Novel metal-template assembled highly-functionalized cyanoporphyrazine ytterbium and vanadium complexes for potential photonic and optoelectronic applications

Article abstract of DOI:10.1039/b821667c

A novel facile synthetic route to the metal-template assembly of a tetrapyrrollic framework from tetracyanoethylene (TCNE) and tricyanovinylbenzene (TCNVB) structural units through reaction of TCNE and TCNVB with metal π-sandwich complexes has been develo

Full text of DOI:10.1039/b821667c

View Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
Novel metal-template assembled highly-functionalized cyanoporphyrazine
ytterbium and vanadium complexes for potential photonic and
optoelectronic applications
a
b
Larisa G. Klapshina, William. E. Douglas, Ilya S. Grigoryev,a Alexey I. Korytin,c
Sergey A. Lavrentiev,c Michael A. Lopatin,a Andrey Yu. Lukyanov,d Vladimir V. Semenov,a
Philippe Gerbierb and Valeriy M. Treushnikove
*
*
Received 3rd December 2008, Accepted 17th March 2009
First published as an Advance Article on the web 20th April 2009
DOI: 10.1039/b821667c
A novel facile synthetic route to the metal-template assembly of a tetrapyrrollic framework from
tetracyanoethylene (TCNE) and tricyanovinylbenzene (TCNVB) structural units through reaction of
TCNE and TCNVB with metal p-sandwich complexes has been developed. The reactions occur under
extremely mild conditions, the porphyrazine macrocycles being assembled in high yield from TCNE
and TCNVB building-blocks by VO2+ or Yb3+-template synthesis. The redox behaviour of the novel
complexes has been investigated. The vanadyl octacyanoporphyrazine complex was found to be a rare
example of a highly-absorbing dye combining significant electron-acceptor properties with a band gap
unusually narrow for an organic semiconductor (ca. 1.1 eV). The preparation is described of a novel
highly emissive ytterbium complex with a proposed unusual structure obtained by reaction of TCNVB
with bis(indenyl)ytterbium(II) in THF. The analytical, spectral and electrochemical investigations of
the obtained ytterbium complex indicate it to be a binuclear adduct with Yb(TCNVB)3 species in which
a single doubly-reduced TCNVB molecule bridges two Yb3+ cations. The formation of a disordered
polynuclear coordination polymer network including a macrocyclic structure and metal cations bridged
through the nitrile nitrogen atoms is proposed. The complex is readily soluble and is compatible with
a variety of polymeric matrices giving doped polymeric glasses and films which are highly luminescent
in the biologically relevant optical window covering the visible and near-infrared (NIR) range
(640–1000 nm). In addition, doped polymeric glasses and films highly emissive at the
telecommunication wavelength (1540 nm), including a composition consisting of the novel ytterbium
complex and an equimolar ratio of the novel ytterbium complex and a per se non-luminescent erbium
chelate, have been obtained. The complex is found to be an extraordinarily strong sensitizer of NIR
Er3+ emission.
semiconductors can be engineered by chemical design methods.
Introduction
Furthermore, organic semiconductors present advantages such
as low-cost synthesis and easy manufacture of thin film devices
by solution casting or printing technologies.
Organic and organometallic semiconductors are very promising
for a variety of photonic and optoelectronic applications.1,2 They
can show high absorption coefficients making them good chro-
mophores for organic-based photovoltaic elements. On the other
hand they can also possess intense luminescence over a wide
range of wavelengths. The electronic band gap of organic
Recently, we reported a novel facile synthetic route for the
metal template assembly of a tetrapyrrollic framework from
TCNE structural units by the reaction of TCNE with p-sand-
wich metal complexes.3 This approach provided the first prepa-
ration of metalloporphyrazine dyes bearing eight peripheral
strongly electron-withdrawing CN substituents (Fig. 1).
Knowledge of the redox behaviour of these novel compounds
should be of great interest from the aspects of both fundamental
studies and practical applications. It motivated us to investigate
the electrochemical properties of I and II and we report here the
results of this study. Furthermore, we have applied this new
synthetic method to other TCNE derivatives, and we also report
here the preparation of two novel metal complexes designed from
tricyanovinylbenzene (TCNVB) molecules as the structural units
including an ytterbium tetracyanotetraphenylporphyrazine
complex strongly luminescent over a wide range of wavelengths
for which we propose an unusual binuclear structure IV.
aG. A. Razuvaev Institute of Organometallic Chemistry, Russian Academy
of Sciences, Tropinin Street 49, GSP-445, Nizhny Novgorod, 603600,
Russia. E-mail: klarisa@iomc.ras.ru; Fax: +7 8132 661497; Tel: +7
8132 664370
bLabo. CMOS, CNRS UMR 5253, Institut Gerhardt, Universite
ꢀ
Montpellier II, Place E. Bataillon, 34095, Montpellier cedex 5, France.
