Non-innocence of 1,4-dicyanamidobenzene bridging ligands in dinuclear ruthenium complexes.

Article abstract of DOI:10.1021/ic401316g

Four dinuclear complexes, [{Ru^II(ttpy)(bpy)}₂(μ-L) ][PF₆]₂, where bpy is 2,2′-bipyridine, ttpy is 4-(tert-butylphenyl)-2,2′:6′,2″-terpyridine, and L is 2,5-dimethyl-, 2,5-dichloro-, 2,3,4,5-tetrachloro- and unsu

Full text of DOI:10.1021/ic401316g

Article
pubs.acs.org/IC
Non-Innocence of 1,4-Dicyanamidobenzene Bridging Ligands in
Dinuclear Ruthenium Complexes.
Mohommad M. R. Choudhuri,† Wolfgang Kaim,‡ Biprajit Sarkar,§ and Robert J. Crutchley*,†
†Chemistry Department, Carleton University, Ottawa, Ontario K1S 5B6, Canada
‡Institut fur Anorganische Chemie, Universitat Stuttgart, Pfaffenwaldring 55, DE 70550 Stuttgart, Germany
̈
̈
§Inorganic Chemistry, Freie Universitat Berlin, Fabeckstrasse 34-36, DE 14195 Berlin, Germany
S
* Supporting Information
ABSTRACT: Four dinuclear complexes, [{RuII(ttpy)(bpy)}2(μ-L)][PF6]2, where bpy is 2,2′-bipyridine, ttpy is 4-(tert-
butylphenyl)-2,2′:6′,2″-terpyridine, and L is 2,5-dimethyl-, 2,5-dichloro-, 2,3,4,5-tetrachloro- and unsubstituted 1,4-
dicyanamidebenzene dianion have been synthesized and characterized. Electron paramagnetic resonance (EPR) spectroscopy
of electrogenerated [{Ru(ttpy)(bpy)}2(μ-L)]3+ ions shows largely ligand centered spin and thus the complexes’ oxidation states
are best formulated as [Ru(II), L•−, Ru(II)]3+. Visible-NIR and IR spectra of [{Ru(ttpy)(bpy)}2(μ-L)]3+,4+ ions were also
obtained by spectroelectrochemical methods. For the [{Ru(ttpy)(bpy)}2(μ-L)]3+ ions, the significant variations in the spectra
were rationalized in terms of an increased ruthenium contribution to the singly occupied molecular orbital with increasing
number of chloro substituents on the bridging ligand L.
tetramethyl-3,5-heptanedionate anion, possesses two Ru(II)
ions bridged by a radical dicyd− ligand. Why does replacing the
ammines with trpy and thd cause the destabilizing of dicyd2−
relative to Ru(II)? Clearly, the solvent donor−acceptor
interaction for [{Ru(trpy)(thd)}2(μ-dicyd)]+ is not as great
compared to dinuclear ammine complexes, and Ru(II) is
stabilized by the trpy and thd− coordination sphere. This
stabilization should be even greater if thd− is replaced by the
neutral acceptor 2,2′-bipyridine (bpy), and it is thus likely that
the previously reported [{Ru(trpy)(bpy)}2(μ-dicyd)]3+ also
possesses a mostly ligand-localized SOMO.7 The highest
occupied molecular orbital (HOMO) of the dicyd2− bridging
ligand can be significantly perturbed by electron donating or
withdrawing substituents and this creates an opportunity to
study a perturbed, mostly ligand-localized SOMO by
spectroscopic methods.
INTRODUCTION
■
Interest in non-innocent (redox-active) ligands stems not only
from examples in bioinorganic chemistry1 but also with respect
to catalysis2 and the design of electro-optic materials.3 For the
latter, there are many examples of mixed-valence complexes
whose properties (electrochemistry, dipole moment, NIR
absorption) are determined by the degree of noninnocence of
a bridging ligand.4 Purposeful design of donor−acceptor
systems and materials require that the expression and
recognition of these properties be understood. Research into
the complexes of the redox active 1,4-dicyanamidebenzene
dianion (dicyd2−) bridging ligand is illustrative.
[{Ru(NH3)5}2(μ-dicyd)]3+,4+,5+ ions have been shown5 by
electron paramagnetic resonance (EPR) and 1H NMR evidence
to involve a singly occupied molecular orbital (SOMO) of
mostly metal character with the bridging ligand remaining
dicyd2−. Density functional theory (DFT) calculations of
[{Ru(NH3)5}2(μ-dicyd)]3+ in the gas phase gave a mostly
ligand-based SOMO, in disagreement with experiment, and
could only be reconciled with the EPR and NMR data by
invoking a strong donor−acceptor interaction between solvent
molecules and the ammine protons and cyanamide groups. In
contrast, Bonvoisin et al.6 provided EPR and crystallographic
evidence that the complex [{Ru(trpy)(thd)}2(μ-dicyd)]+,
where trpy is 2,2′;6′,2″-terpyridine and thd− is the 2,2,6,6-
For this study, four dinuclear complexes, [{Ru(ttpy)-
(bpy)}2(μ-L)][PF6]2, where ttpy is 4-(tert-butylphenyl)-
2,2′:6′,2″-terpyridine and L is 2,5-dimethyl-, 2,5-dichloro-,
2,3,4,5-tetrachloro- and unsubstituted 1,4-dicyanamidebenzene
dianion have been synthesized and characterized. EPR
spectroscopy was used to establish unambiguously the
Received: May 26, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ic401316g | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
oxidation states of the [{Ru(ttpy)(bpy)}2(μ-L)]3+ complexes
and the complexes’ visible-NIR and IR spectra were derived by
spectroelectrochemical methods. The spectral variation that is
observed can be related to an increase in metal character of the
SOMO with electron-withdrawing substituents on the dicyd2−
bridging ligand.
