CO)3 (6a).
1
"MLn"
or
+
MLn
NMe2
OTf
N
NMe2
N
H
N
[1+H]OTf
MLn= RuCp*+ (5b·PF6), Cr(CO)3 (6b).
MLn
SM * yield (%) **
ratio
5a·PF6 : 5b·PF6= 1 : 5
5a·PF6 : 5b·PF6= 5 : 1
6a : 6b= 1 : 5
-
"MLn"= [RuCp*(NCMe)3]PF6 (4·PF6) /
CH2Cl2 / r.t.
RuCp*
1
71
80
10
0
Cr(CO)6 / OBu2 + THF / 120˚C
[1-H]OTf
1
Cr(CO)3
[1-H]OTf
* starting material. ** total yield.
Scheme 3.

M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
95
Table 1
1H NMR data for pH-indicator complexesa,b
Compound (solvent)
Cp*
Coordinated Ar
Non-coordinated Ar
NMe2
3.13
5a Æ PF6 (CD2Cl2)
5b Æ PF6 (CD2Cl2)
5a0 Æ PF6 (CD2Cl2)
1.88
1.94
1.90
6.40 (2H, d, 6.2), 5.97 (2H, m), 5.85 (1H, m)
6.23 (2H, d, 7.6), 5.51 (2H, d, 7.6)
6.41 (1H, d, 6.4), 6.10 (2H, m), 6.04 (1H, d, 5.1)
7.86 (2H, d, 7.2), 6.76 (2H, d, 7.2)
7.7 (2H, m), 7.4 (3H, m)
7.76 (2H, d, 9.2), 7.38 (4H, m), 7.17–7.24
(6H, m), 7.03 (2H, d, 9.2)
6a (acetone-d6)
6b (acetone-d6)
7 Æ PF6 (CD3CN)
6.29 (2H, d, 6.8), 5.86 (2H, m), 5.70 (1H, m)
6.64 (2H, d, 7.6), 5.59 (2H, d, 7.6)
5.95 (2H, d, 6.7), 5.91 (2H, d, 6.7), 5.31
(2H, d, 6.7), 5.25 (2H, d, 6.7)
5.86 (2H, d. 6.8), 5.77 (2H, d, 6.8), 5.25
(2H, d, 6.8), 5.18 (2H, d, 6.8)
7.79 (2H, d, 9.4), 6.84 (2H, d, 9.4)
7.82 (2H, m,), 7.55 (3H, m)
7.59 (1H, d, 8.4), 7.17 (1H, d, 8.6), 7.12
(1H, s), 6.86 (2H, d, 9.0), 6.61 (2H, d, 9.0)
7.78 (1H, d, 9.6), 7.25–7.29 (2H, m)
3.13
3.06
3.03, 3.01, 2.87
–
1.65
8 Æ (PF6)2 (CD3CN)
10 (C6D6)
1.63
–
3.07, 2.98
5.66–5.62 (2H, m), 4.18 (1H, d, 5.7)
7.95 (2H, d, 8.8), 7.42 (2H, d, 8.8),
6.60 (2H, d, 9.0), 6.42 (2H, d, 9.0)
7.21–7.03 (8H, m)
2.51, 2.45, 2.23
12 (DMSO-d6)
13a (CDCl3)
–
–
5.36 (1H, s), 5.28 (1H, d, 6.8), 4.97 (1H, d, 6.8)
6.14 (1H, d, 6.1), 5.00 (1H, s), 4.67 (1H, d, 5.6)
ꢀ3.10 (br.)
7.70 (2H, d, 8.6), 6.99 (2H, d, 8.6),
6.71 (2H, d, 8.6), 6.55 (2H, d, 8.6)
7.67 (2H, d, 8.3), 6.96 (2H, d, 8.3),
6.70 (2H, d, 8.5), 6.54 (2H, d, 8.3)
7.12 (2H, d, 8.6), 6.95 (2H, d, 8.6),
6.70 (2H, d, 8.6), 6.64 (2H, d, 8.6)
7.11 (2H, d, 8.4), 6.95 (2H, d, 8.4),
6.65–6.57 (4H, m)
3.30, 3.23, 2.95, 2.91
3.24, 2.93, 2.89
3.10, 3.02, 2.88, 2.86
2.90–2.84
13b (CDCl3)
14a (CD3CN)
14b (CDCl3)
16 (CDCl3)
a
–
6.31 (1H, d, 6.6), 5.30 (1H, d, 6.6), 4.76 (1H, s)
5.51 (1H, d, 6.6), 4.63 (1H, d, 6.1), 4.29 (1H, s)
5.77 (1H, s), 5.27 (1H, m), 3.96 (1H, d, 5.8)
–
–
1.60
5.62 (1H, d, 7.4), 5.43 (1H, d, 7.4), 4.79
(1H, d, 7.4), 4.75 (1H, d, 7.4)
8.00 (1H, d, 7.6), 7.71–7.50 (3H, m),
6.82 (2H, d, 8.6), 6.69 (2H, d, 8.6)
–
dH in ppm. Coupling pattern and coupling constant (in Hz) are shown in parentheses.
For other signals: 10: cod signals: 3.66–3.44 (4H, m), 2.45–2.23 (8H, m). 13a: piperidine: 2.43–1.17. 13b: PEt3: 1.80 (6H, m, CH2), 0.87 (9H, m, CH3).
12: 6.21 (CH). 14a: 7.83 (1H, s, NH), 5.86 (1H, s, CH), 3.18–1.31 (piperidine). 14b: 5.37 (1H, s), 1.32 (6H, CH2), 0.88 (9H, CH3).
b
3). The compositions (1:1 adducts) were readily determined
on the basis of the intensities of the H NMR signals for
can be interpreted as follows. The direct reaction with 1
1
results in coordination to the electron-rich b ring bearing
the NMe2 group. Protonation of 1 occurs at the nitrogen
atom attached to the a ring to decrease the electron density
of the b ring and, therefore, the RuCp*+ fragment should
be attached to the more electron-rich a ring in the proton-
ated form ([1+H]+) to give 5a Æ PF6. Contribution of the
the MY and Cp* ligands (Table 1), and the coordination
sites were also readily determined on the basis of the
assignments of the aromatic signals shifted to higher field
(Fig. 1 and Table 1). It is established that coordination
of an aromatic group to a transition metal species in g6-
fashion causes upfield-shifts of the aromatic proton signals
[4–6]. As shown in Fig. 1, the signals for the a and b rings
are assigned on the basis of the coupling patterns; a ring:
three multiplet signals for the o-, m-, and p-hydrogen
atoms; b ring: a pair of coupled doublets (dH 6.76, 7.91
(d, J = 7.2 Hz)). In the case of 5a Æ PF6, the signals assigned
to the a ring are shifted to higher field leading to the assign-
ment to the a ring adduct. On the other hand, the 1H NMR
spectrum of 5b Æ PF6 contains the doublet pairs shifted to
higher field leading to the assignment to the b ring adduct.
Thus it is revealed that the major product 5b Æ PF6 results
from coordination to the C6H4–NMe2 part (b ring). The
assignments have been verified by X-ray crystallography
(Fig. 2a and b). A remarkable difference is noted for
UV–Vis spectra. The a ring adduct 5a Æ PF6 shows a visible
absorption at 504 nm with the intensity comparable to that
of 1, whereas the intensity of the 500 nm absorption
observed for 5b Æ PF6 is much weaker that those of 5a Æ PF6
and 1 presumably because of the negligible contribution of
the colored, quinoidal form (see F0 in Scheme 9).
+
quinoidal form ([1+H]0 in Scheme 2), which cannot be
coordinated in a g6-fashion, should also promote coordi-
nation to the aromatic a ring.
For comparison sake, the NPh2 derivative of 5 Æ PF6
(5a0 Æ PF6) was also prepared. In this case, the adduct of
the a ring (5a0 Æ PF6) was obtained as the major product
(5a0 Æ PF6:5b0 Æ PF6 = 5:2) even from the non-protonated
precursor 10, presumably because (i) the lone pair electrons
on the NPh2 part are delocalized over the NAr3 part to
decrease electron density of the a ring (compared to 1)
and (ii) the bulky Ph substituents may hinder approach
of the bulky RuCp* fragment to the b ring. Reaction with
[10-H]OTf improved the selectivity for 5a0 Æ PF6 (5a0 Æ
PF6:5b0 Æ PF6 = 8:1). The adduct of the NPh2 part was
not detected at all.
Preparation of the RuCl2(L) adduct was also attempted
but reaction of 1 with Ru(cod)(naphthalene) did not afford
the desired g6-arene complex, (g6-MY)Ru(cod), as
observed for CVL (see below). Because azobenzene with-
out the NMe2 group gave the 1:1 adduct, (g6-azoben-
1
It is notable that the analogous reaction of the proton-
ated form of 1, [1+H] Æ OTf, inverted the isomer ratio to
give 5a Æ PF6 as the dominant product (5:1). These results
zene)Ru(cod), as judged by H NMR, the failure in the
formation of the 1-adduct should be ascribed to the
NMe2 substituent, which might work as a r-donor.

