Charge-transfer interactions in tris-donor-tris-acceptor hexaarylbenzene redox chromophores

Article abstract of DOI:10.1002/chem.201104020

Symmetric- and asymmetric hexaarylbenzenes (HABs), each substituted with three electron-donor triarylamine redox centers and three electron-acceptor triarylborane redox centers, were synthesized by cobalt-catalyzed cyclotrimerization, thereby forming compounds with six- and four donor-acceptor interactions, respectively. The electrochemical- and photophysical properties of these systems were investigated by cyclovoltammetry (CV), as well as by absorption- and fluorescence spectroscopy, and compared to a HAB that only contained one neighboring donor-acceptor pair. CV measurements of the asymmetric HAB show three oxidation peaks and three reduction peaks, whose peak-separation is greatly influenced by the conducting salt, owing to ion-pairing and shielding effects. Consequently, the peak-separations cannot be interpreted in terms of the electronic couplings in the generated mixed-valence species. Transient-absorption spectra, fluorescence-solvatochromism, and absorption spectra show that charge-transfer states from the amine- to the boron centers are generated after optical excitation. The electronic donor-acceptor interactions are weak because the charge transfer has to occur predominantly through space. Moreover, the excitation energy of the localized excited charge-transfer states can be redistributed between the aryl substituents of these multidimensional chromophores within the fluorescence lifetime (about 60a ns). This result was confirmed by steady-state fluorescence-anisotropy measurements, which further indicated symmetry-breaking in the superficially symmetric HAB. Adding fluoride ions causes the boron centers to lose their accepting ability owing to complexation. Consequently, the charge-transfer character in the donor-acceptor chromophores vanishes, as observed in both the absorption- and fluorescence spectra. However, the ability of the boron center as a fluoride sensor is strongly influenced by the moisture content of the solvent, possibly owing to the formation of hydrogen-bonding interactions between water molecules and the fluoride anions. Copyright

Full text of DOI:10.1002/chem.201104020

FULL PAPER
DOI: 10.1002/chem.201104020
Charge-Transfer Interactions in Tris-Donor–Tris-Acceptor Hexaarylbenzene
Redox Chromophores
Markus Steeger and Christoph Lambert*[a]
Abstract: Symmetric- and asymmetric
hexaarylbenzenes (HABs), each substi-
tuted with three electron-donor triaryl-
amine redox centers and three elec-
tron-acceptor triarylborane redox cen-
ters, were synthesized by cobalt-cata-
lyzed cyclotrimerization, thereby form-
ing compounds with six- and four
donor–acceptor interactions, respec-
tively. The electrochemical- and photo-
physical properties of these systems
were investigated by cyclovoltammetry
(CV), as well as by absorption- and flu-
orescence spectroscopy, and compared
to a HAB that only contained one
neighboring donor–acceptor pair. CV
measurements of the asymmetric HAB
show three oxidation peaks and three
reduction peaks, whose peak-separa-
tion is greatly influenced by the con-
ducting salt, owing to ion-pairing and
shielding effects. Consequently, the
peak-separations cannot be interpreted
in terms of the electronic couplings in
the generated mixed-valence species.
Transient-absorption spectra, fluores-
cence-solvatochromism, and absorption
spectra show that charge-transfer states
from the amine- to the boron centers
are generated after optical excitation.
The electronic donor–acceptor interac-
tions are weak because the charge
transfer has to occur predominantly
through space. Moreover, the excita-
tion energy of the localized excited
charge-transfer states can be redistrib-
uted between the aryl substituents of
these multidimensional chromophores
within the fluorescence lifetime (about
60 ns). This result was confirmed by
steady-state
fluorescence-anisotropy
measurements, which further indicated
symmetry-breaking in the superficially
symmetric HAB. Adding fluoride ions
causes the boron centers to lose their
accepting ability owing to complexa-
tion. Consequently, the charge-transfer
character in the donor–acceptor chro-
mophores vanishes, as observed in both
the absorption- and fluorescence spec-
tra. However, the ability of the boron
center as a fluoride sensor is strongly
influenced by the moisture content of
the solvent, possibly owing to the for-
mation of hydrogen-bonding interac-
tions between water molecules and the
fluoride anions.
Keywords: boranes · charge trans-
fer · chromophores · donor–accept-
or systems · symmetry
Introduction
are twisted relative to the central benzene ring. The twist
angle depends on the steric hindrance of the p systems.[1]
Owing to this geometric arrangement, HABs are suitable
models for investigating transition dipole–dipole interactions
between chromophores in the weak interaction regime.[2] On
the basis of such HAB scaffolds, Gust and co-workers de-
signed an artificial antenna–reaction-center complex[3] that
mimics the ring-like assembly of the chromophores that are
found in natural antenna systems LH1 and LH2 in light-har-
vesting purple bacteria.[4] Moreover, a molecular photo-
chemical transistor has also been synthesized and investigat-
ed by using HAB.[5] Furthermore, we have used a pyrene-
substituted HAB to investigate exciton-coupling effects.[6]
Besides through-space interactions, the propeller arrange-
ment of HAB p-systems might enable an electronic coupling
of redox centers by direct overlap of the pz orbitals of the
ipso-C atoms that are directly bound to the central benzene
ring. Whether (and to what degree) through-space- or direct
orbital interactions dominate the electronic communication
between the chromophore branches depends on the system.
For this reason, various HABs that are functionalized with a
variety of redox centers have been synthesized and their
electronic properties have been examined to study these in-
teractions.[1,7] In addition to their aforementioned electronic
Herein, we are focusing on the fundamental principles of
energy- and charge transfer in multidimensional chromo-
phores by considering substituted hexaarylbenzenes (HABs)
1 and 2. In these covalently linked chromophore aggregates,
triarylborane and triarylamine act as functional chromo-
phore units that are arranged in the 1,3,5- (1) and 1,2,4-posi-
tions (2) relative to the central benzene ring. Because these
functional units are linked in this HAB framework through
a central benzene ring, their molecular axes form an angle
of 608 relative to each other. Moreover, they are arranged in
a propeller-like manner because the six functional p-systems
[a] M. Steeger, Prof. C. Lambert
Institut fꢀr Organische Chemie
Universitꢁt Wꢀrzburg and Wilhelm–Conrad–Rçntgen Research
Center
for Complex Material Systems
Am Hubland, 97074 Wꢀrzburg (Germany)
Fax : (+49)931-31-87218
E-mail: christoph.lambert@uni-wuerzburg.de
Supporting information for this article, including all synthetic proce-
dures, is available on the WWW under http://dx.doi.org/10.1002/
chem.201104020.[a] Excitation at 340 nm; [b] Excitation at 416 nm.
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

properties, HABs are used as
proton conductors[8] and as
precursors for molecular gra-
phenes.[9]
In the molecules studied
herein, we use optically in-
duced charge-transfer (CT)
processes to investigate the
energy-transfer pathways be-
tween the CT states by per-
forming steady-state anisotro-
py
fluorescence
measure-
ments.[10] These pathways are
interpreted and compared
more easily for transitions with
known transition-moment di-
rections, such as CT states in
which the donor–acceptor
vector is parallel to the transi-
tion-moment direction. There-
fore, we combined three triar-
ylamine electron donors[11] and
three triarylborane electron
acceptors in the HAB frame-
Figure 1. Possible through-space CT interaction pathways (red arrows) in HABs 1–3 and through-bond interac-
tions in previously investigated carbazole derivative A.
work. Because of their electron-donor properties and their
very low internal reorganization energy, triarylamines have
been widely used as hole-transport materials in optoelec-
tronic applications.[12] On the other hand, owing to the unoc-
cupied boron pz orbital, the acceptor properties of triaryl-
boranes have been used in optoelectronic devices as elec-
tron-transporting and/or -emitting layer,[13] as well as in stud-
ies of their linear-[10,14] and nonlinear optical proper-
ties[14f,h,15] and in sensor systems for fluoride[14k,16] and
cyanide anions.[17] Because of these favorable characteristics,
we have incorporated triarylboranes into polymeric structur-
es[14e] and into a multidimensional chromophore (A),[10,18] in
combination with carbazole as the electron-donating redox
center in both cases. In case of the multidimensional chro-
mophore A, the excitation energy of the optically induced
CT states is redistributed between the different chromo-
phore branches of the molecule; this property initiated the
work presented herein, in which HABs with through-space
CT transitions are in focus.
which can then be used for ion-sensing through colorimetric,
fluorescence, or electrochemical responses.[16b] Typically, the
binding constants of triorganoboranes depend on the Lewis
acidity of the boron center and on the steric hindrance of
the substituents. It has been shown that their Lewis acidity
can be amplified by linking multiple trivalent boron centers
together.[14j] With that, the binding constant of the first com-
plexation is enhanced whereas further binding constants are
characterized by a negative binding cooperativity, which can
be explained by a gradual loss of Lewis acidity. The question
that arises is if the triarylborane moieties in HABs 1 and 2
are subject to the same effects. A further point concerns the
redox properties of HABs 1 and 2. Because both of them
contain reducible- and oxidizable redox centers, they may
form different mixed-valence compounds,[20] that is, species
that contain both oxidized- and neutral (or reduced and
neutral) redox centers of the same chemical nature. Thus,
we also investigated HABs 1–3 by using electrochemical
methods.