4 days during which time the solution became yellow and then cloudy
with the precipitation of the product. The reaction mixture was filtered
and the light-green product washed with copious water and allowed to
air-dry. Recrystallization from acetone/water (4:1) yielded 3.0 g,
(26%), of light-green flaky crystals, mpt. 194−195 °C. Anal. Calcd for
C25H23N3: C, 82.16; H, 6.34; N, 11.50. Found: C, 82.13; H, 6.43; N,
1
11.57. H NMR (CDCl3): 1.38 (9H, s), 7.33 (2H, dd), 7.52 (2H, d),
7.85 (2H, dd), 7.86 (2H, d), 8.66 (2H, d), 8.72 (2H, d), 8.75 (2H, s),
in agreement with literature.11
EXPERIMENTAL SECTION
■
Preparation of Ru(ttpy)Cl3·H2O. A mixture of RuCl3·3H2O (2.4
g) and ttpy (3.2 g) in absolute ethanol (600 mL) was refluxed 14 h.
The dark-brown reaction mixture was cooled and then filtered. The
dark-brown product was washed with water, followed by absolute
ethanol and then diethylether and vacuum-dried. Yield: 4.2 g, 78%.
Anal. Calcd for C25H25N3OCl3Ru: C, 50.81; H, 4.26; N, 7.11. Found:
C, 50.53; H, 3.83; N, 6.94.
Preparation of [Ru(ttpy)(bpy)Cl][PF6]. Ru(ttpy)Cl3·H2O (2 g)
and 2,2′-bipyridine (0.52 g) were added to 400 mL of water/ethanol
(2:1) and refluxed under argon for 14 h. The reaction solution was
allowed to cool slightly before adding LiCl (2 g) and refluxing for a
further 1 h. The reaction solution was filtered and to the hot filtrate
was added 4 g of NH4PF6, precipitating the desired product which was
immediately filtered off, washed with water, and allowed to air-dry.
Recrystallization by diffusing ether into an acetone solution of the
complex yielded dark brown crystals. Yield 2.0 g (74%). Anal. Calcd
for C35H31N5ClPF6Ru: C, 52.34; H, 3.89; N, 8.72. Found: C, 52.16; H,
3.78; N, 8.71. 1H NMR (DMSO-d6): 10.12 (d, 2H), 9.16 (s, 4H), 8.92
(d, 6H), 8.65 (d, 2H), 8.37 (t, 2H), 8.24 (d, 4H), 8.10−8.00 (m, 6H),
7.79 (t, 2H), 7.72 (d, 4H), 7.64 (d, 4H), 7.44−7.37 (m, 6H), 7.12 (t,
2H), 1.42 (s, 18H). Filtering off the complex from the hot solution
keeps most impurities in solution and yields an almost pure reagent
complex.
Reagents. The organic solvents used for spectroscopy and
electrochemistry were distilled under reduced pressure and stored
under argon. All solvents were dried with an appropriate reagent.
Acetonitrile (HPLC grade, Aldrich) was distilled in the presence of
phosphorus pentoxide. N,N′-dimethylformamide (DMF) and propy-
lene carbonate were dried overnight and distilled in the presence of
aluminum oxide (neutral grade) which had been previously activated
by heating to 300 °C for 3 h. The mononuclear complex,
[Ru(trpy)(bpy)(2,4-Cl2pcyd)][PF6], where trpy is 2,2′;6′,2″-terpyr-
idine, bpy is 2,2′-bipyridine, and 2,4-Cl2pcyd− is 2,4-dichlorophenyl-
cyanamide has been previously prepared.8 TlNO3 (BDH) (Caution!:
highly toxic) and 2,2′-bipyridine (Aldrich) were used as received. 1,4-
Dicyanamidebenzene (dicydH2) and its derivatives 2,5-dimethyl-
(Me2dicydH2), 2,5-dichloro- (Cl2dicydH2), and 2,3,5,6-tetrachloro-
1,4-dicyanamidobenzene (Cl4dicydH2) were prepared by literature
methods.9 Both Cl2dicydH2 and Cl4dicydH2 possessed a significant
impurity of the guanidine dimer as shown by a strong IR υ(CN)
band at approximately 1680 cm−1. This dimer impurity reverts to a
monomer in basic solutions and does not affect the isolation of the
thallium salt of the ligand as discussed previously.10 Thallium salts
(Caution: highly toxic) of dicyd2− and its substituted derivatives were
prepared by using the general method described below for
Tl2[Me2dicyd]) and were used without further purification. These
salts are only slightly soluble in strong donor solvents, but this is
sufficient for the reactions described below.
Preparation of Thallium 2,5-Dimethyl-1,4-dicyanamideben-
zene Dianion, Tl2[Me2dicyd]. Me2dicydH2 (0.4 g) was dissolved in
100 mL of gently boiling 3:1 acetone:water and then filtered.
