Fe(ii), Ru(ii) and Re(i) complexes of endotopic, sterically non-hindering, U-shaped 8,8′-disubstituted-3,3′-biisoquinoline ligands: Syntheses and spectroscopic properties

Article abstract of DOI:10.1039/b712834g

The redox behaviour, optical-absorption spectra and emission properties of U-shaped and elongated disubstituted biisoquinoline ligands and of derived octahedral Fe(ii), Ru(ii), and Re(i) complexes are reported. The ligands are 8,8′-dichloro-3,3′-biisoquin

Full text of DOI:10.1039/b712834g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Fe(II), Ru(II) and Re(I) complexes of endotopic, sterically non-hindering,
U-shaped 8,8ꢀ-disubstituted-3,3ꢀ-biisoquinoline ligands: syntheses and
spectroscopic properties†
Barbara Ventura,*a Francesco Barigelletti,a Fabien Durola,b Lucia Flamigni,a Jean-Pierre Sauvageb and
Oliver S. Wengerb
Received 20th August 2007, Accepted 9th October 2007
First published as an Advance Article on the web 1st November 2007
DOI: 10.1039/b712834g
The redox behaviour, optical-absorption spectra and emission properties of U-shaped and elongated
disubstituted biisoquinoline ligands and of derived octahedral Fe(II), Ru(II), and Re(I) complexes are
reported. The ligands are 8,8ꢀ-dichloro-3,3ꢀ-biisoquinoline (1), 8,8ꢀ-dianisyl-3,3ꢀ-biisoquinoline (2), and
8,8ꢀ-di(phenylanisyl)-3,3ꢀ-biisoquinoline (3), and the complexes are [Fe(2)3]2+, [Fe(3)3]2+,
[Ru(1)(phen)2]2+, [Ru(2)3]2+, [Ru(3)3]2+, [Re(2)(py)(CO)3]+, and [Re(3)(py)(CO)3]+. For the ligands, the
optical properties as observed in dichloromethane are in line with expectations based on the
predominant 1pp* nature of the involved excited states, with contributions at lower energies from 1np*
and 1ILCT (intraligand charge transfer) transitions. For all of the Fe(II), Ru(II), and Re(I) complexes,
studied in acetonitrile, the transitions associated with the lowest-energy absorption band are of 1MLCT
(metal-to-ligand charge transfer) nature. The emission properties, as observed at room temperature and
at 77 K, can be described as follows: (i) the Fe(II) complexes do not emit, either at room temperature or
at 77 K; (ii) the room-temperature emission of the Ru(II) complexes (φem > 10−3, s in the ls range) is of
mixed 3MLCT/3LC character (and similarly at 77 K); and (iii) the room-temperature emission of the
Re(I) complexes (φem ∼3 × 10−3, s < 1 ns) is of 3MLCT character and becomes of 3LC (ligand-centered)
character (s in the ms time scale) at 77 K. The interplay of the involved excited states in determining the
luminescence output is examined.
chelates in d6 transition-metal complexes have appeared but their
Introduction
number is very limited.16
Bidentate chelates of the 2,2ꢀ-bipyridine (bipy) or 1,10-phenan-
Recently, we have reported the syntheses and the preliminary
throline (phen) type play a crucial role as ligands in inorganic
structural properties of non-sterically-hindering but endocyclic
photochemistry.1–4 Their ruthenium(II), rhenium(I) and, to a
lesser extent, osmium(II) and iridium(III) complexes have been
extensively used and continue to be much utilised as photo- and
ligands 1, 2 and 3, illustrated in Scheme 1, and their com-
plexes [Fe(2)3]2+, [Fe(3)3]2+, [Ru(2)3]2+, [Re(2)(py)(CO)3]+, and
[Re(3)(py)(CO)3]+.17,18 The key building block for such systems
electro-active species, in relation to light-energy conversion.5–9
Numerous derivatives of these bidentate chelate archetypes have
been prepared and investigated, leading to a wide range of
electronic properties for the corresponding ligands and their
complexes.10–13 As early as 1988, Balzani and his colleagues
published a review article in which they gathered the various
complexes of this class of compounds.14 In the same publication,
they discuss the properties of hundreds of complexes containing
bidentate ligands of the bipy and phen family. Among the many
ligands mentioned, 3,3ꢀ-biisoquinoline appears as one of the less
frequently used bidentate chelates.10,15 Since then, a few additional
publications dealing with the incorporation of 3,3ꢀ-biisoquinoline
is a 3,3ꢀ-biisoquinoline, the 8 and 8ꢀ positions of which have been
functionalised by various aromatic groups. We now report on the
syntheses of [Ru(1)(phen)2](PF6)2 and [Ru(3)3](PF6)2 and on
the electrochemical and spectroscopic properties in acetonitrile
of the series of Fe(II), Ru(II) and Re(I) complexes; a similar char-
acterization of the ligands has been carried out in dichloromethane
for comparison purposes.
Experimental
Syntheses
The following chemicals were obtained commercially and
were used without further purification: trifluoromethanesul-
fonic anhydride (Acros), [1,2-bis(diphenylphosphino)ethane]-
dichloronickel(II) (Aldrich), trifluoroacetic acid (Aldrich).
