Substituent and solvent effects on the excited state deactivation channels in anils and boranils

DOI: 10.1002/chem.201404669

Full Paper

&

Photochemistry

Substituent and Solvent Effects on the Excited State Deactivation

Channels in Anils and Boranils

[a]

Jacek Dobkowski,^[a]Paweł Wnuk,^{[a, b]}Joanna Buczynska, Maria Pszona,

´

Graz˙yna Orzanowska,^[a]Denis Frath,^[c]Gilles Ulrich,^[c]Julien Massue,^[c]Sandra Mosquera-

Vꢀzquez,^[d]Eric Vauthey,*^[d]Czesław Radzewicz,^[b]Raymond Ziessel,*^[c]and Jacek Waluk*^{[a, e]}

Abstract: Differently substituted anils (Schiff bases) and their

boranil counterparts lacking the proton-transfer functionality

have been studied using stationary and femtosecond time-

resolved absorption, fluorescence, and IR techniques, com-

bined with quantum mechanical modelling. Dual fluores-

cence observed in anils was attributed to excited state intra-

molecular proton transfer. The rate of this process varies

upon changing solvent polarity. In the nitro-substituted anil,

proton translocation is accompanied by intramolecular elec-

tron transfer coupled with twisting of the nitrophenyl group.

The same type of structure is responsible for the emission of

the corresponding boranil. A general model was proposed

to explain different photophysical responses to different

substitution patterns in anils and boranils. It is based on the

analysis of changes in the lengths of CN and CC bonds link-

ing the phenyl moieties. The model allows predicting the

contributions of different channels that involve torsional dy-

namics to excited state depopulation.

Introduction

excited-state deactivation, either in a competing or coopera-

tive way. The best known examples include proton-coupled

electron transfer^[1–8]and excited state electron transfer accom-

panied by structural changes.^[9–13]It has been recently pro-

posed that also photoinduced proton transfer can be coupled

to a conformational change, involving mutual twisting of

proton donor and acceptor moieties.^[14]

Understanding of the elementary steps of ultrafast photoin-

duced phenomena, such as electron or proton/hydrogen trans-

fer or conformational changes is the prerequisite for compre-

hensive description of basic natural processes: photosynthesis

and vision. It is also crucial for applications in numerous areas:

materials science, optoelectronics, information storage, design

of sensors, biology, medicine, and environmental protection. A

particular case, challenging both for experimentalists and theo-

reticians, arises when more than one channel participates in

In this work, we analyse the experimental and theoretical re-

sults obtained for a series of anils (aniline-imines) and boranils

(Scheme 1). These compounds can formally be derived from N-

salicylideneaniline (SA), the most investigated molecule of the

Schiff base family, famous for extremely complex behaviour in-

volving ultrafast excited state intramolecular proton transfer

(ESIPT), torsional dynamics, photochromism, thermochromism,

and solvatochromism.^[15–44]Upon photoexcitation, the enol

form of SA undergoes ultrafast intramolecular proton transfer

from the hydroxy group to the nitrogen atom. This leads to

a cis-keto form, which is subsequently converted into a photo-

chromic product (trans-keto tautomer).

´

[a] Dr. J. Dobkowski, Dr. P. Wnuk, Dr. J. Buczynska, M. Pszona, G. Orzanowska,

Prof. J. Waluk

Institute of Physical Chemistry, Polish Academy of Sciences

Kasprzaka 44, 01-224 Warsaw (Poland)

E-mail: waluk@ichf.edu.pl

[b] Dr. P. Wnuk, Prof. C. Radzewicz

Institute of Experimental Physics, Faculty of Physics

University of Warsaw, Hoz˙a 69, 00-681 Warsaw (Poland)

[c] Dr. D. Frath, Dr. G. Ulrich, Dr. J. Massue, Prof. R. Ziessel

Laboratoire de Chimie Organique et Spectroscopies Avancꢀes

UMR7515 au CNRS, ꢁcole de Chimie

Polymꢂres Matꢀriaux de Strasbourg, 25 rue Becquerel

67087 Strasbourg, Cedex 02 (France)

E-mail: ziessel@unistra.fr

[d] Dr. S. Mosquera-Vꢃzquez, Prof. E. Vauthey

Department of Physical Chemistry, University of Geneva

30 quai Ernest-Ansermet, 1211 Geneva 4 (Switzerland)

E-mail: eric.vauthey@unige.ch

[e] Prof. J. Waluk

Faculty of Mathematics and Natural Sciences

´

College of Science, Cardinal Stefan Wyszynski University

Dewajtis 5, 01-815 Warsaw (Poland)

Supporting information for this article is available on the WWW under

http://dx.doi.org/10.1002/chem.201404669.