We synthesized HABs 1 and 2, which each consist of
three triarylboranes and three triarylamines, as well as the
corresponding model compound (3), which only contains
one donor–acceptor pair (Figure 1). Because of their general
susceptibility to hydrolysis, the boron centers have to be ki-
netically shielded by sterically demanding methyl groups at
the ortho position of the aryl rings.[15f,19] Compounds 1–3
were all characterized by using steady-state- and time-re-
solved absorption- and emission spectroscopy, as well as by
fluorescence-anisotropy measurements, to investigate the
energy-transfer process. Furthermore, we performed fluo-
ride-titration experiments with HABs 1–3. Because of the
very small size of the fluoride and cyanide anions, they can
selectively form complexes with the trivalent boron centers,
Results and Discussion
Synthesis: HABs 1 and 2 were easily synthesized by cobalt-
catalyzed cyclotrimerization of the appropriate tolan precur-
sor (4, Scheme 1).[21] Although complexes of other metals
are also known to catalyze this type of reaction, cobalt com-
plexes rank among the most effective.[22] Tolan derivative 4
was synthesized by lithiation of known triarylamine deriva-
tive 5,[12a] followed by reaction with Mes2BF.
The cyclotrimerization reaction was performed with
Co2(CO)8 as the catalyst, which yielded symmetric (1) and
asymmetric isomers (2). Because of the reaction mechanism,
&
2
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Charge-Transfer in Tris-Donor–Tris-Acceptor Chromophores
FULL PAPER
Scheme 2. Synthesis of model HAB 3: 1) Tetraphenylcyclopentadienone,
diphenyl ether, reflux; 2) tBuLi, THF, ꢀ788C; 3) Mes2BF, THF, ꢀ788C.
for the electrochemical measurements but we were able to
measure the oxidation- and reduction potentials of HABs 2
and 3 by CV (Table 1).
To measure the very negative reduction potentials of the
Table 1. Redox potentials of HABs 2 and 3 (versus Fc/Fc< M+ >).[a]
2[b]
3[c]
Eox [mV]
E1red [mV]
E2red [mV]
E3red [mV]
316
300[d]
ꢀ2453
ꢀ2586
ꢀ2726
ꢀ2517
–
–
[a] Compound 1 is insufficiently soluble in THF for the electrochemical
measurements; [b] 0.65m THF/[Bu4N][ClO4]; [c] 0.5m THF/[Bu4N]-
[ClO4]; [d] determined by DPV.
G
E
ACHTUNGTRENNUNG
Scheme 1. Synthesis of HABs 1 and 2: 1) tBuLi, THF, ꢀ788C; 2) Mes2BF,
THF, ꢀ788C; 3) Co2(CO)8, 1,4-dioxane, reflux.
AHCTUNGTRENNUNG
boron redox centers, our conventional electrochemical cell
had to be upgraded by adding an intrinsic aluminum-oxide
column, as recommended by Kiesele,[24] to further dry the
electrolyte. Thus, the measureable potential window could
be expanded from ꢀ3.1 to ꢀ3.4 V in THF. The best results
were obtained when the electrolyte and the added substance
was dried over the aluminum-oxide column until all of the
peaks were well-separated from the accessible potential
limit of the solvent. This procedure requires substances that
do not degrade on the column. Unfortunately, compound 3
is not stable under these conditions and, in particular, the
oxidation was not well-resolved in the CV. Thus, we applied
differential pulse techniques (DPV), which compensate
better for electrochemical processes of the solvent. A one-
electron oxidation of the triarylamine and a reversible one-
electron reduction of the triarylborane moiety of model
compound 3 could be observed at 300 mV and ꢀ2517 mV,
respectively (Table 1). These values are in the typical range
for this type of donor–acceptor compounds.[25] The CV of
compound 2 shows three reversible reduction processes and
one reversible oxidation process (Figure 2) at around the
same potential as those in model compound 3. Unfortunate-
which proceeds through metallacyclopentadiene intermedi-
ates,[23] two different isomers with chromophore substituents
at the 1,3,5- and 1,2,4-positions are formed. Yields of 17%
(for 1) and 62% (for 2) are in agreement with the statistical-
ly expected product ratio of 1:3.[22] Because the ratio of the
products of such reactions are influenced by substituents
that differ significantly in size, the steric hindrance of the
two different chromophores can be assumed to be nearly
identical.[22]
The synthesis of model compound 3 was also based on tri-
arylamine derivative 5 (Scheme 2), which was transformed
into HAB 6 by using a Diels–Alder reaction with tetraphe-
nylcyclopentadienone. The boron acceptor center was con-
structed in the final step, by using the same procedure to
that mentioned above, because it decomposed under high-
temperature Diels–Alder reaction conditions.
Electrochemistry: HABs 1 and 2 each possess three electron
acceptors and three electron donors, which should lead to,
in total, seven accessible redox states: ꢀ3, ꢀ2, ꢀ1, 0, +1,
+2, and +3. Unfortunately, HAB 1 was not soluble enough
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

M. Steeger and C. Lambert
For better comparison, all of the redox peaks in the DPV
were fitted with Voigt functions, with the exception of the
oxidation peak in THF/ACHTNURTGENNUG[Bu4N]ACHTUGNTREN[NUNG ClO4], in which the poten-
tials are too close to get a stable fit. In this case, the use of
three Voigt functions yields the two outermost oxidation
processes with a maximum peak-separation of 60 mV, whilst
the third in-between process is fitted equally well for varia-
ble potential values. In contrast, the two outermost oxida-
tion processes are separated by 135 mV in THF/
[B(C6F5)4]. For both electrolyte salts, three reduction peaks
are visible in the DPVs with a somewhat larger outermost
peak-separation of 264 mV in THF/[Bu4N][ClO4] compared
to that in THF/[Bu4N][B(C6F5)4] (196 mV). Both phenom-
ACHTUNGTRENNUNG[Bu4N]-
AHCTUNGTRENNUNG
N
ACHTUNGTRENNUNG
G
ACHTUNGTRENNUNG
[ClO4], n=100 mVsꢀ1
Figure 2. CV of HAB 2 in 0.65m THF/ACHTUNRGTENNUG[Bu4N]ACHTUTGNRENNUNG .
ena are consistent with a decreased donor strength of the
borate counteranion compared to the perchlorate anion,
which leads to less ion-pair formation and poorer shielding
of the oxidized triarylamine moieties. As a result, the elec-
trostatic repulsion between a given oxidized amine and
other positive charges within the molecule is increased and,
consequently, the next oxidation is hampered. Furthermore,
ion-pair formation of the borate and the tetrabutylammon-
ium cation is also decreased, thereby resulting in increased
acceptor strength of the tetrabutylammonium cation. Fol-
lowing the same train of thought, the second- and third re-
ductions are facilitated and, hence, the peak-separation is
narrowed. Two further remarks have to be made: 1) The
redox potentials of the DPVs differ from those of the CVs
by up to 20 mV, which can be ascribed to slow interfacial
electron-transfer kinetics in the DPV.[27] However, because
all of the potentials are shifted in the same direction, the
above-mentioned differences in the peak-separations are
ly, as mentioned above, the solubility of compound 1 in
THF is insufficient for electrochemical studies.