1H and 13C NMR spectra were recorded with a Bruker
AVANCE 300 [300 MHz (1H); 75 MHz (13C)] spectrometer,
using deuterated solvent as the lock. The spectra were collected
at 25 ◦C and the chemical shifts were referenced to residual
aIstituto per la Sintesi Organica e la Fotoreattivita` (ISOF), Consiglio
Nazionale delle Ricerche (CNR), Via P. Gobetti 101, 40129, Bologna, Italy.
E-mail: bventura@isof.cnr.it
bLaboratoire de Chimie Organo-Mine´rale, LC3 du CNRS, Institut de
Chimie, Universite´ Louis Pasteur, 4 rue Blaise Pascal, 67070, Strasbourg
Cedex, France
† Electronic supplementary information (ESI) available: Experimental
details of the synthesis of 4, 6 and 7. See DOI: 10.1039/b712834g
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 491–498 | 491
©

View Article Online
powder (0.4 g, 6.1 mmol), potassium bromide (280 mg, 2.4 mmol)
and 8-chloroisoquinolin-3-yl trifluoromethanesulfonate (200 mg,
0.64 mmol) were stirred in 5 mL of dry and degassed THF
for 5 hours at 70 ◦C. The solvent was evaporated and the
residue dissolved in dichloromethane (10 mL), distilled water
(10 mL) and a 32% ammonium hydroxide solution (1 mL). The
organic phase was separated and the aqueous phase extracted
twice with dichloromethane. The combined organic phases were
washed once with distilled water and then evaporated. The crude
product was purified by chromatography on silica gel by using
dichloromethane–methanol (98 : 2) as the eluent to give the
title compound (white solid, 26 mg, 25%). 1H NMR (CDCl3,
300 MHz): d = 9.77 (s, 2H), 8.95 (s, 2H), 7.94-7.91 (m, 2H),
7.66–7.73 (m, 4H). ES-MS m/z = 325.0373 (calculated 325.0299
for C18H10Cl2N2 + H+).
[Ru(1)(phen)2](PF6)2. 17 mg (0.020 mmol) of [Ru(phen)2-
(MeCN)2](PF6)2 and 7.0 mg (0.022 mmol) of 8,8ꢀ-dichloro-3,3ꢀ-
biisoquinoline 1 were dissolved in 2 mL of ethylene glycol. The
solution was heated to 140 ◦C under argon for 3 hours and
then allowed to cool to room temperature. The crude product
was precipitated by addition of a saturated aqueous solution
of potassium hexafluorophosphate and cold distilled water. The
orange precipitate was purified by column chromatography (silica;
eluent: acetone, water, saturated aqueous solution of potassium
nitrate, 80 : 5 : 0.5 (v/v/v); re-precipitation with a saturated
aqueous solution of potassium hexafluorophosphate in water).
This procedure yielded 20 mg (0.019 mmol; 91%) of pure
[Ru(1)(phen)2](PF6)2 as an orange powder. 1H NMP (CD2Cl2.
300 MHz): d = 9.50 (s, 2H), 8.90–8.80 (m, 6H), 8.65 (s, 2H), 8.55
(dd, 2H, J = 5.1, 1.2 Hz), 8.45 (dd, 4H, J = 10.2, 8.7 Hz), 8.23
(d, 2H, J = 8.1 Hz), 7.93–7.80 (m, 8H). EM-MS m/z = 931.07
(calculated 931.03 for C42H26Cl2F6N6PRu+).
Scheme 1 Schematic structures of ligands.
solvent protons as internal standards. 1H: CDCl3 7.27 ppm,
CD2Cl2 5.32 ppm, CD3CN 1.96 ppm, DMSO-d6 2.50 ppm. Mass
spectra were obtained with a VG ZAB-HF spectrometer (FAB)
and a VG-BIOQ triple quadrupole in positive or negative mode
(ES-MS).
All silica-column chromatographies were performed using
Merck Silicagel 60 (0.063–0.200 mm).
Electrochemical measurements were performed with a three-
electrode system consisting of a platinum working electrode, a
platinum wire-counter electrode, and a silver wire as a pseudo-
reference electrode (ferrocene was then used as a reference).
All measurements were carried out under Ar, in degassed
spectroscopic-grade acetonitrile, using 0.1 M n-Bu4NBF4 solu-
tions as supporting electrolyte. An EG&G Princeton Applied
Research model 273A potentiostat connected to a computer was
used (software from Princeton Applied Research).
[Ru(3)3](PF6)2. 211 mg (0.340 mmol) of 8,8ꢀ-di(phenylanisyl)-
3,3ꢀ-biisoquinoline were suspended in 300 mL of 1,2-dichloro-
ethane. With the addition of 0.5 mL trifluoroacetic acid this
mixture turned into a yellow homogenous solution. After adding
49 mg (0.100 mmol) of ruthenium(II) tetra(dimethylsulfoxide)
dichloride the solution was refluxed under argon for 24 hours
whereby it turned deep red. Then the solvent was evaporated, the
residue was taken up in tetrahydrofuran, and the crude product
was precipitated by addition of a saturated aqueous solution
of potassium hexafluorophosphate. The orange precipitate was
purified by column chromatography (silica; eluent: acetonitrile,
water, saturated aqueous solution of potassium nitrate, 200 : 10 : 1
(v/v/v); re-precipitation with a saturated aqueous solution of
potassium hexafluorophosphate in water). This procedure yielded
130 mg (0.058 mmol; 58%) of pure [Ru(3)3](PF6)2 as an orange
powder. 1H NMR (CD3CN): d = 8.88 (s, 6H; H1), 8.35 (s, 6H; H4),
7.75 (d, 6H, J = 8.1 Hz; H7), 7.27 (d, 12H, J = 8.7 Hz), 7.20–7.02
(m, 36H), 6.79 (d, 12H, J = 8.1 Hz), 3.91 (s, 18H, OCH3). ES-MS
m/z = 981.3194 (calculated 981.3230 for C132H96N6O6Ru2+).