Scheme 1. The investigated anils and boranils.

Chem. Eur. J. 2014, 20, 1 – 17

1

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

However, both experiment

and theory suggest that several

other channels may be compet-

ing with or following the proton

transfer (Scheme 2). These in-

clude twisting about CN or CC

1

bonds, pp*! np* internal con-

version, and singlet–triplet inter-

system crossing. Using boranils,

compounds devoid of proton-

transfer functionality, allows to

focus on the role of torsional dy-

namics not competing with tau-

tomerization, since the possibili- Scheme 3. Synthetic pathways for anils and boranils.

ty of twisting about the C6N1

and C7N1 bonds is retained in

these molecules. Moreover, functionalizing SA with electron-

donating and -accepting groups allows the study of the role of

electronic structure on the kinetics and thermodynamics of

both proton transfer and rotational isomerism. Finally, substitu-

tion stabilizes the pp* versus np* states, reducing the possible

contribution from the internal conversion channel. We demon-

strate that the ESIPT rate in anils can be controlled by the

nature of substituent and by the polarity of the environment.

For both anils and boranils, we show the existence of an effec-

tive non-radiative excited state depopulation channel which is

activated in polar environment.

78 and 85%. All these compound have been consistently char-

1

acterized by H, ¹³C, ¹¹B NMR, mass spectroscopy and elemental

analysis.

Figure 1 shows room temperature electronic absorption and

fluorescence spectra of 1, 1a, and 1b, recorded in a nonpolar

and a polar solvent, toluene and acetonitrile, respectively.

Figure 1. Left: absorption; right: fluorescence of 1, 1a, and 1b in: a) tolu-

ene, and b) acetonitrile at 293 K.

Scheme 2. Possible channels of excited state deactivation in SA. CI1 and CI2

indicate structures corresponding to the region of S₁/S₀conical intersection.

The absorption spectra in two solvents are very similar for

all three compounds. This claim is also true for the emission

profile of 1a and 1b in both solvents. In contrast, fluorescence

of anil 1 is strikingly different from the emission of boranils 1a

and 1b: it consists of two bands, of which the higher intensity,

low energy one (F₂) is located at similar energies in toluene

and acetonitrile. The high energy band (F₁) is barely detectable

in toluene, but is clearly observed in a polar environment. The

F₁/F₂intensity ratio is much higher for acetonitrile than for tol-

uene. The total fluorescence quantum yield is quite low (10^ꢀ3

or less, see Table 1); it is about four-times higher in toluene.

The single fluorescence observed for 1a and 1b is much

more intense than the emission of 1. It is definitely stronger

for toluene than for acetonitrile for 1a and only slightly so in

1b. The intensity and lifetime ratios are similar.

Results

The synthetic pathway leading to anils^[45]and boranils^[46,47]is

described in Scheme 3. Preparation of the anils 1–3 can easily

be achieved by treatment of salicylaldehydes A with the appro-

priate aniline B. This condensation occurs in refluxing ethanol

with a catalytic amount of p-TsOH. The desired imines precipi-

tate pure out of the reaction mixture and can be purified by

filtration with yield from 38 to 85%. Subsequent boron(III)

complexation is achieved in 1,2-dichloroethane with BF₃·Et₂O

under basic conditions to form boranils 1a–3a with yield rang-

ing between 43 and 72%. Boranils 1b and 3b are prepared

from the BPh₃precursor in hot toluene with respective yield of

&

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

2

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

is similar to that observed for 1,

but in 2 the F₁band is definitely

more intense with respect to F₂

in each solvent. As was the case

for 1, the dual emission in 2 is

assigned to the ESIPT process.

The dual emission profile of 2

has been recently reported and

interpreted in the same way.^[44]

Figure 3 presents the spectra

of the nitro derivatives 3, 3a,

and 3b. The emission pattern of

the anil 3 is quite different from

that exhibited by 1 and 2. A

single fluorescence band peak-

ing at 514 nm is observed in tol-

uene solution. In contrast, dual

emission is detected in acetoni-

trile, with the maximum of the

dominant, low energy band

peaking at 680 nm. A similar be-

haviour is observed for butyroni-

trile solution of 3, but the maxi-

mum of F₂lies at 652 nm, about

30 nm blue-shifted compared to

acetonitrile. Such behaviour is

different from that of 1 and 2

which exhibit F₂at similar wave-

lengths in both polar and non-

polar solvents. The F₁emission is

barely detectable at room tem-

perature in anhydrous acetoni-

trile and in anhydrous butyroni-

trile. In acetonitrile its intensity

increases with respect to F₂over

a period of several days, which

suggests that it may be due to

Table 1. Spectral and photophysical room temperature characteristics.