Each reduction wave in the CV of compound 2 corre-
sponds to the reduction of one boron center, whereas the
oxidation of the three nitrogen centers occurs at almost the
same potential. Because the triarylamine- and triarylborane
centers have almost the same linkage and distance relative
to each other, the different redox behaviors of the amine-
and borane centers is mainly ascribed to the effects of the
electrolyte. Barriere and Geiger showed that the donor/ac-
ceptor number and polarity of the solvent, as well as the
ion-pairing ability of the conducting salt, may have a strong
influence on the separation of the potentials in multistep
redox processes.[26] To investigate these effects in HAB 2,
we replaced the tetrabutylammonium perchlorate conduct-
ing salt by [Bu4N][BACTHNUGTRNEUNG(C6F5)4], which has a very poorly coor-
dinating anion. Indeed, the use of the non-coordinating
anion leads to a small shift in the reduction potentials of
compound 2 to a less-negative value and a shift of the oxida-
tion potentials to less-positive values. Even more remarka-
ble is a change in the redox-potential splittings within the
oxidation- and reduction signals, as observed by differential
pulse voltammetry (DPV, Figure 3: the reduction signals are
not perturbed. 2) By using the [Bu4N][BACHTGNUTERNU(NG C6F5)4] conducting
salt, the reductions of the borane moieties become chemical-
ly irreversible, for unknown reasons.
These results nicely demonstrate that the redox-potential
splitting of multi-redox chromophores depends on the type
of electrolyte salt in the electrochemical experiment and
no definite conclusions about possible mixed-valence
interactions in the differently charged species can be drawn
from the redox-potential splitting. On the other hand, the
observation of discrete reduction waves and at-least-
closer together in [Bu4N][B
nals are now split (visible shoulder) in contrast to
THF/[Bu4N][ClO4].
ACHTNUGRTEN(NGNU C6F5)4], whilst the oxidation sig-
A
ACHTUNGTRENNUNG
partially resolved oxidation waves of compound
THF/[Bu4N][B(C6F5)4] shows that a series of mixed-valence
states (22ꢀ, 2ꢀ, 2+, and 22+) can be selectively generated.
2 in
A
ACHTUNGTRENNUNG
Absorption- and emission spectroscopy: The UV/Vis spectra
of compounds 1, 2, and 3 in CH2Cl2 are characterized by an
intense absorption band at 32000, 31700, and 31900 cmꢀ1, re-
spectively (Figure 4). This band can be attributed to overlap-
ꢀ
ping p p* excitations, which are localized at the triaryla-
mine- and triarylborane centers.[14d] As expected, the extinc-
tion coefficient of this band in model compound
3
(43000mꢀ1 cmꢀ1) is around three times weaker than those of
compounds 1 (116000mꢀ1 cmꢀ1) and 2 (129000mꢀ1 cmꢀ1) be-
cause it only contains one donor–acceptor pair instead of
three. Furthermore, in all three spectra, the lowest-energy
absorption band is observed as a weak shoulder at around
Figure 3. DPV of HAB 2 in 0.46m THF/
ACHTUGTNRENNGU[Bu4N]ACHTUGNTREN[NUGN ClO4] (blue) and in 0.36m
THF/[Bu4N][B(C6F5)4] (red) normalized to the integral of the reduction.
A
ACHTUNGTRENNUNG
&
4
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Charge-Transfer in Tris-Donor–Tris-Acceptor Chromophores
FULL PAPER
Table 2. Extinction coefficients and absorption- and emission maxima of
HABs 1–3 in solvents of different polarity.
Solvent
v˜abs [cmꢀ1] (e (ꢃ104) [mꢀ1 cmꢀ1])
v˜fl [cmꢀ1
]
1
2
3
1
2
3
[a]
[a]
MeCN
PhCN
–
32000 (12.7) 32300 (4.4)
–
15600 15600
31700 (11.0) 31300 (11.9) 31600 (4.4) 16500 16600 16600
CH2Cl2 32000 (11.6) 31700 (12.9) 31900 (4.3) 17800 17600 17500
THF
32100 (–)[a] 31800 (12.5) 32000 (4.8) 18300 18300 18200
32200 (–)[a] 31800 (12.3) 31900 (4.6) 19700 19700 19600
32200 (–)[a] 31800 (12.4) 31900 (4.8) 20500 20600 20600
32100 (–)[a] 31700 (13.2) 31700 (4.5) 22200 22000 22200
MTBE
nBu2O
C6H12
[a] Insufficient solubility.
3. Whilst this property would be no surprise for asymmetric
HAB 2, there would have to be breaking of the symmetry in
the first excited state of the formally symmetric HAB 1.
The fluorescence excitation spectra of compounds 1–3
agree very well with their absorption spectra in these sol-
vents, thus indicating a complete energy-transfer from
higher excited states to the lowest excited states. The fluo-
rescence quantum yields of the three compounds were de-
termined in PhCN (Table 3) and were in the expected range
for this type of donor–acceptor systems.[14d]
Figure 4. Absorption spectra of HABs 1 (blue), 2 (red), and 3 (green) in
CH2Cl2. The absorption spectrum of HAB 3 is multiplied by three.
27000 cmꢀ1. This band can be attributed to CT transitions
whose intensity varies with the number of possible transi-
tions. In symmetric HAB 1, there are, in total, six possible
CT interactions (Figure 1), whilst there are four possible CT
interactions in asymmetric trimer 2 and only one possible
CT interaction in model compound 3. Accordingly, the CT
absorption band of HAB 1 is the most intense, whilst that of
compound 3 is the least intense.
Table 3. Fluorescence quantum yields (ff), fluorescence lifetimes (tf),
rate constants for the fluorescence and non-radiative deactivation proc-
esses (kf and knr), and transition moments of the fluorescence (mfl) of
HABs 1–3 in PhCN.
The fluorescence spectra of all of the compounds show a
large positive solvatochromism, which supports the assign-
ment of the lowest excited state as a CT state (Figure 5,
ff[a]
tf [ns][b]
kf (ꢃ106) [sꢀ1
]
knr (ꢃ107) [sꢀ1
]
mfl [D]
1
2
3
0.30
0.36
0.35
58
59
63
5.2
6.1
5.6
1.2
1.1
1.0
1.11
1.25
1.18
Time-resolved fluorescence spectroscopy: The fluorescence
lifetimes of HABs 1–3 were obtained by measuring their flu-
orescence decays in PhCN by using the time-gated techni-
que (Table 3). The decays were single exponential in all
cases. With the quantum yields (ff) and fluorescence life-
times (tf) in hand, the rate constants for the fluorescence
(kf) and non-radiative deactivation processes (knr) were cal-
culated by using Equations (1) and (2).
Figure 5. Emission maxima of HABs 1 (red squares), 2 (green circles),
and 3 (blue triangles) in solvents of different polarity (from left to right:
C6H12, nBu2O, MTBE, THF, CH2Cl2, PhCN, MeCN). f(D)ꢀ0.5f(n2) is the
1 ꢀ ꢀf
ð1Þ
kf ¼
tf
solvent-polarity function, with f(D)=(Dꢀ1)/
(2D+1) and f(n2)=
ACHTUNGTRENNUNG
(n2ꢀ1)/
ACHTUNGTRENNUNG
(2n2+1). The molecules were excited at 29400 cmꢀ1. MTBE=
1 ꢀ ꢀ
f ZS > ð2Þ
methyl tert-butyl ether.
knr
¼
tf
Table 2; also see the Supporting Information, S1–S3). From
this observation, we conclude that the molecules have a
small- or vanishing dipole moment (for symmetric HAB 1)
in the ground state and a high dipole moment in the dipolar
excited CT.[10] Because the fluorescence solvatochromism of
all of the HABs is remarkably similar, it follows that the
lowest excited state of multichromophore systems 1 and 2
must be localized within one donor–acceptor pair of the
molecule, thus resembling the situation in model compound
Notably, the fluorescence lifetimes and quantum yields of
compounds 1–3 are very similar. Again, this result supports
the conclusion that, in all cases, the photophysically active
unit is a single donor–acceptor pair, which requires symme-
try-breaking in compound 1.
The lifetimes for all of the HABs (around 60 ns) are con-
siderable longer (by a factor of 3 to 20) than those of other
known triarylamine–triarylborane donor–acceptor com-
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

M. Steeger and C. Lambert
pounds.[10,14h,28] Consequently, both rate constants kf and knr
are significantly lower in HABs 1–3. The relatively small kf
values can be explained by the small fluorescence-transition
moments (which should be equal to the absorption-transi-
tion moments in the absence of structural changes in the ex-
cited state), which were estimated by using the Strickler–
Berg equation[29] [Eq. (3)].
R
3he0
9
ge
2 gg
v~ꢀ3Ifdv~
Ifdv~
m2fl ¼
kf
R
ð3Þ
16 ꢁ 106p3
nðn2 þ 2Þ
Here, ge and gg are the degree of degeneracy of the excited-
and ground states, n is the refractive index of the solvent,
and If is the fluorescence intensity at a given wavenumber.