E-mail: douglas@univ-montp2.fr; Fax: +33 467143852; Tel: +33
467143848
cInstitute of Applied Physics, Russian Academy of Sciences, Uljanov Street
46, Nizhny Novgorod, 603600, Russia. E-mail: alex@ufp.appl.sci-nnov.ru
dInstitute for Physics of Microstructures of Russian Academy of Sciences,
GSP-105, Nizhny Novgorod, 603950, Russia. E-mail: luk@ipm.sci-nnov.ru
eNPP Reper-NN, Block 23, Office 7, Gagarin Prospect 37, Nizhny
Novgorod, 603009, Russia. E-mail: reper@reper.ru
3668 | J. Mater. Chem., 2009, 19, 3668–3676
This journal is ª The Royal Society of Chemistry 2009

View Online
mmol) in THF (2 ml). After 5 h the reaction mixture was
extracted with dry THF affording a dark-blue solution which
was separated and evaporated to dryness. The resulting residue
was taken up in dry benzene. Careful removal of benzene under
vacuum afforded II in 60% yield. IR (KBr): 2205, 2122, 1650,
1597, 1560, 1470, 1417, 1300, 1220, 1167, 1129, 1105, 1029, 765,
753, 743, 703, 682 cmꢀ1. UV-vis (THF) lmax, nm: 335, 400, 608,
1
1335. H NMR (200 MHz, CD3CN, ppm) d 7.1–7.9 (5H, m);
d 1.81, 3.65–3.69 (8H (THF), m). Elemental analysis found: C%
60.38, H% 4.79, N% 13.45, Yb% 15.86 (EDAX), calcd. for
C
109H99N21O8Yb2: C% 60.13, H% 4.58, N% 13.51, Yb% 15.90.
Preparation of doped polymer glasses
A solution of either II or a mixture of II with Er(AcFAcF)3 in
THF was added to a mixture of oligocarbonatedimethacrylate
(M ¼ 418.15) and 2,2-dimethoxy-phenylacetophenone as pho-
toinitiator under intense stirring for 12 h in a vessel made of dark
glass. After the formation of a transparent blue solution the THF
was carefully removed under vacuum. The procedure of layer-
by-layer UV polymerization was performed as described.4c
Fig. 1 Structures of complexes prepared by reaction of V(C6H6)2 with
TCNE (I) and with TCNVB (III) and by reaction of bis(indeny-
l)ytterbium(II) with TCNE (II).
Experimental
Powder diffraction measurements
The starting material (C9H7)2Yb(THF)2 was synthesized by the
method reported previously.4a V(C6H6)2 was a commercial
product (Cardinal Industries) additionally purified by sublima-
tion. 2-(Phenyl)-1,1,2-tricyanoethylene (tricyanovinylbenzene,
TCNVB), was synthesized according to the literature method4b
and purified by sublimation. THF was distilled from sodium
benzophenone ketyl; acetonitrile, benzene and toluene were
distilled over P2O5. I$C6H6$3H2O and II$4CH3C6H5$3THF
were synthesized according to previously published procedures.3
The UV/Vis spectra were recorded on a Perkin-Elmer Lambda
25 spectrophotometer. Infrared spectra were taken on an FSM
1201 FT spectrophotometer by using Nujol mulls on NaCl plates
or as KBr pellets. The steady-state fluorescence investigation was
performed with a Perkin Elmer LS 55 luminescence spectrometer.
Data were collected in reflection mode on an R-Axis Rapid
diffractometer using copper radiation,
a graphite mono-
chromator, and a 0.5 mm incident beam collimator.
Time-resolved spectroscopy
For excitation of the investigated samples we used an Evolution-
30 laser (Spectra Physics Lasers) delivering 100 ns pulses at 527
nm with a repetition rate of 1 kHz. The fluorescence spectra of
the samples were measured by using a Princeton Instruments
SpectraPro-2300 monochromator equipped with a Hamamatsu
R4220 photomultiplier tube (PMT). The signal from the PMT
was supplied by a Stanford Research SR830 lock-in amplifier.
The time dependence of the fluorescence of the investigated
samples was analyzed with an Agilent Technologies DSO6054A
digital oscilloscope. PMT signals were acquired relative to sync
pulses from the control unit of the laser. Excitation and fluo-
rescence pulses were rejected and stored independently by using
an appropriate colour optical filter set. The excitation pulse was
deconvoluted from the fluorescence data to yield undisturbed
decays which were fitted by an iterative least-squares procedure
based on the conventional three-level model.
Preparation of III$H2O
A solution of TCNVB (740 mg, 5.80 mmol) in THF (20 ml) was
added to bis(benzene)vanadium(0) (154 mg, 1.45 mmol) under
vacuum with stirring. The reaction mixture was kept for 2 h
under vacuum and exposed to moist air for 5 h. The dark-green
solution was separated, evaporated to dryness under vacuum,
and the resulting residue was taken up in dry benzene. Careful
removal of benzene under vacuum afforded III$H2O in 55%
yield. IR (THF): 2200, 1619, 1596, 1562, 1260, 1158, 1098, 1036,
1014, 804, 757, 696. UV-vis (THF) lmax, nm: 360, 645. 1H NMR
(200 MHz, CD3CN, ppm) d 7.1–7.7 (5H, m); d 1.8–2.2 (2H
(H2O), m). Elemental analysis found: C% 65.28, H% 2.98, N%
21.34, V% 6.70, calcd. for C44H22N12O2V: C% 65.92, H% 2.77,
N% 20.97, V% 6.35. MALDI+ m/z: 802 (M + H+), MALDIꢀ m/z:
800 (M ꢀ H) where M ¼ C44H22N12O2V.