96
M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
Fig. 1. 1H NMR spectra for 1, 5a Æ PF6, and 5b Æ PF6 observed in CD2Cl2 at 400 MHz (s: residual protio solvent signals; : impurities).
*
2.1.1.2. Cr(CO)3 complexes, [(g6-MY)Cr(CO)3]PF6
(6). Reaction of 1 with Cr(CO)6 in Bu2O–THF at
120 ꢁC in a glass autoclave gave an isomeric mixture of
the g6-adduct 6a and 6b in a low yield accompanying
formation of a large amount of black precipitates which
did not show m(CO) vibration (Scheme 3). Reaction at
higher temperature (>170 ꢁC) caused decomposition of
the product, and reaction of the protonated precursor
[1+H]OTf did not afford 6 presumably because of its
low solubility in the reaction medium. The two regioi-
somers 6a,b could not be separated by neither recrystalli-
zation nor chromatography but by hands, because the
crystal shapes were considerably different (6a: needles;
6b: plates). The two isomers were readily characterized
by analyzing the shifted H NMR signals for the coordi-
1
nated arene parts in a manner similar to 5 Æ PF6 (Table 1).
The presence of the Cr(CO)3 auxiliary was confirmed by
the characteristic two strong m(CO) vibrations and the 13C
NMR signals (dC 233). Molecular structures of the two
isomers were determined by X-ray crystallography
(Fig. 2d and e).
2.1.2. Crystal violet lactone complexes
Crystal violet lactone (CVL; 2) opens the lactone
ring under acidic conditions through protonation at the
COO moiety to show violet color based on the resonance

M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
97
Fig. 2. Molecular structures of the methyl yellow complexes drawn with thermal ellipsoids at the 30% probability level. (a) 5a+, (b) 5b+, (c) 5a0+, (d) 6a,
and (e) 6b.
structures C0, C00, etc. (Scheme 2). Synthesis and character-
ization of the CVL–Ru complexes are described below [9].
We also examined the reaction with Cr(CO)6 but no char-
acterizable product was obtained.
2.1.2.1. Cationic RuCp* complexes, [(g6-CVL)RuCp*]PF6
(7 Æ PF6) and [(g6-CVL)(RuCp*)2](PF6)2 (8 Æ (PF6)2)
(CVL = crystal violet lactone). Treatment of 2 with
4 Æ PF6 in CH2Cl2 caused immediate color change from
Me2N
NMe2
α
α
PF6
Ru
C
+
MeCN
NCMe
O
N
β
C
4·PF6
O
Me
Me2N
CH2Cl2
2 (0.9 equiv.)
Ru
Ru
Ru
Ru
Ru
Ru
Me2N
NMe2
Me2N
NMe2
Me2N
NMe2
Me2N
NMe2
+
C
C
O
C
C
O
O
O
O
O
O
PF6
O
PF6
(PF6
)
(PF6
)
2
2
Me2N
Me2N
Me2N
Me2N
7·PF6
(72 %)
8·(PF6)2
7'·PF6
8'·(PF6)2
(8 %)
H
NEt3
H
NEt3
Ru
Ru
Ru
Me2N
NMe2
Me2N
NMe2
C
C
OH
O
OH
O
[7+H]2+
[8+H]3+
Me2N
Me2N
Scheme 4.

98
M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
yellow to deep blue (Scheme 4). Chromatographic separa-
tion (alumina) gave two pale blue products, 7 Æ PF6 (72%)
and 8 Æ (PF6)2 (9%), which were characterized to be mono-
and dinuclear adducts, respectively, as described below.
The 1:2 adduct 8 Æ (PF6)2 was obtained as the major prod-
uct by the 1:2 reaction, while a 1:1 reaction with the pro-
tonated CVL ([2+H]+ Æ OTf) afforded the mononuclear
product preferentially (7 Æ PF6/8(PF6)2 = 20:1).
similar to the MY complexes 5 Æ PF6 (Fig. 3 and Table 1).
As shown in Fig. 3a, the pair of two doublet signals for 2
at dH 6.62 and 7.13 was assigned to the signals for the a-
ring and the remaining aromatic signals to the b-ring.
Upon coordination, half of the doublet pair was shifted
to higher field (dH 5–6), whereas no significant shift was
observed for the remaining half and the signals for the b-
ring (Fig. 3b). These spectral changes revealed that the
RuCp* fragment was coordinated to one of the two a-rings.
In the case of 8 Æ (PF6)2 (Fig. 3c), the doublet pair was com-
The number of the attached RuCp* fragments and the
coordination site were also readily determined in a manner
Fig. 3. 1H NMR spectra for 2 (a), 7+ (b), and 82+ (c) observed in CD2Cl2 at 400 MHz (s: residual protio solvent signals; : impurities).
*

M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
99
pletely shifted to the higher field, indicating coordination
of the RuCp* fragments to the two a-rings. Coordination
of the bulky RuCp* fragments hinders rotation around
the C–(g6-Ar)Ru axis to cause separation of the a ring sig-
nals into two pairs of doublets. Thus the RuCp*+ fragment
is attached to the more electron-rich a ring in a selective
manner as also confirmed by X-ray crystallography
(Fig. 4a and b, see below).
The NMR data confirmed the compositions of the
adducts 7 Æ PF6 and 8 Æ (PF6)2 but the structure of the lac-
tone moiety (open or closed) could not be deduced from
the NMR data alone. On the other hand, IR spectra of
the adducts contained not only the mC@O vibrations assign-
able to the closed lactone group (1759 (7 Æ PF6), 1775 cmꢁ1
(8 Æ (PF6)2)) but also weak absorptions assignable to the
open carboxylate group (1557 (70 Æ PF6), 1565 cmꢁ1
(80 Æ (PF6)2)), suggesting formation of both of the closed
and open forms. In addition, UV–Vis spectra (see for
example, Fig. 5) contained weak absorptions around
600 nm attributable to the open forms 70 Æ PF6 and
80 Æ (PF6)2. These data suggested that the two forms were
equilibrated in solutions and, judging from the intensities
of the UV–Vis absorptions weaker than those of the pro-
tonated open forms, [7+H]2+ and [8+H]3+, the closed
forms should be the dominant species in solutions. The
equilibrium observed for 7+ and 82+ is in sharp contrast
to the non-coordinated CVL (2), for which the colored
open form is not detected (even by the naked eyes). This
difference sounds strange, because attachment of a cationic
fragment to the a-ring(s) in 2 should destabilize the open
form with the cationic charge on the p-conjugated system
coordinated by the cationic RuCp* fragment(s). The differ-
ence could be interpreted in terms of the X-ray structures
discussed below.
The solid-state structures of the open forms of 7 Æ PF6
and 8 Æ (PF6)2 were determined by X-ray crystallography
(Fig. 4a and b). For the CVL parts no significant difference
is noted when compared with 2 [10]. Molecule 7+ Æ PF6 sits
on a crystallographic mirror plane passing through Ru1,
C1, O1, C11 and so on, and the C2–O2 moiety being
refined with the occupancy of 0.5 is disordered with respect
to the mirror plane. The structures are consistent with the
above-mentioned spectroscopic characterization including
the number of the attached RuCp* fragments and the coor-
Fig. 4. Molecular structures of the ruthenium complexes derived from crystal violet lactone drawn with thermal ellipsoids at the 30% probability level. (a)
7+ (cationic part), (b) 82+ (cationic part), (c) 13a, and (d) 14a0.

100
M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
Fig. 5. Addition of acid to pH-indicators and their complexes as monitored by UV-spectroscopy (in CH2Cl2; (a)–(e) CH3SO3H; (f) CF3SO3H).
([7+H]+): see text.
*
dination sites (a-ring). One of the key features of 7 Æ PF6 is
the coplanar geometry of the Ru1–C11–C1–O1 moiety
(dihedral angle = 0ꢁ) associated with the close contact
inatively saturated and then we sought to prepare a more
reactive system, i.e. (g6-arene)RuCl2(L). Because, however,
preparation of [(g6-arene)Ru(l-Cl)]2-type complexes with a
functionalized arene molecule has few precedents, synthesis
of an [(g6-arene)Ru(l-Cl)]2 complex with a pH-indicator, a
highly functionalized molecule, is a challenging problem
[11].
˚
between the Ru1 and O1 atoms (3.520(7) A). The similar
feature is noted for 8 Æ (PF6)2 [Ru1–O1: 3.864(3), Ru2–
˚
O1: 3.937(3) A]. Such short contacts should result from
electrostatic, attractive interactions between the cationic
ruthenium center and the electronegative lactone oxygen
(g6-Arene)RuCl2(L)-type complexes can be prepared by
treatment of [(g6-arene)Ru(l-Cl)]2 complexes with appro-
priate donors (L) [12]. The precursors, [(g6-arene)Ru(l-
Cl)]2, have been prepared mainly by two methods: (i)
arene-exchange reaction of [(g6-p-cymene)Ru(l-Cl)]2 con-
ducted by heating at the boiling point (or above the melting
point) of the arene to be introduced [13]; and (ii) HCl-treat-
ment of (g6-arene)Ru(g4-cod), which is obtained by treat-
ment of the arene with (g6-naphthalene)Ru(g4-cod) (9) or
(g6-cot)Ru(g4-cod) [14]. We examined these reactions and
found that the desired compound could be prepared by
method (ii) starting from 9. Method (i) is not applicable
to solid aromatic substances and the reaction at the melting
point of CVL causes its decomposition.
+
atom(s). The equilibria mentioned above (7+ M 70 ,
82+ M 802+; Scheme 4) could be explained on the basis of
the electrostatic interaction (Scheme 5), which may facili-
tate the ring opening and stabilize the resultant negative
charge developed on the oxygen atom, although no signif-
˚
icant elongation of the C1–O1 bonds (7: 1.45(1) A; 8:
˚
1.455(4) A) is observed and the ruthenium centers are
coordinatively saturated. Other structural features will be
discussed below.
2.1.2.2.
RuCl2(L)
complexes,
(g6-CVL)RuCl2(L)
(13). The [Cp*Ru(g6-arene)]+ species are stable enough
to be fully characterized, as described above, and easy to
deal with as the initial synthetic targets. But they are coord-
Reaction of 2 with 9 in THF in the presence of MeCN
followed by separation by chromatography under Ar gave
the yellow, air-sensitive product 10, which was character-
1
ized only by H NMR because of its sensitivity to the air
O
O
(Scheme 6). The 1:1 stoichiometry was confirmed by the
intensities of the CVL and cod signals. The coordinated
arene ring was readily determined to be the b-ring on the
basis of the upfield-shifted aromatic ring signals (Table
1). It is notable that the adduct formation on the b ring
contrasts with the addition to the a-ring observed for 7+
Ru
Ru
δ-
O
O
C
Me2N
Me2N
C
7+
Ar
Ar
Scheme 5.