The small transition moments are due to the essentially
through-space character of the CT process. This latter as-
signment is affirmed by structure-optimization of compound
1 (see the Supporting Information, S4), which shows very
small molecular-orbital coefficients at the central benzene
ring in the three highest-energy HOMOs and in the three
lowest-energy LUMOs.
Transient-absorption spectroscopy: Transient-absorption
spectra were recorded on the nanosecond timescale for
HABs 1–3 in PhCN (Figure 6) to characterize the excited
state in more detail. In general, the transient-absorption sig-
nals increase immediately with 416 nm laser-light excitation
(instrument response function: 8 ns) and decay to the
ground state with the same time constants (58 ns for HAB 1
and 59 ns for HABs 2 and 3, measured at the low-energy
transient-absorption-band maximum) as the fluorescence
lifetimes (see the Supporting Information, S5–S7). Conse-
quently, all of these transient signals originate from the first
excited singlet state and no intersystem crossing to any trip-
let states is observed. The transient spectra of compounds
1–3 all feature an absorption band at 13400, 13500, and
13400 cmꢀ1, respectively. These bands can be assigned to an
overlap of the SOMO–LUMO transitions of the triarylbor-
ane radical anion (the absorption maximum is expected at
14500 cmꢀ1)[30] and of the triarylamine radical-cation moiet-
ies (absorption maximum is typically located at
13000 cmꢀ1).[11] Furthermore, in all of the transient spectra, a
broad absorption band can be observed at the high-energy
end of the spectra, at around 26000 cmꢀ1, which is accompa-
nied by a shoulder at 22800–23400 cmꢀ1. The exact maxima
of these bands cannot be determined, but they can be attrib-
uted to HOMO–SOMO transitions of the boron radical
anion and the nitrogen radical cation. These assignments are
based on spectroelectrochemical studies on the 4,4’-bis(di-
mesitylboryl)biphenyl radical anion compared to the isoelec-
tronic tetramethylbenzidinium radical cation reported by
Kaim and co-workers.[30] The UV/Vis spectra of both species
show exceptionally similar band structures, albeit slightly
shifted. Although the maxima of the HOMO–SOMO transi-
tions of triarylboranes are expected to be in the near-UV
region, they can be shifted to lower energies (up to
Figure 6. Transient-absorption spectra of HABs 1–3 in PhCN; excitation
at 24000 cmꢀ1 (for the transient-absorption decay curves, see the Support-
ing Information, S5–S7). The spectra are displayed in 7 ns steps (blue to
orange) and 33 ns to 66 ns (red).
23500 cmꢀ1) by conjugation with p-acceptor centers. Like-
wise, this effect can be observed in triarylamines.[30] All of
this information supports the formation of a CT state be-
tween the triarylamine and triarylborane moieties with pro-
nounced radical-cation and radical-anion character in the
first excited state of HABs 1–3.
Steady-state fluorescence-anisotropy spectroscopy: To gain
further insight into the symmetry-breaking effects in HABs
1 and 2, we performed steady-state fluorescence-anisotropy
&
6
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Charge-Transfer in Tris-Donor–Tris-Acceptor Chromophores
FULL PAPER
measurements at room temperature in sucrose octaacetate.
The latter compound forms a rigid glass matrix upon heating
and subsequent cooling to room temperature, which is nec-
essary to avoid rotation of the chromophores during the flu-
orescence lifetime. By measuring the fluorescence anisotro-
py (r), as defined by Eq. (4), the relative orientation (b) of
the excitation- and fluorescence-transition dipole moments
can be studied.
value of ꢀ0.02 in the range of the second absorption band,
ꢀ
which is attributed to localized p p* transitions (see
above). The interpretation of this value is difficult because
several absorption bands are overlapping and their transi-
tion moments are not precisely known. In comparison, the
lower anisotropy of compounds 1 and 2 in the range of the
CT transitions is remarkable and can be explained by the
“red-edge excitation effect”.[32] This effect can be observed
in symmetry-broken or asymmetric chromophores, in which
energy-transfer between chromophoric branches occurs.[10]
Owing to symmetry-breaking in HAB 1, which is induced
by the inhomogeneous environment (solvent or matrix), and
to the inherently asymmetric HAB 2, the six- and four possi-
ble CT transitions in compound 1 and 2, respectively, differ
slightly in energy between each other. According to Kasha’s
rule, fluorescence will only be emitted from the lowest-
energy CT state. On the low-energy side of the CT absorp-
tion bands, only the lowest-energy CT state will be excited
and the transition moments for absorption and fluorescence
are parallel, thereby resulting in a high anisotropy value of
(ideally) 0.4. Excitation on the high-energy side of the CT
bands leads to energy-transfer to the lowest-energy CT
state, followed by fluorescence emission. Thus, the transition
moments for absorption and fluorescence are no longer par-
allel, which leads to lower anisotropies. Because of the addi-
tivity of anisotropies, one should ideally observe an aniso-
tropy value of 0.1 if there is rapid energy-transfer between
all of the CT states in two dimensions. Overlap of the CT
ꢀ
ꢁ
2
5
3 cos2 bꢀ1
ð4Þ
r ¼
2
With a parallel orientation (b=08) of the transition dipole
moments, a maximum anisotropy of r=0.4 should be ob-
served, whilst with a perpendicular orientation (b=908), a
minimum value of ꢀ0.2 should be obtained. In the special
cases of 3D- and 2D depolarization of the fluorescence
light, anisotropy values of 0 and 0.1 are expected, respec-
tively.[31] In particular, the anisotropy for 3D- and 2D depo-
larization is of great significance in HABs 1 and 2 because
excitation into one donor–acceptor CT state of the molecule
could be followed by energy-redistribution within all of the
possible donor–acceptor CT combinations of the chromo-
phore, thus leading to depolarization within two (or three-)
dimensions. The topology of the HAB framework and the
CT nature of the transition are ideal for this kind of study
because excitation- and fluorescence-transition moments of
the CT processes should be parallel and oriented in the
plane of the central benzene ring whereby the analysis of
the polarized steady-state fluorescence data is simplified.
The anisotropy curves for the excitation spectra of com-
pounds 1, 2, and 3 are shown in Figure 7.
ꢀ
bands with the more-intense p p* transitions may, in fact,
lead to an observed complete depolarization (rꢂ0) in that
spectral range. The different anisotropy behavior of the CT
band of HABs 1 and 2 on the one hand with that of HAB 3
on the other hand nicely demonstrates that energy-transfer
between the different CT states occurs in HABs 1 and 2
within the fluorescence lifetime.
The excitation anisotropy of model compound 3 has a
high value of 0.34 in the energy range of the lowest absorp-
tion band, the CT transition. This value is somewhat lower
than the expected r=0.4 for parallel transition moments,
which may be due to residual molecular motion in the su-
crose octaacetate matrix at room temperature. At higher en-
ergies, the anisotropy decreases and has a rather constant
Titration with fluoride: Fluoride-sensing is an important
issue in the detection of chemical weapons that are based
on phosphorous fluorides.[33] To probe the applicability of
HABs 2 and 3 for fluoride-sensing (HAB 1 is not soluble
enough in THF), titration experiments were performed in
THF by using TBAF as the anion source. Complexation of
the triarylborane centers with fluoride anions was observed
by both UV/Vis- and fluorescence spectroscopy. On com-
plexation of the fluoride anion to the unoccupied pz orbital,
the boron center loses its character as an acceptor[16b] and
the CT states and their associated transitions vanish in the
absorption- and fluorescence spectra. In the absorption
spectra of HABs 2 and 3 (Figure 8), fluoride-titration leads
ꢀ
to a decrease in the intensities of the p p* transitions at
about 33000 cmꢀ1 and of the CT shoulder at 27000 cmꢀ1.
Likewise, upon the addition of fluoride to HABs 2 and 3,
the charge-transfer fluorescence at about 18000 cmꢀ1 de-
creases and a new band appears at about 23000 cmꢀ1.
The latter band at 23000 cmꢀ1 is due to localized fluores-
cence from the triarylamine chromophores. This high-energy
fluorescence is more intense than the CT fluorescence in
Figure 7. Excitation anisotropy spectra in sucrose octaacetate (triangles)
and normalized absorption spectra in nBu2O (solid lines) of HABs 1
(blue), 2 (green), and 3 (orange).