l_max[nm]

abs

f ^[a]

Decay time [ps]

FU^[b]

fl

em

SPC^[c]

TA^[d]

1

toluene

acetonitrile

butyronitrile

1a

372

430/553

445/554

0.0011

0.00025

0.25/61

0.41/45

56 (550)

48 (500, 550)

0.04/57^[e]

0.4/40^[e]

0.7/46^[e]

toluene

acetonitrile

butyronitrile

1b

396

395

447

453

0.046

0.0085

120 (450, 500)

20 (450)

0.6/150

0.4/25

0.82/36

0.45/20

toluene

acetonitrile

2

409

488

480

0.11

0.07

760 (500)

670 (500)

toluene

cyclohexane

acetonitrile

butyronitrile

valeronitrile

THF

398

460/570

0.0019

0.83/236

0.49/160

2.2/160

220 (570)

200 (480)

0.9/210^[e]

398

399

482/567

476/568

0.0018

0.0016

2.5/150^[e]

3.0/175^[e]

2.8/190^[e]

400

471/559

461

0.0017

2a

toluene

acetonitrile

butyronitrile

THF

416

411

468

466

0.50

1510 (460)

2.7/85/1500

0.7/6/21

0.4/4/33

0.0083

0.010

0.015

0.77/17

0.7/13/90

3

514

toluene

acetonitrile

butyronitrile

3a

428

431

525/680

514/652

0.010

0.0005

0.0068

1.2/13

0.22/4.2

260 (525)

35 (600)

1610

8.6/260/>3000

0.15/0.45/5

0.5/1.5/33

toluene

acetonitrile

butyronitrile

3b

toluene

acetonitrile

426

429

477

494/650

504/630

0.53

0.0006

0.0045

0.3/4.5/2000

0.3/5.1

1/2/18

0.5/2.6

30 (610, 660)

2730^[f]

440

444

509

518/660

0.61^[f]

0.0007

0.27/80

[a] Estimated error: ꢁ20–30% for quantum yields larger than 0.001 and ꢁ50% for lower values. [b] Fluores-

cence upconversion. [c] Single photon counting; in parentheses, detection wavelength [nm]. [d] Transient ab-

sorption. [e] Long-lived component also detected, see Figures S4 and S6 in the Supporting Information. [f] Ref-

erence [46].

Because of the difference in the fluorescence patterns in the

anil and boranils, the origin of the dual emission of 1 can be

readily assigned to the excited state proton transfer, a process

responsible for the lower energy emission band F₂. The large

increase of the F₁/F₂intensity ratio in acetonitrile indicates that

the reaction barrier is higher in a polar environment, which

may be a consequence of a larger excited state dipole

moment in the initially excited form. As discussed below, this

hypothesis was confirmed by calculations.

The emission spectra of 2 and 2a, shown in Figure 2, are

qualitatively similar to those of 1 and 1a, but the quantitative

differences are significant. Again, a single fluorescence is ob-

served for the boranil derivative 2a. It is very intense in tolu-

ene (quantum yield of 50%), but drops dramatically in acetoni-

trile, as well as in less polar solvents such as THF and butyroni-

trile (Table 1). Analogous decrease is observed for the fluores-

cence decay time. The anil derivative 2 exhibits a dual emis-

sion profile in both polar and nonpolar solvents. The F₁/F₂

intensity ratio is strongly solvent-dependent: 0.2 in toluene,

0.6 in THF and butyronitrile, and 0.9 in acetonitrile. This trend

Figure 2. Left: absorption; right: fluorescence of 2 and 2a in: a) toluene,

and b) acetonitrile at 293 K.

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

3

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

The temperature dependence of the single fluorescence ob-

served in 2a (Figure S2 in the Supporting Information) is very

different in toluene versus THF and butyronitrile. In toluene,

the emission is intense already at room temperature and does

not change much as the temperature decreases. On the con-

trary, in THF and BuCN the emission intensity strongly increas-

es at low temperature. As a result, practically the same fluores-

cence quantum yield values are determined at 173 K: 0.27

(BuCN), 0.24 (THF), 0.21 (toluene). At 77 K in toluene, no phos-

phorescence is observed, contrary to the behaviour of 2.