Electrochemical measurements
Amperometric measurements were carried out by using a Volta-
lab PGZ100 potentiostat. A conventional three-electrode cell
was used. The working electrode was an Au disk of diameter 0.5
mm inserted into a glass tube. The auxiliary electrode was
a coiled Au wire. A classical calomel electrode saturated with
KCl (SCE) was used as the reference, being connected to the cell
through a separator assembly containing DMF and the sup-
porting electrolyte tetrabutylammonium tetrafluoroborate. The
electrolyte solution was purged with Ar gas for 30 min before
each measurement to minimize oxygen concentrations. A blanket
of Ar was maintained over the electrolyte solution during the
Preparation of IV$8THF
A solution of TCNVB (442 mg, 2.5 mmol) in dry THF (10 ml)
was added to bis(indenyl)ytterbium(II)$2THF (270 mg, 0.5
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 3668–3676 | 3669

View Online
entire electrochemical investigation. Aꢁll measurements were
Optoelectronics; 20 mW) working in the cw mode at 660 nm was
used as the pump source. The intensity-modulated pump radia-
tion (516 Hz) was focussed through a hole in the ellipsoidal
mirror 2 onto the sample 3. The sample emission was collected
and focussed to the input split of the monochromator MDR-41
(LOMO) (4). In the wavelength range 800–1700 nm an InGaAsP
p-i-n photodiode 5 was used as the receiver. The signal from the
photoreceiver was recorded by the lock-in amplifier 6 and dis-
played with the computer 7.
carried out at room temperature (ca. 20 C).
Magnetic measurements
Magnetic measurements down to 2 K were carried out in
a Quantum Design MPMS-XL-7T SQUID susceptometer. The
magnetic investigations were performed on a polycrystalline
sample. The molar susceptibility was corrected for the sample
holder and for the diamagnetic contribution of all atoms by using
Pascal’s tables.4d
Results and discussion
NIR luminescence registration
The redox properties of complexes I–IV were investigated by
cyclic voltammetry in DMF vs. SCE. The cyclic voltammograms
(CV) for both complexes exhibit closed current–voltage loops
showing that the electrogenerated intermediates are stable in the
time-frame of the experiment. A typical CV for complex I is
shown in Fig. 3.
The luminescence spectra registration setup is shown in
Fig. 2. A semiconductor laser 1 (ADL-66Z01HU, ARIMA
The cyclic voltammetric behaviour of I shows that it is able to
accept successively at least four electrons. A fairly well-defined
reversible first reduction is observed at ꢀ0.08 V. This value of the
formal first reversible reduction potential is unusually ‘‘positive’’
for a tetrapyrrollic macrocycle and its complexes. In particular,
the vanadyl tetra-tert-butyl-porphyrazine metal complex has its
first reduction potential 700 mV more negative than that for I.5
Thus, the introduction of multiple strongly electron-withdrawing
CN substituents into the porphyrazine peripheral framework has
caused a remarkable increase in the electron-acceptor properties
(Table 1). To the best of our knowledge the only previous
example of a similar alteration in redox behaviour arising from
the presence of peripheral substituents was that reported for
Zn(II) and Ni(II) tetrakis(2,6-dichlorophenyl)-b-heptani-
troporphyrines prepared by nitration of the corresponding tet-
rakis(2,6-dichlorophenyl)-porphyrines.6 The first reduction step
for I can be assigned, presumably, not to the macrocycle core but
rather to the peripheral substituents since TCNE fragments
framing the macrocycle, as has been noted,3 partially retain the
p-acceptor properties of free TCNE. The lowest unoccupied
molecular orbital (LUMO) of I calculated from the linear rela-
tionship between the ionization and redox potentials of molec-
ular organic semiconductors7a,b has an energy of 3.9 eV relative to
free space.
Fig. 2 The near-IR luminescence spectra registration setup.
In spite of its very high electron affinity, two irreversible
oxidations at 0.71 and 0.94 V are also observed for I. The
potentials given for these processes are the irreversible first and
second anodic peaks in the CV trace. However, according to
Forrest et al.,7a for materials exhibiting closed CV loops an
irreversible peak potential can correspond to within 100 mV of
the reversible oxidation potential. Thus, the energy ca. 5.0 eV
Fig. 3 Cyclic voltammogram of complex I.
Table 1 Reduction and oxidation potentials vs. saturated calomel electrode (SCE) in DMF and molecular orbital energies of I, II, III and IV
Molecular orbital energy/eV
Complex
Reduction potentials/V
Oxidation potentials/V
HOMO
LUMO
Band gap
Q-band/nm
I
II
III
IV
ꢀ0.08, ꢀ0.70, ꢀ1.28, ꢀ1.87
ꢀ0.39, ꢀ1.07, ꢀ1.26
ꢀ0.61, ꢀ1.51
0.71, 0.94
1.19
0.80, 1.38
0.86, 1.41
ꢀ5.0
ꢀ5.7
ꢀ5.1
ꢀ5.2
ꢀ3.9
ꢀ3.6
ꢀ3.1
ꢀ2.2
1.1
2.1
2.0
3.0
678, 630
672, 620
673, 610
582
ꢀ1.21, ꢀ1.75
3670 | J. Mater. Chem., 2009, 19, 3668–3676
This journal is ª The Royal Society of Chemistry 2009

View Online
UV/visible absorption spectrum for porphyrazine metal
complexes being very similar to the spectrum of I (Fig. 4c) and
exhibiting Soret- and Q bands with high molar absorbtivities at
358 nm (log 3 ¼ 4.6) and 673 nm (log 3 ¼ 4.3), respectively.