M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
101
Me2N
Me2N
Me2N
NMe2
H
C
NMe2
Cl
C
C
NMe2
Cl
O
+
Ru
9
COOH
Cl
COOH
Cl
O
Me2N
2
Me2N
Me2N
Ru
Ru
Ru
Ru
THF-CH3CN
1 M HCl aq.
NMe2
NMe2
Cl
112+·Cl2
Cl
Cl
Cl
HOOC
HOOC
36 % HCl aq.
acetone / -78˚C
acetone / -78˚C
H
C
Me2N
C
Me2N
12
Me2N
NMe2
α
β
α
NMe2
NMe2
NMe2
NMe2
L / base
L
C
O
Me2N
Me2N
O
Me2N
H
C
Ru
10
C
O
OH
O
O
Me2N
Cl
Me2N
Cl
Ru
Ru
Cl
Cl
L
L
L / base= piperidine / piperidine 13a (27 %)
PEt3 / NEt3 13b (38 %)
L= piperidine 14a (57 %)
PEt3 14b (47 %)
Scheme 6.
and 82+, and the different regiochemistry can be formally
interpreted in terms of the difference in the oxidation state.
The Ru(0) center in 10 may prefer coordination to the aro-
matic ring, where back donation is most effective, and,
therefore, the b ring bearing the electron-withdrawing car-
boxyl group should be the preferential reaction site [15].
Subsequent HCl-treatment of 10 gave two different
products depending on the concentration of the HCl solu-
tion. Addition of 36% HCl aqueous solution diluted with
acetone (1:100) caused immediate precipitation of orange
solid 11 Æ Cl2, whereas treatment with 1 M HCl aqueous
solution diluted with acetone (1:20) gave red product 12
(Scheme 6).
The cationic complex 11 Æ Cl2 was characterized as the
dicationic, dimeric, open-protonated form with the Cl
bridges but its detailed characterization was hampered by
its low solubility in organic solvents and formation of dia-
stereomers. But its 1H NMR spectrum observed in DMSO-
d6 contained signals around dH 6.5–4.5 suggesting retention
of the (g6-arene)Ru interaction, and the ESI-MS peaks
around m/z = 1143 (11ꢁCl) supported the formulation.
The resultant green color suggested formation of an open
structure, which should be formed by the protonation at
the lactone group by excess HCl present in the mixture
(Scheme 2). Characterization of the other neutral product
12 will be described below.
actions could be effected by a combination of deprotona-
tion and coordination of a 2e-donor. Piperidine, an
amine, should achieve both of the functions and, in fact,
treatment of 11 Æ Cl2 with piperidine gave the mononuclear
adduct 13a (Scheme 6). Coordination site of the Ru frag-
ment was determined to be the b-ring on the basis of the
shifted signals assigned to that part (Table 1), and the
appearance of the piperidine signals confirmed its coordi-
nation. The closed structure was suggested by the mC@O
vibration at 1777 cmꢁ1 and the lack of mC@O vibration for
a carboxylate group. Reaction of with 112+ Æ Cl2 with
PEt3 in the presence of NEt3 also gave the corresponding
adduct 13b, which showed spectroscopic features similar
to those of 13a (Table 1).
The spectroscopic features of 13a are consistent with the
molecular structure determined by X-ray crystallography
(Fig. 4c). The N–H moiety forms hydrogen-bonding inter-
action with the two chloro ligands.
The red product 12 obtained by treatment of 10 with a
diluted HCl solution was also characterized after conver-
sion to the piperidine (14a) and PEt3 adducts (14b) because
of the low solubility of 12 in common organic solvents and
formation of the diastereomers (Scheme 6). The adduct 14
showed four 1H NMR signals in the region where the coor-
dinated g6-arene signals appeared. Such a data was not
consistent with the desired product 112+ (with three g6-
Ar signals) but 14 could not be characterized by the spec-
troscopic data alone. Attempted crystallization of 14a from
Conversion of 11 Æ Cl2 into mononuclear ring-closed
products through cleavage of the Ru–Cl–Ru bridging inter-

102
M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
ble in organic solvents to hamper solution chemistry. For
Me2N
NMe2
H
C
example, reaction of 3 with 4 Æ PF6 suspended in CH2Cl2
in a manner analogous to the synthesis of 7 Æ PF6 resulted
in a very low conversion and the product could not be iso-
lated from the mixture (Scheme 8). Then we changed the
synthetic method from the g6-coordination (to 4+) to neu-
tralization reaction with a basic precursor, [Cp*Ru(l-
OMe)]2 (15). It was expected that the neutralization was
so fast as to drive dissolution of 3 in CH2Cl2 and finally
led to a high conversion. As we expected, treatment of 3
with an equimolar amount of 15 in MeOH gave the neutral
1:1 adduct 16 (Scheme 8) [16]. Although an analytically
pure sample could not be obtained, the product was satis-
factorily characterized as the a ring adduct on the basis of
O
O
14a'
Me2N
Cl
Ru
NCMe
H
N
Scheme 7.
acetonitrile gave a small amount of orange red crystals
14a0, which was subjected to X-ray crystallography. As a
result, 14a0 turned out to be the zwitterionic reduced prod-
uct resulting from Cl-replacement by MeCN (Scheme 7).
The most striking feature is that a hydrogen atom is
attached to the central carbon atom as confirmed by its
hybridization characterized by the sum of the three C–C–
C angles (336.5ꢁ; cf. 13a: 360ꢁ). On the basis of the struc-
ture of 14a0, complexes 12 and 14 were characterized to
be the l-Cl dimer complex bearing the reduced arene
ligand and the donor-coordinated mononuclear complex,
respectively. In accord with the structures, the lactone
C@O vibration observed for 13 (ꢀ1770 cmꢁ1) disappears
1
its H NMR data (Table 1).
Preparation of the corresponding RuCl2(L) and
Cr(CO)3 complexes was hampered by the lack of appropri-
ate basic precursors.
2.2. Molecular structures of the pH-indicator complexes
Of the complexes obtained by the present study, the
seven Ru complexes (7 Æ PF6, 8 Æ (PF6)2, 13a, 14a0, 5a Æ PF6,
5b Æ PF6, and 5a0 Æ PF6) and the two Cr complexes (6a and
6b) were characterized by X-ray crystallography. Selected
bond lengths are summarized in Tables 2 and 3.
1
in 14 and the singlet H NMR signals around dH 5.9 (14)
The g6-coordination of the arene moieties is evident from
the M–C distances in the ranges of the bonding interactions
are assigned to the central H–CAr3 part. Although the for-
mation mechanism of 14 is not clear, the reduced ligand
might be formed via electron transfer to the open trityl cat-
ion-type intermediate followed by H-abstraction.
˚
(2.1–2.4 A), and the C–C distances of the coordinated arene
parts are slightly longer than those of the non-coordinated
arene parts owing to the back donation from the metal cen-
ters [15]. The coordination is not always symmetrical, and
the complexes are divided into two groups. In the case of
the (g6-C6HxNMe2)M complexes (5b Æ PF6, 6b, 7 Æ PF6,
8 Æ (PF6)2, 13a, and 14a0), the differences between the longest
and shortest M–Ar distances (D in Tables 2 and 3) are in the
2.1.3. Phenolphthalein–RuCp* complexes
In contrast to CVL, phenolphthalein (PP; 3) exhibits
pink color under basic conditions through deprotonation
of the phenolic hydrogen atom (Scheme 2). Compared to
the other pH-indicators studied herein, 3 is sparingly solu-
HO
OH
α
α
C
O
3
β
O
4·PF6
[RuCp*(μ-OMe)]2 15
(1 equiv.)
Ru
Ru
Ru
O
O
HO
OH
O
OH
H
H
C
C
C
O
O
O
O
O
O
16
(81 %)
[16+H]+
[16-H]-
Scheme 8.

