C. Lambert, M. Kaupp et al.
of 125000mꢀ1 cmꢀ1. This difference could be due to the
charge separation that occurs in TA2 + + and TACN1+ + that
leads to the two amine centres being oxidised, whereas this
cannot take place in TA1 + + in which the single amine and
the squaraine are oxidised.
In the case of TA3, we can only compare the monocation
with the other compounds because for reasons discussed
above, dication formation is rapidly followed by trication
formation. The trication is then the only species present at
the given electrode potential. In fact, careful inspection of
the spectro-electrochemistry data did not give any evidence
for the presence of a sizeable TA3 + + concentration during
stepwise oxidation of TA3 + to TA3 + + +. For the trication
TA3 + + +, two strong absorption bands are found at 9700
and 13300 cmꢀ1 with extinction coefficients of just below
100000 and 144000mꢀ1 cmꢀ1, respectively.
Figure 7. Molecular orbitals of the neutral and charged species TA1-3
and TACN1. The HOMO, LUMO and so forth descriptions are only
valid for the neutral species: xX=compound; the subscripts define the
spin multiplicity; os=open shell.
Even though we obtained characteristic absorption spec-
tra of the oxidised species, the interpretation of the ob-
served bands remained difficult. Therefore we performed
DFT and time-dependent (TD)-DFT calculations (see com-
putational details) for the neutral systems, the monocations
and the dications (and the trication of TA3) to compare the
calculated with experimental excitation energies and to
reach a conclusion about the electronic structure in case of
favourable agreement. Several questions such as the locali-
sation/delocalisation of positive charges or the spin multi-
plicity of the doubly charged species shall be addressed. In
the latter case, both the closed- and open-shell singlet (treat-
ed as broken-symmetry solution) as well as the triplet spe-
cies were evaluated to assess which one is in fact generated
in the spectro-electrochemistry experiment. In the case of
TA3, we calculated the doublet and quartet species of the
trication (the doublet again as broken-symmetry state).
These calculations provide dipole moments, excitation ener-
gies, the corresponding transition moments and oscillator
strengths as well as the contributions of various orbitals to
the excitations.
The TD-DFT-B3LYP/SVP computed data are presented
as stick spectra together with the experimental spectra. In
these diagrams, we normalised the strongest measured ab-
sorption band to the oscillator strength of the corresponding
computed transition. The spectra of TA1, TA2, TACN1 and
TA3 are shown in Figures 10–13 (below), respectively, and
an overview of some of the relevant orbitals is given in
Figure 7. For easier comparison, we retain the assignment of
the HOMOs and LUMOs of the neutral molecules, even for
the charged species and even though it is incorrect. The
spectra of SQ1 and CN1 are shown in the Supporting Infor-
mation (Figures S2 and S3). In the latter two cases, theory
and experiment agree very nicely for the neutral compounds
and the cations, but not at all for the dications. However, for
both dications the calculations reveal that the singlet is 80–
90 kJmolꢀ1 lower in energy than the triplet. This is due to
the highly delocalised charge in these two molecules relative
to the other molecules. Therefore, the second oxidation
takes place in the same orbital without charge separation
and thus favours the singlet state.
For all neutral compounds, the computed excitation ener-
gies are in excellent agreement with experiment for the
lowest-energy absorption. This transition is in all cases due
to a HOMO!LUMO excitation. The transition at higher
energy is also at the appropriate energy but its oscillator
strength comes out too large relative to the first excitation.
For the monocations, the agreement with experiment is also
very good, but excitation energies for the lowest-energy
transitions are somewhat too low. These transitions at
lowest energy (measured around 6000 cmꢀ1) are due to
HOMOꢀ1!HOMO excitations, whereas the one at next
higher energy is mainly due to a HOMO!LUMO excita-
tion, slightly redshifted relative to the equivalent transition
in the neutral chromophores (for TACN1+ there is a transi-
tion in between, which will be discussed below). The striking
similarity of this transition to that of the neutral compounds
is particularly apparent for TA1 +, TA2 + and TACN1+. For
the monocations, the computed oscillator strength of the
lowest-energy transition comes out too large on a relative
level, especially for TA2+ and TACN1+. Inspection of the
orbitals that are involved in these transitions shows that the
monocations are delocalised, and the lowest-energy transi-
tion is due to a degenerate charge transfer from the triaryl-
AHCTUNGTRENNUNG
amine to the squaraine bridge in the case of TA2 + and
TACN1+. In this respect, “degenerate” refers to diabatic ex-
cited states that may be defined as having one positive
charge on either of the two amine moieties. These two
states mix and yield two non-degenerate adiabatic states,
which can both be observed separately in TACN1+: when
comparing the lowest-energy absorption bands of TA2+ and
TACN1+, we see that there is only one for TA2 +
(6200 cmꢀ1) but two for TACN1+ + (5800 and 8300 cmꢀ1).
The latter corresponds to the HOMOꢀ2!HOMO transi-
tion. The analogous transition has no oscillator strength in
TA2 + due to symmetry reasons: squaraine TA2 + has Ci
symmetry. The appropriate orbitals are of au (HOMOꢀ2), ag
(HOMOꢀ1) and again au (HOMO) symmetry (Figure 8).
For a transition to be allowed, the direct product of the in-
volved MOs and of the transition dipole moment vector has
to include the totally symmetric descriptor ag. This holds for
14156
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 14147 – 14163