Pentacyano-N,N-dimethylaniline in the excited state. only locally excited state emission, in spite of the large electron affinity of the pentacyanobenzene subgroup

Article abstract of DOI:10.1021/jp108804q

Pentacyano-N,N-dimethylaniline (PCDMA) does not undergo an intramolecular charge transfer (ICT) reaction, even in the strongly polar solvent acetonitrile (MeCN), in clear contrast to 4-(dimethylamino)benzonitrile (DMABN). Within the twisted ICT (TICT) model, this is unexpected, as the electron affinity of the pentacyanobenzene moiety of PCDMA is much larger than that of the benzonitrile subgroup in DMABN. According to the TICT model, the energy of the ICT state of PCDMA would be 2.05 eV (～16550 cm^-1) lower than that of DMABN, on the basis of the reduction potentials E(A^-/A) of pentacyanobenzene (-0.29 V vs saturated calomel electrode (SCE)) and benzonitrile (-2.36 V vs SCE), more than enough to compensate for the decrease in energy of the locally excited (LE) state of PCDMA (E(S₁) = 19990 cm^-1) relative to that of DMABN (E(S₁) = 29990 cm^-1). This absence of a LE → ICT reaction shows that the TICT hypothesis does not hold for PCDMA in the singlet excited state, similar to what was found for DMABN, N-phenylpyrrole, and their derivatives. In this connection, the six dicyano-substituted dimethylanilines are also discussed. The energy gap ΔE(S₁,S₂) between the two lowest singlet excited states is, at 7170 cm^-1 for PCDMA in MeCN, considerably larger than that for DMABN (2700 cm^-1 in n-hexane, smaller in MeCN). The absence of ICT is therefore in accord with the planar ICT (PICT) model, which considers a sufficiently small ΔE(S₁,S₂) to be an important condition determining whether an ICT reaction will take place. The fluorescence quantum yield of PCDMA is very small: Φ(LE) = 0.0006 in MeCN at 25 °C, predominantly due to LE → S₀ internal conversion (IC), as the intersystem crossing yield Φ(ISC) is practically zero (<0.01). From the LE fluorescence decay time of 27 ps for PCDMA in MeCN at 25 °C, a radiative rate constant k_f(LE) = 2 × 10⁷ s^-1 results, comparable to the k_f(LE) of DMABN (6.5 × 10 ⁷ s^-1) and 2,4,6-tricyano-N,N-dimethylaniline (TCDMA) (1.2 × 10⁷ s^-1) in this solvent, but clearly larger than the k′_f(ICT) = 0.79 × 10⁷ s^-1 of DMABN in MeCN. The IC reaction with PCDMA in MeCN at room temperature, with a rate constant k_IC of 3.6 × 10¹⁰ s^-1, is much faster than with TCDMA (25 × 10⁷ s^-1) and DMABN (1.3 × 10⁷ s^-1, in n-hexane). This is connected with the nonzero (37°) amino twist angle of PCDMA, which leads to a decrease of the effective LE-S₀ energy gap. The femtosecond excited state absorption (ESA) spectra of PCDMA in MeCN at 22 °C are similar to the LE ESA spectra of TCDMA and DMABN and are therefore attributed to the LE state, confirming that an ICT reaction does not occur. The decay of the LE ESA spectra of PCDMA is single exponential, with a decay time of 22 ps, in reasonable agreement with the LE fluorescence decay time of 27 ps at 25 °C. The spectra decay to zero, showing that there is no triplet or other intermediate.

Full text of DOI:10.1021/jp108804q

J. Phys. Chem. A 2010, 114, 13031–13039
13031
Pentacyano-N,N-Dimethylaniline in the Excited State. Only Locally Excited State Emission,
in Spite of the Large Electron Affinity of the Pentacyanobenzene Subgroup
Klaas A. Zachariasse,*,† Sergey I. Druzhinin,*,† Victor A. Galievsky,†,‡ Attila Demeter,†,§
Xavier Allonas,| Sergey A. Kovalenko,*,⊥ and Tamara A. Senyushkina†
Max-Planck-Institut fu¨r biophysikalische Chemie, Spektroskopie und Photochemische Kinetik,
37070 Go¨ttingen, Germany, B.I. StepanoV Institute of Physics, National Academy of Sciences of Belarus,
Prospekt NezaVisimosti 68, 22072 Minsk, Belarus, Institute of Materials and EnVironmental Chemistry,
Chemical Research Center, Hungarian Academy of Sciences, P.O. Box 17, 1525 Budapest, Hungary,
De´partement de Photochimie Ge´ne´rale, UMR CNRS 7525, UniVersite´ de Haute Alsace, ENSCMu, 3 rue Alfred
Werner, 68093 Mulhouse, France, Institut fu¨r Chemie, Humboldt UniVersita¨t zu Berlin, Brook-Taylor Strasse 2,
12489 Berlin, Germany
ReceiVed: September 15, 2010; ReVised Manuscript ReceiVed: October 30, 2010
Pentacyano-N,N-dimethylaniline (PCDMA) does not undergo an intramolecular charge transfer (ICT) reaction,
even in the strongly polar solvent acetonitrile (MeCN), in clear contrast to 4-(dimethylamino)benzonitrile
(DMABN). Within the twisted ICT (TICT) model, this is unexpected, as the electron affinity of the
pentacyanobenzene moiety of PCDMA is much larger than that of the benzonitrile subgroup in DMABN.
According to the TICT model, the energy of the ICT state of PCDMA would be 2.05 eV (∼16550 cm-1)
lower than that of DMABN, on the basis of the reduction potentials E(A-/A) of pentacyanobenzene (-0.29
V vs saturated calomel electrode (SCE)) and benzonitrile (-2.36 V vs SCE), more than enough to compensate
for the decrease in energy of the locally excited (LE) state of PCDMA (E(S1) ) 19990 cm-1) relative to that
of DMABN (E(S1) ) 29990 cm-1). This absence of a LE f ICT reaction shows that the TICT hypothesis
does not hold for PCDMA in the singlet excited state, similar to what was found for DMABN, N-phenylpyrrole,
and their derivatives. In this connection, the six dicyano-substituted dimethylanilines are also discussed. The
energy gap ∆E(S1,S2) between the two lowest singlet excited states is, at 7170 cm-1 for PCDMA in MeCN,
considerably larger than that for DMABN (2700 cm-1 in n-hexane, smaller in MeCN). The absence of ICT
is therefore in accord with the planar ICT (PICT) model, which considers a sufficiently small ∆E(S1,S2) to
be an important condition determining whether an ICT reaction will take place. The fluorescence quantum
yield of PCDMA is very small: Φ(LE) ) 0.0006 in MeCN at 25 °C, predominantly due to LE f S0 internal
conversion (IC), as the intersystem crossing yield Φ(ISC) is practically zero (<0.01). From the LE fluorescence
decay time of 27 ps for PCDMA in MeCN at 25 °C, a radiative rate constant kf(LE) ) 2 × 107 s-1 results,
comparable to the kf(LE) of DMABN (6.5 × 107 s-1) and 2,4,6-tricyano-N,N-dimethylaniline (TCDMA) (1.2
× 107 s-1) in this solvent, but clearly larger than the k′f(ICT) ) 0.79 × 107 s-1 of DMABN in MeCN. The
IC reaction with PCDMA in MeCN at room temperature, with a rate constant kIC of 3.6 × 1010 s-1, is much
faster than with TCDMA (25 × 107 s-1) and DMABN (1.3 × 107 s-1, in n-hexane). This is connected with
the nonzero (37°) amino twist angle of PCDMA, which leads to a decrease of the effective LE-S0 energy
gap. The femtosecond excited state absorption (ESA) spectra of PCDMA in MeCN at 22 °C are similar to
the LE ESA spectra of TCDMA and DMABN and are therefore attributed to the LE state, confirming that an
ICT reaction does not occur. The decay of the LE ESA spectra of PCDMA is single exponential, with a
decay time of 22 ps, in reasonable agreement with the LE fluorescence decay time of 27 ps at 25 °C. The
spectra decay to zero, showing that there is no triplet or other intermediate.
Introduction
The two other cyano-N,N-dimethylaniline isomers, 2-(dimethyl-
amino)benzonitrile (2DMABN)20 and 3-(dimethylamino)benzoni-
trile (3DMABN),20-23 only show LE emission24,25 in solution, even
in strongly polar solvents20-22 such as acetonitrile (MeCN) or
methanol.