A shoulder to lower wavelength at ca. 610 nm can be attributed
to a vibronic overtone according to Gouterman’s four orbital
model. The IR spectrum of the complex exhibits pyrrole
vibration bands (nC]N 1654 cmꢀ1, 1619 cmꢀ1, nC]C 1597 cmꢀ1
,
1561 cmꢀ1), the typical skeletal stretching of the porphyrazine
macrocycle at 1490 cmꢀ1 and also nVO at 980 cmꢀ1. In the ChN
stretch region only a single band at 2204 cmꢀ1 is observed. This
band differs from nChN in free TCNVB (2233 cmꢀ1) and so can
be assigned to the CN groups belonging to the TCNB fragments
at the periphery of the macrocycle analogically to that in the
vanadyl octacyanoporphrazine complex (Table 2).3 Elemental
analysis of the purified product gives the empirical formula
III$H2O in excellent agreement with the mass-spectrometry
results. The MALDIꢀ and MALDI+ spectra displayed isotopic
clusters at m/z ¼ 800 and m/z ¼ 802, respectively, corresponding
to the ions [M ꢀ H]ꢀ and [M + H]+, where M ¼ III$H2O.
The substitution of four cyano groups by phenyls causes
a strong decrease in the complex electron affinity (Table 1).
Furthermore, like II but unlike I, the first oxidation of complex
III may be assumed to be a macrocycle-centred process since
its band gap (2 eV) is rather close to the Q-band energy of III
(610 nm).
Fig. 4 Absorption spectra of I (a), II (b) III (c) and IV (d) in THF.
estimated from the first oxidation potential (0.71 V) is also the
energy of the highest occupied molecular orbital (HOMO) of I
(Table 1). The energy difference between the HOMO and the
LUMO, ca. 1.1 eV, is substantially lower than the energy of the
long-wavelength p / p* transition (Q-band 1.83 eV) (Fig. 4a),
so the oxidation is, presumably, metal-centred rather than
occurring in the macrocycle. Interestingly, the value of the first
oxidation potential of I is consistent with that for the vanadium
(IV/V) redox couple (0.69 V vs. SCE) in vanadyl Schiff base
complexes with electron-withdrawing NO2 substituents (5-NO2
Salen).8
It should be noted that, in principle, four possible isomers of
III with a different distribution of CN and Ph groups at the
macrocycle periphery can be formed. For the sake of clarity only
isomer III with alternating peripheral Ph and CN groups is
presented in Fig. 1.
The redox properties of complex II are very different from
those of complex I exhibiting three reversible reductions and one
irreversible oxidation (Table 1). The first reduction potential is
substantially more negative for II (ꢀ0.39 V) than for I. This can
be explained by the influence of the axial ligand, TCNEꢂꢀ, which
is able to donate to the porphyrazine p-orbital thus decreasing its
electron affinity. The oxidation potential of II is also much
higher than that for I since YbIII, unlike VOII, is inactive in the
anodic processes so the oxidation of II, presumably, is macro-
cycle-centred. This is confirmed by the UV/visible spectrum of II.
The energy difference between the first oxidation and reduction
processes (2.1 eV) is comparable to the energy of the Q-band
short wavelength shoulder energy (2.0 eV) (Fig. 4b).
The analogous reaction of TCNVB with the ytterbium
p-sandwich complex (bis(indenyl)ytterbium (II)) proceeds very
smoothly under vacuum with the formation of a dye-like dark-
blue product (IV) readily soluble in THF and acetonitrile. In
contrast to III which is rather unstable in the air rapidly changing
colour from green to brown in solution, the Yb-containing
product is very stable not only in the solid state but also in
solution. The IR spectrum of the ytterbium complex shows
porphyrazine macrocycle skeletal (ca. 1470 cmꢀ1) and pyrrole
ring stretching vibration bands (nC]N 1650 cmꢀ1, nC]C 1597,
1560 cmꢀ1). In the ChN stretch region a band at 2205 cmꢀ1
which can be assigned to the CN groups belonging to the
TCNVB fragments at the periphery of the macrocycle is addi-
tionally observed as is also the case for I–III. The UV/visible
electronic absorption spectrum (Fig. 4d) is also rather typical for
porphyrazine metal complexes, particularly for those containing
extra ligands in the axial position of the macrocycle, exhibiting
absorption bands at 582 nm (Q-band) and 324 nm (Soret band).
Furthermore, the spectrum shows an additional electronic tran-
sition in the visible region at 393 nm. Lower intensity absorptions
Metal
template
assembly
of
the
tetraphenylte-
tracyanoporphyrazine framework was achieved by using the
same synthetic approach through reaction of TCNVB with
bis(benzene)vanadium or bis(indenyl)ytterbium.
