Homoleptic Co(II), Ni(II), Cu(II), Zn(II) and Hg(II) complexes of bis-(phenyl)-diisoindol-aza-methene

Article abstract of DOI:10.1039/c0dt01655a

The synthesis of five homoleptic transition metal complexes of bis-(phenyl)-diisoindol-aza-methene is described together with the optical, electrochemical and thermal properties of these compounds. Additionally, crystal structures for the Co and the Zn complex are reported. The Royal Society of Chemistry 2011.

Full text of DOI:10.1039/c0dt01655a

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2011, 40, 3476
www.rsc.org/dalton
PAPER
Homoleptic Co(II), Ni(II), Cu(II), Zn(II) and Hg(II) complexes of
bis-(phenyl)-diisoindol-aza-methene†
Roland Gresser,* Alexander Hoyer, Markus Hummert, Horst Hartmann, Karl Leo and Moritz Riede
Received 29th November 2010, Accepted 25th January 2011
DOI: 10.1039/c0dt01655a
The synthesis of five homoleptic transition metal complexes of bis-(phenyl)-diisoindol-aza-methene is
described together with the optical, electrochemical and thermal properties of these compounds.
Additionally, crystal structures for the Co and the Zn complex are reported.
structural and spectroscopic properties were investigated, in terms
of supramolecular chemistry, light harvesting systems or metallic
Introduction
organic frameworks.6 To the best of our knowledge, in the case
of bidentate metal complexes, with the (aza)-diisoindolmethene
ligands, C (and D), only a copper compound was synthesized,
but no spectroscopic data was given (Fig. 1).8 Recently, we
reported a series of new aza-diisoindolemethene ligands, D, and
the corresponding boron difluoride compounds.7 These complexes
absorb, depending on their substituents, in the red or NIR (Near-
infrared) region and the maximum can be related to an intense
HOMO–LUMO transition. The more pronounced stabilization of
the LUMO upon complexation with boron difluoride, originates
in the bathochromic shift observed in the absorption spectra.7
In this study, we describe the synthesis and systematic character-
ization of five divalent transition metal derivatives of the aza-bis-
(phenyl)-diisoindolmethene ligand,7–9 D (R = Ph, 1), with Co(II),
Ni(II), Cu(II), Zn(II), and Hg(II) as central metal ions, in order to
support a better understanding of aza-BODIPY derivatives in the
context of organic optoelectronics and enable the development of
future applications.
Dipyrromethenes of the general structure A and their aza-methene
analogue ligands B (Fig. 1) are able to form extremely stable
complexes with a variety of metal or pseudo-metal ions, especially
with boric acid derivatives yielding BODIPY and aza-BODIPY
derivatives.1,2 This emerging material class has attracted enormous
interest due to their intense absorption and strong fluorescence,
especially in the red region of the spectrum. These outstanding
properties stimulated the investigation of BODIPYs for various
applications such as fluorescence probes for biological imaging,3 as
emitter materials in organic light emitting diodes,4 or as absorber
materials for organic solar cells.5
For this purpose, the spectroscopic properties of these metal
complexes were investigated and compared to the corresponding
ligand and boron difluoride complex. Further quantum chemical
calculations were performed to gain a deeper insight in the
absorption properties, in particular into the absorption features of
the Zn complex. The electrochemical behavior, which is dependant
on the central metal ions, was determined by cyclic voltammetry.
From single crystals of the Co and the Zn complexes, the molecular
structure is revealed by X-ray diffraction and are presented and
discussed. Since possible applications of these materials may
occur by thermal evaporation in a vacuum, we also examined the
thermal stability by thermogravimetric analysis in order to probe
a potential processability.
Fig. 1 Dipyrromethenes, A, and their aza-methene analogues, B, and the
corresponding (aza-)diisoindole-methenes, C and D.
In contrast to the chemistry of the boron containing BODIPYs
and aza-BODIPYs, the chemistry of the metal complexes of
(aza-)dipyrromethenes, A (and B), developed significantly slower.
However, in recent years, di- and tri-coordinative transition metal
complexes of the compounds A and B were described. These
complexes were prepared with a variety of metals like Ag(I),
Au(I), Cu(I) and Co(II), Ni(II), Cu(II), Zn(II), and Hg(II), and their
Institut fu¨r Angewandte Photophysik, Fakulta¨t fu¨r Physik, Technische
Universita¨t Dresden, George-Ba¨hr-Strasse 1, 01069, Dresden, Germany.
E-mail: roland.gresser@iapp.desxsxst; Fax: (+) 49-(0)351-463-37065
Results and discussion
† Electronic supplementary information (ESI) available: Crystal structure
and refinement data, thermogravimetric analysis of 1, 3 and the Cartesian
coordinates of the optimized structures 1 and 2d together with additional
molecular orbitals. CCDC reference numbers are 801539 (2a) and 801540
(2d). For ESI and crystallographic data in CIF or other electronic format
see DOI: 10.1039/c0dt01655a
Synthesis
For the synthesis of the metal complexes, 2, the corresponding
metal-(II)-acetate salts were reacted with two equivalents of 1 in
either n-butanol or tetrahydrofuran (THF) as shown in Fig. 2.
3476 | Dalton Trans., 2011, 40, 3476–3483
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
and the Hg complexes typically gave spectra in the 0–10 ppm range.