Of the six possible dicyano-N,N-dimethylanilines (Chart 1),
two have been discussed in the literature in connection with
the occurrence of an ICT reaction: 3,4-dicyano-N,N-dimethy-
laniline (34DCDMA) and 3,5-dicyano-N,N-dimethylaniline
(35DCDMA), from an experimental20,22,26-28 as well as from a
computational29-31 point of view. These molecules were reported
to have fluorescence spectra only consisting of a LE fluorescence
band, for 34DCDMA and 35DCDMA in diethyl ether and for
Cyano-substituted N,N-dimethylanilines have played an impor-
tant role in the experimental investigation of intramolecular charge
transfer (ICT) in the excited state.1-19 The phenomenon of ICT
and dual fluorescence from a locally excited (LE) and an ICT state
was first described with 4-(dimethylamino)benzonitrile (DMABN).1,2
* Correspondingauthor.Fax:+49-551-201-1501.E-mail:kzachar@gwdg.de
(K.A.Z.); sdruzhi@gwdg.de (S.I.D.); skovale@chemie.hu-berlin.de (S.K.).
† Max-Planck-Institut fu¨r biophysikalische Chemie, Spektroskopie und
Photochemische Kinetik.
‡ National Academy of Sciences of Belarus.
§ Hungarian Academy of Sciences.
| Universite´ de Haute Alsace.
⊥ Humboldt Universita¨t zu Berlin.
10.1021/jp108804q  2010 American Chemical Society
Published on Web 11/24/2010

13032 J. Phys. Chem. A, Vol. 114, No. 50, 2010
Zachariasse et al.
CHART 1
35DCDMA even in the strongly polar solvent MeCN.20,22,26-28
In addition, their fluorescence decays are single exponential with
nanosecond decay times, without any indication of a second,
shorter picosecond decay.20,26,28 It was hence concluded that the
possibility to undergo an ICT reaction from an initially excited
S1(LE) state to an ICT state with a larger dipole moment did
not exist for 34DCDMA and 35DCDMA.
of the TICT hypothesis that the energy E(ICT) of the ICT state
could be calculated from the difference between the oxidation
potential E(D/D+) and the reduction potential E(A-/A) of their
D and A subunits, or, equivalently, from the difference between
the ionization potential IP(D) and electron affinity EA(A).4a,9,34-36
Such calculations are equivalent to adopting the semiempirical
Weller equation, which was developed for intermolecular
exciplexes (A-D+) in the solvent n-hexane.26,37-40 In these
exciplexes, the A and D molecules are weakly interacting
chromophores, without an appreciable electronic coupling.
It should then be expected that the introduction of an
additional cyano-substituent in DMABN would reduce E(ICT),
as the E(A-/A) of the dicyano-substitued benzenes is consider-
ably smaller than that of benzonitrile: -1.68 V vs saturated
calomel electrode (SCE) for 1,2-dicyanobenzene (the A moiety
of 34DCDMA), and -1.95 V vs SCE for 1,3-dicyanobenzene
(the A moiety of 35DCDMA), for example, as compared with
-2.36 V vs SCE for benzonitrile.41 In contrast to this expecta-
tion, 34DCDMA and 35DCDMA were reported not to undergo
a LE f ICT reaction, as discussed above.20,22,25-28
The appearance of an additional ICT fluorescence with DMABN
and its absence with 2DMABN, 3DMABN, 34DCDMA, and
35DCDMA, has been explained in terms of the magnitude of the
energy gap ∆E(S1,S2), which is substantially smaller for DMABN
than for the other four molecules.20,22,26,28 The central role of
∆E(S1,S2) in ICT reactions is an integral part of the planar ICT (PICT)
model, which states that the D and A subunits in the ICT state are not
electronically decoupled, although it is not required that they are in a
mutually planar configuration.4b,10-12,20,22,25,26,28,42,43
Similarly, only LE emission and single exponential fluores-
cence decays are observed with 2,4,6-tricyano-N,N-dimethyla-
niline (TCDMA) in n-hexane and MeCN, in spite of the strong
increase in electron affinity of the 1,3,5-tricyanobenzene ac-
ceptor moiety (E(A-/A) ) -1.36 V vs SCE) relative to
benzonitrile.25,41 Also in this case, ∆E(S1,S2) is much larger than
that for DMABN: 5370 cm-1 as compared with 2700 cm-1, in
n-hexane at 25 °C.20,25
The pentacyanobenzene acceptor moiety in pentacyano-N,N-
dimethylaniline (PCDMA) has, with E(A-/A) ) -0.29 V vs
SCE,41 an even much larger electron affinity than the tricy-
anobenzene in TCDMA. The occurrence or absence of an ICT
reaction with PCDMA will be discussed in the present article,
in a treatment of the excited state behavior of N,N-dimethyl-
anilines substituted with one, two, three, and five cyano groups.
From computations on 35DCDMA it was deduced that its
two meta-cyano substituents influence the S1 as well as the S2
state, in contrast to DMABN, for which the interaction with
CN mainly takes place in S2.29 The S1 and S2 states of
35DCDMA, with similar dipole moments, are therefore equally
stabilized when the solvent polarity becomes larger. This means
that the energy gap ∆E(S1,S2) between the two lowest excited
singlet states of 35DCDMA, which in the gas phase (calcula-
tions) is, at 5290 cm-1, already much larger than for DMABN
(2520 cm-1 (calculated, ref 30); 3520 cm-1 (experimental, ref
19),31 does not become smaller with increasing solvent polarity,
explaining why dual emission comparable to that of DMABN
does not take place with 35DCDMA. It is concluded that its
fluorescence spectrum is that of a single “locally excited” S1
emission band and that even in very polar solvents a LE state
S1 with planar conformation is favored for 35DCDMA.29 In
another computation on 34DCDMA and 35DCDMA, it was
found that their S1 state has charge transfer (CT) character and
that these molecules have a significantly larger energy gap
∆E(S1,S2) than DMABN, explaining the absence of dual
fluorescence in accordance with the planar ICT (PICT) model.32
In a time-dependent density-functional calculation, without
geometry optimization for the excited states, the absence of dual
fluorescence with 34DCDMA and 35DCDMA is discussed.33
It was again computed that these molecules have a large
∆E(S1,S2) gap, taken as an explanation that an ICT reaction
does not take place (PICT model). The S1 fluorescence band of
34DCDMA is identified as coming from a twisted ICT (TICT)
state, with a dipole moment µe(S1) for emission of 24 D. The
energy E(S1) of 35DCDMA is almost independent of the amino
twist angle. This S1 state has important CT character, as seen
from its µe(S1) of 19 D. The S1 Franck-Condon (FC) state,
reached by excitation of the equilibrated S0 ground state, is an
ICT state with a dipole moment µe(S1,abs) of 17 D for
34DCDMA and 15 D for 35DCDMA. The S2(FC) state is
claimed to be of LE nature, although µe(S2,abs) of both
molecules is, at 16 D, similar to µe(S1,abs). ∆E(S1,S2) is hence
not reduced by increasing the solvent polarity, an explanation
for the absence of an ICT reaction.
Experimental Section
In all three calculations,29,32,33 only the S0 structure was
optimized, and this structure was subsequently used for the
excited states. For the TICT states, a rotation around the
N-phenyl bond was introduced, leaving the rest of the molecular
structure unchanged. Therefore, absorption spectra can be
obtained from the calculations, but the computed states are not
the relaxed S1 and ICT states from which the fluorescence is
emitted.
PCDMA (mp 99.2-101.6 °C) was made by reacting hexacy-
anobenzene (HCB) with NH(CH3)2 ·HCl and NaOH in 1,2-
dimethoxyethane at room temperature.44 HCB was synthesized
from 1,3,5-trichloro-2,4,6-tricyanobenzene and trimethylsilyl-
cyanide in absolute dimethylformamide at 80 °C.44,45 For
PCDMA and HCB, high-performance liquid chromatography
(HPLC) was the last purification step.