Approximately 1.5 mL of triethylammine was added to the filtrate,
followed quickly by a warm solution of 1.3 g of TlNO3 in 25 mL of
water. The slightly blue solution was gently boiled for 5 min forming a
yellow precipitate which was filtered and washed with acetone and
water and finally acetone and allowed to dry. Yield: 0.7 g (54%). Anal.
Calcd for C10H8N4Tl2: C, 20.26; H, 1.36; N, 9.45. Found: C, 20.17; H,
1.16; N, 9.43. 1H NMR (DMSO-d6): the poor solubility and oxidation
to the radical by trace oxygen made the Me2dicyd2− chemical shift
assignment unreliable.
Complex Synthesis. The complexes of Cl2dicyd2− and Cl4dicyd2−
could be purified by chromatography by using alumina (Grade III) and
eluting with DMF. However, under the same conditions, the
complexes of dicyd2− and Me2dicyd2− proved to be air sensitive. For
the latter complexes, fractional crystallization under an argon
atmosphere yielded pure product in good yield.
Preparation of [{Ru(ttpy)(bpy)}2(μ-Me2dicyd)][PF6]2 1. [Ru-
(ttpy)(bpy)Cl][PF6] (1.0 g) and Tl2[Me2dicyd] (0.3 g) were placed in
175 mL of DMF and refluxed under argon for 40 h during which time
the reaction solution changed color from violet-purple to a brownish-
purple and TlCl precipitated. The reaction mixture was gravity filtered
through Celite, and the filtrate’s volume reduced to 5−10 mL. The
crude complex was precipitated by the addition of ether (300 mL),
filtered, and allowed to air-dry. Crude yield: 0.8 g. Separation of the
dinuclear complex from monomer impurity was achieved under an
argon atmosphere by dissolving the complex in 20 mL of CH3CN,
filtering, and to the filtrate slowly adding approximately 40 mL of
toluene. The precipitated dinuclear complex was filtered, washed with
toluene and then ether, and vacuum-dried. Yield: 0.35 g, 33% Anal.
Calcd for C80H70N14P2F12Ru2: C, 55.88; H, 4.10; N, 11.40. Found: C,
55.39; H, 4.36; N, 11.75. IR (KBr), ν(NCN) = 2104 cm−1.
Preparation of [{Ru(ttpy)(bpy)}2(μ-dicyd)][PF6]2·1.6(toluene)·
4H2O 2. This was prepared as for 1 but precipitated with toluene and
water of crystallization as shown by 1H NMR. Yield: 0.30 g, 31%. Anal.
Calcd for C89.2H86.8N14O4P2F12Ru2: C, 56.06; H, 4.58; N, 10.26.
Found: C, 56.11; H, 4.19; N, 10.50. IR (KBr), ν(NCN) = 2156 cm−1.
Preparation of [{Ru(ttpy)(bpy)}2(μ-Cl2dicyd)][PF6]2 3. [Ru-
(ttpy)(bpy)Cl][PF6] (1 g) and Tl2[Cl2dicyd] (0.4 g) were placed in
175 mL of DMF and refluxed under argon for 60 h during which time
the reaction solution changed color from violet-purple to a brownish-
purple and TlCl precipitated. The reaction mixture was gravity filtered
through Celite and the filtrate’s volume reduced to 50 mL. The crude
complex was precipitated by the addition of ether (300 mL), filtered,
and allowed to air-dry. Crude yield: 0.7 g. Chromatography using
alumina (Type III) and elution with CH3CN:toluene 1:1 yielded an
orange band, followed by a purple reagent band and finally a brown
band. The latter was collected and the acetonitrile evaporated off
yielding the desired product suspended in toluene. Yield 0.23 g (21%).
Preparation of Thallium 1,4-Dicyanamidebenzene Dianion,
Tl2[dicyd]. Prepared in the same manner as Tl2[Me2dicyd]. Yield:
77%). Anal. Calcd for C8H41N4Tl2: C, 17.01; H, 0.71; N, 9.92. Found:
C, 17.18; H, 0.58; N, 9.75. H NMR (DMSO-d6): 6.32 (4H, s) ppm.
Preparation of Thallium 2,5-Dichloro-1,4-dicyanamideben-
zene Dianion, Tl2[Cl2dicyd]. Prepared in the same manner as
Tl2[Me2dicyd]. Yield: 60%. Anal. Calcd for C8H4N4Cl2Tl2: C, 15.16;
H, 0.32; N, 8.84. Found: C, 15.19; H, 0.33; N, 8.75. 1H NMR
(DMSO-d6): 6.73 (2H, s) ppm.
Preparation of Thallium 2,3,5,6-Tetrachloro-1,4-dicyanami-
debenzene Dianion, Tl2[Cl4dicyd]·0.25[HN(CH2CH3)3][NO3]. Pre-
pared in the same manner as Tl2[Me2dicyd]. Yield: 48%. Anal. Calcd
for C9.5H4N4.5Cl4 O0.75Tl2: C, 15.34; H, 0.54; N, 8.47. Found: C,
15.43; H, 0.3; N, 8.66. This procedure was repeated twice, and the
elemental analyses gave the same triethylammonium impurity whose
1
presence was confirmed by H NMR.
Preparation of 4-(tert-Butylphenyl)-2,2′:6′,2″-terpyridine
(ttpy). This compound has been prepared previously11 but the
method described below eliminates the need to isolate the penta-1,5-
dione derivative. 4-tert-Butylbenzaldehyde (5 g) and 2-acetylpyridine
(13 g) were added to 800 mL of methanol in a 2 L Erlenmeyer flask.