8-Chloroisoquinolin-3-yl trifluoromethanesulfonate.
A solu-
tion of 8-chloroisoquinolin-3-ol (1.04 g, 5.8 mmol) in dry pyridine
(40 mL) was cooled to 0 ◦C and carefully treated with triflu-
oromethanesulfonic anhydride (1.6 mL, 2.7 g, 9.2 mmol). The
mixture was allowed to warm to room temperature and stirred
over night. The solvent was removed under reduced pressure. The
crude product was purified by chromatography on silica gel by
using pentane–diethyl ether (1 : 4) as the eluent to afford the
title product (colourless crystals, 1.71 g, 95%). 1H NMR (CD2Cl2,
300 MHz): d = 9.52 (s, 1H), 7.93–7.89 (m, 1H), 7.80–7.76 (m,
2H), 7.67 (s, 1H). ES-MS m/z = 312.1150 (calculated 311.9709
for C10H5ClF3NO3S + H+).
Optical spectroscopy
8,8ꢀ-Dichloro-3,3ꢀ-biisoquinoline 1. Zinc powder was activated
by treatment of 20 g in 100 mL of acetic acid for 1 hour. After
filtration, the powder was washed three times with distilled water
and dried under vacuum for 6 hours at 120 ◦C. Dichloro[1,2-
bis(diphenylphosphino)ethane]nickel(II) (32 mg, 0.06 mmol), zinc
Spectrophotometric-grade dichloromethane and acetonitrile at
295 K and at 77 K were used without further purification. Absorp-
tion spectra of dilute solutions (2 × 10−5 M) in dichloromethane
(for the ligands) and acetonitrile (for the complexes) were
492 | Dalton Trans., 2008, 491–498
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
obtained with a Perkin-Elmer Lambda 45 UV-vis spectrometer.
For luminescence experiments, the samples were placed in fluo-
rimetric 1-cm-path cuvettes and purged of oxygen by bubbling
with argon or by evacuating with repeated freeze–pump–thaw
cycles. Uncorrected luminescence spectra were obtained with a
Spex Fluorolog II spectrofluorimeter equipped with a Hamamatsu
R928 phototube. Sample solutions were excited at the indicated
wavelength (see below), and dilution was adjusted to obtain
absorbance values ≤0.15. While uncorrected luminescence-band
maxima are used throughout the text, corrected spectra were
employed for the determination of the luminescence quantum
yields. The correction procedure is based on the use of soft-
ware which takes care of the wavelength-dependent phototube
response. From the wavelength-integrated area of the corrected
luminescence spectra we obtained luminescence quantum yields
φem with reference to a standard with known yield, φr, and by using
eqn (1):19
Results and discussion
Syntheses and characterization
The synthetic route leading to biisoquinoline 1 is represented
in Scheme 2. Isoquinoline 7, functionalized on its 3 and 8
positions, is synthesized following an existing methodology:22
sodium diethoxy acetate 4 is first activated with thionyl chloride
and subsequently condensed with 2-chlorobenzylamine 5 in 59%
yield. The resulting amide 6 cyclizes in concentrated sulfuric
acid to form 8-chloroisoquinolin-3-ol 7 in a so-called Pomeranz–
Fritsch reaction (71% yield). Activation of its alcohol function
using triflic anhydride results in the formation of triflate compound
8 in nearly quantitative yield (95%). Ligand 1 is then obtained in
25% yield by a nickel-catalyzed homocoupling reaction between
two triflate molecules 8. In addition to the relatively poor yield of
this last step, the purification of 1 is difficult because of its very
low solubility, which is why the global strategy for synthesis of
functionalized biisoquinolines has been revised.
In spite of the small amount of 1 that was obtained, one
heteroleptic complex of ruthenium, [Ru(1)(phen)2]2+, has been
synthesised from [Ru(MeCN)2(phen)2]2+ with a very good yield
(91%).
φem
φr
Absr · g2 · (area)
Abs · g2r · (area)r
=
(1)
Ligands 2 and 3 have been described in previous papers16 as
well as their Fe(II), Re(I) complexes and the homoleptic [Ru(2)3]2+
complex.16–18 [Ru(3)3]2+ has been obtained in a relatively different
way than [Ru(2)3]2+ because of the poor solubility of 3. In this
case, addition of an acid is necessary to solubilize the ligand,
and complexation is then possible thanks to the better association
constant of the complex.
where Abs and g are the absorbance values and refractive index of
the solvent respectively. For the ligands kexc was 330 nm and the
reference was quinine sulfate in aerated 1 N sulfuric acid, φr =
0.546.20 For the Fe(II) and Ru(II) complexes, kexc was 450 nm; for
the Re(I) complexes kexc was 390 nm; for both cases the reference
was [Ru(bpy)3]Cl2 in air-equilibrated water, φr = 0.028.21 Band
maxima and relative luminescence intensities were affected by an
uncertainty of 2 nm and 20% respectively. Luminescence lifetimes
were obtained by using an IBH 5000 F single-photon counting
spectrometer. Excitation was performed by using nanoLED
sources at 331 nm for the ligands, and 465 and 373 nm for the
Ru(II) and Re(I) complexes respectively. The Fe(II) complexes
did not show any emission. Analysis of the luminescence-decay
profiles against time was accomplished by using software provided
by the manufacturers. EHMO calculations were performed on
the ligands with standard programs from the CS ChemOffice
package, Cambridge Corporation, MA; the geometry of the
ligands was optimized by using a MM2 approach, with the
chelating rings kept parallel to simulate the coordination arrange-
ment.