In the early stages of the investigations of the ICT reaction
with electron donor (D)/acceptor (A) molecules such as the
cyano-substituted anilines, it was assumed within the context
The measurement and treatment of the absorption and
fluorescence spectra, fluorescence and triplet quantum yields,
triplet-triplet (T-T) absorption spectra, redox potentials, single

PCDMA in the Excited State
J. Phys. Chem. A, Vol. 114, No. 50, 2010 13033
The fluorescence spectra of PCDMA in THF and MeCN at
25 °C consist of a single emission band and can be considered
as mirror images of the lowest-energy S1 absorption band
(Figure 1). This is a first indication that a LE f ICT reaction
does not occur, even in the strongly polar solvent MeCN (ε25
) 36.7). The fluorescence spectra are therefore attributed to
the LE state. We can not, however, exclude with absolute
certainty that a weak ICT emission would be present below
700-800 nm, beyond our detection limit. A single LE emission
band has also been reported with 2DMABN, 3DMABN,
34DCDMA, 35DCDMA, and TCDMA, as mentioned in the
Introduction.20-23,25-28 In the case of 34DCDMA, new experi-
ments presented below show the presence of dual LE + ICT
fluorescence in MeCN at 25 °C.
LE Fluorescence Quantum Yield. The LE fluorescence
quantum yield Φ(LE) of PCDMA in MeCN at 25 °C is
determined relative to that of rubrene with the same optical
density (0.2) at the excitation wavelength in this solvent.
Rubrene was selected in order to be able to excite PCDMA in
the S1 absorption band (480 nm), as it was found that, for shorter
excitation wavelengths, additional (impurity and/or photopro-
duct) emissions appeared in its fluorescence spectrum. Upon
also adopting for rubrene in MeCN a value of Φf ) 1.0
measured for the fluorescence quantum yield of rubrene in
benzene and n-heptane,48,49 Φ(LE) ) 0.0006 results for PCDMA
in MeCN.
Figure 1. Absorption (Abs) and fluorescence (LE) spectra of PCDMA
in (a) THF and (b) MeCN at 25 °C. Excitation wavelength: 480 nm.
Picosecond Fluorescence Decays. The fluorescence decays
of PCDMA in EtCN and MeCN at 25 °C are shown in Figure 2.
Three exponentials (eq 1) are required to fit the curves. The shortest
decay times τ3, 27.4 ps in MeCN (Figure 2a) and 30.7 ps in THF
(Figure 2b), are considered to be the LE f S0 deactivation times
(Table 2). The two much longer decay times τ2 and τ1 are attributed
to intrinsic impurities and photoproducts, as their contribution to
the total decay becomes larger with increasing time between HPLC
purification and the start of the measurements as well as with
irradiation time during the experiment.
photon counting (SPC) decays and femtosecond transient
absorption spectra have been described elsewhere.10,12,19,25,42,46,47
Results and Discussion
Absorption and Fluorescence Spectra of PCDMA. The
absorption spectra of PCDMA in tetrahydrofuran (THF) and
MeCN at 25 °C are shown in Figure 1. The energy gap
∆E(S1,S2) ) ν˜max(S2,abs) - ν˜max(S1,abs) between the maxima
of the S1 and S2 absorption bands is 7040 cm-1 in THF and
7170 cm-1 in MeCN (Table 1). In ethyl cyanide (EtCN), a
similar energy gap of 7110 cm-1 is obtained. These data indicate
that ∆E(S1,S2) becomes somewhat larger with increasing solvent
polarity. The molar absorption coefficients εmax of the S1 and
S2 bands of PCDMA in MeCN (Table 1) are similar to those of
TCDMA25 and DMABN12 in this solvent: 3910 M-1cm-1 (S1)
and 16000 M-1cm-1 (S2) for PCDMA, 4830 M-1cm-1 (S1) and
21610 M-1cm-1 (S2) for TCDMA, and 27990 M-1cm-1 (S2)
for DMABN. When the εmax of the S2 band of PCDMA is
smaller than that of DMABN with a planar dimethylamino group
if(flu) ) A1 exp(-t/τ1) + A2 exp(-1/τ2) + A3 exp(-t/τ3)
(1)
Radiative Rate Constants. From the fluorescence quantum
yield Φ(LE) ) 0.0006 and the decay time τ3(LE) ) 27.4 ps for
PCDMA in MeCN at 25 °C, the radiative rate constant kf(LE)
) Φf/τ3(LE) ) 2.2 × 107 s-1 is obtained (Table 2). Its value is
comparable with that of TCDMA (1.2 × 107 s-1) and somewhat
smaller than that of DMABN (6.5 × 107 s-1), in this solvent,
see Table 2. These kf(LE) rate constants are clearly larger than
k’f(ICT), with a value of 0.79 × 107 s-1 in the case of DMABN
in MeCN, confirming their LE nature.
ε
max(0) because of a nonzero amino twist angle θ, an angle θ )
37.3° is calculated for PCDMA, employing the relation εmax(θ)/
ε
max(0) ∼ cos2 θ.11 Support for this approach comes from
TCDMA, for which via εmax(θ)/εmax(0) a twist angle θ of 29° is
obtained, in very good agreement with the experimental θ )
29° coming from X-ray crystal analysis.25
Intersystem Crossing Yield Φ(ISC) of PCDMA in MeCN.
Measurements with PCDMA in MeCN at 23 °C could not detect
a T-T absorption in the 550-650 nm range. This means that
TABLE 1: Data Obtained from the Fluorescence and Absorption Spectra at 25 °C of PCDMA in THF, EtCN, and MeCN
solvent
THF
7.39
EtCN
29.2
MeCN
36.7
ε25
ν˜max(flu) [cm-1
]
18150 (0.98)a 17380 (1)a
22080 (0.24)b
29120 (1)b
18630 (1)a 17300 (0.89)a
21940 (0.24)b
29050 (1)b
18400 (0.99)a 17280 (1)a
21900 (3910)c
29070 (16000)c
19990
ν˜max(S1,abs) [cm-1
]
]
ν˜max(S2,abs) [cm-1
E(S1) [cm-1
]
20180
7040
19950
7110
d
∆E(S1,S2)max [cm-1
]
7170
e
a Intensity ratio of the vibrational bands in the fluorescence spectrum given in parentheses, see Figure 1. b Intensity ratio of the S1 and S2
absorption bands given in parentheses (Figure 1). c Molar absorption coefficient of the absorption band maximum εmax [M-1 cm-1]. d Crossing
point of the fluorescence and absorption spectra; see Figure 1 for THF and MeCN. e Energy difference ν˜max(S2,abs) - ν˜max(S1,abs).

13034 J. Phys. Chem. A, Vol. 114, No. 50, 2010
Zachariasse et al.
Figure 3. T-T absorption spectra in MeCN at 23 °C of 24DCDMA,
26DCDMA, TCDMA, and PCDMA. The T-T absorption spectrum
of PCDMA is obtained by senzitation (senz), because of its effectively
zero ISC quantum yield Φ(ISC) (see text).
N,N-dimethylaniline (26DCDMA), and TCDMA are shown in
Figure 3. These three spectra were determined by a procedure
employed previously.51 The T-T absorption maximum of
PCDMA is found at 585 nm, within the measured spectra range,
confirming that Φ(ISC) is less than 0.01 for PCDMA in MeCN
(Table 2). With the energy transfer method used to determine
the triplet yield of the other DMABN derivatives, the Φ(ISC)
also was equal to zero (i.e., <0.01).
Internal Conversion S1(LE) f S0. From the fluorescence
yield Φ(LE) ) 0.0006 and the intersystem crossing yield
Φ(ISC) < 0.01 for PCDMA in MeCN, a quantum yield of Φ(IC)
) 1.0 - (Φ(LE) + Φ(ISC)) g 0.99 is calculated for the internal
conversion (IC) process S1(LE) f S0. From Φ(IC) and the
fluorescence decay time τ3(LE), a rate constant kIC ) Φ(IC)/
τ3(LE) of around 3.6 × 1010 s-1 at room temperature results
(Table 2). The IC process from LE to S0 is much slower for
TCDMA (25 × 107 s-1, in MeCN) and DMABN (1.3 × 107
s-1, in n-hexane) than for PCDMA (Table 2). This may be due
to the smaller energy difference (energy gap law)52 between
LE and S0 for PCDMA (19990 cm-1) as compared with
TCDMA (24580 cm-1) and DMABN (31830 cm-1).
Figure 2. Fluorescence decay of PCDMA in (a) MeCN and (b) EtCN
at 25 °C. The decay times τi with the corresponding amplitudes Ai (see
eq 1) are depicted in the figure. The shortest decay times τ3 ) 27.4 ps
(MeCN) and τ3 ) 30.7 ps (EtCN) are attributed to the LE f S0
deactivation. The two other decay times τ2 and τ1 may originate from
impurities or photoproducts. The weighted deviations σ, the autocor-
relation functions A-C and the values for ꢀ2 are also indicated.