The 1H NMR spectrum of the Ni complex is of particular interest,
as it displayed considerable positive and negative peak shifts
between -2 to +40 ppm at room temperature (see experimental
section). Similar 1H NMR characteristics were already described
for a Ni bis-dipyrromethene10 and other Ni complexes.11 The
distortion of the usually favored planar geometry to a pseudo-
tetrahedral geometry due to the sterically demanding ligands in
these complexes is given in the literature as an explanation and
can also be expected to be the reason for 2b. Nickel complexes
exhibiting a pseudo-tetrahedral geometry, have short electron-spin
lifetimes and therefore record well-resolved NMR spectra.10
Crystal structures
Single crystals of the Co and Zn complexes, 2a and 2d, were
grown by slow evaporation of a chlorobenzene or toluene solution.
Attempts to grow X-ray quality crystals for the other metal
complexes, e.g. by slow vapor diffusion of pentane in a THF or
chloroform solution, were not successful and resulted in crystals
which were too small. Both the Co complex 2a and the Zn
Fig. 2 Synthesis of the metal complexes, 2.
This preparation method leads to the appropriate bis-
homoleptic complexes, 2, as crystalline solids in moderate to excel-
lent yields, ranging from 38% to 84%. The Co, Ni, and Zn acetate
salts were refluxed with the bis(phenyl)-aza-diisoindole-methene
ligand 1 in n-butanol, until no starting material could be assessed
by thin layer chromatography (for 2a and 2b) or until the product
precipitated (in the case of the Zn complex 2d). The Cu and Hg
complexes, 2c and 2e, were synthesized by heating the acetate
salts in THF solution with diisopropylethylamine (DIPEA) for
six hours for four days. In all cases, the precipitated products were
purified by re-crystallization from dichloromethane/hexane solu-
tion, yielding pure crystalline complexes. The mass spectra of all
prepared compounds, 2, evidently showed the formation of the bis-
homoleptic metal complexes 2a–2e. As further characterization,
1H NMR spectra were recorded for all compounds except for the
paramagnetic Co and Cu complex, as might be expected. The Zn
¯
complex 2d crystallize in the triclinic space group P1 with one
solvent molecule in the unit cell (Fig. 3). Relevant interatomic
distances and angles are summarized in Table 1. The molecular
structure of both bis-chelates, 2a and 2d, can be regarded as a
pseudo-tetrahedral coordination geometry with the metal atom in
the center. The corresponding angle between the aza-diisoindole
planes are 68.0(6)◦ for the Co complex, 2a, and 66.7(9)◦ for the
Zn complex, 2d, which is significantly distorted from an ideal
D2d symmetry, as shown in Fig. 3a and 3d. The metal–nitrogen
˚
bond lengths (M–N) range between 1.98(1) and 1.99(2) A for the
˚
Co complex, 2a, and are 1.99(1) and 1.99(8) A long for the Zn
complex, 2d.
Fig. 3 Single crystal structure of the Co complex a–c and the Zn complex d–f. (a) and (d) Pseudo tetrahedral conformation; (b) and (e) top view; (c)
and (f) unit cell (solvent molecule truncated for reasons of clarity). Hydrogen atoms are truncated in all cases for reasons of clarity as well.
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 3476–3483 | 3477
©

View Article Online
Table 1 Selected bond lengths and angles obtained from the single crystal
structure data
Absorption characteristics
The UV-Vis absorption properties of all the metal complexes 2a–
2e were investigated and compared with respect to the free ligand,
1, and the boron difluoride complex, 3, which were described
elsewhere.7–9 The spectroscopic data are shown in Fig. 4 and are
summarized in Table 2. All spectra of the ligand, 1, and the
metal complexes, 2a–2e, exhibit an intense absorption band in
the UV region, and an absorption maximum in the visible regime.
In contrast to the spectrum of 1, the absorption maxima of the
metal complexes, 2a–2e, display a broadening and the appearance
of shoulders. The absorption maxima of the metal complexes are
in the range between 632 nm and 674 nm in DCM, depending on
the metal (Table 2). The Co and Ni complexes, 2a and 2b, show
a red-shifted absorption maximum by 24 nm compared to the
ligand, 1, and an absorption tail extending into the NIR region
over 800 nm. The absorption maximum of the Cu complex, 2c,
is red-shifted about 3 nm, with respect to the ligand, 1, whereas
the absorption maximum of the Hg complex, 2e, is at the same
wavelength. The Zn complex, 2d, displays a hypsochromic shift
of the maximum but exhibits an intense red-shifted shoulder at
685 nm, compared to 1. All the metal complexes 2a–2e feature, in
comparison to the maximum, a strong absorbance in the UV part
of the spectra around 300 nm as is also the case for the free ligand,
1. This is in contrast to the boron difluoride complex, where this
absorbance in the UV is weakened compared to the maximum.
The extinction coefficients of 2a–2e are in the range of 46 000–
108 000 Lmol-1 cm-1 for the corresponding absorption maxima
and therefore increased compared to the ligand, 1, except for the
2a (Co)
2e (Zn)
˚
˚
˚
˚
˚
˚
˚
1.99(6) A
M–N1
M–N2
M–N4
M–N5
N1–M–N2
N4–M–N5
C–N3–C
C–N6–C
Plane angle
1.99(1) A
1.99(2) A
1.98(1) A
1.98(3) A
˚
1.99(3) A
˚
1.99(3) A
˚
1.99(8) A
˚
˚
95.1(5) A
94.2(2) A
93.9(4) A
95.3(4) A
126.2(4)◦
127.4(2)◦
68.0(6)◦
128.0(4)◦
127.2(0)◦
66.7(9)◦
These bond lengths are similar to those found for comparable
Zn(II) bis-dipyrromethene complexes,12 but significantly longer
than for the corresponding boron difluoride complex, with the
9
˚
N–B bond being 1.57(9) A long.