The reaction of the vanadium p-sandwich complex with
TCNVB proceeds analogically to that with TCNE: a green solid
soluble in THF and acetonitrile is formed in the presence of
excess TCNVB after contact of the reaction mixture with
air oxygen and moisture. Likewise, the compound has a typical
Table 2 Summary of nChN data for I–IV, TCNE, TCNVB and their radical-anions
Compound
I
II
III
IV
TCNE10
TCNVB
2234 (s)
TCNE radical-anion10
TCNVB radical-anion
nChN (cmꢀ1
)
2215 (s)
2210 (s)
2144 (m,br)
2204 (s)
2204 (s,br)
2122 (s,br)
2262 (s)
2228 (m)
2214 (w)
2183 (m)
2144 (m)
2195 (m)
2185 (sh)
2122 (s)
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 3668–3676 | 3671

View Online
of porphyrazine metal complexes in the optical window between
the Q and Soret bands are usually assigned to charge-transfer
(CT) transitions, Yb/macrocycle and/or axial ligands.9
The presence of a reduced form of TCNVB as an axial ligand is
consistent with the IR spectrum of the purified ytterbium
complex showing additional ChN stretch bands at 2122 cmꢀ1
.
We observed the same frequency for the ChN stretch bands for
the product of TCNVB reduction on a potassium mirror in THF
(Table 2). It should be noted, however, that the nC]N band in the
ytterbium complex is greatly broadened (half-width at half-
height ca. 60 cmꢀ1 cf. ca. 15 cmꢀ1 for free TCNVBꢂꢀ) consistent
with the TCNVB reduced forms being bound to ytterbium in the
axial position. Interestingly, we did not observe any TCNVBꢂꢀ
ESR signal for IV: this may be explained by presence of electron
delocalization within the complex.
Electrochemical investigations of IV revealed an appropriate
decrease in reduction potential as a result of the replacement
of four peripheral cyano groups by four phenyl groups (Table
1). The first reduction potential of IV (ꢀ1.21 V) is consistent
with the second reduction potential for free TCNVB con-
firming the presence of TCNVBꢂꢀ which can undergo reduc-
tion in the complex.11 Thus, a structure analogous to II
might be expected for the TCNVB based ytterbium complex.
However, such a structure was not confirmed by the elemental
analysis data. Substantial differences between the two
complexes can also be deduced from a comparison of their
redox properties (Table 1).
Fig. 5 Near IR absorption of thin film of complex IV.
Adduct IV can be described as the final product of two
processes resulting from oxidation of the bis(indenyl)ytterbiu-
m(II) sandwich p-complex with TCNVB: the metal template
assembly of the porphyrazine macrocycle and the formation of
Yb(TCNVB)3. It should be noted that a species similar to the
latter, Ln(TCNE)3, (Ln ¼ Gd, Dy), has been observed previously
by Raebiger and Miller.13 However, in spite of numerous
attempts, we have been unable to obtain crystals of IV presum-
ably because of its polymeric nature. In addition, a mixture of at
least 4 isomers resulting from the distribution of phenyl and
cyanide substituents should be formed. Powder diffraction
measurements showed complex IV to be amorphous and in the
absence of X-ray structural data the binuclear bridging structure
can only be tentative (Fig. 6). The fragment of Ln(TCNVB)3 is
assumed to form a disordered polynuclear coordination polymer
network with metal cations bridged through the nitrile nitrogen
atoms.
We observed for the TCNVB-based Yb complex two irre-
versible oxidation potentials at 0.86 and 1.41 V. Neither oxida-
tion can be assigned to the metal since Yb3+ is inactive in the
anodic processes nor can they be assigned to the macrocycle since
the energy of the Q band (2.13 eV) is much lower than the energy
difference between the HOMO and LUMO (3.0 eV). The latter
band-gap is very close to 393 nm which is consistent with the
absorption band we observed in the optical window between the
Q and Soret bands, i.e. for charge-transfer between the transition
metal and the axial ligand. Thus, it is reasonable to assume that
the first and second oxidation potentials correspond to oxidation
of doubly and singly reduced TCNVB species, respectively. This
is not inconsistent with the IR data showing two nChN bands for
IV attributed to the macrocycle and CN radical-anion groups:
both bands are very broad and may obscure the bands corre-
sponding to the dianion CN.
The MALDI spectrum of IV shows a large number of peaks
corresponding to ionic species containing more than one ytter-
bium atom including the ionic cluster with a maximum at m/z
In turn, the presence of doubly reduced TCNVB allows one to
postulate the formation of binuclear species formed by its
bridging at least two Yb atoms. Such an assumption is consistent
with the presence of the strong broad absorption band in the near
IR region with a maximum at 1335 nm (Fig. 5) very typical for
binuclear structures. It is known that bi- and polynuclear
complexes containing bridging TCNE ligands have intense
absorption bands in the NIR region (800–1500 nm) corre-
sponding, for example, to p / p* transitions of polymetal-
containing systems in a delocalized description.12
Indeed, assumption of the formula [Ybpz(Ph)4(CN)4]$m-
[Ph(CN)C]C(CN)2]$Yb[Ph(CN)C]C(CN)2]2 (IV) for the
binuclear adduct formed in the system bis(indenyl)ytterbium(II)–
TCNVB gives excellent agreement with the elemental analysis
results. It should be noted that chemical and NMR analysis also
showed the presence of 6–8 molecules of THF in adduct IV.
Fig. 6 Proposed structure for complex IV (single isomer shown for
clarity).