M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
103
Table 2
Selected bond lengths (A) for the methyl yellow complexes
˚
Complex
5a Æ PF6
5b Æ PF6
5a0 Æ PF6
6aa
6b
C1–C2
C1–C6
C2–C3
C3–C4
C4–C5
C5–C6
C7–C8
C7–C12
C8–C9
C9–C10
C10–C11
C11–C12
C1–N1
N1–N2
N2–C7
N3–C4
1.405(9)
1.389(8)
1.38(1)
1.416(9)
1.410(9)
1.40(1)
1.416(9)
1.432(8)
1.41(1)
1.42(1)
1.402(9)
1.41(1)
1.408(9)
1.252(7)
1.437(9)
1.36(1)
1.419(6)
1.405(6)
1.416(6)
1.434(6)
1.431(6)
1.417(6)
1.395(6)
1.397(6)
1.390(7)
1.403(7)
1.368(8)
1.381(7)
1.435(5)
1.260(5)
1.432(5)
1.352(6)
1.40(2)
1.35(2)
1.38(1)
1.38(2)
1.37(2)
1.39(2)
1.34(2)
1.45(2)
1.43(2)
1.38(2)
1.39(2)
1.38(2)
1.55(2)
1.13(2)
1.60(1)
1.42(1)
1.44(2)
1.38(2)
1.31(2)
1.40(2)
1.42(2)
1.34(2)
1.44(2)
1.42(2)
1.41(2)
1.39(2)
1.40(2)
1.37(2)
1.37(2)
1.29(2)
1.41(2)
1.38(2)
1.40(2)
1.42(2)
1.39(2)
1.41(2)
1.42(2)
1.32(2)
1.50(2)
1.35(2)
1.38(2)
1.38(2)
1.42(2)
1.43(2)
1.39(2)
1.26(2)
1.43(2)
1.36(2)
1.384(6)
1.420(6)
1.417(6)
1.439(6)
1.436(6)
1.412(6)
1.393(6)
1.380(6)
1.369(7)
1.376(8)
1.389(7)
1.379(7)
1.429(6)
1.253(5)
1.430(5)
1.333(6)
M–Ar
2.221(6) (C7)
2.208(5) (C8)
2.217(6) (C9)
2.212(7) (C10)
2.235(6) (C11)
2.214(5) (C12)
2.211(4) (C1)
2.220(4) (C2)
2.231(4) (C3)
2.374(4) (C4)
2.235(4) (C5)
2.193(4) (C6)
2.237(9) (C7)
2.20(2) (C8)
2.23(1) (C9)
2.18(1) (C10)
2.22(1) (C11)
2.21(2) (C12)
2.21(1)
2.27(2)
2.19(2)
2.21(2)
2.21(2)
2.25(1)
2.24(1) (C7)
2.23(2) (C8)
2.20(2) (C9)
2.24(2) (C10)
2.19(2) (C11)
2.22(2) (C12)
2.224(5) (C1)
2.208(5) (C2)
2.260(5) (C3)
2.394(4) (C4)
2.258(4) (C5)
2.190(4) (C6)
Db
0.027
0.181
0.057
0.08
0.05
0.204
M–L
2.167–2.205(6) (Cp*)
2.166–2.187(5) (Cp*)
2.16–2.208(2) (Cp*)
1.86(2)
1.87(2)
1.83(1)
1.83(2) (C31)
1.78(2) (C32)
1.83(1) (C33)
1.843(5) (C31)
1.835(5) (C32)
1.817(6) (C33)
a
With two independent molecules.
The difference between the longest and shortest M–C distances.
b
Table 3
Selected bond lengths (A) for the crystal violet lactone complexes
˚
Complex
M–Ar
7 Æ PF6
8 Æ (PF6)2
13a
14a0
2.244(8) (Ru1–C11)
2.203(6) (Ru1–C12)
2.217(6) (Ru1–C13)
2.344(8) (Ru1–C14)
2.235(4) (Ru1–C11)
2.213(4) (Ru1–C12)
2.225(3) (Ru1–C13)
2.360(3) (Ru1–C14)
2.214(3) (Ru1–C15)
2.206(4) (Ru1–C16)
2.234(4) (Ru1–C21)
2.197(4) (Ru1–C22)
2.209(3) (Ru1–C23)
2.384(3) (Ru1–C24)
2.236(3) (Ru1–C25)
2.215(4) (Ru1–C26)
2.20(1) (Ru1–C31)
2.206(3) (Ru1–C31)
2.191(3) (Ru1–C32)
2.201(4) (Ru1–C33)
2.360(4) (Ru1–C34)
2.212(3) (Ru1–C35)
2.135(3) (Ru1–C36)
2.13(1) (Ru1–C32)
2.21(2) (Ru1–C33)
2.33(1) (Ru1–C34)
2.15(2) (Ru1–C35)
2.12(2) (Ru1–C36)
Da
0.141
0.154
0.187
0.21
0.225
C–C
(g6–Ar)
1.418(7) (C11–C12)
1.413(8) (C12–C13)
1.426(7) (C13–C14)
1.416(5) (C11–C12)
1.413(5) (C11–C16)
1.413(5) (C12–C13)
1.424(5) (C13–C14)
1.418(5) (C14–C15)
1.418(5) (C15–C16)
1.412(5) (C21–C22)
1.402(5) (C21–C26)
1.418(5) (C22–C23)
1.439(5) (C23–C24)
1.418(5) (C24–C25)
1.410(5) (C25–C26)
1.41(2) (C31–C32)
1.42(2) (C31–C36)
1.37(2) (C32–C33)
1.47(2) (C33–C34)
1.42(2) (C34–C35)
1.41(2) (C35–C36)
1.427(4)(C31–C32)
1.430(6) (C31–C36)
1.431(5) (C32–C33)
1.430(6) (C33–C34)
1.420(5) (C34–C35)
1.415(5) (C35–C36)
C–C (Ar)
N–C
1.37–1.40(2)
1.382–1.419(6)
1.37–1.47(2)
1.375–1.431(6)
1.35(1) (N1–C14)
1.40(1) (N2–C24)
1.355(5) (N1–C14)
1.373(6) (N3–C34)
1.357(5) (N2–C24)
1.36(2) (N3–C34)
1.41(2) (N1–C14)
1.42(2) (N2–C24)
1.360(5) (N3–C34)
1.412(5) (N1–C14)
1.380(5) (N2–C24)
M–L
2.161–2.201(8)
(Ru1–Cp*)
2.163–2.200 (5)
(Ru1–Cp*)
2.167–2.2068(4)
(Ru2–Cp*)
2.408(4) (Ru1–Cl1)
2.418(4) (Ru1–Cl2)
2.12(1) (Ru1–N4)
2.4223(9) (Ru1–Cl1)
2.160(9) (Ru1–N4)
2.070(3) (Ru1–N5)
a
The difference between the longest and shortest M–C distances.
˚
range of 0.14–0.23 A, and the distances from the metal cen-
(NMe2) groups are found for the carbon atoms p and m with
respect to the NMe2 group, respectively. The distortion
should arise from the p-donation from the NMe2 group
leading to the g5-iminocyclohexadienyl structure D0
ter to the carbon atom attached to the NMe2 group are
always longer than the other M–C distances. The shortest
distances for the g6-C6H5NMe2 and g6-C6H4(COO)-