The nitrogen–metal–nitrogen (N–M–N) bite angle differs for
each bis-isoindole aza-methene ligand in both metal complexes.
In the Co complex, 2a, the angle N1–Co–N2 is 95.15◦ and 93.94
for the N4–Co–N4 angle. In the Zn complex, 2d, the angle is
94.2(2)◦ for the N1–Zn–N2 and 95.3(4) for the N4–Zn–N4. The
bite angle in the metal complexes is significantly smaller compared
to the corresponding boron difluoride complex with 105.7(3)9 and
is consistent with the longer N–metal bond length. Additionally,
the C–N3/6–C angle is widened in the metal aza-diisoindole bis-
chelates◦in the range between 126.2(4)◦ and 128.0(4)◦ compared to
119.7(1) in the boron difluoride complex.9 An interesting aspect
is the sterically induced ligand–ligand interactions between the
phenyl ring of one ligand and the aza-bis-isoindole plane of the
opposite ligand, worthy of discussion in more detail. In the Co
complex, 2a, the minimum distance of the phenyl ring to ◦the
˚
opposite isoindole plane is 2.9(0) A, with an angle of 23.2(2) to
the plane. In the Zn complex, 2d, the distance is slightly increased
to 3.09(4) but the angle is decreased to 20.8(8). This distance of
the phenyl ring to the isoindole plane in 2a and 2d is much shorter
compared to non benzannulated aza-bodipy Zn complexes, having
12
˚
a minimum distance of 3.6(2) A. Therefore, an intermolecular
p–p interaction can be considered to be present in these bis-
chelate complexes. The interaction increases with an enlarging
p-system. Accordingly, the distance of the phenyl ring plane to the
opposite ligand is reduced in the benz-annulated aza-bis-isoindole-
methenes compared to the aza-dipyro-methenes with their smaller,
non-annulated p-system on the backbone.12
Fig.
4
Absorption spectra of the metal complexes, 2a–2e, in
dichloromethane. 1 and 3 are shown for comparison.
Table 2 Absorption characteristics of the metal complexes, 2a–2e, in comparison to 1 and 3
Extinction [Lmol-1 cm-1]
lmax [nm]
Extinction [Lmol-1 cm-1]
lshou [nm]
Extinction [Lmol-1 cm-1]
a
b
c
lUV [nm]
1 (H)
299
298
288
290
297
260
305
34 000
37 500
27 000
38 000
43 000
24 000
28 000
653
674
674
656
632
653
715
56 000
58 500
46 500
73 500
108 500
60 500
106 000
—
—
2a (Co)
2b (Ni)
2c (Cu)
2d (Zn)
2e (Hg)
3 (BF2)
725
645
—
685
626
—
50 500
35 500
—
80 500
49 500
—
Extinction coefficients are given fora the most intense absorption in the UV region of the spectra, b the absorption maxima in the visible regime and c the
most pronounced shoulder of the absorption maxima.
3478 | Dalton Trans., 2011, 40, 3476–3483
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Table 3 Energy, oscillator strength and principal orbital contribution in
the ligand, 1, and the Zn complex, 2d
Ni complex, 2b (Table 2). While the absorbance of the Co- and Ni
complexes, 2a and 2b, is weak (with 46 500–58 500 Lmol-1 cm-1),
the extinction increases from the Hg, 2e, to the Cu, 2c, and again
to the Zn chelate, 2d, from 60 500 to 73 500 and up to 108 500
Lmol-1 cm-1. This intense absorbance is comparable to the boron
difluoride complex (as shown in our previous results)7,9 or even
stronger in the case of the Zn complex, 2d. The Zn complex, 2d,
displays the strongest absorbance for this ligand system so far.
All bis-(phenyl)-diisoindol-aza-methene metal complexes, 2a–
2e, are non-emissive and show no fluorescence, which is in contrast
to 3, since the boron difluoride compounds show an emission.9
Similar behavior was observed for (aza-dipyrromethene)-zinc(II)
and -mercury(II) complexes,12 but these aza-dipyrromethenes
offer four rotatable phenyl-rings. However, since the more rigid
compounds, 2, do likewise not emit, the quenching is due to the
heavy-atom effect, induced by the central transition metal ion.