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

M. Steeger and C. Lambert
were employed in the fluores-
cence experiments, whereas
absorption measurements at
much-higher
chromophore
concentrations (ꢃ10ꢀ5 m) are
affected much less. Thus, this
“deactivation” of fluoride can
neither be caused by impurities
in the analyte nor in the chro-
mophore. However, it is possi-
ble that residual moisture in
the THF solvent affects the
measurements by hydrogen
bonding to the fluoride anions.
By analyzing the absorption
measurements with the global-
fit routine of the HypSpec soft-
ware package (Figure 9), the
complexation constant of HAB
3 was determined to be 1.5ꢃ
105. The fluorescence titration
could be fitted by both Hyp-
Spec software and by a simple
mathematical treatment that
involved coupled chemical
equilibria (see the Supporting
Information, Figures S8 and
S9), from which complexation
constants of 3.6ꢃ105 and 5ꢃ
105 could be obtained, respec-
tively. Therefore, the deactiva-
tion process is described by a
1:1 complex of fluoride with a
Figure 8. UV/Vis (left) and fluorescence (right) spectra of HABs 2 (up) and 3 (down) in THF upon titration
with TBAF under ambient conditions. The number of equivalents of TBAF added is shown in the legend. The
UV/Vis spectra were measured at concentrations of 1ꢃ10ꢀ5 m (2) and 4ꢃ10ꢀ5 m (3); the fluorescence spectra
were measured at a concentration of 5ꢃ10ꢀ7 m (2 and 3).
model HAB 3, whilst it is much weaker in HAB 2. For a
quantitative analysis of the spectroscopic changes, it is im-
portant to account for the decreasing absorption in the
second unknown substrate. Both the substrate and the “de-
activating” complex do not interact with light over the given
range of the spectrum. The complexation constants that are
obtained in this way are similar to those that have been re-
ported previously for other triarylboranes that were con-
ꢀ
range of the CT- and p p* transitions from 23000–
37000 cmꢀ1, where the chromophores were excited (lex =
340 nm). However, the quantitative analysis of the fluores-
cence titration experiments in THF at ambient conditions is
further complicated by the instability of the complexed spe-
cies, which is noticeable by a slow decrease in its fluores-
cence over a time range of a few hours. This result is pre-
sumably due to rearrangement reactions, which are known
for four-coordinated organoboron compounds.[34] Further-
more, there is another process that competes with the com-
plexation of the fluoride by the boron center. In the case of
model compound 3, which only contains one triarylborane
moiety, this competing process can be seen in the fluoride-
binding curve in the fluorescence spectra, which does not
rise instantaneously with a finite slope on the addition of
fluoride, as would be expected for a 1:1 ratio of binding
sites/ligands, but rather is sigmoidal, thus indicating that the
added fluoride is partially “deactivated” prior to boron-com-
plexation (see the Supporting Information, Figure S8). In
particular, this effect interferes with the complexation at the
very low chromophore concentrations (about ꢃ10ꢀ7 m) that
Figure 9. Normalized fluoride-binding curve of HAB 3 at an absorption
of 31900 cmꢀ1 under an argon atmosphere (dark red squares) in dry THF
and at 31800 cmꢀ1 in THF under ambient conditions (dark green squares)
upon adding incremental amounts of TBAF. The linear fit of the former
is shown in red and the fit of the latter (HypSpec software) is shown in
bright green.
&
8
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Charge-Transfer in Tris-Donor–Tris-Acceptor Chromophores
FULL PAPER
nected to a donor center.[14d] For further investigation of the
“deactivation” process, compound 3 (4.4ꢃ10ꢀ5 m) was titrat-
ed under inert-gas conditions in dry THF (water content<
0.03%), as monitored by UV/Vis spectroscopy. In this ex-
periment, “deactivation” of fluoride could no longer be de-
tected and the complexation constant increased so strongly
that it could not be determined by a fit of the data (for the
binding curve and spectra, see the Supporting Information,
Figures S9 and S10, respectively) and only an upper limit of
K>1ꢃ107 could be estimated. From this observation, we
conclude that, by choosing moisture-free conditions, the re-
activity of the fluoride is no longer impeded by hydrogen
bonding to water, which leads to a higher complexation con-
stant.
It is remarkable that the fluoride-binding curve at
31800 cmꢀ1 absorption decreases almost immediately after
the addition of fluoride, whereas the CT fluorescence at
17600 cmꢀ1 maintains a constant high intensity until 10
In the case of HAB 2, which contains three binding sites,
the fitting of the binding curve from the absorption spectra
in THF under ambient conditions is a much-more-demand-
ing task owing to the fact that the extinction coefficients of
the complex species involved are unknown and the binding
constants are too similar to obtain well-defined complexa-
tion steps. For this reason, the automatic global-fitting rou-
tine of HypSpec was not able to yield physically reasonable
results from the absorption spectra. However, with manual
optimization, the arguable best-fit is obtained with binding
constants of 6ꢃ105, 2ꢃ105, and 0.7ꢃ105 for the first-,
second-, and third complexations, respectively (for the fit
and a further discussion about the quality of the fit, see the
Supporting Information, Figures S11 and S12). These values
agree with the statistical ratio of 3:1:0.33 for the subsequent
complexation of the analyte with three noninteracting li-
gands.[40] However, these findings disagree with the negative
binding cooperativity that is found in many systems with
more than one boron center.[16b,e] A reason for this disagree-
ment could be the poor electronic coupling of the neighbor-
ing redox centers in HAB 2. This hypothesis is also support-
ed by the good agreement between the complexation con-
stants of HAB 2 and that of model compound 3. Thus, the
greater steric hindrance provided by the HAB framework in
compound 2 does not affect the binding ability of the triaryl-
boranes to the small fluoride anion. Under moisture-free
conditions, a binding curve with a linear decrease was ob-
served (see the Supporting Information, Figure S13), which
was impossible to be fitted with the given procedures. Full
complexation is reached after adding four equivalents of
TBAF. At this point, we stress that a commercially available
solution of TBAF contains about 5 wt.% water, which defi-
nitely hampers its Lewis base reactivity. Nevertheless, both
observations indicate that the binding constants exceed
those measured under ambient conditions, which agrees
with the findings in HAB 3. Fluorescence-titration experi-
ments at concentrations of 7ꢃ10ꢀ7 m under ambient condi-
tions gave no definite fit of the binding curves with all of
the parameters optimized, owing to the increased complexi-
ty of the system and the significant “fluoride-deactivation”
process at these low concentrations. However, with some
constraint parameters, a fit was possible (see below).
Figure 10. Normalized fluoride-binding curves of HAB 2 at 17600 cmꢀ1
fluorescence (red), 23100 cmꢀ1 fluorescence (black), and at 31800 cmꢀ1
absorption (blue) in THF under ambient conditions upon adding incre-
mental amounts of TBAF.
equivalents of fluoride have been added (Figure 10). This
observation shows that, as long as there is one unblocked
CT state present in the molecule, fluorescence will be emit-
ted from this unblocked state at constant intensity. This
property requires energy-transfer in between the chromo-
phore branches to take place before emission.
With the assumptions of energy-transfer and no coopera-
tivity effects, the binding curve at 23100 cmꢀ1 fluorescence
could be fitted with binding constants of K=1.8ꢃ106, 6ꢃ
105, and 2ꢃ105 for the first, second and third complexations,
respectively, by using the Equations given in the Supporting
Information (Figure S14). These values are very similar to
those that were obtained by UV measurements at higher
substrate concentrations.
Conclusion
In comparison with model compound 3, redox chromo-
phores 1 and 2 allow us to investigate their versatile redox
behavior, as well as their optically induced charge-transfer
processes. By using CV and DPV, we demonstrated that
electronic coupling between redox centers cannot be directly
estimated from the separation of the redox potentials. As
Barriere and Geiger have already pointed out, this separa-
tion depends on the solvent, as well as on the electrolyte
salt.[26] HAB 2 was a suitable model compound to support
their findings, because three electron-donor redox centers
and three electron-acceptor redox centers were present
within one molecule. Changing the electrolyte salt from
[Bu4N]ACHTNUGTRENNUG[ClO4] to [Bu4N][BCAHTUTGNERN(NGUN C6F5)4] led to a shift in both the
anodic and cathodic peaks, as well as to changes in their rel-
ative potential splittings. These findings are due to electro-
static interactions between the charged redox centers, which
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!

M. Steeger and C. Lambert
are shielded by coordinating (small) counterions, such as the
perchlorate ion. Irrespective of the splitting of the redox po-
tential, one can switch HAB 2 from triply negatively charg-
ed to triply positively charged. Four of the redox states are
mixed-valent, that is, they contain both charged- and neutral
centers of the same chemical type.
conditions are discussed. However, this aspect should be
paid more attention because fluoride sensors in several ap-
plications will definitely be used under humid conditions.