To the stirred solution were added 330 mL of concentrated aqueous
ammonia and 65 mL of 3.75 M NaOH. The Erlenmeyer flask was
covered with a watch glass, and the reaction solution allowed to stir for
B
dx.doi.org/10.1021/ic401316g | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Anal. Calcd for C78H64N14Cl2P2F12Ru2: C, 53.22; H, 3.66; N, 11.14.
Found: C, 53.36; H, 3.43; N, 11.46. IR (KBr), ν(NCN) = 2141 cm−1.
Preparation of [{Ru(ttpy)(bpy)}2(μ-Cl4dicyd)][PF6]2·0.5tol-
uene·CH3CN 4. This was prepared in the same manner as 3.
Recrystallized from CH3CN/toluene. Yield 15%. Anal. Calcd for
C83.5H69N15Cl4P2F12Ru2: C, 52.33; H, 3.63; N, 10.96. Found: C, 52.08;
H, 3.54; N, 11.10. IR (KBr), ν(NCN) = 2156 cm−1.
analyses were consistent with their formulation. Assignments
of the ttpy chemical shifts were determined using COESY
analysis (Supporting Information, Figures S5−S8 and Table
S1).
Figure 1 shows the cyclic voltammograms of the [{Ru-
(ttpy)(bpy)}2(μ-L)](PF6)2 compounds, and Table 1 summa-
Physical Measurements. Electrochemistry. Cyclic voltammetry
studies were performed using a Metrohm Autolab potentiostat/
galvanostat PGSTAT30. DMF (Sigma-Aldrich, ChromosolvPlus,
99.9%, HPLC grade) was used for the studies. A three electrode
arrangement consisting of a platinum disk electrode working electrode
(BAS 1.6 mm diameter), a platinum wire auxiliary electrode, and a
silver-wire quasi-reference electrode was used. The electrochemical cell
consisted of a double jacketed glass container with an internal volume
of 15 mL. Ferrocene (E°= 0.665 V versus NHE)12 was used as internal
reference, and TBAH (0.1M) was the supporting electrolyte. Argon
gas was bubbled into the solutions for 10−15 min to degas them
before scans were recorded. An optically transparent thin-layer
electrochemical (OTTLE) cell was used to conduct spectroelec-
trochemical studies.7,13−15 ITO (indium−tin oxide) coated glass
served as working and counter electrodes and AgCl/Ag was used as
reference electrode. The solvents and supporting electrolyte were the
same as those used for cyclic voltammetry. Potential was varied
between 0.00 and 1.10 V during the experiment, and spectra were
collected on a UV−vis-NIR Cary 5 spectrophotometer.
NMR Studies. 1H NMR spectra were obtained at ambient
temperature by using Bruker AMX-400 NMR or Bruker 300 Ultra
Shield spectrometers and referenced to TMS (0.00 ppm)
EPR Spectroscopy. EPR spectra of the complexes were recorded in
DMF at room temperature to 4 K by using a Bruker system EMX, and
a continuous flow cryostat ESR 900 of Oxford Instruments was used
for this purpose.
Semi-Empirical Calculations. PM3 calculations of orbital energies
and display of wave functions were obtained with Spartan V.10
software.
RESULTS AND DISCUSSSION
The complexes 1−4 were prepared by metathesis in refluxing
DMF as shown in Scheme 1. The precipitation of TlCl removes
■
Figure 1. Cyclic voltammetry of [{Ru(ttpy)(bpy)}2(μ-L)]2+ in DMF,
0.1 M TBAH; L = Me2dicyd2− 1; L = dicyd2− 2; L = Cl2dicyd2− 3; L =
Cl4dicyd2− 4.
Scheme 1
a
Table 1. Cyclic Voltammetry Data of [{Ru(ttpy)(bpy)}2(μ-
L)]2+ Complexes, [Ru(ttp)(bpy)Cl]+ and [AsPh4]2dicyd
compound
L(−1/−2)
L(0/−1)
Ru(III/II)
c
d
1
2
3
4
0.23 (71)
0.72 (71)
0.80 (73)
c
d
0.33 (83)
c
d
d
0.54 (66)
0.96 (106)
1.03 (110)
c
0.66 (66)
[Ru(ttpy)(bpy)Cl]+
1.02 (85)
b
b
[AsPh4]2dicyd
−0.20
0.47
a
Data in volts vs NHE, 0.1 M TBAH in DMF, scan rate 100 mV/s,
b
c
anodic and cathodic peak. In acetonitrile, reference 9. Assignment
supported by EPR studies. Assignment supported by semiempirical
chloride from the reaction solution and ensures a reasonable
yield of complex. Unfortunately, the purification of 1 and 2
proved problematic because of their tendency to oxidize during
chromatography on alumina. This was solved by the fortuitous
discovery that the dinuclear complexes were far less soluble in
acetonitrile/toluene solvent mixtures as compared to mono-
nuclear complexes and could therefore be preferentially
precipitated by the addition of toluene to an acetonitrile
d
calculations.
rizes the electrochemical data. Two oxidation couples are
observed ranging from −0.3 to 1.4 V versus NHE in DMF. For
each couple, the anodic and cathodic waves are of equal current
but with peak to peak separation greater than 58 mV. The latter
occurs when heterogeneous electron transfer is slow and
identifies these one-electron oxidations as quasi-reversible
1
solution of the crude product. The H NMR spectra of 1−4
(Supporting Information, Figures S1−S4) and elemental
C
dx.doi.org/10.1021/ic401316g | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
processes. These couples shift to more positive potentials as
electron-withdrawing substituents are added to the dicyd2−
ligand.