Oscillator
Energy [eV] (nm) strength
Principal orbital contributions
1
2.19 (567.21)
3.14 (394.17)
0.788
0.016
HOMO→LUMO
HOMO→LUMO+1
HOMO-1→LUMO
HOMO-2→LUMO
HOMO→LUMO+1
HOMO-1→LUMO
HOMO→LUMO+1
HOMO→LUMO
3.15 (392.80)
2.14 (577.16)
2.24 (553.26)
3.35 (369.48)
0.065
0.552
0.878
0.011
2do,a
HOMO-1→LUMO+1
HOMO-4→LUMO+1
HOMO-3→LUMO
HOMO-1→LUMO+4
HOMO→LUMO+5
HOMO-4→LUMO
HOMO-3→LUMO+1
HOMO-1→LUMO+5
HOMO→LUMO+4
HOMO-1→LUMO
HOMO→LUMO+1
HOMO→LUMO
3.36 (368.92)
0.021
Quantum chemical calculations
2dc,b
2.10 (588.63)
2.29 (540.12)
3.35 (369.48)
0.382
0.878
0.014
In order to study the absorption properties in more detail, the
absorption spectra of the ligand, 1, and the corresponding Zn
complex, 2d, regarded as a model system, were calculated with
time-dependent density functional theory (TD-DFT, b3lyp/6-
31g*).13,14 For this purpose, the geometries of both structures
1 and 2d were optimized (ESI†) and the optical transitions
up to the first ten excited states were calculated. Although the
prediction of the absorption maximum deviated about 0.28 eV, the
agreement with the observed spectra is good for 1 and 2d, since
both absorption features were predicted correctly.15 All absorption
features in the aza-methene ligand system 1 result from a p–p*
transition, as one might expect. The calculation of the absorption
maximum in the visible regime for 1 clearly showed the principal
orbital contribution for this absorption to be a HOMO→LUMO
transition. In the ligand, 1, the next occupied levels lower in
energy (HOMO-1, -6.35 eV) and the next unoccupied levels
higher in energy (LUMO+1, -0.89 eV) differ significantly in energy
compared to the HOMO (-4.67 eV) and LUMO (-2.54 eV) energy
values. Consequently, optical transitions from these lower lying
occupied orbitals to the LUMO or LUMO+1 are involved for
the high energy absorption in the UV (Table 3), in case of the
ligand, 1. The calculations on the optimized structure of the Zn
complex, 2d, showed degenerated HOMO energies (-4.56 eV)
and nearly similar LUMO and LUMO+1 energies (-2.47 eV
and -2.36 eV). This energetic situation, which is different to
1, leads to several allowed optical transitions with absorption
energies very close to each other, observed in the optical spectra as
shoulders. The TD-DFT calculations revealed, that the absorption
maximum including the shoulder in the visible region arises from
two contributions. First, transitions from the HOMO→LUMO
and HOMO-1→LUMO+1 lead to the most intense absorption.
Secondly, the intense red-shifted shoulder arises due to transitions
from the HOMO-1 to LUMO and the HOMO to LUMO+1. The
relative intensities of these transitions were predicted in agreement
with the experiment, with a lower extinction for the red-shifted
shoulder in the case of the Zn complex, 2d. For the high energy
absorption in 2d, detected in the UV region, deeper lying orbitals
are involved, as for 1, which are listed in detail in Table 3.
HOMO-1→LUMO+1
HOMO-4→LUMO+1
HOMO-3→LUMO
HOMO-1→LUMO+4
HOMO→LUMO+5
HOMO-4→LUMO
HOMO-3→LUMO+1
HOMO-1→LUMO+5
HOMO→LUMO+4
3.37 (366.88)
0.021
Calculation with the optimized ligand structure, 1,a with the optimized Zn
complex structure, 2do, and b with the single crystal structure, 2dc.
quasi-degeneration of the HOMO and HOMO-1 energy levels
with a deviation of only 0.0232 eV. This break-up of degeneracy
compared to the optimized structure is most likely due to
variations from an ideal D2d symmetry, induced by solid state
packing effects. However, the absorption features were predicted
in the same way as for the optimized structure without any
significant changes (Table 3). A detailed analysis of the frontier
orbitals of the Zn complex, 2d, (Fig. 5) displayed the HOMO and
HOMO-1 orbital to be the only ligand located. The LUMO and
LUMO+1 instead are partially metal located with a clearly visible
alternation of the orientation of the metal d-orbital in the center.
This indicates, that the optical transitions in 2d, leading to the
absorption maximum including the shoulder, have a partial ligand
to metal L(p)→M(d*) charge transfer character (LMCT).
Electrochemical characterization
Since the redox potentials and the resultant HOMO and LUMO
energies are important for (opto-)electronic applications, the
electrochemical properties of all prepared complexes, 2a–2e,
were examined with cyclic voltammetry (CV), to obtain the
corresponding oxidation and reduction potentials. The results
were compared to the potentials of the free ligand, 1, and the
boron difluoride complex, 3. The measurements were done in
dichloromethane, using an Ag/AgCl reference electrode with
the ferrocene/ferrocinium couple (Fc/Fc+) as the internal redox
standard.16 The metal compounds, 2a–2e, exhibit a complex
electrochemical behavior. All reversible redox states were only
Calculations based on the single crystal structure of 2d support
the findings of the optimized structure. The results showed a
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 3476–3483 | 3479
©

View Article Online
Table 4 Electrochemical data of the metal complexes, 2a–2e, the ligand,
1, and the boron difluoride complex, 3, vs. Fc/Fc+
Compound
Er1ed2
Er1ed1
Eo1x1
Eo1x2
2
2
2
2
1 (H)
—
-1.44a
-1.52a
-1.36b
-0.78b
-1.62a
-1.49b
-1.12a
0.18a
0.02a
-0.01a
-0.14a
0.02a
0.04a
0.43a
—
2a (Co)
2b (Ni)
2c (Cu)
2d (Zn)
2e (Hg)
3 (BF2)
-1.78a
—
0.26a
0.21b
0.44b
0.26a
0.26b
—
—
-1.78a
-1.59b
—
Half wave potentials vs. Fc/Fc+ of the first reduction and oxidation and (if
present) the second reduction and oxidation; measured in CH2Cl2/Bu4PF6
(0.1 M) vs. Ag/AgCl, scan rate 100 mV s-1, with Fc/Fc+ as internal
standard.17a Reversible redox waves. b Irreversible redox waves.