Very strong temperature dependence was observed for 3 in

BuCN (Figure 4). Upon lowering the temperature from 293 to

123 K, the intensity increases by more than two orders of mag-

nitude. Drastic changes in intensity are observed in a narrow,

low temperature range below 140 K, that is, when butyronitrile

becomes glassy (Figure 4D). These changes are accompanied

by spectacular blue shifts of the emission: upon lowering the

temperature by just five degrees, from 128 to 123 K, the fluo-

rescence maximum shifts from 592 to 544 nm (Figure 4C).

Below 123 K, the quantitative measurements are difficult, due

to fluctuations of intensity associated with long-term instability

of butyronitrile glass. However, it can be safely concluded that

both intensity and shape of fluorescence do not change much

in the low temperature region. The low temperature emission

consists of one band with maximum at 537 nm. This band is

barely visible at 293 K, where the fluorescence is dominated by

a band with maximum around 650 nm.

Figure 3. Left: absorption; right: fluorescence of 3, 3a, and 3b in: a) tolu-

ene, and b) acetonitrile at 293 K. The asterisks mark artefacts in the emission

spectra in acetonitrile due to non-perfect removal of solvent Raman lines.

formation of complexes with water (cf. the section below de-

voted to the spectra in protic solvents).

The boranils 3a and 3b exhibit single emission in toluene,

but in acetonitrile a second, low intensity band corresponding

to F₂is observed. Its location is similar to that of F₂emission of

3.

In a similar fashion as for 2a, huge differences in quantum

yields and lifetimes are observed for 3a and 3b in toluene and

acetonitrile. In the latter fluorescence intensity drops by over

two orders of magnitude.

The above results demonstrate the influence of viscosity on

the photophysics of 3 in a polar solvent, suggesting a confor-

mational change involving a large amplitude motion and

a high dipole moment of the excited species.

The emission behaviour of 3 suggests a different pathway of

S₁deactivation than in 1 and 2. The observed huge shift of the

emission maximum with changing solvent polarity points to

a charge transfer character of the excited state. Moreover, con-

trary to the case of fluoro and cyano derivatives, dual emission

is observed for 3a and 3b, analogous to that of 3. One should

note, however, very different F₁/F₂intensity ratios observed for

the anil and boranils, suggesting that the photophysical prop-

erties are sensitive to the presence of intramolecular hydrogen

bonding, which may lead to geometry changes or to different

relative positions of low-lying electronic states.

The temperature dependence of the emission of boranil de-

rivative 3a in butyronitrile (Figure S3 in the Supporting Infor-

mation) is similar to that observed for 3. At room temperature,

weak dual fluorescence is observed, with the maxima located

at around 490 and 630 nm. The intensity and position of the

lower energy band strongly depend on temperature: at 143 K,

the intensity increases by a factor of twenty with respect to

the value measured at 296 K. Similarly to 3, the shape of the

emission changes dramatically in a narrow temperature range

between 138 and 133 K. (Figure S3C). Further decrease of tem-

perature does not lead to significant changes; the fluorescence

consists of a single band with a maximum at 487 nm, very

close to that of F₁fluorescence at room temperature.

Temperature dependence of the emission

Comparison of the emission properties of 3 and 3a indicates

that in both molecules the initially excited structure is trans-

formed into a strongly polar transient species that undergoes

a viscosity-dependent relaxation process.

Stationary fluorescence spectra were recorded as a function of

temperature in a number of solvents. For toluene, THF, and bu-

tyronitrile solutions of 2, similar changes in the F₁/F₂intensity

ratio are observed (Figure S1 in the Supporting Information):

the intensity of both bands increases about five times at lower

temperature. Interestingly, the F₁/F₂ratio initially decreases and

then increases upon lowering of temperature; at 294 and

180 K this ratio is practically the same for all three solvents. At

lower temperatures in butyronitrile the F₁fluorescence intensi-

ty significantly increases with respect to F₂. At 77 K, phosphor-

escence is observed: for toluene, its intensity is similar to that

of F₁, whereas in butyronitrile it is about three-times weaker.

Protic solvent solutions

We have also recorded absorption and fluorescence in protic

solvents, methanol, and n-propanol. While the absorption

spectra resemble those obtained in other solvents, the emis-

sion profile is quite different. Indeed, multiple bands are ob-

served with relative intensities dependent on the excitation

wavelength. This behaviour may have several origins. The first

&

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

4

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

In contrast to the behaviour

of boranils, the F₂decay in the

anils 1 and 2 is insensitive or

only weakly dependent on the

nature of the solvent. This is not

the case for the nitro derivative

3, which exhibits shorter decay

in a polar environment.