EPR spectroscopy of the dinuclear complexes establishes the
first oxidation couple, [{Ru(ttpy)(bpy)}2(μ-L)]3+/2+, as ligand
centered L(−1/−2) while the second oxidation couple is
formally assigned to a L(0/−1) couple based on semiempirical
calculations as discussed below.
K (frozen solution) results in the loss of hyperfine information
but reveals a small g anisotropy (g1, g2, g3 = 2.025, 2.011, and
2.00, respectively). The calculated gav = 2.012 close to the free
electron value of 2.0023 and the Δg = g1 − g3 = 0.025 indicate
little mixing of ruthenium ions with their high spin−orbit
coupling contributions in the ligand-localized SOMO.18 Similar
results have been reported for the complex [{Ru(tpy)-
(thd)}2(μ-dicyd)]+, where thd− is 2,2,6,6-tetramethyl-3,5-
heptanedione.5
Figure 2 shows the broad low-temperature EPR spectrum of
the mononuclear complex [Ru(trpy)(bpy)(2,4-Cl2pcyd]2+ that
Chloro substitution on dicydn− should induce greater mixing
of metal and ligand orbitals upon oxidation of the complex.
This is illustrated by the EPR spectrum of 4+ at 110 K (Figure
4; no room temperature signal) which exhibits greater g
Figure 2. EPR spectrum of oxidized mononuclear [Ru(trpy)(bpy)-
(2,4-Cl2pcyd)][PF6] in DMF at 110 K.
is typical of a low spin d5 Ru(III) ion.16 The g components are
g1 = 2.34, g2 = 2.10, and g3 = 1.92, and the average g value gav =
2.13. Oxidation of the dinuclear complex [{Ru(ttpy)(bpy)}2(μ-
dicyd)]2+ (2) in DMF solution gave the EPR spectra shown in
Figure 3. In contrast to [Ru(trpy)(bpy)(2,4-Cl2pcyd)]2+, 2+
Figure 4. EPR spectrum of [{Ru(ttpy)(bpy)}2(μ-Cl4dicyd)]3+ 4+ in
DMF solution at 110 K. A weak signal from an organic impurity is
observed at g = 2.003.
anisotropy at g1, g2, g3 = 2.06, 2.03, and about 2.02, respectively,
and a larger gav = 2.037 and Δg = 0.04. These values are still
consistent with a dominant contribution to the SOMO by the
bridging ligand and so the formal assignment for all
[{Ru(ttpy)(bpy)}2(μ-L)]3+/2+ couples is suggested to be a
L•−/L2− reduction couple.
EPR spectroscopy is the method of choice for evaluating the
noninnocence of metal complexes containing redox-active
ligands and indeed, in this study, EPR has shown that
[Ru(trpy)(bpy)(2,4-Cl2pcyd)]2+ is a Ru(III) species with
metal centered spin while 1+−4+ are radical complexes with
ligand centered spin. However, cost and availability of EPR
instrumentation to researchers are significant disadvantages,
and some radical complexes can be EPR silent even at very low
1
temperatures. In these cases, paramagnetic H NMR spectros-
copy may provide evidence of SOMO parentage. With this in
mind, the oxidized complexes 1+−4+ were prepared by reacting
−
1−4 with NO+ BF4 in acetonitrile solution, checking for
completeness by comparison to the spectra of 1+-4+ as obtained
by spectroelectrochemical methods (vide infra). Evaporation of
acetonitrile left the crude oxidized complexes which were used
without further purification to investigate their paramagnetic
1H NMR spectroscopy. The spectra of 1+−4+ showed broad,
poorly resolved signals with aromatic chemical shifts belonging
to ttpy between 6 and 15 ppm that could not be assigned
unambiguously. However, the chemical shift of the t-butyl
group for a given complex could be readily identified because it
was largely unaffected by oxidation, being shifted downfield by
about 0.02 ppm relative to its diamagnetic chemical shift range
from 1.39 to 1.40 ppm (in the complexes of this study). The
situation is quite different for the complex [Ru(ttpy)(bpy)Cl]2+
which was prepared in the same manner as that for 1+−4+. For
this complex, the chemical shift of the t-butyl group was shifted
upfield to 1.09 ppm from its diamagnetic chemical shift of 1.42
Figure 3. EPR spectra of [{Ru(ttpy)(bpy)}2(μ-dicyd)]3+ 2+ in DMF
solution at (a) room temperature and (b) at 110 K (a weak signal from
an organic impurity is observed at g = 2.003).
gave a room temperature isotropic spectrum (Figure 3a) at g =
2.009 which is relatively narrow and exhibits several hyperfine
lines split by approximately 1.7 G. Although the partial
resolution did not allow us to fully analyze the hyperfine
coupling pattern, the splitting is similar to the EPR spectrum of
the potassium salt of the dicyd anion radical with g = 2.0023
and hyperfine splitting varying between 1 and 5 G.17 These
results strongly support an oxidized (radical) bridging ligand in
2+. As shown in Figure 3b, a lowering of the temperature to 110
D
dx.doi.org/10.1021/ic401316g | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
ppm. Assuming that pseudocontact shift contribution to the
isotropic shift is small,19 the contact shift must result from spin
density being placed on the t-butyl hydrogens through a
combination of σ and π spin polarization of ttpy from spin
density originating on Ru(III).