Electrochemistry
ox
The obtained oxidation potentials E1/2 for the complexes are
collected in Table 1, and are discussed below in connection with
the spectroscopic properties. The solubility of [Ru(3)3](PF6)2 in
MeCN is about 1 mg in 5 mL (about 5 mg in 5 mL for the other
complexes), and, due to the low solubility, the value given in the
table should be taken with some care.
Absorption spectra
The absorption spectra of ligands 1, 2, and 3 in dichloromethane
are shown in Fig. 1, and absorption data are listed in Table 2. The
high-intensity bands peaking below 300 nm (e ∼5 × 104 M−1 s−1)
Scheme 2 The synthesis of 4, 6, and 7 is outlined in the ESI.†
The Royal Society of Chemistry 2008
This journal is
Dalton Trans., 2008, 491–498 | 493
©

View Article Online
are ascribed to allowed 1pp* transitions, while for the region
300–400 nm transitions of different character overlap.23,24 In this
region, contributions from both weaker 1np* and intraligand
charge transfer (1ILCT) transitions might be expected, the
latter due also to the presence of electron-releasing (methoxy)
and electron-accepting (chloride) subunits interacting with the
biisoquinoline ligand frame. The spectra appear reasonably
homogeneous within the series, except for a progressive extension
of the absorption tail in the range 350–400 nm on passing from
1 to 3. This is ascribable to the elongation of the peripherical
arms of the biisoquinoline unit, which leads to an increased
delocalisation of the p orbitals and a smaller HOMO–LUMO
gap (see below for a discussion on this point).
Table 1 Half-wave oxidation potentials of the Fe(II), Ru(II) and Re(I)
complexesa
E1/2ox/V vs. SCE
[Fe(2)3]2+
+1.01
+1.00
+1.32
+1.18
(+1.62)b
+1.93
+1.75
[Fe(3)3]2+
[Ru(1)(phen)2]2+
[Ru(2)3]2+
[Ru(3)3]2+
[Re(2)(py)(CO)3]+
[Re(3)(py)(CO)3]+
a In degassed acetonitrile, using 0.1 M n-Bu4NBF4 solutions as supporting
electrolyte. b Value affected by a high uncertainty due to the poor solubility
of the complex.
The absorption spectra of the examined Fe(II), Ru(II) and Re(I)
complexes in acetonitrile are displayed in Fig. 2–4 respectively.
Absorption data are collected in Table 2. For discussing the general
features of the absorption spectra of the complexes, one may take
into account the spectral portions below and above 400 nm, as
dealt with in the following.
Fig. 1 Absorption spectra of ligands 1 (thick solid line), 2 (thin solid line)
and 3 (dashed line) in dichloromethane at room temperature.
Table 2 Absorption dataa
Fig. 2 Absorption spectra of complexes [Fe(2)3]2+ (thin solid line) and
[Fe(3)3]2+ (dashed line) in acetonitrile at room temperature.
kmax/nm
e/(105 M−1 cm−1
)
1
2
3
258
334
253
336
269
336
378
258
361
458
267
364
458
263
342
413
255
347
405
266
357
433
259
391
268
390
0.54
0.23
0.55
0.34
0.48
0.28
0.07 (sh)
1.37
0.83
0.16
2.28
1.06
0.17
0.76
0.26
0.14
1.31
1.14
0.25
2.18
1.09 (sh)
0.18
0.36
0.14
0.66
0.24
[Fe(2)3]2+
[Fe(3)3]2+
[Ru(1)(phen)2]2+
[Ru(2)3]2+
[Ru(3)3]2+
Fig. 3 Absorption spectra of complexes [Ru(1)(phen)2]2+ (thick solid
line), [Ru(2)3]2+ (thin solid line) and [Ru(3)3]2+ (dashed line) in acetonitrile
at room temperature.
[Re(2)(py)(CO)3]+
[Re(3)(py)(CO)3]+
(i) For the Fe(II)25,26 and Ru(II)14 complexes, the bands below
400 nm can be ascribed to transitions mainly centered on the
ligands, which likely include 1LC and 1ILCT transitions, as
a In dichloromethane solvent for the ligands and acetonitrile for the
complexes, at room temperature.
494 | Dalton Trans., 2008, 491–498
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
orbital for a M → L CT transition), with respect to what happens
for the phen ligand. Results from EHMO (Extended Hu¨ckel
Molecular Orbitals) calculations are summarized in Fig. 5, where
for comparison purposes the cases of bpy and i-biq are also given;
the bottom plot reports the estimated LUMO levels. The fact
1
that for [Ru(2)3]2+, and [Ru(3)3]2+ the lowest-lying MLCT band
is blue-shifted with respect to what happens for [Ru(phen)3]2+ is
apparently in contrast with the predictions based on the trend for
the ligand LUMO levels. This discrepancy might be a consequence
of a complex balance of electrostatic factors that could not
be investigated in detail because we were unable to observe
the ligand-based reduction processes. With regard to the metal-
based oxidation potentials E1/2ox (Table 1), these appear consistent
with the absorption properties discussed above. For instance, for
the series [Fe(2)3]2+, [Ru(2)3]2+, and [Re(2)(py)(CO)3]+, E1/2
=
ox
Fig. 4 Absorption spectra of complexes [Re(2)(py)(CO)3]+ (thin solid
line) and [Re(3)(py)(CO)3]+ (dashed line) in acetonitrile at room
temperature.