In order to probe shorter fluo-

rescence decays and to monitor

time-resolved spectra, the tech-

niques of fluorescence upcon-

version (FU) and transient ab-

sorption (TA) were used. Exem-

plary time-wavelength maps ob-

tained using FU are shown in

Figure 5, whereas Figure 6 pres-

ents the time evolution of fluo-

rescence of 2, 3, and 3a. Table 1

summarizes the decay parame-

ters obtained for all compounds

in various solvents. The decays

are biexponential, the shorter

occurs in the subpicosecond

range. The decay time values ob-

tained by fluorescence upcon-

version correspond well to the

Figure 4. A–B) Temperature dependence of fluorescence of 3 in butyronitrile. C) Fluorescence spectra recorded at

296, 128, 123, and at 77 K. D) Integrated fluorescence intensity relative to the value at 296 K; black and blue sym-

bols indicate measurements on different spectrofluorimeters (Jasny and Edinburgh, respectively).

is trivial for boranils: they can be transformed back to anils by

boron decomplexation. Indeed, we have observed that, for in-

stance, 1b is rapidly converted to 1. The other source of com-

plicated spectra may be due to the formation of hydrogen-

bonded complexes with alcohols. This should, in particular,

affect the spectra and photophysics of anils if the alcohol

breaks the intramolecular OH···N hydrogen bond. The complex

behaviour in alcohols definitely requires further studies.

kinetic parameters extracted from transient absorption studies.

The components of the dual emission of the anils 1 and 2

reveal a characteristic kinetic pattern. The decay of F₁fluores-

cence, occurring in a few hundred femtoseconds to a few pico-

seconds, is definitely shorter in nonpolar solvents. This finding

is in agreement with stationary measurements which show

that the F₁/F₂intensity ratio increases in acetonitrile with re-

spect to toluene (Figures 1 and 2). The F₂emission decay is

much longer and not strongly solvent-dependent, as already

noted on the basis of TR-SPC experiments. Time-resolved emis-

sion spectra clearly show the ultrafast decay of F₁. In order to

check whether the precursor–successor relationship exists be-

tween F₁and F₂, we looked for the rise time in the latter that

would correspond to the decay of the former. However, no

such rise time could be detected. This result does not exclude

the F₁emitting state as a precursor of F₂, but it may indicate

a much lower radiative constant in the latter. Additionally, the

rise time may be masked by spectral overlap between the two

emission bands.

Kinetic studies

Fluorescence studies

The temporal resolution of our TC-SPC setup, estimated as 20–

30 ps, allowed the reliable measurement of the longest com-

ponents of fluorescence decays (Table 1). Nanosecond decay

times are observed in nonpolar environments for the boranils

2a, 3a, and 3b, whereas for 1b and, in particular, for 1a the

fluorescence decays faster (760 and 120 ps, respectively). In

polar solvents, fluorescence lifetimes decrease dramatically

with 1b being the only exception. For the fluoro and cyano

derivatives this behaviour correlates well with the correspond-

ing changes in quantum yields, suggesting that the structure

responsible for the emission is similar in nonpolar and polar

solvents. However, in the nitro anil 3a the ratio of quantum

yields in toluene and acetonitrile is much larger than that of

the lifetimes. This indicates a smaller value of the radiative con-

stant in the species emitting in a polar solvent, a hypothesis

that was corroborated by time-resolved emission experiments

discussed below.

For 2, we have also studied the isotope effect. In the OD-

deuterated 2, the F₁decay time in acetonitrile increases from

2.2 to about 9 ps. Deuteration does not affect the F₂lifetime.

This result strongly suggests that the decay of F₁is associated

with the OH proton translocation. The definitive argument for

photoinduced tautomerization was provided by time-resolved

IR studies, discussed below.

The kinetic behaviour of 3 is completely different from that

observed for the other two anils. Now, both decay compo-

nents are shorter in acetonitrile. A similar behaviour, a huge

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

5

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

signed to the decay of the ini-

tially excited species via proton

transfer. We note that no recov-

ery of the bleaching is observed

on the short time scale. This

means that proton transfer does

not compete with other fast de-

population channels, which was

postulated for SA.^[41,42,48]On the

contrary, the second component,

about 200 ps, assigned to the

decay of the phototautomer, ap-

pears also at wavelengths corre-

sponding to bleaching, indicat-

ing that the rate of ground state

back proton transfer is faster

than 5ꢂ10⁹s^ꢀ1

.

The third decay corresponds

to the lifetime which is too long

(>15 ns) to be measured accu-

rately with the femtosecond

Figure 5. Maps of time-resolved fluorescence of: a) 2, b) 2a, c) 3, and d) 3a. The spectra were recorded at 293 K

using acetonitrile as solvent.

decrease of the decay parameters in polar solvents is observed

for 3a and 3b.