Electronic and IR spectroelectrochemistry were performed
on the [{Ru(ttpy)(bpy)}2(μ-L)]2+ complexes to gain a greater
understanding of these complexes with non-innocent ligands
and, more specifically, to see if the introduction of more metal
character to the SOMO of [{Ru(ttpy)(bpy)}2(μ-L)]3+ results
in significantly perturbed electronic and IR spectra. The
spectroelectrochemical studies of the complexes of this study
showed reversible processes with at least 95% recovery of the
initial spectrum except where otherwise noted. Electronic
absorption data has been compiled in Table 2.
Table 2. Quantitative Electronic Absorption Data of
[{Ru(ttpy)(bpy)}2(μ-L)]2+/3+/4+ Complexes in DMF, 0.1 M
TBAH
complex
wavelength in nm (molar absorption coefficient M−1 cm−1
)
1
364 (sh,29300), 498 (21000)
1+
12+
2
399 (sh,23900), 491 (22300), 1023 (13300)
463 (24200), 918(17100)
364 (sh,30600), 493(23400)
2+
22+
3
3+
32+
4
383 (sh,27800), 487 (25000), 1083(17700)
457 (26700), 959 (27500)
Figure 5. Visible/NIR spectroelectrochemical oxidation of [{Ru-
(ttpy)(bpy)}2(μ-dicyd)]2+ 2 to form (a) [{Ru(ttpy)(bpy)}2(μ-
dicyd)]3+ 2+ and (b) [{Ru(ttpy)(bpy)}2(μ-dicyd)]4+ 22+; in DMF,
0.1 M TBAH.
364 (sh,29800), 482 (22800)
358 (sh,30300), 426 (30000), 482(27000), 1265(18800)
482(27000), 855 (9900)
364 (sh, 38100), 485 (22200)
4+
42+
430 (37000), 489 (sh,26300), 1376 (25200)
489 (26000), 869 (8500), 1376 (6300)
Figure 6 shows the visible/NIR spectroelectrochemical
oxidation of 3 to the 3+ and 32+ complexes. For 4, the first
and second oxidations gave similar spectral changes to those
shown in Figure 5 except that the second oxidation was
incomplete because of poor reversibility. The data have been
placed in Supporting Information, Figure S12. Similar to Figure
5a, in Figure 6a, an intense NIR absorption band appears at λmax
= 1265 nm with the formation of 3+ which we again assign to a
ligand centered transition. However, the second oxidation
(Figure 6b) results in a relatively large shift of the band to
shorter wavelengths with λmax = 855 nm, a significant decrease
in the intensity of this band, and the appearance of a low energy
absorption (1200−1800 nm). This is very different from the
spectral changes associated with the formation of 22+ in Figure
5b and may result from a significant contribution from
[Ru(III), L•−, Ru(II)]4+ to 32+.
Figure 5 shows the spectroelectrochemical oxidation of 2 to
the singly oxidized 2+ and doubly oxidized 22+ complexes.
These results are similar to those of complex 1, which have
been placed in Supporting Information, Figure S9. In Figure 5a,
oxidation of the bridging ligand results in the appearance of a
strong NIR absorption with λmax = 1083 nm. Because EPR
spectroscopy supports a [Ru(II), L•−, Ru(II)] oxidation state
description, the band could be assigned to Ru(II) dπ to dicyd
pπ, metal-to-ligand charge transfer (MLCT) transition or a
dicyd ligand centered transition. DFT calculations were
attempted to settle this issue; however, the calculations would
not converge.20 Semiempirical calculations show a HOMO that
is mostly ligand centered in agreement with EPR studies and
possible low energy transitions from filled pπ MOs of the
dicyd•− ligand thereby supporting a mostly ligand centered
transition (Supporting Information, Figure S10). In Figure 5b,
further oxidation increases the intensity and slightly shifts this
band to higher energies with λmax = 959 nm. Semiempirical
calculations of 22+ assuming a singlet ground state show mostly
ligand centered HOMO and LUMO (Supporting Information,
Figure S11) which arises from the oxidation of the radical
bridging ligand to dicyd0. N,N′-Dicyanoquinonediimine (dicyd0
or DCNQI in the literature) and its substituted derivatives have
been isolated as diamagnetic yellow compounds with no NIR
absorption.21 The band with λmax = 959 nm is therefore not
easily explained by a ligand centered transition, and it is
suggested that, if 22+ can be formally represented by [Ru(II),
L0, Ru(II)]4+, this band is a mostly Ru(II) to L0 MLCT
transition.
Further evidence for the participation of ruthenium in the 3+
complex’s SOMO can be found in IR spectroelectrochemical
studies.
Figure 7 shows spectral changes that occur in the cyanamide
stretching region with the formation of [Ru(trpy)(bpy)(2,4-
Cl2pcyd)]2+ from the monocation in DMF. Upon oxidation to
Ru(III) there occurs a loss of the band at 2170 cm−1 and the
appearance of two weak bands centered at 2040 and 2270
cm−1. The band at 2270 cm−1 is associated with decomposition
of the complex, and indeed only 85% reversibility to the
original dication spectrum was observed. The weak band at
2040 cm−1 is expected from IR studies of [Ru(NH3)5(pcyd)]2+
and [Ru(NH3)3(bpy)(pcyd)]2+ where pcyd is a phenyl-
cyanamido ligand.22 These studies showed that oxidation of
Ru(II) to Ru(III) shifts ν(NCN) to lower frequencies because
of a shift of cyanamide resonance forms as shown below
E
dx.doi.org/10.1021/ic401316g | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Me2dicyd2− results in a decrease in intensity and a shift of
ν(NCN) to higher frequencies.