Excitation wavelength: 272 nm. Emission wavelength: 565 nm. Time
resolution: 0.496 ps/channel, with a time window of 1000 effective
channels.
TABLE 2: Fluorescence Quantum Yield Φ(LE),
Fluorescence Lifetime τ3(LE) and Radiative Rate Constant
kf(LE) at 25 °C for the LE State of PCDMA and TCDMA in
MeCN and DMABN in n-Hexane
PCDMA
TCDMAa
DMABNb
Internal Conversion of D/A Molecules Comparable with
PCDMA. With 9-cyano-10-(dimethylamino)anthracene (CDA)
in an alkane solvent such as cyclopentane, only LE fluorescence
is observed.53 For this electron donor (D)/acceptor (A) molecule
CDA, with an energy of E(S1) ) 20140 cm-1, which is only
slightly larger than that of PCDMA, the fluorescence quantum
yield Φ(LE) at 25 °C is likewise very small (0.0002), as IC
also has become the dominant deactivation process, with Φ(IC)
) 0.95. Similar to PCDMA, this large IC yield is accompanied
by a very short LE fluorescence decay time τ of 12 ps (at -70
°C). The occurrence of the dominant IC process in CDA has
been connected with the S0 amino twist angle of 56° (crystal)
relative to the anthracene plane of CDA.53
solvent
Φ(LE)
MeCN
0.0006
27.4c
2.2
<0.01
0.99
MeCN
0.021
1790
1.2
0.53 ( 0.02
0.45
n-hexane
0.127
3430
3.7
0.83 ( 0.05
0.04
τ3(LE) [ps]
kf(LE) [107 s-1
Φ(ISC)d
Φ(IC)
]
kIC [107 s-1
]
]
3600
19990
25
24580
1.3
E(S1) [cm-1
31830e
a Reference 25. b Reference 51a. c 30.7 ps in EtCN at 25 °C; see
Figure 2b. d Measurements (at 23 °C) as in ref 51. Φ(ISC) ) 0.0
(<0.02) for PCDMA in diethyl ether. e E(S1) ) 29990 cm-1 in
MeCN.
the intersystem crossing yield Φ(ISC) equals zero, within an
accuracy range of 0.01 (Table 2). As an outcome of an
effectively zero yield is always problematic, the T-T absorption
spectrum was measured, to make sure that the correct spectral
range had been investigated. By using naphthalene as sensitizer
(308 nm excitation), the absolute T-T absorption spectrum of
PCDMA could be obtained (Figure 3). Utilizing the molar
absorption coefficient ε ) 24500 M-1 cm-1 for naphthalene at
415 nm,50 εmax ) 5800 M-1 cm-1 is calculated for PCDMA at
the T-T absorption maximum of 585 nm, following the
procedure from ref 50. For comparison, T-T absorption spectra
of 2,4-dicyano-N,N-dimethylaniline (24DCDMA), 2,6-dicyano-
A second example of a D/A molecule with strongly enhanced
IC deactivation of LE, is 1-(dimethylamino)-4-cyanonaphthalene
(14DMCN).54 In n-hexane at 25 °C (E(S1) ) 27600 cm-1),
Φ(IC) is 0.99 and τ is 5 ps, giving kIC ) Φ(IC)/τ ) 1.72 ×
1011 s-1. This ultrafast IC process is attributed to the presence
of a substantial amino twist angle θ (48°, AM1) in S0, as
compared with a largely planar structure in the relaxed S1 state.
The IC reaction pathway has been assumed to pass through a
conical intersection of the S1 and S0 potential energy surfaces,
existing as a consequence of a horizontal shift of the S1 surface,
coming closer to that of S0. This shift is caused by the different
amino twist θ angles (and bond lengths) in these states, which

PCDMA in the Excited State
J. Phys. Chem. A, Vol. 114, No. 50, 2010 13035
TABLE 3: Reduction Potentials E(A-/A) and Adiabatic
Electron Affinities aEA(A) for Cyano-Substituted Benzenes
and the Oxidation Potential E(D/D+) for CH3N(C2H5)2
reduces the energy gap between the relaxed S1 state and the
corresponding (planarized) FC ground state, leading to a faster
IC deactivation.54
E(A-/A)a [V vs SCE] aEA(A)b [eV]
A similar IC enhancement connected with a decrease in amino
twist angle θ between S0 and S1 is observed with 1-(dimethyl-
amino)naphthalene (1DMAN).55 For 1DMAN, this antitwist
motion leads to a decrease of θ from 50° in S0 to around 30° in
S1,55b thereby reducing the energy gap between the relaxed S1
and its FC S0 state, inducing a fast IC deactivation in accordance
with the energy gap law. In conclusion, with CDA, 14DMCN,
and 1DMAN, the IC efficiencies are determined by the structural
differences (amino twist angle) between S1 and S0. Also in the
present case of PCDMA, the fast IC reaction from LE to S0
will be caused by the planarization of the amino twist angle θ,
when going from S0 (37.3°, AM1) to the equilibrated S1(LE)
state.
benzonitrile
-2.36
-1.68
-1.95
-1.63
-1.36
-0.61
(-0.29)c
-0.02
0.874
1.551
1.470
1.671
1.8878
2.711
3.0216
3.270
1,2-dicyanobenzene
1,3-dicyanobenzene
1,4-dicyanobenzene
1,3,5-tricyanobenzene
1,2,4,5-tetracyanobenzene
pentacyanobenzene
HCB
E(D/D+) [V vs SCE]
CH3N(C2H5)2
0.91
a In MeCN (see ref 46a). b Reference 58. c Interpolated value,
from a plot of E(A-/A) versus aEA(A).
Dual LE + ICT Fluorescence with Cyano-Substituted
N,N-Dimethylanilines. For the 16 N,N-dimethylanilines sub-
stituted by two to five cyano groups, discussions of the
possibility of an ICT reaction have appeared in the literature
only for 34DCDMA, 35DCDMA, and TCDMA, as mentioned
in the Introduction.20-22,25-28 From the observation of single
exponential fluorescence decays with picosecond time resolution
for 34DCDMA in DEE and 35DCDMA in DEE and MeCN at
different temperatures, it was concluded than a LE f ICT
reaction does not take place with these molecules. The fact that
an ICT emission was not detected in the fluorescence spectra
supported this conclusion. Also in the case of TCDMA,25 the
LE fluorescence decays and spectra do not give an indication
of the appearance of an ICT reaction. CASPT2/CASSCF
calculations on TCDMA likewise lead to the conclusion that
the lowest excited singlet state S1 determines its photophysical
behavior, without the occurrence of a LE f ICT reaction, in
the sense that the initially excited LE state already has a strong
CT character, and there is no equilibrium between two electronic
states with strongly different electronic structures (i.e., LE and
ICT with very different dipole moments) leading to dual (LE
+ ICT) fluorescence.25
mined from the experimental enthalpy difference ∆H for the
electron transfer reaction from 1A* or 1D* to (A-D+) in
n-hexane at room temperature: E(A-D+) ) E(S1) + ∆H, where
E(S1) is the energy of the excited singlet state of 1A* or 1D* in
the S1 state.
Although eq 2 is strictly only valid for electronically
decoupled A and D subgroups, as in intermolecular exciplexes
(A-D+), this expression has been used to calculate the energy
E(ICT) of the intramolecular ICT state of D/A molecules such
as DMABN within the context of the TICT model.4a,9,34-36 The
adoption of eq 2 for the calculation of E(ICT) is in fact a logical
consequence of the basic TICT assumption that, in the ICT state
of D/A molecules such as DMABN, the D and A subgroups
are in a mutually perpendicular configuration. This configuration
leads to an electronic decoupling of D and A (principle of
minimum overlap),9 similar to that present in intermolecular
exciplexes.
Energies E(A-D+) for Cyano-Substituted N,N-Dimethyl-
anilines. For the calculation of the energy E(A-D+) via eq 2,
E(D/D+) - E(A-/A) are needed. In the case of the cyano-
substituted N,N-dimethylanilines DMABN, 2DMABN, 3DMABN,
the six xyDCDMAs, TCDMA, and PCDMA with electronically
decoupled D and A subgroups, the D moiety is approximated with
N(CH3)3, whereas for the A subunits the corresponding cyanoben-
zenes are taken. With 4-(diethylamino)benzonitrile (DEABN),
CH3N(C2H5)2 serves as the D subgroup.