Whilst many electronic and optical properties were practi-
cally identical for symmetric HAB 1 and asymmetric HAB
2, there was a distinct difference with regard to solubility,
which was much better for the asymmetric isomer (2), and
with regard to the formation of amorphous materials. In
contrast to the symmetric isomer (1), the asymmetric isomer
(2) shows a glass transition at 1518C (see the Supporting In-
formation, S15). Thus, HAB 2 appears to be the more-prom-
ising candidate in terms of possible applications in optoelec-
tronic devices.
The combination of triarylamine- and triarylborane redox
centers at the positions ortho to the central benzene ring
allow for CT interactions from the electron-rich amines to
the electron-deficient boron moieties. The CT character is
evident by both a large positive fluorescence solvatochrom-
ism and by the transient-absorption spectra, which display
the characteristic, isoelectronic amine-radical-cation- and
boron-radical-anion absorption bands. However, this elec-
tronic CT interaction is weak, because the intensity of the CT
absorption is low and the lifetime of the back-electron trans-
fer is long compared to similar donor–acceptor compounds.[10]
With the CT state verified as the lowest excited singlet
state, the investigation of energy-transfer in HABs 1 and 2
becomes greatly simplified. Steady-state fluorescence-aniso-
tropy measurements were performed to probe energy-hop-
ping between the six- and four CT states in HABs 1 and 2,
respectively. The anisotropy data were compared with that
of the model compound (3), in which only one CT state ex-
isted. By observing a red-edge excitation effect in the spec-
tra of compounds 1 and 2, we confirmed that both molecules
were asymmetric in the excited state; thus, in the case of
compound 1, breaking of the symmetry must occur. Further-
more, the CT states differ slightly in energy and there is
energy redistribution from high-energy CT states to the
lowest state within the fluorescence lifetime. It remains to
be clarified by which mechanism the energy is transferred.
Time-dependent anisotropy measurements of similar HABs
suggest transition dipole–dipole interactions, as formulated
by Fçrster, whilst the poor orbital overlap hampers Dexter
energy-transfer.[2b]
The weak interactions between the triarylborane centers
were also confirmed by fluoride-complexation of compound
2 (with TABF in THF), which showed additive effects (that
is, a lack of cooperativity) on the absorption spectra.
Adding fluoride to HABs 2 and 3 inhibits the formation of
CT states by complexation of the boron redox centers,
which can then no longer act as acceptors. Instead, a new
high-energy fluorescence band appears, which originates
from local transitions in the triarylamine moieties. The
changes in the absorption spectra of compound 3 could be
analyzed quantitatively to obtain a fluoride binding constant
of K=1.5ꢃ105 in THF under ambient conditions. However,
the moisture content of the solvent plays a critical role in
the quantitative determination of the binding constant. Flu-
oride has a high hydration enthalpy, which explains its lower
reactivity when going from water-free THF to about 0.03%
water content in commercially available THF (spectroscopic
grade). Under these two conditions, the binding constant de-
viates by at least two orders of magnitude. This deviation
questions the published binding constants that are given in
many publications, in which no details concerning water-free
Experimental Section
Electrochemistry: CV and DPV were performed under an argon atmos-
phere in a flame-dried cell with the standard three electrode setup on a
BAS CW-50W potentiostat. Platinum wires served as the counter elec-
trode and the pseudo-reference electrodes and a platinum wire with a di-
ameter of 1 mm was used as the working electrode. The ferrocene/ferro-
cenium redox couple served as an internal standard. The electrochemical
cell was equipped with an aluminum-oxide column (dried in vacuo at
3008C for 30 min), which allowed in-situ drying of the solvent/solute/elec-
trolyte mixture. THF was dried over Na and freshly distilled prior to use.
The samples were measured at a concentration of 3ꢃ10ꢀ4–1ꢃ10ꢀ3 m.
[Bu4N][B
ACHTUNGTRENNUNG
from Li[BAHCUTNGTRENNUNG
CH2Cl2 after use.
UV/Vis spectroscopy: UV/Vis absorption spectra were measured in 1 cm
quartz cuvettes on JASCO V-570 and JASCO V-670 UV/Vis/NIR spec-
trometers. All solvents were spectroscopic grade and used as received.
Transient-absorption measurements: An Edinburgh LP 920 Laser Flash
spectrometer was used to measure transient-absorption spectra on the
nanosecond timescale. All solvents were spectroscopic grade and used as
received. The samples were measured in 1 cm quartz cuvettes and de-
gassed with argon for 10 min. A Continuum Minilite II Nd:YAG laser
produced 5 ns laser pulses at 28200 cmꢀ1 (355 nm; third harmonic of the
1064 nm fundamental wavelength) at a repetition rate of 10 Hz. For exci-
tation of the sample, the laser output was shifted to a lower energy of
24000 cmꢀ1 (416 nm) by means of a 50 cm hydrogen-charged (50 bar)
Raman shifter and the appropriate wavelength was selected by using a
Pellin–Broca prism. The white-light probe pulse was provided by a
pulsed Xe flash lamp. The instrument-response function (IRF) of the
setup was about 8 ns and was obtained by measuring the laser pulse that
was scattered with an empty quartz cell. The decay curves were deconvo-
luted with the measured IRF and the quality of the fit was evaluated by
using the residuals and the c2 values.
Fluorescence spectroscopy: Steady-state fluorescence spectra were meas-
ured on a PTI (Photon Technology International) QM-2000–4 fluores-
cence spectrometer with a cooled photomultiplier (R928 P). All solvents
were of spectroscopic grade and used as received. The samples (about
ꢃ10ꢀ6 m) were measured in 1 cm quartz cuvettes and freed of oxygen by
degassing with argon for 5 min prior to use. Eq. (5) was used to deter-
mine the fluorescence quantum yield, where If, OD, and n denote the in-
tensity of the fluorescence, the optical density of the solution at the exci-
tation wavelength, and the refractive index of the solvent, respectively.
R
ꢀ
ꢁ
Ifðv~Þ ODref n2
R
ꢀf ¼ ꢀf;ref
ꢃ
ꢃ
ð5Þ
n2ref
OD
Ifðv~Þref
A solution of quinine sulfate in 1n sulfuric acid with a quantum yield of
0.546 was used as the reference.[37]
&
10
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Charge-Transfer in Tris-Donor–Tris-Acceptor Chromophores
FULL PAPER
Fluoride-titration measurements: Fluorescence- and UV/Vis fluoride-ti-
tration measurements were performed on the steady-state spectrometers
mentioned above. Tetrabutylammonium fluoride (TBAF) was purchased
from Sigma–Aldrich and used as received. THF was either dried over Na
and freshly distilled before use or obtained from Acros in spectroscopic
grade with a water content of<0.03% and used without further purifica-
tion. The samples were measured in 1 cm quartz cuvettes at concentra-
R. Rathore, Org. Lett. 2006, 8, 5041–5044; f) B. Traber, J. J. Wolff, F.
Rominger, T. Oeser, R. Gleiter, M. Goebel, R. Wortmann, Chem.
Eur. J. 2004, 10, 1227–1238; g) R. Rathore, C. L. Burns, S. A. Abdel-
wahed, Org. Lett. 2004, 6, 1689–1692; h) R. Rathore, C. L. Burns,
M. I. Deselnicu, Org. Lett. 2001, 3, 2887–2890.
[8] L. Jimꢄnez-Garcꢅa, A. Kaltbeitzel, V. Enkelmann, J. S. Gutmann, M.
Klapper, K. Mꢀllen, Adv. Funct. Mater. 2011, 21, 2216–2224.
[9] L. Zhi, K. Mullen, J. Mater. Chem. 2008, 18, 1472–1484.
tions of about 5ꢃ10ꢀ7 m (fluorescence measurements) and 1–4ꢃ10ꢀ5
m
(UV/Vis measurements) and TBAF was added in portions of 10–20 mL
with concentrations of 1–3ꢃ10ꢀ4 m and 0.6–1ꢃ10ꢀ3 m, respectively. The
fluorescence spectra were normalized by the OD at the excitation wave-
length of 29400 cmꢀ1 prior to fitting. For measurements under inert-gas
conditions, the cuvettes were filled under an argon atmosphere and
sealed with septa. The HypSpec software package, which is part of the
Hyperquad suite,[38] was used to fit the titration curves.
[10] R. Stahl, C. Lambert, C. Kaiser, R. Wortmann, R. Jakober, Chem.