Figure 8 shows the IR spectral changes that occur upon
oxidation of oxidation of 2 to the 2+ and 22+ cations in DMF
Figure 6. Visible NIR spectroelectrochemical oxidation of [{Ru(ttpy)-
(bpy)}2(μ-Cl2dicyd)]2+ 3 to form a) [{Ru(ttpy)(bpy)}2(μ-
Cl2dicyd)]3+ 3+ and b) [{Ru(ttpy)(bpy)}2(μ-Cl2dicyd)]4+ 32+, in
DMF, 0.1 M TBAH.
Figure 8. IR spectroelectrochemical oxidation of [{Ru(ttpy)-
(bpy)}2(μ-dicyd)]2+ 2 to form (a) [{Ru(ttpy)(bpy)}2(μ-dicyd)]3+ 2+
and (b) [{Ru(ttpy)(bpy)}2(μ-dicyd)]4+ 22+, in DMF, 0.1 M TBAH.
solution. In Figure 8a and similar to Me2dicyd (Supporting
Information, Figure S13a), the initial spectrum shows at least
two overlapping ν(NCN) bands which are probably due to a
mixture of syn- and anti-conformations in solution. Crystal
structures of dinuclear ruthenium complexes bridged by dicyd
ligands have shown both syn and anti-conformations, but the
most common structure is that of a planar anti-conforma-
tion.5,8,13,23 Upon oxidation to 2+, the intensity of the ν(NCN)
bands decreases with only a slight shift to higher frequencies.
Based on the EPR data, the single cyanamide stretching
frequency must be a consequence of a SOMO possessing
equivalent contributions from the cyanamide groups of the
bridging ligand. Further oxidation to 22+ (Figure 8b), slightly
shifts the ν(NCN) band to lower energy and increases its
intensity. The IR spectroelectrochemical oxidations of 1 and 3
are very similar to these results and have been placed in
Supporting Information, Figures S14 and S15. Oxidation of 4
shows significantly different behavior.
Figure 7. IR spectral changes associated with the oxidation of
[Ru(trpy)(bpy)(2,4-Cl2pcyd)](PF6) in DMF, 0.1 M TBAH. Only 85%
recovery of initial spectrum.
IR spectroelectrochemistry showing the oxidation of the free
dianion ligand Me2dicyd2− in DMF has been placed in
Supporting Information, Figure S11. A DMF solution of
Me2dicyd2− shows a strong ν(NCN) band at 2070 cm−1 and a
far weaker band at 2110 cm−1. Cyanamide groups can adopt
syn- and anti-conformations and so a multiplicity of ν(NCN)
bands is not unexpected. Upon oxidation to the anion radical
Supporting Information, Figure S13a, the intensity of the band
at 2070 cm−1 drops and new bands appear at 2120 and 2090
cm−1. Oxidation to Me2dicyd0 (Supporting Information, Figure
S13b) sees a further decrease in intensity and new bands
appearing at 2170 and 2220 cm−1. Overall, oxidation of
In Figure 9, the oxidation of 4 to 4+ in DMF causes a
decrease in intensity of the ν(NCN) band at 2150 cm−1 and the
growth of a new band at 2100 cm−1. This shift to lower
frequencies is very different from that of 2+ in Figure 8a, and it
is suggested to be due to a greater contribution of ruthenium
character to the SOMO. A single band ν(NCN) at 2100 cm−1
in Figure 8 is consistent with a delocalized state, and indeed
2100 cm−1 is the average of Ru(II) and Ru(III) cyanamide
F
dx.doi.org/10.1021/ic401316g | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(4) (a) Low, P. J. Coord. Chem. Rev. 2013, 257, 1507. (b) Das, A. K.;
Sarkar, B.; Fiedler, J.; Zalis, S.; Hartenbach, I.; Strobel, S.; Lahiri, G. K.;
Kaim, W. J. Am. Chem. Soc. 2009, 131, 8895. (c) Chakraborty, S.; Laye,
R. H.; Paul, R. L.; Gonnade, R. G.; Puranik, V. G.; Ward, M. D.; Lahiri,
G. K. J. Chem. Soc., Dalton Trans. 2002, 1172. (d) Ward, M. D.;
McCleverty, J. A. J. Chem. Soc., Dalton Trans. 2002, 275.
(5) Naklicki, M. L.; Gorelsky, S. I.; Kaim, W.; Sarkar, B.; Crutchley,
R. J. Inorg. Chem. 2012, 51, 1400.
(6) Fabre, M.; Jaud, J.; Hliwa, M.; Launay, J.-P.; Bonvoisin, J. Inorg.
Chem. 2006, 45, 9332.
(7) Rezvani, A. R.; Evans, C. E. B.; Crutchley, R. J. Inorg. Chem. 1995,
34, 4600.
(8) Mosher, P. J.; Yap, G. P. A.; Crutchley, R. J. Inorg. Chem. 2001,
40, 550.