A recent investigation56 of the excited state behavior of
all six x,y-dicyano-substituted N,N-dimethylanilines xyD-
CDMA (23DCDMA, 24DCDMA, 25DCDMA, 26DCDMA,
34DCDMA, and 35DCDMA) reveals that the fluorescence
spectrum of these molecules in n-hexane at 25 °C indeed
consists of a single LE emission band. In the polar solvent
MeCN at 25 °C, a single LE fluorescence band is also
observed for 23DCDMA, 25DCDMA, 26DCDMA, and
35DCDMA, whereas a small amount of ICT fluorescence is
found with 34DCDMA in MeCN (Φ′(ICT)/Φ(LE) ) 0.34)
and more ICT with 24DCDMA (Φ′(ICT)/Φ(LE) ) 1.23) (see
Table 4).
Energetics of A/D Molecules Having a Hypothetical ICT
State with Electronically Decoupled A and D Subgroups.
As already outlined in the Introduction, the energy E(A-D+) of
intermolecular exciplexes (A-D+) with weakly interacting A and
D chromophores, can be calculated from the reduction and
oxidation potentials E(A-/A) and E(D/D+) via the semiempirical
Weller equation (eq 2), valid for the solvent n-hexane.26,37-40
For E(D/D+) the following values are employed: 1.05 V57 vs
SCE for N(CH3)3 and 0.91 V vs SCE for CH3N(C2H5)2. The
E(A-/A) data for the cyanobenzenes are collected in Table 3.
The reduction potential of pentacyanobenzene (-0.29 V vs SCE)
is obtained by interpolation from a plot of the experimental
E(A-/A) data against the adiabatic electron affinities aEA(A)58
(Table 3).
Energies E(ICT) of DMABN in MeCN and n-Hexane. For
DMABN in MeCN at 25 °C, the energy E(ICT) of the ICT
state above the equilibrated S0 ground state is known from
experiments. Its value of 27730 cm-1 has been determined from
the enthalpy difference of the LE f ICT reaction ∆H ) -27.0
kJ/mol (-2260 cm-1) and the 0-0 energy of the LE state E(S1)
) 29990 cm-1: E(ICT) ) E(S1) + ∆H.12,19 For DMABN in
n-hexane at 25 °C, E(ICT) ) 31590 cm-1, similarly calculated
from the extrapolated ∆H ) -2.9 kJ/mol (-240 cm-1) and the
E(A-D+) ) E(D/D+) - E(A-/A) + 0.15 ( 0.10 eV
(2)
0-0 energy of the LE state E(S1) ) 31830 cm-1 12,19
It follows
.
The factor 0.15 in eq 2 is obtained by comparing experimental
data for E(A-D+) with the difference in redox potentials
E(D/D+) - E(A-/A).38 The energy E(A-D+) has been deter-
from these data that, when going from n-hexane to MeCN, the
E(ICT) of DMABN is lowered by 3860 cm-1 (0.48 eV). This
energy difference of around 0.5 eV is adopted to calculate the

13036 J. Phys. Chem. A, Vol. 114, No. 50, 2010
Zachariasse et al.
energy E(A-D+) in MeCN from that in n-hexane obtained via
eq 2 (Table 4).
ICT Energetics for A/D Molecules. As already pointed out
previously,19,20,22,25-28 an analysis of the outcome of eq 2 in
relation to E(S1) leads to the conclusion that the TICT approach
can not explain the absence of an ICT reaction with 2DMABN,
3DMABN, and TCDMA. For the discussion of the ICT
energetics of PCDMA, the energies E(A-D+) (eq 2) and E(S1)
in n-hexane and MeCN are presented in Table 4 for a series of
cyano-substituted N,N-dimethylanilines.
DMABN, DEABN, 2DMABN, and 3DMABN: E(A-D+)
and E(S1) (Table 4). Of the D/A molecules DMABN, DEABN,
2DMABN, and 3DMABN with one CN-substituent appearing
in this article, only DMABN and DEABN undergo a LE f
ICT reaction, in sufficiently polar solvents, from dialkyl ethers
to MeCN.1-23,25-28,43,46a In nonpolar alkanes such as n-hexane,
practically no ICT emission is found for DMABN, whereas a
minor ICT emission band could be detected with DEABN, with
a small ICT/LE fluorescence quantum yield ratio Φ′(ICT)/
Φ(LE) of 0.04.46a In contrast, 2DMABN and 3DMABN do not
show an ICT reaction with dual LE + ICT fluorescence at any
condition of solvent polarity or temperature.20-23
From inspection of the ICT energetics in Table 4 it becomes
clear, however, that for DMABN in n-hexane, E(A-D+) is, at
3.56 eV, substantially lower than E(S1) ) 3.95 eV, which would
predict an efficient ICT reaction, contrary to observation.9,19,51a
The same is the case for DEABN in n-hexane (E(A-D+) )
3.42 eV, E(S1) ) 3.84 eV), although ICT emission only just
starts to appear in this solvent.46a
For 2DMABN in n-hexane, the calculated E(A-D+) is, at
3.56 eV, slightly larger than E(S1) ) 3.52 eV, whereas with
3DMABN, E(A-D+) ) E(S1) ) 3.56 eV. These results can still
be considered to be marginally in accordance with the absence
of ICT emission. In MeCN, however, E(A-D+) is clearly smaller
than E(S1) for both molecules, with an energy difference of
0.30 eV (-29 kJ/mol) for 2DMABN and 0.26 eV (-25 kJ/
mol) for 3DMABN, which predicts that an ICT reaction
would be energetically feasible and also efficient, different
from the experimental finding of a single LE emission. It
should be noted that, for DMABN in MeCN with a
comparable ∆H of -27.0 kJ/mol, a fast (4 ps) LE f ICT
reaction occurs, with Φ′(ICT)/Φ(LE) ) 39.5.12
xyDCDMA, TCDMA. For the six xyDCDMAs in n-hexane,
the calculated E(A-D+) is smaller than E(S1), with an energy
difference between 0.59 eV for 34DCDMA and 0.10 eV for
26DCDMA. In the fluorescence spectra of the six xyDCDMAs,
an ICT emission could not be detected next to the LE emission
band, which again shows that the TICT approach leads to
incorrect results with these D/A molecules. In MeCN, the energy
difference E(A-D+) - E(S1) is larger than in n-hexane (Table
4), now ranging between 0.83 eV for 34DCDMA and 0.47 eV
for 26DCDMA, predicting a very efficient (and fast) ICT
reaction for all six isomers. From the appearance of a single
LE fluorescence band for 23DCDMA, 25DCDMA, and
26DCDMA in MeCN at 25 °C, it follows that an ICT reaction
does not take place, contrary to the TICT prediction. Only
with 24DCDMA and 34DCDMA is dual ICT + LE fluores-
cence observed in MeCN at 25 °C, indicating that a LE f
ICT reaction occurs with these molecules. The ratio of
Φ′(ICT)/Φ(LE) is larger for 24DCDMA (1.23) than for
34DCDMA (0.34), although the other way around would have
been expected from the relative values of E(A-D+) - E(S1):
0.66 eV (24DCDMA) and 0.83 eV (34DCDMA) (see Table
4).

PCDMA in the Excited State
J. Phys. Chem. A, Vol. 114, No. 50, 2010 13037
Also for TCDMA, E(A-D+) is considerably smaller than
E(S1), with a difference of 0.63 eV in n-hexane, increasing in
MeCN to 0.99 eV (Table 4). Nevertheless, against this predic-
tion, an ICT reaction is not observed with TCDMA in these
solvents.25 These results show that, from a calculation of
E(A-D+) based on the TICT hypothesis of perpendicularly
twisted and hence electronically decoupled A and D moieties,
it cannot be predicted when, under experimental conditions
E(A-D+) e E(S1), the energetic requirement for a LE f ICT
reaction will be fulfilled for a D/A system.
PCDMA. Likewise for PCDMA in MeCN, the calculated
E(A-D+) is substantially smaller than E(S1): 0.99 eV as
compared with 2.48 eV, a very large difference of 1.47 eV (142
kJ/mol). As seen from Figure 1, only LE emission is observed
with PCDMA in MeCN at 25 °C, making it clear that an ICT
reaction does not take place. Again, the TICT approach evidently
is not applicable to PCDMA.