+1.01, +1.18, and +1.93 V respectively (vs. SCE). For this series,
therefore, given the metal-to-ligand CT nature of the transitions
in the lowest-energy region, and by noticing that the same ligand
ox
mentioned above. These exhibit a bathochromic shift with respect
to the case of corresponding ligands, as suggested by comparison
of Fig. 1 and 2. This is likely a consequence of the stabilization of
the orbitals of the ligand upon coordination to a positive metal
ion, in turn leading to a reduced energy gap between HOMO
and LUMO orbitals. The perturbation effect of the metal centre
on the electronic transitions of the ligands can further account
for the observed changes in the extinction coefficient (e, Table 2).
For the Re(I) complexes,27 the same line of reasoning seems to
hold, however, the LC-absorption region is somewhat confined
to higher energy, i.e. below 350 nm. The MLCT (metal-to-ligand
charge transfer) absorption transitions fall in the 350 to 400 nm
region, in agreement with previously reported cases.28,29 This is
due to the presence of the electron-withdrawing CO groups which
render the metal centre more electron-deficient than its formal +1
charge.
2 is involved, a correlation between E1/2 values and the energy
of the 1MLCT transitions is expected and found to (qualitatively)
hold, kmax being 458, 405, and 391 nm (Table 2) respectively.
(ii) For the Fe(II) and Ru(II) complexes studied here, the
band systems in the region 400–600 nm can be assigned to
1MLCT transitions, in agreement with previous reports on related
complexes.14,25 In particular, the 1MLCT peak is found at 458 nm
for the Fe(II) complexes and in the 405 to 433 nm range for
the Ru(II) complexes, with extinction coefficients, e ∼ 2
1 ×
104 M−1 cm−1, in line with expectation for MLCT transitions.
With regard to the latter complexes, it is interesting to recall
the behaviour of a previously investigated series of heteroleptic
species, [Ru(bpy)n(i-biq)3−n]2+, with n = 1, 2, 3, and where i-biq is
3,3ꢀ-biisoquinoline.10,15 In that series, the lowest-energy absorption
peak is at 452 nm for [Ru(bpy)3]2+, 392 nm for [Ru(i-biq)3]2+ and
at intermediate positions for the intermediate cases, [Ru(bpy)2(i-
biq)]2+ and [Ru(bpy)(i-biq)2]2+. This was explained by admitting
that Ru → bpy CT (charge transfer) states are lower in energy
than the corresponding Ru → i-biq CT levels. Given the close
1
Fig. 5 Results of EHMO calculations for the indicated ligands with a
planar arrangement of the chelating subunits. Bottom panel, LUMO level;
top panel, LUMO–HOMO gap.
The results from EHMO calculations provide another useful
indication. As illustrated in the top panel of Fig. 5, the HOMO–
LUMO gap (D) for i-biq, 1, 2, and 3 is smaller than for the
cases of bpy and phen (D = 1.55 vs. 2.6 eV respectively). Along
the same series of ligands, the LUMO orbital (the one bound
to play as accepting site for the promoted electron during the
MLCT transition) undergoes a smaller stabilization, by ca. 1.30 eV.
According to these results, the LC excited levels for i-biq, 1, 2, and
3 might be expected to be low enough in energy to approach the
corresponding MLCT levels, an occurrence that could result in
a mixed LC/MLCT character for the emission of the complexes
(see below).
similarity of bpy and phen ligands,14 these findings suggest that
1
for [Ru(1)(phen)2]2+ the lowest-lying MLCT transitions, kmax
=
413 nm (Table 2), could involve the phen ligand and not ligand 1.
Consistent with this, for [Ru(2)3]2+ and [Ru(3)3]2+ the lowest-lying
1MLCT band likewise appears blue-shifted (Table 2) with respect
to the case of [Ru(phen)3]2+, kmax = 477 nm.14
It is interesting to notice that for the series of biisoquinoline
ligands 1, 2, and 3, the increased size is expected to result in
a significant stabilization of the LUMO orbital (the accepting
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 491–498 | 495
©

View Article Online
Luminescence and photophysics
high fluorescence quantum yields (0.11–0.56) are in agreement
with values observed for similar compounds.30 It is interesting to
note that along the 1, 2, 3 series an increase of the fluorescence
quantum yield is observed (Table 3). Since 3 also presents the
smallest lifetime in the ligand series, s = 2.8 ns, its radiative rate
constant (kr = φem/s) is rather high, kr = 2 × 108 s−1.31
The luminescence properties of ligands 1, 2, and 3 in aerated
dichloromethane both at room temperature and at 77 K are
reported in Table 3 and illustrated in Fig. 6. For these ligands, the
room-temperature emission spectra can reasonably be assigned
to the lowest singlet level.30 At this temperature, a bathochromic
shift is registered on passing from 1 to 3, the latter presenting
a tail extending up to 600 nm. This trend is in accordance with
the observed broadening of the absorption spectra toward low
energies. Lifetimes of the order of 3–4 ns (Table 3) and quite
The 77 K fluorescence spectra of ligands 1 and 3 are red-shifted
1
compared to the room-temperature cases, as expected for LC
emission of predominantly pp* character,31 whereas for ligand 2,
emission peaks are found to be practically coincident. Ligand 1
fluorescence shows the largest red-shift (42 nm) on passing from
fluid to rigid solvent whereas for 3 a smaller bathochromic shift
(9 nm) is found (Fig. 6 and Table 3). No phosphorescence could be
detected for 1, 2, and 3. Ligand 1 differs from the other two ligands
also because its lifetime increases from 4.0 to 11.0 ns in passing
from room temperature to 77 K, whereas for ligands 2 and 3 the
lifetimes appear almost identical at the two temperatures (Table 3).