Fig. 6 Reversible one- and two-electron redox waves for the Co and Zn
complex, 2a and 2d, measured in CH2Cl2/Bu4PF6 (0.1 M) vs. Ag/AgCl,
scan rate 100 mV s-1, with Fc/Fc+ as internal standard.16
less favored for 2d (Zn) by 0.10 V. This characteristic is represented
in the LUMO energies in the Fig. 6.18
The one-electron oxidation shows a trend in the CV measure-
ment with a decreasing ionization potential from complex 2a (Co)
to 2b (Ni) by 0.03 V, to 2c (Cu) by 0.16 V, and further increases
in the potential for complex 2d (Zn) and 2e (Hg) by 0.02 V. On
the one hand, this behavior is in contrast to the trend of the pure
transition metals M(0), where a higher oxidation potential for Cu
than for Co is expected, due to the higher ionization energy. One
the other hand, this trend can be qualitatively understood with
the simple Pearson’s concept of hard and soft Lewis acid and
bases (HSAB).19 In the HSAB concept, the hardness of transition
metal ions (M2+) is decreasing (or the softness is increasing) in
the periodic table by going from Co(II) via Ni(II) to Cu(II).20
Consistently, the oxidation potentials of 2a–2c are continuously
shifted towards lower potentials. The higher oxidation potential
of the Zn and the Hg compound, 2d and 2e, can be considered to
be due to the additional stability of the electronic configuration
for a closed s-atomic orbital shell.20
Fig. 5 Frontier orbitals of the optimized Zn complex, 2d, involved in the
lowest energy absorption. Note the ligand only character of the HOMO-1
and HOMO and the partial metal located LUMO and LUMO+1.
The electrical gap of the metal complexes, 2a–2e, obtained from
the CV measurements (Fig. 7) does not correlate with the observed
measured for the Co and Zn complexes, 2a and 2d, with two
reversible oxidation and reduction waves (Fig. 5) in each case. The
Co complex, 2a, displayed potentials at -1.52 V vs. Fc and at
-1.78 V vs. Fc for the first and the second reduction, respectively.
The oxidation waves were observed at 0.02 V and 0.26 V vs. Fc for
the first and second oxidation wave respectively (Fig. 6). The Zn
complex, 2d, showed the same redox potentials as the Co complex,
2a, apart from the first reduction wave which was shifted by -100
mV to lower potentials. The complexes 2b (Ni), 2c (Cu) and 2e (Hg)
displayed only reversible one-electron oxidation waves, whereas
the second oxidation and the one electron reduction (and two
electron reduction for 2e) processes appeared to be irreversible. All
measured redox potentials are listed in Table 4. The one-electron
reduction potentials of the Co, Ni, Cu, and Zn complexes, 2a–
2d, reflect the periodic trend of the electron affinity for the first
row transition metal elements. In this trend one would expect an
increase of the electron affinity from Co to Ni to Cu, and a decrease
for Zn. The experimental reduction in fact becomes more facile
from complex 2a (Co) to 2b (Ni) by 0.16 V, to 2c (Cu) by 0.74 V, and
Fig. 7 Calculated HOMO and LUMO energy values from the obtained
redox potentials; EHOMO(Fc) = -4.78 eV, EHOMO = -4.78 + (Eo1x(Fc) - (Eo1x)),
2
2
ELUMO = -4.78 + (Eo1x (Fc) - (Er1ed)).17
2
2
3480 | Dalton Trans., 2011, 40, 3476–3483
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
absorption properties (Fig. 4). According to the electrochemical
data, the Cu complex, 2c, should have the most red-shifted ab-
sorption, assuming a pure HOMO–LUMO transition. However,
this is neither the case for the absorption maximum nor for the
red-shifted shoulder occurring in the spectrum. For this reason,
the electrochemical measurements confirm that no pure HOMO–
LUMO transitions are present in the absorption process, as was
evaluated by quantum chemical calculations for the Zn complex,
2d.
shown for a related Hg complex.12 It is worth mentioning, that the
Co and the Zn complex, 2a and 2d, are both evaporable in a high
vacuum (10-6 mbar) at moderate temperatures (~200 ◦C), offering
an easy processability for further investigations.
Conclusions
The synthesis and systematic characterization of Co, Ni, Cu,
Zn, and Hg complexes of aza-bis-(phenyl)-diisoindolmethene
is described. The molecular structures of the Co and the Zn
complexes, 2a and 2d, adopt a pseudotetrahedral geometry in
the solid state, with a significant distortion from D2d symmetry,
as obtained from single crystal X-ray diffraction.
Thermogravimetric analysis
To probe a possible processability of these materials by thermal
vacuum sublimation, the thermal stability of the complexes,
2a–2e, was investigated by differential thermal analysis (DTA)
and thermogravimetric analysis (TGA).21 The mass loss of the
metal complexes was measured by TGA in an inert nitrogen gas
atmosphere, as shown in Fig. 8.
The absorption spectra of all complexes, 2a–2e, are charac-
terized by a broadening of the absorption maximum together
with shoulders. The extinction coefficients are increased compared
to the ligand, 1, in particular for the Zn complex, 2d, with the
highest extinction for this ligand system so far. All complexes 2a–
2e are nonemissive. TD-DFT calculations of the Zn complex, 2d,
revealed that the absorption behavior originates from degenerate
HOMO and HOMO-1 energies and transition from these orbitals
into the LUMO and LUMO+1, with a partial ligand-to-metal
charge transfer character. Cyclic voltammetry measurements
showed a trend of decreasing oxidation and increasing reduction
potential for the first-row transition metal complexes, 2a–2d.