Energy Gap ∆E(S1,S2) and ICT. As an integral part of the
PICT model, the energy gap ∆E(S1,S2) is of prime importance
as a diagnostic and predictive tool for the occurrence or absence
of a LE f ICT reaction.20,22,26,28 Its usefulness has been
established for the aminobenzonitriles DMABN,12,19 4-(diiso-
propylamino)benzonitrile (DIABN),11,25,46a,59 and other D/A
molecules, such as N-phenylpyrroles and fluorazenes.46b,60
For DMABN, ∆E(S1,S2) is 2700 cm-1 in n-hexane and around
2400 cm-1 for DEABN in this solvent (Table 4). The two other
monocyano-substituted N,N-dimethylanilines have considerably
larger energy gaps ∆E(S1,S2) in n-hexane: 7710 cm-1 for
2DMABN and 7940 cm-1 for 3DMABN. In MeCN, the gaps
are somewhat larger: 8060 cm-1 (2DMABN) and 8070 cm-1
(3DMABN) (see Table 4). As discussed above, the increase of
∆E(S1,S2) for 2DMABN and 3DMABN as compared with
DMABN and DEABN is accompanied by the disappearance
of ICT fluorescence.
Figure 4. PCDMA in MeCN at 360 nm excitation. (a) Transient
absorption spectra and (b) ESA spectra, at pump-probe delay times
between 1.0 and 80 ps, after subtraction of the BL and SE. The BL
and SE (LE, cf. Figure 1b) spectra are also depicted. The spectral range
is 270-690 nm. (c) Band integrals for the transient absorption spectra
(TrA) in panel a. A global analysis of the band integrals BI(485,550),
BI(320,365), and BI(400,485) results in a decay time τ1 of 22 ps.
BI(320,365) and BI(400,485) show a decrease in absorption (negative
amplitude A1, eq 3) due to BL. The offset A0, equal to zero, is also
presented (eq 3). (d) Band integrals for the ESA spectra in panel b.
From a global analysis of the band integrals BI(275,300), BI(370,410),
and BI(275,300), a decay time τ1 of 22 ps is likewise obtained. The
observation that A0 ) 0 in panels c and d indicates that LE f S0 is the
only deactivation process. m∆OD is the optical density/1000.
For the six xyDCDMAs, in the order of increasing ∆E(S1,S2),
the following values are obtained for n-hexane/MeCN (Table
4): 24DCDMA, 4840/4870 cm-1; 34DCDMA, 4360/4610 cm-1;
26DCDMA, 6635/6858 cm-1; 35DCDMA, 7200/7450 cm-1
;
.
23DCDMA, 8560/8950 cm-1; 25DCDMA, 9270/10040 cm-1
From inspection of these data, it is seen that an ICT reaction is
observed for the two DCDMAs with the smallest energy gaps:
24DCDMA and 34DCDMA.
In the case of TCDMA, ∆E(S1,S2) is 5240 cm-1 in n-hexane
and 5370 cm-1 in MeCN. With PCDMA in MeCN, a somewhat
larger gap of 7170 cm-1 is found. The energy differences
∆E(S1,S2) are much larger for TCDMA and PCDMA than for
DMABN, DEABN, 24DCDMA, and 34DCDMA (Table 4),
which is in accordance with the finding that TCDMA and
PCDMA do not undergo an ICT reaction. It clearly follows from
the discussion presented here that the magnitude of ∆E(S1,S2)
is indeed important in connection with the occurrence or absence
of an ICT reaction with D/A molecules.
Femtosecond Transient Absorption Spectra: PCDMA in
MeCN. The femtosecond transient absorption spectra of PCDMA
were measured in MeCN over the spectral range 270-690 nm.
Experiments in alkane solvents such as n-hexane were not possible,
because of too low solubility of the compound.
The transient absorption spectra of PCDMA in MeCN at 22
°C for pump-probe delay times between 1.0 and 80 ps are
shown in Figure 4a. After correction for bleaching (BL) and
stimulated emission (SE), the excited state absorption (ESA)
spectra are obtained (Figure 4b). The ESA spectra consist of a
major peak at 290 nm and smaller absorption maxima at 380
and 520 nm (Table 5).
TABLE 5: ESA Maxima (in nm) of PCDMA in MeCN at
22 °Ca
PCDMA
TCDMAb
DMABNc
290, 380, 520 (LE)
310, 335, 510, 580 (LE)
320, 355, 440, 710 (LE)
315,425 (ICT)
a See Figures 4 and 5. Data for TCDMA and DMABN are
included for comparison. b Reference 25. c Reference 12.
The band integral BI(485,550) between 485 and 550 nm in
the transient absorption (TrA) spectra in Figure 4c shows a decay
(positive amplitude A1 (eq 3)), whereas BI(320,365) and
BI(400,485) present a growing-in negative amplitude A1. A
global analysis of the three BIs results in a single decay time τ1
) 22 ps. The growing-in of BI(320,365) and BI(400,485) is
caused by the return of the bleached PCDMA molecules from
LE to the ground state S0. The observation that the offset A0 is
equal to zero indicates that LE f S0 is the only deactivation

13038 J. Phys. Chem. A, Vol. 114, No. 50, 2010
Zachariasse et al.
process, without the intermediacy of a triplet or an ICT state.
A global analysis of ESA band integrals BI(275,300), BI(370,410),
and BI(490,550) (see Figure 4d) results in a single decay time
of τ1 ) 22 ps (eq 3), the same as that found for the band integrals
in Figure 4c. Also, these BIs decay to zero: a direct decay from
LE to S0. The ESA decay time of 22 ps is in reasonable
agreement with the 27 ps obtained from the LE fluorescence
decay of PCDMA in MeCN at 25 °C (Figure 2a).
Conclusions
The D/A molecule PCDMA was synthesized to investigate
whether the strong increase in the electron affinity of the
acceptor subunit A leads to an enhancement of the LE f ICT
reaction with respect to DMABN. The difference in electron
affinity of the subunits A in PCDMA and DMABN follows from
their reduction potentials: E(A-/A) ) -0.29 V vs SCE for
pentacyanobenzene and E(A-/A) ) -2.36 V vs SCE for
benzonitrile. In the TICT model, this difference in electron
affinity should lead to a lowering of the energy E(ICT) of
PCDMA by 2.05 eV (16550 cm-1), which is clearly more than
sufficient to compensate for the decrease by 10000 cm-1 of E(S1)
BI ) A1 exp(-t/τ1) + A0
(3)
A proviso should be made for the in principle possible
situation that the decay time τ′o(ICT) would be effectively
shorter than 1 ps, the smallest pump-probe delay time in Figure
4. Such a short-lived ICT state would not have been detected
in the present transient absorption spectra, although this extreme
condition has not been encountered with other D/A molecules.
Comparison of LE ESA Spectra for DMABN, TCDMA,
and PCDMA in MeCN. The ESA spectrum of PCDMA in
MeCN at 22 °C in Figure 5a resembles the ESA spectrum of
the LE state of TCDMA in this solvent (Figure 5b; see Table
5). Both ESA spectra show similarities with the LE ESA
spectrum of DMABN in MeCN (Figure 5c), but not with its
ICT ESA spectrum, which has a major band at 318 nm and a
small peak at 425 nm, but no appreciable absorption above 600
nm (Table 5).12 The ESA spectrum of PCDMA in MeCN is
therefore attributed to the LE state, supporting the conclusion
deduced from the fluorescence spectra that a LE f ICT reaction
does not take place with this molecule.
for PCDMA (19990 cm-1) relative to DMABN (29990 cm-1
)
in MeCN. Therefore, when the TICT approach would hold for
PCDMA, an efficient LE f ICT reaction should have been
observed. This is not the case, however, as the fluorescence
spectrum of PCDMA in the strongly polar MeCN consists of a
single LE emission band. The role of the redox potentials of
the D and A subunits in predicting the possibility of an ICT
reaction is discussed for a series of D/A molecules such as the
six dicyano-substituted N,N-dimethylanilines.
The fluorescence quantum yield Φ(LE) of PCDMA in MeCN
at 25 °C has a value of 0.0006 and the LE f S0 deactivation
predominantly takes place by IC, as Φ(ISC) is practically zero.
The fluorescence decays in MeCN and EtCN have major short
LE decay times of 27 ps in MeCN and 31 ps in EtCN at 25 °C.