Inspection of time-resolved emission spectra (Figure 6) clear-

ly shows higher radiative rates for F₁than for F₂emissions. This

is true both for F₂occurring from a proton-transferred species

and in the case when the origin of the low energy emission

must be different (3a).

Transient absorption studies

Representative examples of femtosecond transient absorption

(TA) spectra are shown in Figures 7, 8, 9, 10. Table 1 contains

the decay parameters extracted from the spectra. Other exam-

ples of TA, together with decay associated spectra are shown

in Figures S4–S9 in the Supporting Information.

The TA spectrum of 2 (Figure 7 and Figure S6 in the Sup-

porting Information) consists of several bands. The negative

features at 400 and 500 nm correspond to ground-state

bleaching and stimulated F₁emission, respectively. Due to

a much lower radiative constant responsible for F₂, its contri-

bution is not observed in the TA signal. The kinetic profiles

measured at 500 nm are in excellent agreement with the fluo-

rescence upconversion results (Table 1). In particular, the F₁

fluorescence decay times, longer in acetonitrile than in tolu-

ene, are reproduced.

The transient absorption features a strong peak at 450 nm

that overlaps with both bleaching and stimulated fluorescence,

so that the kinetic analysis is rather complex. It clearly contains

the slower decay component associated with F₂(209 and

156 ps in toluene and acetonitrile, respectively), but the faster

component (0.86/2.5 ps) is also detected. Thus, the TA in this

region contains contributions from the species responsible for

both F₁and F₂emissions.

Figure 6. Time-resolved fluorescence spectra of 2, 3, and 3a obtained by FU

for acetonitrile solutions at 293 K.

spectrometer. The observation that the TA signal persists after

decay of both emissions indicates the presence of a long-lived

species. From the magnitude of the bleaching it can be esti-

The time evolution of TA, measured in four solvents may in

each case be fitted with three components. The shortest is as-

&

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

6

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

an ultrafast proton transfer, the

molecule has no time to reveal

the cis-enol fluorescence shift. A

significant difference between 2

and 2a is the lack of any transi-

ent signal in acetonitrile at

100 ps delay in 2a. Thus, contra-

ry to the case of the anil ana-

logue, no long-lived species is

now formed. This result suggests

that the long-lived transient in 2

is rather due to the photochro-

mic form (trans-keto) than to the

triplet, since the latter should be

also populated in 2a.

Similar results were obtained

for 1 and 1a (Figures S4–S5 in

the Supporting Information).

Also in this case, the significant

difference between the anil

1 and boranil 1a is that the

former exhibits a long-lived TA

signal (peaking at 420 and

530 nm), whereas the latter does

not: its decay corresponds to

that of fluorescence.

Figure 7. TA of 2 in toluene (left) and acetonitrile (right), recorded at short and long delays (top and bottom, re-

spectively). Dotted lines: inverted stationary absorption and emission.

mated that about 50% of the initially excited population does

not return to the initial ground state form on the time scale of

at least 15 ns. The long-lived species, with the absorption

peaks at 460 and 580 nm, may correspond to the photochro-

mic products and/or to the triplet state.

The TA spectra of 2a (Figure 8 and Figure S7 in the Support-

ing Information) exhibit a longer

The TA behaviour of 3 and 3a is quite complicated (Fig-

ures 9 and 10, Figures S8–S9 in the Supporting Information).

For 3 in toluene, the strong TA band at 459 nm evolves into

a band with a maximum at 481 nm within a few picoseconds.

It should be stressed that this evolution corresponds not to

a band shift, but to the decay of the high energy feature and

component that can be correlat-

ed with the decay time of fluo-

rescence. Its value drastically de-

creases with solvent polarity, in

parallel with quantum yield

(Table 1). In contrast to the be-

haviour of 2, evolution of the

transient absorption and stimu-

lated emission signals is ob-

served on the short time scale.

The kinetic profiles in this range

can be fitted to either mono- or

biexponential decay functions.

We attribute this evolution to vi-

brational relaxation and to the

stabilization of a large S₁dipole

moment by a polar solvent. The

calculations predict a value of

12.6–15.9D (depending on the

method, see Table S1 in the Sup-

porting Information) for the ini-

tially excited 2a. Also for 2, the

S₁the dipole moment is calculat-

Figure 8. TA of 2a in toluene (left) and acetonitrile (right), recorded at short and long delays (top and bottom, re-

ed to be large, but, because of spectively). Dotted lines: inverted stationary absorption and emission.