104
M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
2.3. Protonation behavior of the pH-indicator complexes
MLn
MLn
NMe2
NMe2
The changes caused by protonation were monitored by
1H NMR and UV–Vis spectroscopy. For accurate compar-
ison with the parent compounds we attempted determina-
tion of the pKa values for the pH indicator complexes on
the basis of the titration curves. But the insolubility of
the metal complexes in aqueous media hampered the
attempts, and the measurements were made in organic sol-
vents. In general, owing to the attachment of the Lewis
acidic metal fragments use of a stronger acid is needed
for the protonation of the metal complexes. For example,
CVL (2) can be readily protonated by CH3COOH, while
the mono- (7 Æ PF6) and di-cationic species (8 Æ (PF6)2) are
not protonated by CH3COOH but by CF3COOH and
CF3SO3H, respectively.
D
D'
MLn
MLn
N
NMe2
NMe2
N
NMe2
NMe2
N
N
E
E'
MLn
MLn
N
N
N
N
F
F'
Scheme 9.
Protonation behavior of the MY–Ru complexes
5a,b Æ PF6 and 5a0 Æ PF6 was studied in detail by means of
1H NMR and UV–Vis spectroscopy in CH2Cl2/CD2Cl2.
Addition of CH3SO3H to the a ring adduct 5a Æ PF6 caused
(Scheme 9). In the case of the (g6-C6Hx–N@N)M complexes
(5a Æ PF6, 5a0 Æ PF6, and 6a), the differences in the M–C
distances are less than 0.08 A and no systematic distortion
is found indicating that the electronic effect of the azo group
is not significant.
˚
1
broadening of the H NMR spectrum suggesting occur-
rence of a fluxional process, while protonation of the
NPh2 derivative 5a0 Æ PF6 brought about separation of the
doublet pair for the p-N2-C6H4-NPh2 part into two doublet
pairs indicating desymmetrization of the p-phenylene ring
signals. Protonation of the b ring adduct 5b Æ PF6 caused
shifts of the signals to lower field. Although the proton-
ation site cannot be determined by the data obtained so
far, it is thought that the attachment of the cationic RuCp*
fragment destabilizes some of the possible resonance struc-
tures (H0 for 5a Æ PF6/5a0 Æ PF6 and I0 for 5b Æ PF6) leading
to preferential protonation (a for a ring adducts (5a Æ PF6/
5a0 Æ PF6) and b0 for b ring adduct (5b Æ PF6)). Thus, for the
a ring adducts, the quinoidal form G0 hinders free rotation
around the N–C6H4NR2 bond to lead to the desymmetriza-
tion, and the fluxional process should be related to the
rotation of the @N–C single bond (see Scheme 10).
A notable feature observed for the MY complexes is the
orientation of the central C–N@N–C moieties. The MY
moiety in the b ring adduct 5b is distorted from a planar
structure as indicated by the s-trans C–C–N1–N2 and
N1–N2–C–C dihedral angles: 5b Æ PF6: 148.3(4)ꢁ,
148.0(4)ꢁ; cf. 5a Æ PF6: 171.5(5)ꢁ, 174.1(5)ꢁ, 5a0 Æ PF6:
178(1)ꢁ, 174(1)ꢁ; 6a: 165(1)ꢁ, 176(1)ꢁ/166(1)ꢁ, 179(1)ꢁ (with
two independent molecules); 6b: 173.8(6)ꢁ, 165.7(4)ꢁ. In the
case of the a ring adducts, the planar structure should be
stabilized by the quinoidal form E0. On the other hand,
in the case of the b ring adduct, contribution of the quinoi-
dal form F0 is negligible because of the quinoid part (a 4e-
donor) leading to a 16e configuration. The conformation of
the Cr complex 6b appears to be determined by intermolec-
ular p–p stacking of the a rings.
Ru+
Ru+
H
N
N
N
N
NR2
N
NR2
H'
N
H
N
N
N
N
G'
5a·PF6 / 5a'·PF6
Ru+
Ru+
Ru+
H
N
b
NR2
N
NR2
NR2
NR2
a
b
N
H
N
G
H
a
Ru+
Ru+
Ru+
Ru+
H
N
b'
NR2
N
NR2
a'
b'
N
H
N
I
J
a'
Ru+
H
N
5b·PF6
NR2
NR2
J'
N
H
I'
Scheme 10.

M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
105
The changes of the absorptions brought about by addi-
tion of CH3SO3H are monitored by UV–Vis spectroscopy
and shown in Fig. 5. The parent compounds 1 and 10 show
typical chromic behavior with the substantial separations
of the absorption maxima for the two colored forms (1:
>110 nm; 10: 136 nm). The RuCp* adducts also show
smooth spectral change with isosbestic points but the sep-
arations of the absorption maxima for the two colored
form bare smaller than those of the parent compounds
(5a Æ PF6: 24 nm; 5b Æ PF6: 30 nm; 5a0 Æ PF6: 64 nm). Addi-
tion of an excess amount of the acid leads to saturated
spectra, which can be regarded as the spectra for the pro-
tonated forms. On the basis of the intensities of the non-
protonated and protonated forms the equilibrium con-
stants for the protonation equilibrium can be calculated
as shown in Scheme 11. K for 5b Æ PF6 cannot be estimated,
because absorption maxima of the two species are too close
to be separated. Coordination to the RuCp*+ fragment
in a manner similar to the parent CVL molecule (Scheme
2). Acidification caused appearance of the absorption at
605 nm. Because, however, (1) the intensity was variable
and (2) it was very similar to that of CVL, we could not
eliminate the possibility of partial decomposition under
the acidic conditions. We have no evidence for formation
of the diprotonated forms. Similar behavior was noted
for the RuCl2(L) adducts 13. For example, protonation
of the PEt3 complex 13b with CH3SO3H in CD3OD +
CDCl3 (2:1) caused (1) a significant change of the 1H
NMR spectrum, (2) disappearance of the band at 270 nm
assignable the p–p* transition, and (3) appearance of the
band at 602 nm of variable intensity.
The PP–RuCp*+ complex 16 undergoes reversible
deprotonation-protonation upon treatment with NaOH
and CH3COOH, respectively, as revealed by the UV-
change (16: kmax 310 nm (e = 1.9 · 103 Mꢁ1 cmꢁ1) !
[16+H]+: 325 nm (e = 2.0 · 103 Mꢁ1 cmꢁ1), 557 nm (e =
5.5 · 102 Mꢁ1 cmꢁ1)) and 1H NMR monitoring (Scheme 8).
causes a decrease of the K values by the order of 10ꢁ1
–
10ꢁ2 owing to the decreased basicity of MV. The MV–Cr
complexes 6a,b decomposed upon acidification.
3. Conclusions
Protonation of the [(g6-CVL)RuCp*]+ complexes,
7 Æ PF6 and 8 Æ (PF6)2, with CF3COOH or CF3SO3H in
CD2Cl2 causes down field shifts of the signals assigned to
the non-coordinated a ring and the b ring; [7+H]2+: dH
ꢀ5.2, ꢀ5.9 (2H · 2, coordinated a ring) 7.5–8.0 (7H, m,
b and non-coordinated a rings); [8+H]3+: dH ꢀ5.4, ꢀ5.9
(2H · 2, coordinated a ring) 8.35 (2H, s), 8.48 (1H, s, b
ring), while those assigned to the coordinated a ring
remained virtually unaffected. This result may be inter-
preted in terms of the resonance structures of the proton-
ated forms, [7+H]2+ and [8+H]3+ (Scheme 4), where the
electronic structures of the coordinated arene parts are lit-
A series of ruthenium and chromium complexes bearing
pH indicators as the g6-arene ligand, (g6-X)(MLn)y
[X = methyl yellow (1), crystal violet lactone (2), phenol-
phthalein (3); MLn = RuCp*+
,
RuCl2(L), Cr(CO)3;
y = 1,2] is prepared and characterized by spectroscopic
and crystallographic methods. Of the plural arene rings
in the indicator molecules, a specific arene ring can be suc-
cessfully coordinated to the metal center in a selective man-
ner under appropriate conditions (i.e. use of the metal
precursors of different oxidation states and reaction with
the non-protonated and protonated pH indicator). The
obtained indicator complexes show halochromic behavior
depending on pH as observed for the parent molecules.
Although most of the obtained complexes are insoluble
in water and thus cannot be used as usual pH indicators,
the transition pH ranges measured in organic solvents are
shifted to the more acidic side compared to the parent indi-
cator molecules because of the attachment of the electron-
withdrawing metal fragments, which lower the basicity of
the attached pH indicator moieties.
tle affected. Addition of NEt3 to [7+H]2+ and [8+H]3+
,
regenerated 7+ and 82+, respectively, indicating reversibil-
ity of the protonation processes. Because NMR provides
information concerning the equilibrated mixture of the
protonated and non-protonated forms, which are intercon-
verted at a rate faster than the NMR timescale, the proton-
ation was monitored by UV–Vis spectroscopy. Complexes
7+ and 82+, show the UV–Vis bands around 270 nm, which
are assigned to the p–p* transition of the closed structure.
Upon addition of CF3SO3H, the intensity of the absorp-
tion decreased and finally disappeared indicating that acid-
ification of 7+ and 82+ containing the closed forms as the
major species caused the ring opening of the CVL moiety
4. Experimental
4.1. General methods
All manipulations were carried out under an inert atmo-
sphere by using standard Schlenk tube techniques. THF,
ether (Na–K alloy), CH2Cl2, acetone, CH3CN (P2O5),
and MeOH (Mg(OMe)2) were treated with appropriate
K
X
5a+
5b+
5a'+
1
K
X
+
H·A
[H-X]+A-
4.1 x 102
a
-
[H-X]+A-
1
drying agents, distilled, and stored under argon. H, 13C
K=
3.1 x 10
1.9 x 104
4.0 x 102
[X]·[H·A]
and 31P NMR spectra were recorded on Bruker AC-200
(1H, 200 MHz; 31P, 81 MHz) and JEOL EX-400 spectrom-
eters (31P, 162 MHz; 13C, 100 MHz). Solvents for NMR
measurements containing 0.5% TMS were dried over
molecular sieves, degassed, distilled under reduced pres-
1'
a see text.
Scheme 11.