This makes it possible to tune the electrochemical potential and
hence the HOMO and LUMO energies simply by choosing the
appropriate metal ion. Together with a good thermal stability (up
to 460 ◦C for the Zn complex, 2d) the metal complexes of the
aza-bis-(phenyl)-diisoindolmethene offers a simple processabiltiy
by thermal evaporation. Therefore, a further investigation for the
application of these materials in organic solar cells is intended.
Experimental
Fig. 8 Thermogravimetric analysis of the metal complexes, 2a–2e, in the
temperature regime of 50 to 500 ◦C.
All starting materials were obtained from commercial sources and
were used without further purification. Tetrahydrofurane (THF)
Since the ligand, 1, in these systems remained the same,
the measured stability is related to the stability of the metal–
nitrogen bond and the thermal activation of the metal-(II)-ions
and associated decomposition reactions. Although there is a mass
drift present in a few samples, the decomposition proceeds very
abruptly and hence allows for an objective analysis of the data.
All of the investigated compounds, except the Hg complex, 2e,
exhibit a good thermal stability and a significantly increased
stability compared to the free ligand, 1.21 For compounds 2a–
2d incorporating first-row transition elements (Co, Ni, Cu and
Zn), the thermal stability correlates with the oxidation potential
of these complexes. The most stable compounds were those with
Co and Zn as central metal atoms, 2a and 2d. These materials
showed a negligible mass loss (<5%) until 420 ◦C but a rapid
mass loss in higher temperature regimes. In terms of thermal
stability, the next most stable compound is the Ni complex, 2b,
with a decomposition temperature around 385 ◦C, followed by
the Cu complex, 2c, with decomposition occurring at 345 ◦C. The
Hg complex, 2e, is the thermally most fragile compound with
degradation occurring at 286 ◦C combined with a very rapid mass
loss. This material behavior may be related to the larger Hg(II)-
ion and hence a longer, and therefore weaker Hg–N bond, as was
was dried over sodium and distilled right before use or stored over
1
˚
˚
mole sieve 4 A. Butanol was dried over mole sieves 4 A. H and
13C spectra were recorded at 25 ◦C with a Bruker DRX 500 P in-
strument at 500.13 and 125.76 MHz, respectively. The assignment
of quaternary C, CH, CH2 and CH3 was completed using DEPT
spectra. Elemental analysis achieved with an Eurovektor Hekatech
EA-3000 elemental analyzer. The UV-VIS spectra were measured
with a Perkin Elmer l 25 spectrometer. Mass spectra were recorded
with a Bruker Esquire-LC 00084 instrument, at 10 Volt with
electrospray ionization with methanol (0.1% ammonium acetate).
Melting points and DTA-TG were measured simultaneously with
the thermal analyzer STA 409PC Luxx, direct mass spectroscopy
were done with a capillary coupling to QMS 403 C Ae¨los, both
from NETZSCH.
Cyclic voltammetry was recorded with a Metrohm m-Autolab
potentiostat in a single-component cell under a nitrogen atmo-
sphere. A typical three electrode configuration with an inlaid
platinum disk as working electrode, platinum wire as counter
electrode and a silver rod coated electrochemically with AgCl.
Potentials were measured versus SCE and referenced to ferrocene
as an internal standard (E◦(Fc/Fc+) = -4.78 eV to vacuum). The
measurements were performed with a scan rate of 100 mV s-1 in
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 3476–3483 | 3481
©

View Article Online
de-gassed dichloromethane (HPLC quality) and addition of tetra-
n-butylammonium hexafluorophosphate (TBAPF, 0.1 mol L-1) as
electrolyte. Compound 1 was synthesized as described in ref. 7.
J = 7.56 Hz, 4H), 6.83 (t, J = 7.56 Hz, 2H). 13C NMR (126 MHz,
CDCl3): d (ppm) = 153.2, 141.5, 137.2, 132.8, 130.3, 128.6, 128.0,
127.4, 127.3, 124.9, 121.5, 120.6. ESI-MS m/z (%): [M+H]+ 857.3.
Elemental analysis calcd. for C56H36N6Zn: C 78.36, H 4.23, N 9.79;
found: C 77.53, H 4.95, N 9.53.
Cobalt-di-aza-bis-(phenyl)-diisoindolmethene (2a)
Compound 1 (1.00 g, 2.52 mmol) and 1.3 e.q. cobalt-(II)-acetate
(0.54 g, 3.09 mol) is refluxed n-butanol (20 mL) with 2 e.q.
diispropylethylamine (DIPEA) (0.7 mL, 4.95 mmol) until the
product starts to precipitate from solution. The cold reaction
mixture is filtered and the so obtained product is re-crystallized
from dichloromethane/hexane (4 : 1), yielding the zinc-di-aza-
bis-(phenyl)-diisoindolmethene as yellow needles (0.40 g, 37%),
m.p. 383 ◦C, (dichloromethane/hexane 4 : 1). 1H NMR: not
possible. 13C NMR: not possible. ESI-MS m/z (%): [M+H]+ 852.3.
Elemental analysis calcd. for C56H36N6Co: C 78.96, H 4.26, N 9.87;
found: C 78.84, H 4.17, N 9.75.