This peculiar behaviour can most likely be related to a smaller
rotational freedom in the glass for the bulkier 2 and 3 ligands
as compared to 1. Actually, free rotation about the central C–C
single bond can presumably affect the excited-state deactivation.
This could account for the low fluorescence quantum yield of 1 at
room temperature as compared to 2 and 3 (Table 3). Accordingly,
the fluorescence intensity of 1 becomes strongly enhanced in the
frozen solvent (about three times, according to what is suggested
by the lifetimes), whereas ligands 2 and 3, already strong emitters
at room temperature (φem = 0.39 and 0.56 respectively, Table 3),
are less sensitive to the rigidification effect.
No luminescence can be registered for [Fe(2)3]2+ and [Fe(3)3]2+,
either at room temperature or at 77 K. This is due to the fact that
the lowest-lying excited levels for polyimine complexes of the Fe(II)
centre are non-emissive, being of MC (metal-centered) nature, with
the 3MLCT levels lying higher in energy, as is well documented in
the literature.25
The Ru(II) and Re(I) complexes studied here exhibit weak
triplet luminescence features, with room-temperature lumines-
cence quantum yields falling in the range 7.5 × 10−3–3 × 10−4.
A summary of their luminescence properties is reported in Table 3
and the spectra are displayed in Fig. 7 and 8. We examine first the
simpler case, that of the Re(I) metal centre.
Fig. 6 Arbitrarily scaled luminescence spectra of ligands 1 (top), 2
(middle) and 3 (bottom) in dichloromethane at room temperature (solid
line) and at 77 K (dotted line). Excitation at 330 nm.
For the Re(I) complexes, the presence of the electron-
withdrawing CO groups renders the metal centre more
Table 3 Luminescence properties of ligands and Ru(II) and Re(I) complexesa
298 K
77 K
kem/nmb
φem
s/ns
kem/nmb
s/ns
c
1d
2d
3d
378
389
397
586
566
566
470
530
0.11
0.39
0.56
0.0075 (0.0034)
0.0027 (0.0005)
0.0018 (0.0004)
0.0003
4.0
4.2
2.8
260 (80)
920 (120)
870 (130)
<0.5
420
390
406
581
581
581
572
577
11.0
4.2
2.9
[Ru(1)(phen)2]2+
,
5.2 × 103
5.1 × 103
4.6 × 103
1.5 × 106
1.6 × 106
e
[Ru(2)3]2+
[Ru(3)3]2+
,
,
e
e
[Re(2)(py)(CO)3]+,f
[Re(3)(py)(CO)3]+,f
0.0029
<0.5
a In degassed dichloromethane solvent for the ligands and acetonitrile for the complexes; within brackets, values for air-equilibrated solvents. b Emission
maxima. c Luminescence quantum yields. d Excitation at 330 and 331 nm for the luminescence spectra and lifetime determinations respectively. e Excitation
at 450 and 465 nm for the luminescence spectra and lifetime determinations respectively. f Excitation at 410 and 373 nm for the luminescence spectra and
lifetime determinations respectively.
496 | Dalton Trans., 2008, 491–498
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
electron-deficient than its formal +1 charge, as also suggested
by the high values for the oxidation step (Table 1).27–29 The
room-temperature emission, of 3MLCT nature as it happens
for this type of complexes, falls in the high-energy region of
the visible range (Table 3). In particular, for [Re(2)(py)(CO)3]+
the emission maximum, kem, falls at higher energy than for
[Re(3)(py)(CO)3]+ (Fig. 8 and Table 3). In the latter case, the
presence of a more delocalised ligand appears to be a stabilizing
factor. It is also to be noted that [Re(3)(py)(CO)3]+ is easier
to oxidize than [Re(2)(py)(CO)3]+ (E1/2 = +1.75 vs. +1.93 V
ox
respectively, Table 1). For both complexes, the luminescence
lifetimes are shorter than 1 ns (Table 3) and all of this is consistent
with a room-temperature 3Re → L CT emission.27–29
Previous reports on the luminescence of Re(I) complexes
indicated that their 3MLCT luminescent level can be strongly
destabilized at 77 K, where the solvent is frozen. The blue-shift of
the 3MLCT energy level can be as high as 2700 cm−1.28 At odds with
expectations, for both [Re(2)(py)(CO)3]+ and [Re(3)(py)(CO)3]+
the emission spectra at 77 K are well resolved, peak at very
close values (kem = 572 and 577 nm respectively), and are red-
shifted by ∼3800 and 1500 cm−1 respectively, with respect to what
happens at room temperature (Table 3 and Fig. 8). In addition,
time-resolved determinations reveal that these emissions have
lifetimes of 1.5 and 1.6 ms respectively. All of this suggests that
3
the expected displacement to high energy of the MLCT levels
unveils the presence of 3LC levels.27 These levels are centered on the
coordinated biisoquinoline ligands 2 and 3, for [Re(2)(py)(CO)3]+
and [Re(3)(py)(CO)3]+ respectively, and are responsible for the
detected emission.