Chem. Eur. J. 2014, 20, 1 – 17 www.chemeurj.org ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

7

&

These are not the final page numbers! ÞÞ

Full Paper

formation of the lower energy

one. The band at 481 nm and

the stimulated emission at

542 nm decay in 260 ps, which is

the fluorescence decay time

measured by TC-SPC. Interest-

ingly, the intensity of the bleach-

ing signal does not change

much during this period. It re-

mains practically constant up to

1300 ps, the longest delay used,

during which time, a build-up of

TA absorption in the region

above 550 nm is observed. An

analogous pattern, interconver-

sion of two TA bands, has been

observed for ethyl ether, THF,

and butyronitrile solutions. The

difference with respect to tolu-

ene is that a significant time

shift of the stimulated emission

maximum is now observed, its

value reaching 551 nm in di-

ethylether, 579 nm in THF, and

700 nm in butyronitrile. Even

a larger shift, to 727 nm, is ob-

served for acetonitrile (Figure 9).

In this solvent, only one TA band

is observed, shifting from

487 nm, observed immediately

after excitation, to 482 nm at

longer delays.

A shoulder at

higher energies is detected at

early delays, suggesting that the

initially excited species decays in

acetonitrile faster than in other

solvents. In contrast to anils

1 and 2, no long-lived transient

could be observed for both ace-

tonitrile and butyronitrile solu-

tions.

Figure 9. TA of 3 in toluene (left) and acetonitrile (right) for selected time windows. Dotted lines: inverted station-

ary absorption and emission.

The boranil derivative 3a (Figure 10 and Figure S9 in the

Supporting Information) exhibits a behaviour similar to 3 re-

garding the temporal evolution of the stimulated emission,

but some differences are observed in the region of the TA

band. These differences are difficult to quantify because of

strong overlap of absorption and fluorescence. Another com-

plication is introduced by very different fluorescence intensities

of 3 and 3a in nonpolar solvents.

The analysis of the strong negative signal due to stimulated

fluorescence in acetonitrile yielded the kinetic parameters of

0.6 and 0.7 ps for 3 and 3a, respectively. These values resem-

ble the relaxation time of acetonitrile,^[49]suggesting that the

emission initially occurs from a non-relaxed state and that its

decay is solvent controlled.

Femtosecond transient IR studies

Figure 11 shows the IR transient spectra of 1 taken after excita-

tion at 400 nm. At short time delays, the transient spectra are

dominated by four ground-state bleach bands at 1510, 1585,

1607, and 1635 cm^ꢀ1, and a positive band at 1545 cm^ꢀ1. This

latter band decays completely within about 150 ps, whereas

the bleach bands recover only partially, and new positive

bands at about 1500, 1600 and 1640 cm^ꢀ1appear. From about

150 ps onward, the spectra remain essentially constant up to

1.8 ns, the upper limit of the experimental time-window.

Global analysis of the transient IR absorption data was per-

formed using the sum of three exponential functions, and

yielded 2.9 ps, 45 ps and >10 ns time constants and the

decay-associated spectra shown in Figure S10 in the Support-

ing Information. The two longest time constants values are in

&

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

8

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

1630 cm^ꢀ1and positive transient

bands at 1530, 1550 and

1595 cm^ꢀ1, growing within a few

picoseconds. In comparison with

1, where the transient band at

1545 cm^ꢀ1is present within the

time resolution of the experi-

ment, the signal, which is as-

signed to intramolecular proton

transfer, rises more slowly in the

cyano derivative 2. Again, this

finding is in good agreement

with the results of fluorescence

upconversion and transient ab-

sorption experiments.

Target analysis using the same

scheme as for 1 was performed

for 2 (Figure S12 in the Support-

ing

Information),

providing

a time constant of 4 ps, compa-

rable to 2.2/2.5 ps extracted

from fluorescence upconversion

and transient-absorption studies.

In this case, species A and B can

be ascribed to the excited enol

and keto forms of 2, respectively.

Quantum chemical calcula-

tions were performed to assign

the transient bands observed in

the time-resolved IR experi-

ments. Figure 13 shows the

comparison of the calculated dif-

ference of IR spectra of the keto

and enol forms of 1 with the ex-

perimental difference spectra.

After excitation, positive bands

due to the proton transfer are

expected. The growth of the

Figure 10. TA of 3a in toluene (left) and acetonitrile (right) for selected time windows. Dotted lines: inverted sta-

tionary absorption and emission.

good agreement with those extracted from the transient visi-

ble absorption experiments performed for the same molecule.