The radiative rate constant kf(LE) of PCDMA in MeCN equals
2 × 107 s-1, a value similar to that of the kf(LE) for DMABN
and TCDMA. The LE f S0 IC reaction of PCDMA in MeCN
is very fast at 25 °C, with a rate constant kIC )3.6 × 1010 s-1
,
which is much faster than that for TCDMA (25 × 107 s-1) and
DMABN (1.3 × 107 s-1, in n-hexane). This is due to the
relatively small value of E(S1) of PCDMA (19990 cm-1), as
compared with TCDMA (24580 cm-1) and, in particular,
DMABN (31830 cm-1). The substantial amino twist angle θ )
37.3° of PCDMA in S0 leads to an enhancement of the IC
deactivation process LE f S0, similar to what has been found
with the D/A molecules CDA and 14DMCN, as well as with
1DMAN.
The femtosecond transient spectra of PCDMA in MeCN at
22 °C originate from the LE state, indicating that an ICT reaction
does not take place. From the ESA spectra, decay times of 22
ps are determined for the decay of the LE ESA absorption as
well as that of the ground-state BL. This shows that the
deactivation of the LE state directly proceeds to S0, without
passing through an intermediate ICT or triplet state. That there
is no triplet intermediate is also concluded from the fact that
direct T-T absorption is negligible (Φ(ISC) < 0.01). The ESA
decay time of 22 ps is in reasonable agreement with the 27 ps
obtained from the LE fluorescence decay at 25 °C.
The observation that a LE f ICT reaction does not occur
for PCDMA in MeCN means that, in the ICT state, the
dimethylamino and pentacyanobenzene subunits are not elec-
tronically decoupled by a perpendicular (90°) twist, as required
by the TICT hypothesis, as otherwise E(ICT) would be
substantially lower than E(S1). The absence of an ICT reaction
is attributed to the relative large energy gap ∆E(S1,S2), in
accordance with the PICT model.
Acknowledgment. We are grateful to Prof. K. Friedrich
(Universita¨t Freiburg) for a gift of 1,3,5-trichloro-2,4,6-tricy-
anobenzene and helpful advice on the synthesis of PCDMA.
Many thanks are due to Prof. N. P. Ernsting (Humboldt
University Berlin) for the use of the femtosecond equipment in
Figure 5. LE ESA spectra in MeCN at 22 °C of (a) PCDMA, (b)
TCDMA, and (c) DMABN, at pump-probe delay times of 1.0, 10,
and 0.2 ps, respectively.

PCDMA in the Excited State
J. Phys. Chem. A, Vol. 114, No. 50, 2010 13039
(29) Gedeck, P.; Schneider, S. J. Photochem. Photobiol. A: Chem. 1999,
the investigations reported here. We are grateful to Mr. W.
Bosch for the synthesis of PCDMA. We also thank Mr. J.
Bienert for carrying out HPLC purifications and Mr. H. Lesche
for technical support.
121, 7.
(30) (a) Go´mez, I.; Mercier, Y.; Reguero, M. J. Phys. Chem. A 2006,
110, 11455. (b) Go´mez, I.; Reguero, M.; Boggio-Pasqua, M.; Robb, M. A.
J. Am. Chem. Soc. 2005, 127, 7119.
(31) In ref 30, an energy gap ∆E(S1,S2) of 2520 cm-1 is computed for
DMABN in the gas phase, whereas in ref 19 ∆E(S1,S2) ) 3520 cm-1 is
determined from experiments and extrapolations.
Note Added after ASAP Publication. This article posted
ASAP on November 24, 2010. Figure 1 has been revised. The
correct version posted on December 1, 2010.
(32) Parusel, A. B. J.; Schamschule, R.; Kohler, G. Z. Phys. Chem. 2002,
216, 361.
(33) Jamorski Jo¨dicke, Ch.; Luthi, H.-P. Chem. Phys. Lett. 2003, 368,
561.
References and Notes
(34) Grabowski, Z. R.; Dobkowski, J. Pure Appl. Chem. 1983, 55, 245.
(35) Grabowski, Z. R.; Rotkiewicz, K.; Siemiarczuk, A.; Cowley, D. J.
NouV. J. Chim. 1979, 3, 443.
(1) Lippert, E.; Lu¨der, W.; Boos, H. In AdVances in Molecular
Spectroscopy; European Conference on Molecular Spectroscopy, Bologna,
Italy, 1959; Mangini, A., Ed.; Pergamon Press: Oxford, U.K., 1962; p 443.
(2) Lippert, E.; Lu¨der, W.; Moll, F.; Na¨gele, W.; Boos, H.; Prigge,
H.; Seibold-Blankenstein, I. Angew. Chem. 1961, 73, 695.
(3) Rettig, W. Angew. Chem., Int. Ed. Engl. 1986, 25, 971.
(4) (a) Leinhos, U.; Ku¨hnle, W.; Zachariasse, K. A. J. Phys. Chem.
1991, 95, 2013. (b) Il’ichev, Yu. V.; Ku¨hnle, W.; Zachariasse, K. A. J.
Phys. Chem. A 1998, 102, 5670.
(36) Rettig, W. Angew. Chem., Int. Ed. Engl. 1986, 25, 971.
(37) Weller, A. In The Exciplex; Gordon, M., Ware, W. R., Eds.;
Academic Press: New York, 1975; p 23.
(38) Weller, A. Z. Phys. Chem. 1982, 133, 93.
(39) The Coulombic term C(A-D+) determining E(A-D+) in eq 2 (refs
37 and 38), obviously depends on the charge delocalization in D+ and A-.
This delocalization can lead to substantial differences in C(A-D+) for
exciplexes with aromatic amines as compared with aliphatic amines, such
as the electron donor D, as the charges are much less delocalized in the
radical cations of the latter molecules. Due to the localization of the negative
charge on the 9- and 10-positions of the radical anion of anthracene, for
example, C(A-D+) is always larger for anthracene exciplexes than for those
with other aromatic hydrocarbons. C(A-D+) is not a constant term, but for
a particular donor D decreases as A becomes larger (such as from benzene
to tetracene), as a result of the corresponding increased charge delocalization
in the radical anion A-. For the intramolecular ICT states of aminoben-
zonitriles, C(A-D+) will depend on the specific charge distribution and is
therefore different for DMABN than for 2DMABN and 3DMABN (ref 20).
(40) Zachariasse, K. A. Spectrum 2006, 19, 22.
(5) Lommatzsch, U.; Brutschy, B. Chem. Phys. 1998, 234, 35.
(6) Okamoto, H.; Inishi, H.; Nakamura, Y.; Kohtani, S.; Nakagaki, R.
J. Phys. Chem. A 2001, 105, 4182.
(7) (a) Ma, C.; Kwok, W. M.; Matousek, P.; Parker, A. W.; Phillips, D.;
Toner, W. T.; Towrie, M. J. Phys. Chem. A 2002, 106, 3294. (b) Kwok, W. M.;
Ma, C.; George, M. W.; Grills, D. C.; Matousek, P.; Parker, A. W.; Phillips, D.;
Toner, W. T.; Towrie, M. Photochem. Photobiol. Sci. 2007, 6, 987.
(8) (a) Fuꢁ, W.; Pushpa, K. K.; Rettig, W.; Schmid, W. E.; Trushin,
S. A. Photochem. Photobiol. Sci. 2002, 1, 255. (b) Fuꢁ, W.; Schmid, W. E.;
Pushpa, K. K.; Trushin, S. A.; Yatsuhashi, T. Phys. Chem. Chem. Phys.
2007, 9, 1151.
(9) Grabowski, Z. R.; Rotkiewicz, K.; Rettig, W. Chem. ReV. 2003,
103, 3899.
(41) See Table 3 in the text.
(42) Druzhinin, S. I.; Kovalenko, S. A.; Senyushkina, T.; Zachariasse,
K. A. J. Phys. Chem. A 2007, 111, 12878.
(43) Zachariasse, K. A.; Grobys, M.; von der Haar, Th.; Hebecker, A.;
Il’ichev, Yu. V.; Morawski, O.; Ru¨ckert, I.; Ku¨hnle, W. J. Photochem.
Photobiol. A: Chem. 1997, 105, 373.
(44) Friedrich, K.; Steiert, P. J. Prakt. Chem. 1995, 337, 313.
(45) Friedrich, K.; Oeckl, S. Chem. Ber. 1973, 106, 3803.