Fig.
7 Arbitrarily scaled luminescence spectra of complexes
[Ru(1)(phen)2]2+ (thick solid line), [Ru(2)3]2+ (thin solid line) and
[Ru(3)3]2+ (dashed line) in acetonitrile at room temperature (top) and 77 K
(bottom). Excitation at 450 nm.
At room temperature, the ruthenium series shows some peculiar
properties with respect to what is commonly observed in the vast
family of Ru(II)–polyimine complexes.10,14,15 To begin with, the
emission of [Ru(1)(phen)2]2+ is blue-shifted, kem = 586 nm, and
its quantum yield is φem = 0.0075 (Table 3), less than half that of
[Ru(phen)3]2+ (kem = 604 nm, φem = 0.02).14,32 A similar trend in
the emission energy is apparent for [Ru(2)3]2+ and [Ru(3)3]2+, with
kem = 566 nm in both cases, i.e. further blue-shifted with respect
to what happens for [Ru(1)(phen)2]2+. In addition, the spectral
profiles for [Ru(2)3]2+ and [Ru(3)3]2+ appear narrower than that
for [Ru(1)(phen)2]2+ (see Fig. 7, room-temperature case, upper
panel). The lifetime of these complexes is ∼0.9 ls (Table 3), as
for [Ru(phen)3]2+ and [Ru(bpy)3]2+.14 Interestingly, for [Ru(2)3]2+
and [Ru(3)3]2+ the radiative rate constant, kr ∼ 2–3 × 103 s−1,
3
is much lower than that for the MLCT emitters [Ru(phen)3]2+
or [Ru(bpy)3]2+, kr ∼ 8 × 104 s−1.33 In conclusion, all of this
suggests a LC contribution to the room-temperature emission for
[Ru(1)(phen)2]2+, [Ru(2)3]2+ and [Ru(3)3]2+. In fact, for the closely
related [Ru(i-biq)3]2+ species, the LC nature of the emission has
been since long established.15
Low-temperature results for these Ru(II) complexes support a
predominant LC nature for the emission, Table 3 and Fig. 7.
This conclusion is based on the observations that (i) only for
[Ru(1)(phen)2]2+ is a small blue-shift of emission peak (5 nm)
observed upon passing from room temperature to 77 K; instead,
a red-shift for the peak (15 nm) is registered for [Ru(2)3]2+ and
[Ru(3)3]2+, (ii) for [Ru(1)(phen)2]2+, [Ru(2)3]2+, and [Ru(3)3]2+, the
77 K emission spectra peak at the same value, kem = 581 nm,
(iii) the emission profiles at 77 K are narrow and well resolved for
all the three complexes. While these data support an LC nature
Fig. 8 Arbitrarily scaled luminescence spectra of complexes [Re(2)(py)-
(CO)3]+ (solid line), and [Re(3)(py)(CO)3]+ (dashed line) in acetonitrile at
room temperature (top) and 77 K (bottom). Excitation at 390 nm.
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 491–498 | 497
©

View Article Online
for the emission properties at 77 K, it remains to be said that
7 M. Hissler, J. E. McGarrah, W. B. Connick, D. K. Geiger, S. D.
Cummings and R. Eisenberg, Coord. Chem. Rev., 2000, 208, 115.
8 O. Johansson, H. Wolpher, M. Borgstrom, L. Hammarstrom, J.
Bergquist, L. C. Sun and B. Akermark, Chem. Commun., 2004, 194.
9 P. W. Du, J. Schneider, P. Jarosz and R. Eisenberg, J. Am. Chem. Soc.,
2006, 128, 7726.
10 A. Juris, F. Barigelletti, V. Balzani, P. Belser and A. von Zelewsky, Inorg.
Chem., 1985, 24, 202.
11 F. Barigelletti, A. Juris, V. Balzani, P. Belser and A. von Zelewsky, Inorg.
Chem., 1987, 26, 4115.
3
the lifetime values are typical of a MLCT emission, s = 4.6–
5.2 ls, Table 3. This apparent discrepancy might be the result of a
complex interplay of excited states, an occurrence sometimes met
for closely lying LC and MLCT levels.34–38
Conclusions
Substituted 3,3ꢀ-biisoquinolines have virtually never been used
in coordination photochemistry. In the present study we have
made and studied transition-metal complexes of various 8,8ꢀ-
disubstituted-3,3ꢀ-biisoquinolines, the two substituents attached
on the 8 and 8ꢀ positions of the ligand being Cl atoms, p-anisyl
groups or p-(4ꢀ-methoxy-1,1ꢀ-biphenyl) substituents. The general
shape of the chelating unit is such that, although its coordination
site is endotopic, i.e. directed towards the concave part of the
molecule, it does not interfere from a steric viewpoint with the
complexed metal and its close surrounding, or only very weakly.
The complexes containing the Fe(II), Ru(II) or Re(I) centres and
these endotopic, sterically non-hindering, U-shaped ligands have
been characterized from a photophysical viewpoint. For all of the
complexes the lowest-energy absorption band is of 1MLCT nature.