Mercury-di-aza-bis-(phenyl)-diisoindolmethene (2e)
Compound 1 (1.00 g, 2.52 mmol) and 1.3 e.q. mercury-(II)-
acetate (1.08 g, 3.41 mmol) is refluxed THF (10 mL) with 3
e.q. DIPEA for 30 min. The solution is refluxed until no starting
material is detectable with TLC. The reaction mixture is cooled
to room temperature and the solid filtered. Re-crystallization
from dichloromethane/hexane (4 : 1) gave the copper-di-aza-bis-
(phenyl)-diisoindolmethene as violet plates (0.66 g, 53%), m.p.
286 ◦C, (dichloromethane/hexane 4 : 1). 1H NMR (500 MHz,
CDCl3): d (ppm) = 8.10 (d, J = 7.92 Hz, 2H), 7.59 (d, J =
7.90 Hz, 2H), 7.45 (t, J = 7.98 Hz, J = 7.04 Hz, 2H), 7.42 (d,
J = 7.92 Hz, 4H), 7.26 (t, H, J = 6.99 Hz, 2H), 6.98 (m, 6H). 13
C
Nickel-di-aza-bis-(phenyl)-diisoindolmethene (2b)
NMR (126 MHz, CDCl3): d (ppm) = 153.8, 140.9, 137.8, 133.6,
130.2, 129.0, 128.4, 127.3, 124.8, 121.5, 120.5. ESI-MS m/z (%):
[M+H]+ 993.3. Elemental analysis calcd. for C56H36N6Hg: C 67.70,
H 3.65, N 8.46; found: C 67.24, H 3.36, N 8.92.
Compound 1 (1.00 g, 2.52 mmol) and 1 e.q. nickel-(II)-acetate
(0.44 g, 2.52 mol) is refluxed n-butanol (20 mL) with 2 e.q. DIPEA
(0.7 mL, 4.95 mmol) until the product starts to precipitate from
solution. The cold reaction mixture is filtered and the so obtained
product is re-crystallized from dichloromethane/hexane (4 : 1),
yielding the zinc-di-aza-bis-(phenyl)-diisoindolmethene as yellow
needles (0.58 g, 57%), m.p. 378 ◦C, (dichloromethane/hexane
Acknowledgements
The authors gratefully acknowledge the financial support by the
Federal Ministry for Education and Research (BMBF) within the
Innoprofile program (03IP602) and the computational resources
from the Center for Information Services and High Performance
Computing (ZIH), TU Dresden. For the thermal analysis, the
authors would like to thank Prof. Dr Peer Schmidt and Hannelore
Dallmann.
1
4 : 1). H NMR (500 MHz, CDCl3): d (ppm) = 40.55 (s, 4H),
13.75 (s, 2H), 6.71 (s, 2H), 6.42,(s, 4H), 5.49 (s, 2H), 3.93 (s, 2H),
-1.96,(s, 2H); 13C: not possible due to low solubility. ESI-MS m/z
(%): [M+H]+ 851.2. Elemental analysis calcd. for C56H36N6Ni: C
78.98, H 4.26, N 9.87. Elemental analysis found: C 78.94, H 4.29,
N 9.49.
Copper-di-aza-bis-(phenyl)-diisoindolmethene (2c)
Notes and references
Compound 1 (1.00 g, 2.52 mmol) and 1.3 e.q. copper-(II)-
acetate (1.08 g, 3.18 mmol) is refluxed THF (10 mL) with 3
e.q. DIPEA for 30 min. The solution is refluxed until no starting
material is detectable with TLC. The reaction mixture is cooled
to room temperature and the solid filtered. Re-crystallization
from dichloromethane/hexane (4 : 1) gave the copper-di-aza-bis-
(phe◦nyl)-diisoindolmethene as violet plates (0.72 g, 66%), m.p.
13C NMR: not possible. ESI-MS m/z (%): [M+H]+ 856.3. Elemen-
tal analysis calcd. for C56H36N6Cu: C 78.53, H 4.24, N 9.81; found:
C 78.41, H 4.41, N 9.79.
1 (a) R. Ziessel, G. Ullrich and A. Harriman, New J. Chem., 2007, 31, 496–
501; (b) A. Loudet and K. Burgess, Chem. Rev., 2007, 107, 4891–4932;
(c) T. E Wood and A. Thompson, Chem. Rev., 2007, 107, 1831–1861.
2 (a) F. Bergstrom, I. Mikhalyo and P. Hagglof et al., J. Am. Chem. Soc.,
2002, 124(2), 196–204; (b) S. Diring, F. Puntoriero and F. Nastasi et al.,
J. Am. Chem. Soc., 2009, 131(17), 6108; (c) A. Harriman, L. J. Mallon
and K. J. Elliot et al., J. Am. Chem. Soc., 2009, 131(37), 13375–13386.
3 A. Hepp, G. Ulrich, R. Schmechel, H. von Seggern and R. Ziessel,
Synth. Met., 2004, 146(1), 11–15.
4 (a) T. Rousseau, A. Cravino, T. Bura, G. Ulrich, R. Ziessel and J. Roncal,
Chem. Commun., 2009, 1673–1675; (b) T. Rousseau, A. Cravino and T.
Bura et al., J. Mater. Chem., 2009, 19(16), 2298–2300.
5 (a) T. P. Ivanov and A. Draganov, Mh. Chem., 1968, 99, 1990–2003;
(b) JP 11092479.