(46) (a) Galievsky, V. A.; Druzhinin, S. I.; Demeter, A.; Jiang, Y.-B.;
Kovalenko, S. A.; Pe´rez Lustres, L.; Venugopal, K.; Ernsting, N. P.; Allonas,
X.; Noltemeyer, M.; Machinek, R.; Zachariasse, K. A. Chem. Phys. Chem.
2005, 6, 2307. (b) Druzhinin, S. I.; Kovalenko, S. A.; Senyushkina, T. A.;
Demeter, A.; Machinek, R.; Noltemeyer, M.; Zachariasse, K. A. J. Phys.
Chem. A 2008, 112, 8238. Erratum: J. Phys. Chem. A 2009, 113, 520.
(47) Karunakaran, V.; Kleinermann, K.; Improta, R.; Kovalenko, S. A.
J. Am. Chem. Soc. 2009, 131, 5839.
(10) Yoshihara, T.; Galievsky, V. A.; Druzhinin, S. I.; Saha, S.;
Zachariasse, K. A. Photochem. Photobiol. Sci. 2003, 2, 342.
(11) Zachariasse, K. A.; Druzhinin, S. I.; Bosch, W.; Machinek, R. J. Am.
Chem. Soc. 2004, 126, 1705.
(12) Druzhinin, S. I.; Ernsting, N. P.; Kovalenko, S. A.; Pe´rez Lustres,
L.; Senyushkina, T.; Zachariasse, K. A. J. Phys. Chem. A 2006, 110, 2955.
(13) Druzhinin, S. I.; Galievsky, V. A.; Zachariasse, K. A. J. Phys. Chem.
A 2005, 109, 11213.
(14) Lee, J.-K.; Fujiwara, T.; Kofron, W. G.; Zgierski, M. Z.; Lim, E. C.
J. Chem. Phys. 2008, 128, 164512.
(15) Pigliucci, A.; Vauthey, E.; Rettig, W. Chem. Phys. Lett. 2009, 469,
115.
(16) Gustavsson, T.; Coto, P. B.; Serrano-Sandres, L.; Fujiwara, T.; Lim,
E. C. J. Chem. Phys. 2009, 131, 031101.
(17) Zachariasse, K. A.; Druzhinin, S. I.; Mayer, P.; Kovalenko, S. A.;
Senyushkina, T. Chem. Phys. Lett. 2009, 484, 28.
(18) Zachariasse, K. A.; Druzhinin, S. I.; Kovalenko, S. A.; Senyushkina,
T. J. Chem. Phys. 2009, 131, 224313.
(19) Druzhinin., S. I.; Mayer, P.; Stalke, D.; von Bu¨low, R.; Noltemeyer,
M.; Zachariasse, K. A. J. Am. Chem. Soc. 2010, 132, 7730.
(20) Zachariasse, K. A.; von der Haar, Th.; Hebecker, A.; Leinhos, U.;
Ku¨hnle, W. Pure Appl. Chem. 1993, 65, 1745.
(48) (a) Berlman, I. B. J. Phys. Chem. 1973, 77, 562. (b) Bowen, E. J.;
Williams, A. H. Trans. Faraday Soc. 1939, 35, 765.
(49) Forster, L. S.; Livingston, R. J. Chem. Phys. 1952, 20, 1315.
(50) Bensasson, R.; Land, E. Trans Faraday Soc. 1971, 67, 1904.
(51) (a) Druzhinin, S. I.; Demeter, A.; Galievsky, V. A.; Yoshihara, T.;
Zachariasse, K. A. J. Phys. Chem. A 2003, 107, 8075. (b) Yoshihara, T.;
Druzhinin, S. I.; Demeter, A.; Kocher, N.; Stalke, D.; Zachariasse, K. A. J.
Phys. Chem. A 2005, 109, 1497.
(21) Rettig, W.; Bliss, B.; Dirnberger, K. Chem. Phys. Lett. 1999, 305, 8.
(22) Zachariasse, K. A. Chem. Phys. Lett. 2000, 320, 8.
(23) Oshima, J.; Shiobara, S.; Naoumi, H.; Kaneko, S.; Yoshihara, T.;
Mishra, A. K.; Tobita, S. J. Phys. Chem. A 2006, 110, 4629.
(24) The definition of the term ‘LE state’. When an electron donor (D)/
acceptor (A) molecule only emits a single fluorescence band, this is called
a LE fluorescence, irrespective of the dipole moment µe(LE) of this state.
In such a case, we say that an ICT reaction does not take place. We speak of
an ICT reaction only when a new excited ICT state is produced from the relaxed
initially excited S1(LE) state as the precursor. For this ICT state, the condition
µe(ICT) > µe(LE) generally holds. A detailed discussion of the limits of the LE
nomenclature is presented in ref 25.
(25) Zachariasse, K. A.; Druzhinin, S. I.; Galievsky, V. A.; Kovalenko,
S.; Senyushkina, T. A.; Mayer, P.; Noltemeyer, M.; Boggio-Pasqua, M.;
Robb, M. A. J. Phys. Chem. A 2009, 113, 2693.
(26) von der Haar, Th.; Hebecker, A.; Il’ichev, Yu., V.; Jiang, Y.-B.; Ku¨hnle,
W.; Zachariasse, K. A. Recl. TraV. Chim. Pays-Bas 1995, 114, 430.
(27) von der Haar, Th.; Hebecker, A.; Il’ichev, Yu.; Ku¨hnle, W.;
Zachariasse, K. A. In Fast Elementary Processes in Chemical and Biological
Systems: Proceedings of the 54th International Meeting of Physical
Chemistry; Tramer, A., Ed.; AIP Conference Proceedings 364; AIP Press:
New York, 1996; p 295.
(52) (a) Siebrand, W. J. Chem. Phys. 1967, 47, 2411. (b) Englman, R.;
Jortner, J. Mol. Phys. 1970, 18, 145. (c) Kakitani, T.; Mataga, N. J. Phys.
Chem. 1986, 90, 993. (d) Bixon, M.; Jortner, J.; Cortes, J.; Heitele, H.;
Michel-Beyerle, M. E. J. Phys. Chem. 1994, 98, 7289.
(53) Druzhinin, S. I.; Demeter, A.; Niebuer, M.; Tauer, E.; Zachariasse,
K. A. Res. Chem. Intermed. 1999, 25, 531.
(54) Suzuki, K.; Demeter, A.; Ku¨hnle, W.; Tauer, E.; Zachariasse, K. A.;
Tobita, S.; Shizuka, H. Phys. Chem. Chem. Phys. 2000, 2, 981.
(55) (a) Ru¨ckert, I.; Demeter, A.; Morawski, O.; Ku¨hnle, W.; Tauer,
E.; Zachariasse, K. A. J. Phys. Chem. A 1999, 103, 1958. (b) Lahmani, F.;
Zehnacker-Rentien, A.; Coudert, L. H.; Zachariasse, K. A. J. Phys. Chem.
A 2003, 107, 7364.
(56) Galievsky, V. A.; Druzhinin, S. I.; Zachariasse, K. A., in preparation.
(57) Mann, C. K. Anal. Chem. 1964, 36, 2424.
(58) Mirek, J.; Buda, A. Z. Naturforsch. 1984, 39A, 386.
(59) (a) Demeter, A.; Druzhinin, S.; George, M.; Haselbach, E.; Roulin,
J.-L.; Zachariasse, K. A. Chem. Phys. Lett. 2000, 323, 351. (b) Techert, S.;
Zachariasse, K. A. J. Am. Chem. Soc. 2004, 126, 5593.
(60) (a) Druzhinin, S. I.; Kovalenko, S. A.; Senyushkina, T. A.; Demeter,
A.; Zachariasse, K. A. J. Phys. Chem. A 2010, 114, 1621. (b) Yoshihara,
T.; Druzhinin, S. I.; Zachariasse, K. A. J. Am. Chem. Soc. 2004, 126, 8535.
(c) Druzhinin, S. I.; Kovalenko, S. A.; Senyushkina, T. A.; Demeter, A.;
Januskevicius, R.; Stalke, D.; Machinek, R.; Zachariasse, K. A. J. Phys.
Chem. A 2009, 113, 9304.
(28) Zachariasse, K. A.; Grobys, M.; von der Haar, Th.; Hebecker, A.;
Il’ichev, Yu. V.; Jiang, Y.-B.; Morawski, O.; Ku¨hnle, W. J. Photochem.
Photobiol. A: Chem. 1996, 102, 59. Erratum: J. Photochem. Photobiol. A:
Chem. 1998, 115, 259.
JP108804Q