3672 | J. Mater. Chem., 2009, 19, 3668–3676
This journal is ª The Royal Society of Chemistry 2009

View Online
Excitation with light in the wavelength range 530–600 nm gives
intense visible luminescence at 650 nm. This emission can be
readily observed under visible light for the doped glasses (Fig. 8;
see Table of Contents for colour version).
The emission and excitation spectra of Sample 1 (Fig. 9) show
a small Stokes shift that is very typical for dye fluorescence.
Time-resolved spectroscopy was used for the luminescence life
time measurements. For excitation we used an Evolution-30 laser
(Spectra Physics Lasers) delivering 100 ns pulses at 527 nm with
a repetition rate of 1 kHz. The short luminescence life-time (6.6
ns for Sample 1 and 5.1 ns for Sample 2) defined from the
emission pulse delay with respect to the excitation pulse also
Fig. 7 Results of magnetic studies on IV: cꢀ1 vs. T and cT vs. T plots
(see text).
1636 for the [M ꢀ {Ph(CN)C]C(CN)2} ꢀ H]+ ion where M ¼
IV$3THF.
Magnetic measurements have been made on IV from 2 to
300K under a magnetic field of 5000 Oe. As expected from the
proposed structure (Fig. 6), this compound displays para-
magnetic behaviour indicating that the ytterbium atoms are
present as Yb(III) ions in the complex. The cT value recorded at
300 K was 4.71 cm3 K molꢀ1 being somewhat smaller than the
expected value for two isolated Yb(III) centres (5.12 cm3
K
molꢀ1).14a This may be explained by the antiferromagnetic
interactions between the TCNVB radical anions and the Yb(III)
centre. The cꢀ1 vs. T curve (Fig. 7) follows rather well the
expected shape for two weakly interacting Yb(III) centres with
the observation of two linear regions (2–40 K and 65–300 K) with
different slopes.14a Weak antiferromagnetic interactions are seen
in the low temperature region, all the paramagnetic centres
(Yb(III) and radical-anions) appearing to be weakly antiferro-
magnetically coupled. However, a comparison of the magnetic
behaviour of IV with that of previously reported14b,c 2:1 metal-to-
ligand neutral adducts (Cp*2Yb)2(m-L), where L are dianionic/
diamagnetic bridging ligands, indicates that IV may be composed
of binuclear adducts in which the bridging ligand does not
facilitate significant magnetic exchange between the two Yb(III)
centres at low temperatures. Indeed, such behaviour is rather
typical for binuclear adducts where L has a free ligand second
reduction potential which is less negative than ca. ꢀ2.0 V vs. SCE
(cf. ca. ꢀ1.2 V for TCNVB11) when the energy of the ligand
LUMO is not properly poised to facilitate metal-metal electronic
interaction.14c
Fig. 8 PCMA luminescent glass doped with 10ꢀ4 mol lꢀ1 of IV
(Sample 2) (a) in transmitted light; (b) in reflected light (see Table of
Contents for colour version).
Similarly to the other cyanoporphyrazine complexes reported
previously,3 adduct IV readily forms optically-transparent solid
solutions and films with various polymers such as poly-
ethylcyanoacrylate,
polyacrylonitrile,
conjugated
poly-
carbosilanes and others. But the most interesting feature of IV is
the extraordinary light-emitting properties of this compound
incorporated into solid polymer matrices. For example, high
optical quality luminescent polymer films and glasses (100 ꢃ 50
ꢃ 2.3 mm) have been prepared from UV-polymerized oligo-
carbonatedimethacrylate (PCMA) doped with 10ꢀ5 and 10ꢀ4 mol
lꢀ1 of IV (Samples 1 and 2, respectively). They reveal an intense
visible and NIR photoluminescence dependent on the dopant
concentration and the excitation wavelength.
Fig. 9 Excitation ((a), lreg ¼ 650 nm) and emission ((b), lexcit ¼ 590 nm)
spectra of Sample 1 (10ꢀ5 mol lꢀ1 IV).
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 3668–3676 | 3673

View Online
Fig. 10 Time dependence of the excitation (solid line) and fluorescence
Fig. 12 The steady-state emission spectra of Sample 2, (a) lexcit ¼ 590
nm; (b) lexcit ¼ 660 nm.
(dot/dash lines) pulses (arbitrary units).
The change of excitation wavelength to 660 nm resulted in
luminescence only in the NIR (Fig. 12(b)). In addition to emis-
sion at 700 nm, intense NIR luminescence at 977 nm appears in
Sample 2 (Fig. 13). This emission is attributed to narrow-band-
width emission derived from the Yb 2F5/2 / 2F7/2 transition. It is
a typical example of ‘‘antenna’’-type excitation, where energy is
transferred from the excited state of the ligand moiety to the
emissive f-excited state of the lanthanide.
confirmed that the luminescence at 630–650 nm originates in the
porphyrazine framework arising from the p / p* electron
transition in the porphyrazine ligand with the first singlet mac-
rocycle excited state (S1) decaying back to the ground singlet
state (S0) (Fig. 10). The quantum yield of the fluorescence peak
with a maximum at 650 nm measured relative to Rhodamine 6G
ethanol solution (0.95)15 is about 0.13.