Eur. J. 2006, 12, 2358–2370.
[11] S. Amthor, B. Noller, C. Lambert, Chem. Phys. 2005, 316, 141–152.
[12] a) C. Lambert, G. Nçll, J. Schelter, Nat. Mater. 2002, 1, 69–73; b) C.
Lambert, G. Nçll, J. Chem. Soc. Perkin Trans. 2 2002, 2039–2043;
c) A. Heckmann, S. Amthor, C. Lambert, Chem. Commun. 2006,
2959–2961; d) K. Y. Law, Chem. Rev. 1993, 93, 449–486; e) S. Jun-
germann, N. Riegel, D. Mꢀller, K. Meerholz, O. Nuyken, Macromo-
lecules 2006, 39, 8911–8919; f) O. Nuyken, S. Jungermann, V. Wie-
derhirn, E. Bacher, K. Meerholz, Monatsh. Chem. 2006, 137, 811–
824; g) H. Zhao, C. Tanjutco, S. Thayumanavan, Tetrahedron Lett.
2001, 42, 4421–4424; h) Y. Shirota, H. Kageyama, Chem. Rev. 2007,
107, 953–1010; i) Z. Ning, H. Tian, Chem. Commun. 2009, 5483–
5495; j) M. Thelakkat, Macromol. Mater. Eng. 2002, 287, 442–461.
[13] a) G. Zhou, C.-L. Ho, W.-Y. Wong, Q. Wang, D. Ma, L. Wang, Z.
Lin, T. B. Marder, A. Beeby, Adv. Funct. Mater. 2008, 18, 499–511;
b) T. Noda, Y. Shirota, J. Lumin. 2000, 87–89, 1168–1170; c) T.
Noda, Y. Shirota, J. Am. Chem. Soc. 1998, 120, 9714–9715; d) H.
Doi, M. Kinoshita, K. Okumoto, Y. Shirota, Chem. Mater. 2003, 15,
1080–1089; e) S. Yamaguchi, A. Wakamiya, Pure Appl. Chem. 2006,
78, 1413–1424.
[14] a) C. H. Zhao, E. Sakuda, A. Wakamiya, S. Yamaguchi, Chem. Eur.
J. 2009, 15, 10603–10612; b) Z. M. Hudson, S. N. Wang, Acc. Chem.
Res. 2009, 42, 1584–1596; c) S. Yamaguchi, T. Shirasaka, K. Tamao,
Org. Lett. 2000, 2, 4129–4132; d) D.-R. Bai, X.-Y. Liu, S. Wang,
Chem. Eur. J. 2007, 13, 5713–5723; e) D. Reitzenstein, C. Lambert,
Macromolecules 2009, 42, 773–782; f) Z. Yuan, J. C. Collings, N. J.
Taylor, T. B. Marder, C. Jardin, J.-F. Halet, J. Solid State Chem. 2000,
154, 5–12; g) W.-L. Jia, D. Song, S. Wang, J. Org. Chem. 2003, 68,
701–705; h) Z.-Q. Liu, Q. Fang, D. Wang, D.-X. Cao, G. Xue, W.-T.
Yu, H. Lei, Chem. Eur. J. 2003, 9, 5074–5084; i) Y. Sun, S. N. Wang,
Inorg. Chem. 2010, 49, 4394–4404; j) S. B. Zhao, P. Wucher, Z. M.
Hudson, T. M. McCormick, X. Y. Liu, S. N. Wang, X. D. Feng, Z. H.
Lu, Organometallics 2008, 27, 6446–6456; k) F. Jꢁkle, Chem. Rev.
2010, 110, 3985–4022; l) J. B. Heilmann, M. Scheibitz, Y. Qin, A.
Sundararaman, F. Jꢁkle, T. Kretz, M. Bolte, H.-W. Lerner, M. C.
Holthausen, M. Wagner, Angew. Chem. 2006, 118, 934–939; Angew.
Chem. Int. Ed. 2006, 45, 920–925.
Fluorescence-lifetime measurements: Time-dependent fluorescence spec-
tra were recorded on an Edinburgh LP 920 Laser Flash spectrometer.
The fluorescence lifetimes were determined in the same manner as in the
transient-absorption measurements (see above).
Polarized steady-state fluorescence spectroscopy: Fluorescence-anisotro-
py measurements were recorded at RT on a PTI (Photon Technology In-
ternational) QM-2000-4 fluorescence spectrometer with two Glan–
Thompson polarizers (PTI) in an L-setup. The analyte was embedded in
a sucrose-octaacetate (SOA) matrix. SOA (Acros Organics) was recrys-
tallized from EtOH prior to use. The sample was prepared according to a
literature procedure.[39] A solution of SOA and the chromophore in
CH2Cl2 (Merck, Uvasol) was filtered through a syringe PTFE filter with
a pore-size of 0.1 mm (Whatman) to remove any traces of lint and dust
and degassed with argon for 10 min. The solution was concentrated in
vacuo until a viscous oil was obtained and then filled into a 1 cm fluores-
cence quartz cuvette. To remove the remaining CH2Cl2, the cuvette was
heated in an oven to 1008C for about 1 h and to 1308C for 4 h. The exci-
tation anisotropy measurements were recorded at an emission energy of
21200 cmꢀ1 (472 nm).
Differential scanning calorimetry (DSC): DSC measurements were re-
corded on a TA Q1000 calorimeter with a heating/cooling rate of 308C
minꢀ1. Two heating–cooling cycles were performed.
Acknowledgements
We thank the Deutsche Forschungsgemeinschaft for financial support
through grant La991/13–1 and the Graduiertenkolleg 1221. We are also
indebted to Prof. H. Braunschweig for a sample of [Bu4N][B
ACHTNUGRTNE(NGNU C6F5)4] and
[15] a) Z. Yuan, N. J. Taylor, T. B. Marder, I. D. Williams, S. K. Kurtz,
L. T. Cheng, J. Chem. Soc. Chem. Commun. 1990, 1489–1492; b) Z.
Yuan, N. J. Taylor, Y. Sun, T. B. Marder, I. D. Williams, L.-T. Cheng,
J. Organomet. Chem. 1993, 449, 27–37; c) Z. Yuan, C. D. Entwistle,
J. C. Collings, D. Albesa-Jove, A. S. Batsanov, J. A. K. Howard, N. J.
Taylor, H. M. Kaiser, D. E. Kaufmann, S.-Y. Poon, W.-Y. Wong, C.
Jardin, S. Fathallah, A. Boucekkine, J.-F. Halet, T. B. Marder, Chem.
Eur. J. 2006, 12, 2758–2771; d) M. Lequan, R. M. Lequan, K. C.
Ching, J. Mater. Chem. 1991, 1, 997–999; e) Z. Yuan, N. J. Taylor, R.
Ramachandran, T. B. Marder, Appl. Organomet. Chem. 1996, 10,
305–316; f) C. D. Entwistle, T. B. Marder, Chem. Mater. 2004, 16,
4574–4585; g) M. Lequan, R. M. Lequan, K. C. Ching, M. Barzou-
kas, A. Fort, H. Lahoucine, G. Bravic, D. Chasseau, J. Gaultier, J.
Mater. Chem. 1992, 2, 719–725; h) M. Charlot, L. Porres, C. D. En-
twistle, A. Beeby, T. B. Marder, M. Blanchard-Desce, Phys. Chem.
Chem. Phys. 2005, 7, 600–606; i) Y. Liu, X. Xu, F. K. Zheng, Y. Cui,
Angew. Chem. 2008, 120, 4614–4617; Angew. Chem. Int. Ed. 2008,
47, 4538–4541; j) J. C. Collings, S. Y. Poon, C. Le Droumaguet, M.
Charlot, C. Katan, L. O. Palsson, A. Beeby, J. A. Mosely, H. M.
Kaiser, D. Kaufmann, W. Y. Wong, M. Blanchard-Desce, T. B.
Marder, Chem. Eur. J. 2009, 15, 198–208; k) C. D. Entwistle, J. C.
Collings, A. Steffen, L.-O. Palsson, A. Beeby, D. Albesa-Jove, J. M.
Burke, A. S. Batsanov, J. A. K. Howard, J. A. Mosely, S.-Y. Poon,
Ms. A.-M. Krause for performing the DSC measurements.
[1] C. Lambert, Angew. Chem. 2005, 117, 7503–7505; Angew. Chem.
Int. Ed. 2005, 44, 7337–7339.
[2] a) P. A. Liddell, G. Kodis, L. de La Garza, A. L. Moore, T. A.