(9) Aquino, M. A. S.; Lee, F. L.; Gabe, E. J.; Bensimon, C.; Greedan,
J. E.; Crutchley, R. J. J. Am. Chem. Soc. 1992, 114, 5130.
(10) Crutchley, R. J. Coord. Chem. Rev. 2001, 219−221, 125.
(11) Constable, E. C.; Harverson, P.; Smlth, D. F.; Whall, L. A.
Tetrahedron 1994, 50, 7799.
Figure 9. IR spectroelectrochemical oxidation of [{Ru(ttpy)-
(bpy)}2(μ-Cl4dicyd)]2+ 4 to form [{Ru(ttpy)(bpy)}2(μ-Cl4dicyd)]3+
4+ in DMF, 0.1 M TBAH.
(12) Gennett, T.; Milner, D. F.; Weaver, M. J. J. Phys. Chem. 1985,
89, 2787.
(13) Evans, C. E. B.; Yap, G. P. A.; Crutchley, R. J. Inorg. Chem. 1998,
37, 6161.
(14) (a) Al-Noaimi, M.; Yap, G. P. A.; Crutchley, R. J. Inorg. Chem.
2004, 43, 1770. (b) Evans, C. E. B.; Ph.D. Thesis, Carleton University,
Ottawa, Canada, 1997.
stretching frequencies from Figure 7. However, as already
discussed, the EPR data do not support a ruthenium-centered
SOMO and indicate the caution which must be invoked before
making state assignments based on IR data alone.
(15) Krejcik, M.; Danek, M.; Hartl, F. J. Electroanal. Chem. 1991, 317,
179.
(16) (a) Patra, S.; Sarkar, B.; Mobin, S. M.; Kaim, W.; Lahiri, G. K.
Inorg. Chem. 2003, 42, 6469. (b) Poppe, J.; Moscherosch, M.; Kaim,
W. Inorg. Chem. 1993, 32, 2640.
CONCLUSIONS
■
EPR spectra of the [{Ru(ttpy)(bpy)}2(μ-L)]3+ ions show
predominately organic radical features, confirming ligand
centered spins and thus the complexes’ oxidation states are
best formulated as [Ru(II), L•−, Ru(II)]. Visible/NIR and IR
spectra of [{Ru(ttpy)(bpy)}2(μ-L)]3+,4+ ions were obtained by
spectroelectrochemical methods. For the [{Ru(ttpy)(bpy)}2(μ-
L)]3+ complexes, spectral variations were rationalized in terms
of the nature of the substituted bridging ligand L and variable
ruthenium d-orbital contributions to the mostly ligand-centered
SOMO.
(17) Gerson, F.; Gescheidt, G.; Mocke1, R.; Aumuller, A.; Erk, P.;
̈
̈
Hunig, S. Helv. Chim. Acta 1988, 71, 1665.
̈
(18) Kasack, V.; Kaim, W.; Binder, J.; Jordanov, J.; Roth, E. Inorg.
Chem. 1995, 34, 1924.
(19) Distance from Ru(III) to the t-butyl group is about 10.5 Å and
pseudocontact shift is proportional to 1/r3.
(20) Serge Gorelsky, personal communication, April 4, 2013. Time-
dependent DFT calculations at the B3LYP/DZVP level of theory
failed to converge.
ASSOCIATED CONTENT
* Supporting Information
(21) (a) Aumuller, A.; Hunig, S. Justus Liebigs Ann. Chem. 1986, 142.
̈
̈
■
(b) Hunig, H.; Bau, R.; Kemmer, M.; Meixner, H.; Metzenthin, T.;
̈
S
1H NMR spectra, COESY, and Table of chemical shifts of all
the complexes (eight figures and a table), Vis-NIR and IR
spectroelectrochemical spectra (four figures) and two figures
showing the orbital energies and frontier molecular orbitals of
2+ and 22+. This material is available free of charge via the
Internet at http://pubs.acs.org.
Peter, K.; Sinzger, K.; Gulbis, J. Eur. J. Org. Chem. 1998, 335.
(22) DeRosa, M. C.; White, C. A.; Evans, C. E. B.; Crutchley, R. J. J.
Am. Chem. Soc. 2001, 123, 1396.
(23) Rezvani, A. R.; Bensimon, C.; Cromp, B.; Reber, C.; Greedan, J.
E.; Kondratiev, V. V.; Crutchley, R. J. Inorg. Chem. 1997, 36, 3322.
AUTHOR INFORMATION
Corresponding Author
*E-mail: robert_crutchley@carleton.ca.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work is supported by Carleton University and the Natural
Sciences and Engineering Research Council of Canada.
■
REFERENCES
■
(1) (a) Kaim, W.; Schwederski, B. Coord. Chem. Rev. 2010, 254, 1580.
(b) Lyons, C. T.; Stack, T. D. P. Coord. Chem. Rev. 2012, 257, 528.
(2) (a) Boyer, J. L.; Rochford, J.; Tsai, W.-T.; Muckerman, J. T.;
Fujita, E. Coord. Chem. Rev. 2010, 254, 309. (b) Lyaskovskyy, V.; de
Bruin, B. ACS Catal. 2012, 2, 270.
(3) Marder, S. R.; Kippelen, B.; Jen, A. K.-Y.; Peyghambarian, N.
Nature 1997, 388, 845.
G
dx.doi.org/10.1021/ic401316g | Inorg. Chem. XXXX, XXX, XXX−XXX