A substantial dependence of emission on the dopant concen-
tration was found for the PCMA doped glasses. In the time-
resolved emission spectrum we observe the fluorescence
maximum occurring at a longer wavelength region for Sample 2
which is one order of magnitude more concentrated than Sample
1 (Fig. 11). In turn, the investigation by steady-state lumines-
cence of Sample 2 (Fig. 12a) showed the appearance of an
additional emission band strongly shifted to the NIR region (710
nm). Furthermore, the appearance of an anti-Stokes lumines-
cence band at about 450 nm in this spectrum is of particular
interest suggesting an increase in photon-phonon energy transfer
in the system that can be indirect evidence for the presence of
Furthermore, additional doping of the composite containing
10ꢀ4 mol lꢀ1 of complex IV with an erbium chelate (erbium tris-
hexafluoroacetylacetonate, Er(AcFAcF)3), Yb:Er molar ratio ¼
1:1 (Sample 3), results in the appearance of an additional
significant emission in the NIR region at 1540 nm (minimum-loss
transmission window of silica fibre in optical telecommunication
networks) and can be attributed to the 4I13/2 / 4I15/2 transition of
Er3+ (Fig. 14).
This phenomenon is not surprising since Yb3+ is a well-known
Er3+ emission sensitizer with a relatively strong absorption at ca.
2
4
975 nm, the F5/2 level of Yb3+ and the I11/2 level of Er3+ being
nearly resonant in energy.16 On the other hand it is known that in
the sensitization process of Er3+ luminescence by Yb3+, the Er3+/
a
disordered coordination polymer network formed by
Yb[Ph(CN)C]C(CN)2]3 fragments. The contribution of the
supramolecular interactions in such a network into energy
transfer processes should obviously increase with the complex
concentration.
Yb3+ distance plays a significant role: the shorter the Er /Yb
distance the more effective the sensitization.13 Thus, it is
3+
3+
^
reasonable to assume that as a result of ligand exchange
Fig. 11 Time-resolved fluorescence spectra of Samples 1, 2 and 3 in
arbitrary units (lexcit ¼ 527 nm).
Fig. 13 NIR photoluminescence spectrum of Sample 2 (lexcit ¼ 660 nm).
This journal is ª The Royal Society of Chemistry 2009
3674 | J. Mater. Chem., 2009, 19, 3668–3676

View Online
network with a macrocyclic structure and metal cations bridged
through the nitrile nitrogen atoms is proposed. The complex is
readily soluble and is compatible with a variety of polymeric
matrices giving doped polymeric glasses and films which are
highly luminescent in the biologically relevant optical window
covering the visible and NIR range (640–1000 nm). In addition,
doped polymeric glasses and films highly emissive at the tele-
communication wavelength (1540 nm), including a composition
consisting of the novel ytterbium complex and an equimolar
ratio of the novel ytterbium complex and a per se non-lumines-
cent erbium chelate, have been obtained. The complex is found to
be an extraordinarily strong sensitizer of NIR Er3+ emission. The
materials developed based on novel porphyrazine vanadyl and
ytterbium complexes formed in high yield by metal-template
assembly from TCNE and TCNVB building-blocks have
promising potential in modern photonic and optoelectronic
devices such as photovoltaic cells, IR amplifiers in photonic
integrated circuits, NIR emission sources in telecommunication
defence applications and bio-imaging.
Fig. 14 NIR photoluminescence spectrum of Sample 3 doped with
Er(AcFAcF)3, (Yb and Er complex molar ratio 1:1), lexcit ¼ 660 nm.
Er(AcFAcF)3 can be incorporated into the coordination polymer
network formed by the Yb(TCNVB)3 fragment of IV. This
causes the short distance between the Yb and Er cations and
facilitates intramolecular Yb–Er energy-transfer. It should be
noted that no luminescence whatsoever of Er(AcFAcF)3 in
a PCMA matrix is observed in the absence of IV. Thus, it may be
expected that ‘‘antenna’’-type excitation applied to IV is more
efficient for energy transfer to Er3+ than is the direct excitation of
Yb3+ applied in the recently reported erbium ytterbium co-crys-
talline complexes exhibiting an increase of as much as ca. 30% in
Er3+ emission in comparison with the neat erbium complex.17 We
report here the first example, to our knowledge, of a lanthanide
porphyrazine complex exhibiting NIR emission of Yb3+ and
efficiently sensitizing the emission of Er3+. Only one other
example of NIR emissive ytterbium and erbium complexes (with
an Yb3+ and Er3+ emission quantum yield ratio of ca. 30:1) based
on a tetrapyrrole macrocycle (tetraphenylporphyrine) has been
reported previously.18
Acknowledgements
ꢀ
We thank Professor Patrick Calas (Universite Montpellier II) for
making the electrochemical measurements, and Dr V. V.
Treushnikov (Reper-NN Company, Nizhny Novgorod) for
preparing the PCMA luminescent glass samples. Financial
support was received from the Russian Foundation, the RAS
Nanotechnology Programme, RF President Grant No.
8017.2006.3 and INTAS (Open Call Project 03-51-5959).
References
1 S. R. Forrest, Organic Electronics, 2003, 4, 45–48.
2 S. Gunes, H. Neugebauer and N. S. Sariciftci, Chem. Rev., 2007, 107,
1324–1338.
3 L. G. Klapshina, I. S. Grigoryev, W. E. Douglas, A. A. Trifonov,
I. D. Gudilenkov, V. V. Semenov, B. A. Bushuk and S. B. Bushuk,
Chem. Commun., 2007, 1942–1944.