Moore, D. Gust, J. Phys. Chem. B 2004, 108, 10256–10265; b) H. S.
Cho, H. Rhee, J. K. Song, C.-K. Min, M. Takase, N. Aratani, S. Cho,
A. Osuka, T. Joo, D. Kim, J. Am. Chem. Soc. 2003, 125, 5849–5860.
[3] G. Kodis, Y. Terazono, P. A. Liddell, J. Andreasson, V. Garg, M.
Hambourger, T. A. Moore, A. L. Moore, D. Gust, J. Am. Chem. Soc.
2006, 128, 1818–1827.
[4] X. Hu, K. Schulten, Phys. Today 1997, 50, 28–34.
[5] A. E. Keirstead, J. W. Bridgewater, Y. Terazono, G. Kodis, S.
Straight, P. A. Liddell, A. L. Moore, T. A. Moore, D. Gust, J. Am.
Chem. Soc. 2010, 132, 6588–6595.
[6] D. Rausch, C. Lambert, Org. Lett. 2006, 8, 5037–5040.
[7] a) D. Sun, S. V. Rosokha, J. K. Kochi, Angew. Chem. 2005, 117,
5263–5266; Angew. Chem. Int. Ed. 2005, 44, 5133–5136; b) V. J.
Chebny, R. Shukla, R. Rathore, J. Phys. Chem. A 2006, 110, 13003–
13006; c) C. Lambert, G. Nçll, Chem. Eur. J. 2002, 8, 3467–3477;
d) S. V. Rosokha, I. S. Neretin, D. Sun, J. K. Kochi, J. Am. Chem.
Soc. 2006, 128, 9394–9407; e) V. J. Chebny, D. Dhar, S. V. Lindeman,
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
11
&
ÞÞ
These are not the final page numbers!

M. Steeger and C. Lambert
W.-Y. Wong, F. Ibersiene, S. Fathallah, A. Boucekkine, J.-F. Halet,
T. B. Marder, J. Mater. Chem. 2009, 19, 7532–7544.
[27] J. C. Curtis, J. A. Roberts, R. L. Blackbourn, Y. H. Dong, M.
Massum, C. S. Johnson, J. T. Hupp, Inorg. Chem. 1991, 30, 3856–
3860.
[28] M. S. Yuan, Z. Q. Liu, Q. Fang, J. Org. Chem. 2007, 72, 7915–7922.
[29] S. J. Strickler, R. A. Berg, J. Chem. Phys. 1962, 37, 814–822.
[30] J. Fiedler, S. Zalis, A. Klein, F. M. Hornung, W. Kaim, Inorg. Chem.
1996, 35, 3039–3043.
[31] a) J. R. Lakowicz, Principles of Fluorescence Spectroscopy, Kluwer
Academic/Plenum, New York, 1999; b) R. D. Hall, B. Valeur, G.
Weber, Chem. Phys. Lett. 1985, 116, 202–205.
[32] a) W. Verbouwe, M. Van der Auweraer, F. C. De Schryver, J. J. Piet,
J. M. Warman, J. Am. Chem. Soc. 1998, 120, 1319–1324; b) A. A.
Demidov, D. L. Andrews, Photochem. Photobiol. 1996, 63, 39–52.
[33] a) H. Sohn, S. Letant, M. J. Sailor, W. C. Trogler, J. Am. Chem. Soc.
2000, 122, 5399–5400; b) S. W. Zhang, T. M. Swager, J. Am. Chem.
Soc. 2003, 125, 3420–3421.
[16] a) A. Pronꢆ, M. Baumgarten, K. Mꢀllen, Org. Lett. 2010, 12, 4236–
4239; b) C. R. Wade, A. E. J. Broomsgrove, S. Aldridge, F. P.
Gabbai, Chem. Rev. 2010, 110, 3958–3984; c) A. Pronꢆ, G. Zhou, H.
Norouzi-Arasi, M. Baumgarten, K. Mꢀllen, Org. Lett. 2009, 11,
3550–3553; d) G. Zhou, M. Baumgarten, K. Mꢀllen, J. Am. Chem.
Soc. 2008, 130, 12477–12484; e) H. C. Schmidt, L. G. Reuter, J. Ha-
macek, O. S. Wenger, J. Org. Chem. 2011, 76, 9081–9085.
[17] a) T. W. Hudnall, F. P. Gabbai, J. Am. Chem. Soc. 2007, 129, 11978–
11986; b) C. W. Chiu, F. P. Gabbai, Dalton Trans. 2008, 814–817;
c) T. Agou, M. Sekine, J. Kobayashi, T. Kawashima, J. Organomet.
Chem. 2009, 694, 3833–3836; d) J. O. Huh, Y. Do, M. H. Lee, Orga-
nometallics 2008, 27, 1022–1025; e) C. R. Wade, F. P. Gabbai, Inorg.
Chem. 2010, 49, 714–720.
[18] a) U. Megerle, F. Selmaier, C. Lambert, E. Riedle, S. Lochbrunner,
Phys. Chem. Chem. Phys. 2008, 10, 6245–6251; b) U. Megerle, C.
Lambert, E. Riedle, S. Lochbrunner, Springer Ser. Chem. Phys.
2009, 92, 316–318.
[34] Y. L. Rao, H. Amarne, S. B. Zhao, T. M. McCormick, S. Martic, Y.
Sun, R. Y. Wang, S. N. Wang, J. Am. Chem. Soc. 2008, 130, 12898–
12900.
[19] a) C. D. Entwistle, T. B. Marder, Angew. Chem. 2002, 114, 3051–
3056; Angew. Chem. Int. Ed. 2002, 41, 2927–2931; b) J. C. Doty, B.
Babb, P. J. Grisdale, M. Glogowski, J. L. R. Williams, J. Organomet.
Chem. 1972, 38, 229–236; c) W. Kaim, A. Schulz, Angew. Chem.
1984, 96, 611–612; Angew. Chem. Int. Ed. Engl. 1984, 23, 615–616.
[20] A. Heckmann, C. Lambert, Angew. Chem. 2012, 124, 334-404.
[21] C. Ester, A. Maderna, H. Pritzkow, W. Siebert, Eur. J. Inorg. Chem.
2000, 2000, 1177–1184.
[35] R. J. LeSuer, C. Buttolph, W. E. Geiger, Anal. Chem. 2004, 76,
6395–6401.
[36] S. Santi, L. Orian, C. Durante, E. Z. Bencze, A. Bisello, A. Donoli,
A. Ceccon, F. Benetollo, L. Crociani, Chem. Eur. J. 2007, 13, 7933–
7947.
[37] W. H. Melhuish, J. Phys. Chem. 1961, 65, 229–235.
[38] P. Gans, A. Sabatini, A. Vacca, Talanta 1996, 43, 1739–1753.
[39] R. C. Dorfman, Y. Lin, M. D. Fayer, J. Phys. Chem. 1989, 93, 6388–
6396.
[22] N. E. Schore, Chem. Rev. 1988, 88, 1081–1119.
[23] A. Steffen, R. M. Ward, W. D. Jones, T. B. Marder, Coord. Chem.
Rev. 2010, 254, 1950–1976.
[40] K. A. Connors, Binding Constants, Wiley, New York, 1987, p. 78.
[24] H. Kiesele, Anal. Chem. 1981, 53, 1952–1954.
[25] a) A. Heckmann, C. Lambert, J. Am. Chem. Soc. 2007, 129, 5515–
5527; b) A. Schulz, W. Kaim, Chem. Ber. 1989, 122, 1863–1868.
[26] F. Barriꢇre, W. E. Geiger, J. Am. Chem. Soc. 2006, 128, 3980–3989.
Received: December 22, 2011
Revised: May 24, 2012
Published online: && &&, 2012
&
12
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Charge-Transfer in Tris-Donor–Tris-Acceptor Chromophores
FULL PAPER
If it ain’t broke: Although superficially
symmetric, a hexaarylbenzene system
(see figure) that contains several
amine- and boron redox centers shows
optical spectroscopic properties that
only marginally differ from those of its
asymmetric analogue, owing to the
breaking of the symmetry in the
excited state. In contrast, the proper-
ties of the materials, such as solubility
or phase-transitions, are strongly influ-
enced by the presence/absence of sym-
metry.
Charge Transfer
M. Steeger, C. Lambert* . . . . &&&& — &&&&
Charge-Transfer Interactions in Tris-
Donor–Tris-Acceptor Hexaarylben-
zene Redox Chromophores
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
13
&
ÞÞ
These are not the final page numbers!