106
M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
sure, and stored under Ar. IR spectra (KBr pellets) were
obtained on a JASCO FT/IR 5300 spectrometer. ESI-
and FD-mass spectra were recorded on a ThermoQuest
Finnigan LCQ Duo and JEOL JMS-700 mass spectrome-
ter, respectively. UV–Vis spectra were obtained with a
JASCO V570 spectrophotometer. Compounds 10 [17],
4 Æ PF6 [8] and 9 [18], 15 [19] were prepared according to
the published method. Other chemicals were purchased
and used as received.
was dried in vacuo and dissolved in CH2Cl2. To the resul-
tant solution was added 4 Æ PF6 (182 mg, 0.36 mmol) dis-
solved in CH2Cl2 via a cannula over 15 min. After the
mixture was stirred overnight, the volatiles were removed
under reduced pressure. The residue was subjected to col-
umn chromatography (alumina). The unreacted 10 was
eluted with CH2Cl2, and elution with CH3CN gave a red
band, crystallization of which from CH2Cl2–ether gave
5a0 Æ PF6 as yellow crystals (129 mg, 0.176 mmol, 80% yield
based on CF3SO3H). dC (CD2Cl2) 152.6, 145.6, 139.5,
129.5, 126.2, 124.9, 119.1, 113.1 (Ar), 97.6 (C5Me5), 87.4,
87.2, 82.0 (g6-Ar), 10.2 (C5Me5). IR (KBr) 1588 (mNN),
839 cmꢁ1 (PF6). ESI-MS: 587 (5a0 Æ PF6). UV (CH2Cl2):
4.2. Preparation of 5a,b Æ PF6
From 1: A mixture of 1 (135 mg, 0.602 mmol) and
4 Æ PF6 (287 mg, 0.568 mmol) dissolved in CH2Cl2
(20 mL) was stirred for 30 min at ambient temperature.
The yellow solution gradually turned into red. After
removal of the volatiles under reduced pressure the residue
was subjected to alumina column chromatography. The
orange band eluted with CH2Cl2 was collected and crystal-
lized from CH2Cl2–ether to give orange crystals (245 mg,
0.404 mmol, 71%), which turned out to be a 1:5 mixture
of 5a Æ PF6 and 5b Æ PF6. From [1-H]OTf: (1-H)OTf was
prepared by treatment of 1 (226 mg, 1.01 mmol) with
TfOH (100 lL, 1.13 mmol) in ether. After cooling the mix-
ture overnight at ꢁ30 ꢁC the supernatant solution was
removed via a pipette and the residue [(1-H)OTf] was dis-
solved in CH2Cl2 (5 mL). To the CH2Cl2 solution was
added dropwise 4 Æ PF6 (293 mg, 0.58 mmol) dissolved in
CH2Cl2 (5 mL) via a cannula. The mixture was stirred
for 1 h and passed through a Celite plug. After removal
of the volatiles under reduced pressure the residue was sub-
jected to alumina column chromatography. Compound 1
(yellow band) was first eluted with CH2Cl2 and elution with
CH3CN gave orange crystals 5 Æ PF6 (190 mg, 0.31 mmol,
54% yield; 5a Æ PF6:5b Æ PF6 = 5:1). The two regioisomers
were separated by recrystallization from CH2Cl2–ether.
5a: dC (CD2Cl2) 153.9, 142.8, 125.8, 111.3 (Ar), 113.9,
87.1, 81.1 (g6-Ar), 97.3 (C5Me5), 39.8 (NMe2), 10.2
(C5Me5). IR: 1600 (mNN), 838 (PF6). ESI-MS: 463
k
max/nm (e/Mꢁ1 cmꢁ1) 491 (2.7 · 104). Anal. Calc. for
C34. 5H35N3F6PClRu (5a0 Æ PF6 Æ (CH2Cl2)0.5): C, 53.60;
H, 4.56; N, 5.43. Found: C, 53.74; H, 4.84; N, 5.33%.
4.4. Preparation of 6a,b
Cr(CO)6 (885 mg, 4.02 mmol) and 1 (1.11 g, 4.92 mmol)
were charged in a glass autoclave. After addition of Bu2O–
THF (10:1, 15 mL) the closed autoclave was heated for
10 h at 120 ꢁC. After removal of the volatiles under
reduced pressure the residue was extracted with THF and
passed through a Celite plug. The filtrate was evaporated
and separated by an alumina column chromatography.
Elution with hexane–Et2O (1:1) gave red crystals composed
of 6a and 6b (132 mg, 0.42 mmol, 10% yield), which were
separated manually. 6a + 6b: FD-MS: 316 (6). Anal. Calc.
for C17H15N3O3Cr: C, 56.51; H, 4.18; N, 11.63. Found: C,
56.18; H, 4.32; N, 11.36%. 6a: dC (acetone-d6) 233.8 (CO),
154.3, 137.9, 126.1, 112.3 (Ar), 124.8, 94.3, 94.1, 90.3 (g6-
Ar), 40.3 (NMe2). IR: 1967, 1870 cmꢁ1 (mCO). 6b: dC (ace-
tone-d6) 234.0 (CO), 152.7, 131.9, 130.0, 123.2 (Ar),
115.1, 112.3, 94.5, 74.9 (g6-Ar), 40.1 (NMe2). IR (KBr)
1942, 1876 cmꢁ1 (mCO).
4.5. Preparation of 7 Æ PF6 and 8 Æ (PF6)2
(5 Æ PF6). UV (CH2Cl2):
k
max/nm (e/Mꢁ1 cmꢁ1
)
480
A mixture of 2 (206 mg, 0.498 mmol) and 4 Æ PF6
(213 mg, 0.422 mmol) dissolved in CH2Cl2 (5 mL) was stir-
red for 2 h at ambient temperature. After removal of the
volatiles under reduced pressure the residue was subjected
to alumina column chromatography. Elution with
CH2Cl2–CH3CN (8:1) gave the recovered 2 followed by
7 Æ PF6 as a blue band. Further elusion with CH3CN gave
8 Æ (PF6)2 as a blue band. The solvents were evaporated
and crystallization of the solids gave 7+ Æ PF6 (242 mg,
0.303 mmol, 72% yield, pale blue crystals) and 8 Æ (PF6)2
(43 mg, 0.037 mmol, 9% yield, pale blue crystals). 7 Æ PF6:
dC (CD3CN) 169.6 (C@O), 151.9, 150.4, 137.4, 126.5,
124.5, 119.1, 111.8, 106.2 (Ar + quart. C), 95.3 (C5Me5),
87.7, 83.6, 82.1, 67.6, 67.5 (g6-Ar), 40.4, 40.1, 39.7
(NMe2), 10.6 (C5Me5). IR 1775 (mC@O), 1622, 1565, 1517,
1444, 1365, 839 (mPF) cmꢁ1. ESI-MS: 652 (7). kmax/nm (e/
(3.6 · 104). Anal. Calc. for C24. 5H31N3F6PClRu [5a Æ P-
F6 Æ (CH2Cl2)1/2]: C, 45.34; H, 4.81; N, 6.47. Found: C,
44.88; H, 4.96; N, 6.41%. 5b Æ PF6: dC (CD2Cl2) 151.7,
132.2, 129.1, 122.6 (Ar), 127.4, 108.1 (g6-Ar), 95.5
(C5Me5), 81.2, 69.7 (g6-Ar), 39.6 (NMe2), 10.4 (C5Me5).
IR (KBr) 1562 (mNN), 837 cmꢁ1 (PF6). ESI-MS: 463
(5 Æ PF6). UV (CH2Cl2):
k ) 337
max/nm (e/Mꢁ1 cmꢁ1
(1.8 · 104), 470 (3.7 · 103). Anal. Calc. for C24H30N3F6
PRu: C, 47.52; H, 4.99; N, 6.93. Found: C, 47.10; H,
5.28; N, 6.60%.
4.3. Preparation of 5a0 Æ PF6
To an ethereal solution of 10 (140 mg, 0.40 mmol) was
added CF3SO3H (20 lL, 0.22 mmol). The resultant mixture
was stored in a refrigerator (ꢁ20 ꢁC) overnight and the
supernatant was removed via a pipette. The obtained solid
M
ꢁ1 cmꢁ1 in CH2Cl2) 275 (1.53 · 104), 349 (1.25 · 103).
Anal. Calc. for C36H44O2N3F6PRu: C, 54.23; H, 5.48; N,