4 (a) A. A. Trifonov, E. A. Fedorova, V. N. Ikorskii, S. Dechert,
H. Schumann and M. N. Bochkarev, Eur. J. Inorg. Chem., 2005,
2812–2818; (b) G. N. Sausen, V. A. Engelhardt and W. J. Middleton,
J. Am. Chem. Soc., 1958, 80, 2815–2822; (c) V. M. Treushnikov,
J. Photochem. Photobiol. A: Chemistry, 2008, 196, 201–209; (d)
O. Kahn, Molecular Magnetism, VCH, New York, 1993.
5 N. Kobayashi, S. Nakajima and T. Osa, Chem. Lett., 1992, 2415–
2418.
Conclusions
The series of new vanadyl and ytterbium(III) cyano-substituted
porphyrazine complexes designed from TCNE and TCNVB
structural units through reaction of TCNE and TCNVB with
metal p-sandwich complexes has been developed. The influence
of the nature of the central metal cation and the macrocycle
peripheral framework on the redox behaviour and structure of
the complex have been demonstrated. The vanadyl octacyano-
porphyrazine complex was found to be a rare example of
a highly-absorbing dye combining significant electron-acceptor
properties with a band gap unusually narrow for an organic
semiconductor (ca. 1.1 eV). The preparation is described of
a novel highly emissive ytterbium complex, for which an unusual
structure is proposed, obtained by reaction of TCNVB with
bis(indenyl)ytterbium(II) in THF. The analytical, spectral and
electrochemical investigations of the ytterbium complex thus
obtained indicate that it exists in the form of a binuclear adduct
consisting of an Yb(TCNVB)3 species in which a single doubly-
reduced TCNVB molecule bridges two Yb3+ cations. The
formation of a disordered polynuclear coordination polymeric
6 K. Ozette, P. Leduc, M. Palacio, J.-F. Bartoli, K. M. Barkigia,
J. Fager, P. Battioni and D. Mansuy, J. Am. Chem. Soc., 1997, 119,
6442–6443.
7 (a) B. W. D’Andrade, S. Datta, S. R. Forrest, P. Djurovich,
E. Polikarpov and M. E. Thompson, Organic Electronics, 2005, 6,
11–20; (b) C. Borek, K. P. Hanson, I. Djurovich, M. E. Thompson,
K. Aznavour, R. Bau, Y. Sun, S. R. Forrest, J. Brooks, L. Michalski
and J. Brown, Angew. Chem. Int. Ed., 2007, 46, 1109–1112.
8 A. H. Kianfar and S. Mohebbi, J. Iran. Chem. Soc., 2007, 46, 1109–1110.
9 P. A. Stuzhin, S. I. Vagin and M. Hanack, Inorg. Chem., 1998, 37,
2655–2662.
10 J. S. Miller, Angew. Chem. Int. Ed., 2006, 45, 2508–2525.
11 H. Bock, W. Seitz, N. Nagel, R. Baur, J. W. Bats, Z. Havlas and
R. F. C. Claridge, Zeit. fur Naturforsch., B: Chem. Sci., 1997, 52,
1125–11383.
12 (a) R. Gross-Lannert, W. Kaim and B. Olbrich-Deussner, Inorg.
Chem., 1990, 29, 5046–5053; (b) W. Kaim and M. Moscherosch,
Coord. Chem. Rev., 1994, 129, 157–193; (c) C. Diaz and
A. Arancibia, Polyhedron, 2000, 19, 137–145.
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 3668–3676 | 3675

View Online
13 J. W. Raebiger and J. S. Miller, Inorg. Chem., 2002, 41, 3308–
3312.
16 M. P. Hehlen, N. J. Cockroft, T. R. Gosnell and A. J. Bruce, Phys.
Rev. B, 1997, 56, 9302.
14 (a) M. Schultz, J. M. Boncella, D. J. Berg, T. D. Tilley and
R. A. Andersen, Organometallics, 2002, 21, 460; (b) C. N. Carlson,
C. J. Kuehl, R. E. da Re, J. M. Veauthier, E. J. Schelter,
A. E. Milligan, B. L. Scott, E. D. Bauer, J. D. Thompson,
D. E. Morris and K. D. John, J. Am. Chem. Soc., 2006, 128, 7230–
7241; (c) D. J. Berg, J. M. Boncella and R. A. Andersen,
Organometellics, 2002, 21, 4622–4631.
17 L. Song, X. Liu, Z. Zhen, C. Chen and D. Zhang, J. Mater. Chem.,
2007, 17, 4586–4590.
18 (a) B. S. Harrison, T. J. Foley, A. S. Knefely, J. K. Mwaura,
G. B. Cunningham, T.-S. Kang, M. Bouguettaya, J. M. Boncella,
J. R. Reynolds and K. S. Schanze, Chem. Mater., 2004, 16, 2938–
2947; (b) T. J. Foley, B. S. Harrison, A. S. Knefely, K. A. Abboud,
J. R. Reynolds, K. S. Schanze and J. M. Boncella, Inorg. Chem.,
2003, 42, 5023–5032.
15 R. F. Kubin and A. N. Fletcher, J. Luminesc., 1982, 27, 455–462.
3676 | J. Mater. Chem., 2009, 19, 3668–3676
This journal is ª The Royal Society of Chemistry 2009