1
344 C, (dichloromethane/hexane 4 : 1). H NMR: not possible.
6 (a) S. R. Halper and S. M. Cohen, Inorg. Chem., 2005, 44, 486; (b) S. R.
Halper, L. Do, J. R. Stork and S. M. Cohen, J. Am. Chem. Soc., 2006,
128, 15255; (c) L. Yu, K. Muthukumaran, I. V. Sazanovich, C. Kirmaier,
E. Hindin, J. R. Diers, P. D. Boyle, D. F. Bocian, D. Holten and J. S.
Lindsey, Inorg. Chem., 2003, 42, 6629.
7 R. Gresser, M. Hummert, H. Hartmann, K. Leo and M. Riede, Chem.
Eur. J., DOI: 10.1002/chem.201002941.
8 H. Bredereck and H. W. Vollmann, Chem. Ber., 1972, 105, 2271–2283.
9 V. F. Donyagina, S. Shimizu, N. Kobayashi and E. A. Lukyanets,
Tetrahedron Lett., 2008, 49, 6152–6154.
10 A. Palma, J. F. Gallagher, H. Mu¨ller-Bunz, J. Wolowska, E. J. L.
McInnes and D. F. O’Shea, Dalton Trans., 2009, 273–279.
Zinc-di-aza-bis-(phenyl)-diisoindolmethene (2d)
Compound 1 (1.00 g, 2.52 mmol) and 1.3 e.q. zinc-(II)-acetate
(0.60 g, 3.28 mol) is refluxed n-butanol (20 mL) with 2 e.q.
diispropylethylamine (DIPEA) (0.7 mL, 4.95 mmol) until the
product starts to precipitate from solution. The cold reaction
mixture is filtered and the so obtained product is re-crystallized
from dichloromethane/hexane (4 : 1), yielding the zinc-di-aza-bis-
(phenyl)-diisoindolmethene as yellow needles (0.91 g, 84%), m.p.
370 ◦C, (dichloromethane/hexane 4 : 1). 1H NMR (500 MHz,
CDCl3): d (ppm) = 8.07 (d, J = 7.88 Hz, 2H), 7.58 (d, J = 7.88 Hz,
2H), 7.40 (t, J = 7.88 Hz, 6H), 7.20 (t, J = 7.88 Hz, 2H), 6.93 (t,
11 (a) E. Szajna, P. Dobrowolski, A. L. Fuller, A. M. Arif and L. M.
Berreau, Inorg. Chem., 2004, 43, 3988; (b) L. Do, S. R. Halper and S. M.
Cohen, Chem. Commun., 2004, 2662; (c) L. Yu, K. Muthukumaran,
I. V. Sazanovich, C. Kirmaier, E. Hindin, J. R. Diers, P. D. Boyle,
3482 | Dalton Trans., 2011, 40, 3476–3483
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
D. F. Bocian, D. Holten and J. S. Lindsey, Inorg. Chem., 2003, 42,
6629.
12 Thomas S. Teets, David V. Partyka, James B. Updegraff
III and Thomas G. Gray, Inorg. Chem., 2008, 47, 2338–
2346.
13 The TD-DFT approach has proven to be superior compared to
semi-empirical methods; see J. Fabian, Dyes Pigm., 2010, 84(1), 36–
53.
(b) See the ESI,† for computational details, cartesian coordinates of
the optimized structures, and frontier orbital visualizations for 1.
15 In order to obtain the most accurate results, benchmark calculations
were performed according to M. Bu¨hl and H. Kabrede, J. Chem. Theory
Comput., 2006, 2, 1282–1290It turned out that the b3lyp functional gave
the best results for the prediction of the absorption energies compared
to the bp86, pbe and the tpss functionals which gave deviations of over
0.6 eV.
14 (a) Gaussian 03, Revision E.01, M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A.
Montgomery Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam,
S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota,
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross,
V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y.
Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G.
Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas,
D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V.
Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov,
G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J.
Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M.
Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong,
C. Gonzalez, and J. A. Pople, Gaussian Inc., Wallingford CT, 2004;
16 J. Heinze, Angew. Chem., 1984, 96, 823–840.
17 By using the half wave oxidation and reduction potentials of the com-
pounds, 2a–e, the HOMO and LUMO energies have been calculated
with the potential of Fc/Fc+ as reference energy of EHOMO = -4.78 eV
from N. G. Connelly and W. E. Geigerm, Chem. Rev., 1996, 96, 877–910.
18 (a) A. F. Holleman, N. Wiberg, in Lehrbuch der Anorganischen Chemie,
de Gruyter, Berlin, 1995, vol. 101, pp. 1202–1204; (b) A. F. Holleman,
N. Wiberg, in Lehrbuch der Anorganischen Chemie, de Gruyter, Berlin,
1995, vol. 101, pp. XXXVIII.
19 (a) R. G. Pearson, J. Am. Chem. Soc., 1963, 85(22), 3533–3539; (b) R. G.
Pearson, J. Chem. Educ., 1968, 45, 581–586; (c) R. G. Pearson, J. Chem.
Educ., 1968, 45, 643–648.
20 A. F. Holleman, N. Wiberg, in Lehrbuch der Anorganischen Chemie, de
Gruyter, Berlin, 1995, vol. 101, p. 1213.
21 A comparison of the thermal stability in the lower temperature regime
between the Hg complex, the free ligand and the corresponding boron
difluoride compound can be found in the ESI†.
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 3476–3483 | 3483
©