M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
107
5.12. Found: C, 54.27; H, 5.57; N, 5.27%. 8 Æ (PF6)2: dC
(CD3CN) 168.3 (C@O), 152.7, 128.2, 126.5, 125.3, 119.4,
106.7, 102.6 (Ar), 95.4 (C5Me5), 81.2, 81.0, 67.9, 67.6 (g6-
Ar), 40.4, 39.6 (NMe2), 10.6 (C5Me5). IR: 1775 (mC@O),
1622, 1565, 1517, 1444, 1365, 839 (mPF) cmꢁ1. ESI-MS:
removed under reduced pressure and the residue was sepa-
rated by silica gel column chromatography eluted with
CH2Cl2–MeOH (20:1) to give 13a (42 mg, 0.062 mmol,
27% yield) as red-brown powders. 13a: IR: 1777 (mCO),
1611, 1562, 1520, 1355 cmꢁ1. ESI-MS: 637 (13aꢁCl), 552
(2.1 · 104). Anal. Calc. for C31.5H41N4O2Cl3Ru (13a Æ
(CH2Cl2)0.5): C, 52.91; H, 5.78; N, 7.83. Found: C, 52.41;
H, 5.69; N, 7.52%.
1033 (8).
k
max/nm (e/Mꢁ1 cmꢁ1 in CH2Cl2) 277
(13aꢁClꢁL).
k
max/nm (e/Mꢁ1 cmꢁ1 in CH2Cl2) 272
(2.7 · 104). Anal. Calc. for C46H59O2N3 F12P2Ru2: C,
46.48; H, 5.02; N, 3.54. Found: C, 46.89; H, 5.05; N, 3.57%.
4.6. Preparation of 10
4.10. Preparation of 13b
A mixture of 2 (584 mg, 1.41 mmol) and 9 (568 mg,
1.68 mmol) were dissolved in THF (30 mL)–CH3CN
(0.2 mL) and stirred for 2 days at room temperature. The
resultant mixture was concentrated to ca. 15 mL under
reduced pressure and passed through an alumina plug.
The volatiles were removed under reduced pressure and
the obtained residue was subjected to alumina column
chromatography under argon. The liberated naphthalene
and the remaining 9 were first eluted with toluene and then
the product 10 was eluted with toluene–THF (25:1). Com-
pound 10 (499 mg, 0.799 mmol, 47% yield) was obtained as
yellow solid after evaporation of the solvent. Because 10
was very sensitive to the air, it was characterized only by
1H NMR and used without further purification.
To a CH2Cl2 suspension (8 mL) of crude 112+ Æ Cl2
(181 mg, 0.144 mmol) prepared as described above were
added PEt3 (40 lL, 0.27 mmol) and NEt3 (40 lL,
0.29 mmol), and the mixture was stirred for 1 h. The color
of the mixture turned from red to yellow green. Separation
as described for 13a gave 13b (77 mg, 0.011 mmol, 38%
yield) as red-brown powders. 13b: 31P NMR (CDCl3/
H3PO4) dP 11.6. IR: 1771 (mCO), 1612, 1569, 1519 cmꢁ1
ESI-MS: 670 (13bꢁCl). UV (MeOH–CH2Cl2 = 9:1):
.
k
max/nm (e/Mꢁ1 cmꢁ1) 270 (2.7 · 104). Anal. Calc. for
C32. 5H45N3O2Cl3PRu (13b Æ (CH2Cl2)0.5): C, 52.18; H,
6.06; N, 5.62. Found: C, 52.45; H, 6.28; N, 5.96%.
4.11. Preparation of 14a
4.7. Preparation of 11
To crude 12 (88 mg, 0.0746 mmol) were added CH2Cl2
(8 mL) and piperidine (40 lL, 0.40 mmol), and the mixture
was stirred for 2 h at room temperature. Removal of the
volatiles under reduced pressure followed by preparative
TLC separation (silica gel; CH2Cl2–MeOH = 8.5:1) gave
14a (54 mg, 0.080 mmol, 57% yield) as orange crystals.
14a: dC (CD3CN) 169.3 (C@O), 150.2, 137.2, 132.6,
132.5, 131.4, 129.9, 113.2, 112.8 (Ar + quart. C), 89.8,
87.8, 54.7, 51.2, 47.5 (g6-Ar), 41.0, 40.8, 40.7 (NMe2),
55.5, 52.6, 28.9, 28.8, 24.5 (piperidine). IR: 1612, 1567
(mCO), 1578, 1348 cmꢁ1. ESI-MS: 639 (14aꢁCl), 554
(14aꢁClꢁpiperidine) [20].
Dropwise addition of 36% aq. HCl (diluted with acetone:
1/100) to an acetone solution (10 mL) of crude 10 (155 mg,
0.248 mmol) cooled at ꢁ78 ꢁC immediately caused precipi-
tation of orange solid. The mixture was stirred for 6 h at
ꢁ78 ꢁC. Then the supernatant solution was removed via a
pipette, and the residue was washed with pentane and dried
under reduced pressure to afford crude 112+ Æ Cl2 (153 mg.
0.122 mmol, quantitative yield) as orange solid, which was
used without further purification. 112+ Æ Cl2: IR: 1781,
1577, 1509, 1467 cmꢁ1. ESI-MS: 1143 (11ꢁCl).
4.8. Preparation of 12
4.12. Preparation of 14b
Addition of 1 M aq. HCl (diluted with acetone: 1/20; 4
equivalents) to an acetone solution (10 mL) of crude 10
(164 mg, 0.263 mmol) cooled at ꢁ78 ꢁC caused color
changes from yellow to green and finally to orange. During
further stirring for 30 min at room temperature red solid
precipitated out of the solution. The volatiles were
removed under reduced pressure and the resultant residue
was washed with acetone and pentane to leave red solid
12 (quantitative yield), which was used without further
purification. 12: ESI-MS: 1145 (12ꢁCl).
To crude 12 (163 mg, 0.138 mmol) were added CH2Cl2
(8 mL) and PEt3 (75 lL, 0.51 mmol), and the mixture was
stirred for 2 h at room temperature. Removal of the vola-
tiles under reduced pressure followed by preparative TLC
separation (silica gel; CH2Cl2–MeOH = 20:1) gave 14b
(88 mg, 0.124 mmol, 47% yield) as orange crystals. 14b:
dC (CDCl3) 169.6, 149.7, 149.0, 131.5, 131.1, 129.4, 128.9,
112.6, 112.5, 88.5, 81.8, 72.0, 71.5, 42.5, 40.6, 40.41,
40.38, 15.3, 15.0, 8.10, 8.06. dP (CDCl3; H3PO4) 20.0. IR:
1612, 1578, 1570 (mCO), 1348 cmꢁ1 [20].
4.9. Preparation of 13a
4.13. Preparation of 16
To crude 112+ Æ Cl2 (138 mg. 0.110 mmol) and piperidine
(40 lL, 0.40 mmol) was added CH2Cl2 (8 mL) and the
resultant mixture was stirred for 1 h. The volatiles were
A MeOH solution (20 mL) containing 15 (135 mg,
0.504 mmol) and 3 (148 mg, 0.464 mmol) was stirred

Table 4
Crystallographic data
Complex solvate
Formula
5a Æ PF6(CH2Cl2)0.5 5b Æ PF6
5a0 Æ PF6
RuC24H30N3F6PRu C34H34N3F6PRu C17H15N3O3Cr C17H15N3O3Cr C36H44N3O2F6PRu C48H63N3O2-
12P2Cl4Ru2
6a
6b
7 Æ PF6
8 Æ (PF6)2(CH2Cl2)2 13a
14a0 Æ (MeCN)2
C24.5H31N3F6PCl
C31H40N4-
O2Cl2Ru
672.66
Monoclinic
P21/n
C37H50N7O2ClRu
F
Formula weight
Crystal system
Space group
649.02
Monoclinic
C2/c
606.56
Monoclinic
P21/c
730.70
Orthorhombic
Pca21
39.33(1)
7.586(1)
10.993(3)
90
361.32
Triclinic
361.32
Triclinic
796.80
Orthorhombic
Pbcm
7.711(1)
21.464(2)
21.171(2)
90
1347.92
Triclinic
761.37
Monoclinic
P21/c
10.732(1)
27.247(4)
12.608(1)
90
101.340(8)
90
3614.8(7)
4
ꢀ
ꢀ
ꢀ
P1
P1
P1
˚
a (A)
26.003(2)
15.421(1)
13.880(1)
90
10.5359(6)
14.519(1)
16.605(2)
90
6.17(2)
15.79(4)
17.58(5)
107.93(9)
97.89(10)
91.45(10)
1610(7)
4
7.602(2)
17.396(6)
6.240(2)
99.60(2)
95.66(2)
87.72(3)
809.5(4)
2
11.1283(3)
13.8350(9)
18.609(1)
102.549(2)
94.029(3)
91.621(3)
2786.9(3)
2
14.97(1)
13.02(1)
15.96(1)
90
92.48(5)
90
˚
b (A)
˚
c (A)
a (ꢁ)
b (ꢁ)
c (ꢁ)
104.963(3)
90
90.984(4)
90
90
90
90
90
3
˚
V (A )
5376.9(8)
8
2539.7(3)
4
3279(1)
4
3504.0(7)
4
3107(4)
4
Z
Temperature (ꢁC) ꢁ60
ꢁ60
25
ꢁ60
25
1.482
0.727
3985
ꢁ60
ꢁ60
ꢁ60
ꢁ60
dcalc (g cmꢁ3
l (mmꢁ1
Number of
)
1.603
0.803
17137
1.586
1.480
0.589
3960
1.490
1.510
1.606
1.438
1.399
)
0.742
19554
0.731
3773
0.562
21718
0.871
21707
0.711
18663
0.551
25250
diffractions
collected
Number of
variable
330
323
406
433
219
244
655
361
430
R1 for data
0.0700 (for
0.0500 (for
4188 data)
0.1366 (for all
5734 data)
0.0445 (for
2362 data)
0.1502 (for
3960 data)
0.0716 (for
1155 data)
0.2254 (for all
3192 data)
0.0625 (for
2322 data)
0.1905 (for all
2848 data)
0.0915 (for
2750 data)
0.2231 (for all
4028 data)
0.0457 (for
9190 data)
0.1419 (for all
11278 data)
0.0880 (for
1553 data)
0.2773 (for all 0.1699 (for all
6287 data) 7436 data)
0.0589 (for
6212 data)
with [I > 2r(I)] 3984 data)
wR2 0.1885 (for all
5943 data)