The emission properties of the Ru(II) and Re(I) complexes show
some peculiar features: the emission of the Ru(II) complexes is of
12 M. Kato, K. Sasano, C. Kosuge, M. Yamazaki, S. Yano and M. Kimura,
Inorg. Chem., 1996, 35, 116.
13 M. I. Azocar, L. Mikelsons, G. Ferraudi, S. Moya, J. Guerrero, P.
Aguirre and C. Martinez, Organometallics, 2004, 23, 5967.
14 A. Juris, V. Balzani, F. Barigelletti, S. Campagna, P. Belser and A. von
Zelewsky, Coord. Chem. Rev., 1988, 84, 85.
15 P. Belser, A. von Zelewsky, A. Juris, F. Barigelletti, A. Tucci and V.
Balzani, Chem. Phys. Lett., 1982, 89, 101.
16 F. Durola, J. P. Sauvage and O. S. Wenger, Chem. Commun., 2006,
171.
17 F. Durola, D. Hanss, P. Roesel, J. P. Sauvage and O. S. Wenger,
Eur. J. Org. Chem., 2007, 125.
18 F. Durola, L. Russo, J.-P. Sauvage, K. Rissanen and O. S. Wenger,
Chem.–Eur. J., 2007, 13, 8749.
19 J. N. Demas and G. A. Crosby, J. Phys. Chem., 1971, 75, 991.
20 D. F. Eaton, Pure Appl. Chem., 1988, 60, 1107.
21 K. Nakamaru, Bull. Chem. Soc. Jpn., 1982, 55, 2967.
22 H. Fukumi and H. Kurihara, Heterocycles, 1978, 9, 1197.
23 D. Collado, E. Perez-Inestrosa and R. Suau, J. Org. Chem., 2003, 68,
3574.
24 N. Kang, A. R. Hlil and A. S. Hay, J. Polym. Sci., Part A: Polym.
3
mixed MLCT/3LC character both at room temperature and at
Chem., 2004, 42, 5745.
77 K, whereas the emission of the Re(I) complexes appears to be
of 3MLCT character at room temperature and of 3LC character at
77 K.
25 P. S. Braterman, J. I. Song and R. D. Peacock, Inorg. Chem., 1992, 31,
555.
26 C. Hamann, A. Von Zelewsky, A. Neels and H. Stoeckli-Evans, Dalton
Trans., 2004, 402.
27 K. S. Schanze, D. B. Macqueen, T. A. Perkins and L. A. Cabana, Coord.
Chem. Rev., 1993, 122, 63.
28 D. A. Bardwell, F. Barigelletti, R. L. Cleary, L. Flamigni, M. Guardigli,
J. C. Jeffery and M. D. Ward, Inorg. Chem., 1995, 34, 2438.
29 L. A. Worl, R. Duesing, P. Y. Chen, L. Dellaciana and T. J. Meyer,
J. Chem. Soc., Dalton Trans., 1991, 849.
30 M. Montalti, A. Credi, L. Prodi and M. T. Gandolfi, Handbook of
Photochemistry, Taylor and Francis, Boca Raton, 2006.
31 A. Gilbert and J. Baggot, Essentials of Molecular Photochemistry,
Blackwell, London, 1991.
Acknowledgements
We thank CNR of Italy (PM P04.010) and CNRS of France. F. D.
thanks the CNRS and the Re´gion Alsace for a fellowship and
O. W. acknowledges a fellowship from the Swiss National Science
Foundation.
32 Y. Kawanishi, N. Kitamura and S. Tazuke, Inorg. Chem., 1989, 28,
References
2968.
33 G. H. Allen, R. P. White, D. P. Rillema and T. J. Meyer, J. Am. Chem.
Soc., 1984, 106, 2613.
34 D. S. Tyson and F. N. Castellano, J. Phys. Chem. A, 1999, 103, 10955.
35 N. D. McClenaghan, F. Barigelletti, B. Maubert and S. Campagna,
Chem. Commun., 2002, 602.
36 A. F. Morales, G. Accorsi, N. Armaroli, F. Barigelletti, S. J. A. Pope
and M. D. Ward, Inorg. Chem., 2002, 41, 6711.
37 X. Y. Wang, A. Del Guerzo and R. H. Schmehl, J. Photochem.
Photobiol., C, 2004, 5, 55.
1 V. Balzani and F. Scandola, Supramolecular Photochemistry, Horwood,
Chichester, 1991, and references therein.
2 V. Balzani, G. Bergamini, F. Marchioni and P. Ceroni, Coord. Chem.
Rev., 2006, 250, 1254.
3 D. G. Cuttell, S. M. Kuang, P. E. Fanwick, D. R. McMillin and R. A.
Walton, J. Am. Chem. Soc., 2002, 124, 6.
4 M. H. Wilson, L. P. Ledwaba, J. S. Field and D. R. McMillin, Dalton
Trans., 2005, 2754.
5 J. H. Alstrum-Acevedo, M. K. Brennaman and T. J. Meyer, Inorg.
Chem., 2005, 44, 6802.
6 A. Hagfeldt and M. Graetzel, Acc. Chem. Res., 2000, 33, 269.
38 N. D. McClenaghan, Y. Leydet, B. Maubert, M. T. Indelli and S.
Campagna, Coord. Chem. Rev., 2005, 249, 1336.
498 | Dalton Trans., 2008, 491–498
This journal is
The Royal Society of Chemistry 2008
©