M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
109
overnight at ambient temperature, and then the volatiles
were removed under reduced pressure. To the residue was
added acetone (35 mL) and the mixture was stand over-
night at ꢁ30 ꢁC. Removal of the supernatant solution gave
16 (207 mg, 0.374 mmol, yield 81%) as cream-yellow pow-
ders. 16: dC (CDCl3) 169.5, 159.0, 151.3, 147.7, 134.4,
130.1, 129.8, 127.3, 126.0, 125.5, 124.2, 115.8 (Ar + quart.
C), 92.7 (C5Me5). 97.7, 89.2, 84.8, 83.2 (g6-Ar), 10.1
(C5Me5). IR (KBr) 1769 cmꢁ1 (mCO). ESI-MS: 555 (16).
UV (CH2Cl2): kmax/nm (e/Mꢁ1 cmꢁ1) 310 (1.9 · 103) [20].
were refined isotropically. 14a0: The piperidine moiety
was found to be disordered and refined by taking into
account the minor component (0.527:0.423). The H0 and
H1 atoms were refined isotropically and hydrogen atoms
attached to the piperidine carbon atoms were not included
in the refinement. The MeCN solvate molecules were
refined isotropically.
Acknowledgement
This work was supported by the Grant-in-Aid for Scien-
tific Research on Priority Areas (No. 16033219, ‘‘Reaction
Control of Dynamic Complexes’’) from Ministry of Educa-
tion, Culture, Sports, Science and Technology, Japan.
4.14. X-ray crystallography
Single crystals were mounted on glass fibers. Diffraction
measurements except for 5a0 Æ PF6 and 6b were made on a
Rigaku RAXIS IV imaging plate area detector with Mo
References
˚
Ka radiation (k = 0.71069 A) at ꢁ60 ꢁC. Indexing was per-
[1] M. Schwartz (Ed.), Encyclopedia of Smart Materials, Wiley, New
York, 2002.
[2] C.O. Oriakhi (Ed.), Encyclopedia of Smart Materials, Wiley, New
York, 2002, p. 172.
[3] A. Moriuchi, K. Uchida, A. Inagaki, M. Akita, Organometallics 24
(2005) 6382;
K. Takano, A. Inagaki, M. Akita, Chem. Lett. 35 (2006) 434.
[4] M.A. Bennett, K. Khan, E. Wenger, in: E.W. Abel, F.G.A. Stone,
G. Wilkinson (Eds.), Comprehensive Organometallic Chemistry II,
vol. 7, Pergamon Press, Oxford, 1995 (Chapter 8).
[5] M.J. Morris, E.W. Abel, in: F.G.A. Stone, G. Wilkinson (Eds.),
Comprehensive Organometallic Chemistry II, vol. 5, Pergamon Press,
Oxford, 1995 (Chapter 8);
formed from two oscillation images, which were exposed
for 5 min. The crystal-to-detector distance was 110 mm
(2hmax = 55ꢁ). In the reduction of data, Lorentz and polar-
ization corrections and empirical absorption corrections
were made [21]. Crystallographic data and results of struc-
ture refinements are listed in Table 4.
Diffraction measurements of 5a0 Æ PF6 and 6b were made
on a Rigaku AFC5R automated four-circle diffractometer
at 25 ꢁC by using graphite-monochromated Mo Ka radia-
˚
tion (k = 0.71069 A). The unit cells were determined and
refined by a least-squares method using 20 independent
reflections (2h ꢀ 20ꢁ). Data were collected with x–2h scan
technique. If r(F)/F was more than 0.1, a scan was repeated
up to three times and the results were added to the first
scan. Three standard reflections were monitored at every
150 measurements. In the reduction of data, Lorentz and
polarization corrections were made. An empirical absorp-
tion correction (W scan) was made.
A.D. Hunter, V. Mozol, S.D. Tsai, Organometallics 11 (1992) 2251.
[6] E.P. Kundig (Ed.), Transition Metal Arene p-Complexes in Organic
¨
Synthesis and Catalysis, Springer, Berlin, 2004;
H. Le Bozec, D. Touchard, P.H. Dixneuf, Adv. Organomet. Chem.
29 (1989) 163;
R.M. Moriarty, U.S. Gill, Y.-Y. Ku, J. Organomet. Chem. 350
(1988) 157.
[7] M. Hojo, T. Ueda, K. Kawamura, M. Yamazaki, Bull. Chem. Soc.
Jpn. 73 (2000) 347;
The structures were solved by a combination of the
direct methods (SHELXS-86 [22]) and Fourier synthesis (DIR-
DIF-94 [23]). Least-squares refinements were carried out
using SHELXL-97 [22] (refined on F2) linked to teXsan. All
non-hydrogen atoms were refined anisotropically. Hydro-
gen atoms were fixed at the calculated positions unless
otherwise stated. Crystallographic data for the structural
analysis have been deposited at the Cambridge Crystallo-
graphic Data Centre: CCDC 608223 (5a Æ PF6), 608224
(5a0 Æ PF6), 608225 (5b Æ PF6), 608226 (6a), 608227 (6b),
608228 (7 Æ PF6), 608229 (8 Æ (PF6)2), 608230 (13a), and
608231 (14a0). Details of the refinements are as follows:
5b Æ PF6: The methyl hydrogen atoms were refined with
the riding models. 6a: A unit cell contained two indepen-
dent molecules. 7: The molecule sat on a crystallographic
mirror plane and was disordered with respect to the
Ru1–C11–C1–O1 plane. The occupancy of C2 and O2
was 0.5 and the hydrogen atom attached to C22 was not
included in the refinement. 8: One of the two CH2Cl2 sol-
vate molecules was found to be disordered and refined by
taking into account the minor component (Cl3–C62–
Cl4:Cl3a–C62a–Cl4a = 0.52:0.48). The disordered parts
G. Rihs, C. Weis, Dyes Pigments 15 (1991) 107.
[8] B. Steinmetz, W.A. Schenk, Organometallics 18 (1999) 943.
[9] We also examined reaction of the related rhodamine B and confirmed
formation of the analogous 1:1 adduct (a ring adduct).
[10] C. Theocharis, W. Jones, J. Crystallogr. Spectrosc. Res. 14 (1984)
121.
[11] G. Bodes, F. Heinemann, G. Marconi, S. Neumann, U. Zenneck, J.
Organomet. Chem. 641 (2002) 90.
[12] H. Werner, R. Werner, Chem. Ber. 115 (1982) 3766;
M.A. Bennett, A. Smith, J. Chem. Soc., Dalton Trans. (1974) 3011.
[13] M.A. Bennett, T.-N. Huang, T.W. Matheson, A.K. Smith, Inorg.
Synth. 21 (1982) 74.
[14] M.A. Bennett, H. Neumann, M. Thomas, X.Q. Wang, P. Pertici, P.
Salvadori, G. Vitulli, Organometallics (1991) 3237.
[15] R.H. Crabtree, The Organometallic Chemistry of the Transition
Metals, fourth ed., Wiley, New York, 2005.
[16] The 1:2 adduct, [(g6-PP)(RuCp*)2](PF6)2, was obtained by the reaction
of 3 with an excess amount of 4 Æ PF6 and was characterized by its
spectral data. The two a rings of the open form are coordinated by the
RuCp*+ fragments. dH (CDCl3) 8.05 (1H, d, 6.3), 7.75–7.66 (2H, m),
7.52 (2H, d, 6.7) (Ar), 5.36 (2H, d, 8.4), 5.03 (2H, d, 8.4), 4.60 (1H, d,
8.4), 4.53 (2H, d, 8.4) (g6-Ar), 1.69 (Cp*); dC (CDCl3) 168.2 (C@O),
150.7, 134.5, 131.1, 126.1, 124.6, 123.8 (Ar), 95.3 (C5Me5), 91.7, 87.2,
83.8, 82.3, 76.5, 76.3 (g6-Ar), 9.7 (C5Me5); IR: 1773 cmꢁ1 (mCO).
[17] T. Kanbara, M. Oshima, T. Imayasu, K. Hasegawa, Macromolecules
31 (1998) 8725.

110
M. Hirasa et al. / Journal of Organometallic Chemistry 692 (2007) 93–110
[18] M.O. Albers, T.V. Ashworth, H.E. Oosthuizen, E. Singleton, Inorg.
Synth. 26 (1989) 68;
[22] (a) G.M. Sheldrick, SHELXS-86: Program for Crystal Structure
Determination, University of Go¨ttingen, Go¨ttingen, Germany, 1986;
(b) G.M. Sheldrick, SHELXL-97: Program for Crystal Structure
Refinement, University of Go¨ttingen, Go¨ttingen, Germany, 1997.
[23] P.T. Beurskens, G. Admiraal, G. Beurskens, W.P. Bosman, S.
Garcia-Granda, R.O. Gould, J.M.M. Smits, C. Smykalla, The
DIRDIF Program System, Technical Report of the Crystallography
Laboratory, University of Nijmegen, Nijmegen, The Netherland,
1992.
P. Powell, J. Organomet. Chem. 65 (1974) 89, see also, Ref. [14].
[19] S.D. Loren, B.K. Campion, R.H. Heyn, T.D. Tilley, B.E. Bursten,
K.W. Luth, J. Am. Chem. Soc. 111 (1989) 4712.
[20] Despite several attempts analytically pure samples could not be
obtained.
[21] T. Higashi, Program for Absorption Correction, Rigaku Corp.,
Tokyo, Japan, 1995.