However, the 2.9 ps lifetime is much longer than that of 0.41–

0.45 ps obtained from the FU and transient visible absorption

measurements. The transient spectra recorded before 500 fs

have not been included in the global analysis of the IR data to

avoid the complications related to cross-phase modulation

and, thus, this ultrafast component, associated with the ESIPT

cannot be detected. To better ascribe the 2.9 ps time constant

a global target analysis assuming an A!B!C!D reaction

scheme has been carried out. Such analysis yielded the same

time constants, as expected, and the species-associated differ-

ence spectra shown in Figure S11 in the Supporting Informa-

tion. The difference spectra associated with A and B are very

similar and, therefore, these species can be ascribed to the un-

relaxed and relaxed ESIPT product forms of 1, respectively.

The same measurements were repeated for 2. The spectra

(Figure 12) show three negative bands at 1510, 1570, and

Figure 11. Upper panel: transient IR spectra measured for 1 in acetonitrile at

different delays after excitation at 400 nm. The dashed line is the inverted

and scaled stationary FTIR spectrum. Lower panel: calculated steady state IR

spectra of enol and keto forms (red and blue, respectively) and their differ-

ence (green). For a better comparison the calculated spectra were scaled

and flipped.

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

9

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

Calculations

Ground- and excited-state geometry optimizations were car-

ried out in order to determine the structure of the emitting

species and the nature of processes responsible for substitu-

ent- and solvent-dependent photophysics. To simplify and

speed up the calculations, dimethylamino substituent was

used instead of diethylamino group. In addition to optimizing

the push–pull structures of all the molecules under study, we

also computed and analysed the parent molecule, SA, and four

model compounds, selected to separately assess the role of

electron-donating and -accepting substituents. These included

SA singly substituted with the dimethylamino group on the

phenol ring (SADMA) and SA with fluorine (SAF), cyano (SACN),

and nitro (SANO₂) groups on the aniline moiety (Figure S15 in

the Supporting Information). The results are presented in

Table S1 in the Supporting Information. Three different func-

tionals were tried: 1) B3LYP; 2) PBE0, which is known to yield

accurate electronic transition energies, and 3) CAM-B3LYP, rec-

ommended for treatment of charge-transfer states.

Figure 12. Upper panel: transient IR spectra measured for 2 in acetonitrile at

different delays [ps] after excitation at 400 nm. The dashed line is the invert-

ed FTIR spectrum. Lower panel: calculated ground state FTIR spectra for the

enol and keto forms (red and blue, respectively), the difference (green), and

the experimental spectrum recorded after 6 ps (black). The numbers corre-

spond to the modes depicted in Figure S13 in the Supporting Information.

band at 1600 cm^ꢀ1in the experimental data coincides with

that calculated for the keto form. A positive band is also ex-

pected around 1540 cm^ꢀ1. The modes calculated for 2 in the

region between 1500 and 1700 cm^ꢀ1(Figure S13 in the Sup-

porting Information) involve the OH or NH bending and the

C7H and CH bending in the aromatic rings where an influence

can be expected when the proton transfer occurs.

In general, no significant differences were obtained for

ground state optimizations using different functionals. The en-

ergies of the lowest singlet electronic transitions were repro-

duced with similar accuracies by B3LYP and PBE0, whereas

using the CAM-B3LYP functional yielded overestimated ener-

gies. For this reason, most calculation were done using B3LYP;

this also ensured compatibility with previous theoretical stud-

ies.^[41,42,50]We also checked that double and triple zeta basis

sets yielded very similar results, as can be seen from Table S1

in the Supporting Information.

Very similar geometries of the dominant, ground state cis-

enol form have been obtained for all the analysed molecules.

Due to the stabilizing effect of the intramolecular hydrogen

bond in anils or boron in boranils, the quasi-six-member ring

assumes a nearly planar geometry. On the other hand, consid-

erable twisting about the N1C6 bond is predicted, with the

values of the angle a (Scheme 2) in the range of 30–40 de-

grees. These predictions are in good agreement with the X-ray

structure of SA.^[51]The calculated ground state dipole moments

increase from about 2 D in SA and SAF, to over 3 D in SADMA,

and over 5 D in SACN and SANO₂. A spectacular push–pull

effect, leading to a considerable increase of the dipole

moment is observed for both anils and boranils. Calculations

systematically yield larger ground state dipole moments for

boranils than for the corresponding anils.

As already reported for some boranils,^[52]the calculations

predict that electronic excitation to S₁results in the increase of

dipole moment. This effect is stronger for anils; no significant

change is obtained for boranils 1a and 1b. For the cyano de-

rivatives, an almost twofold increase is predicted for the anil 2,

whereas the change in 2a is smaller (one should note, though,

that the S₀value is larger for the boranil). A spectacular in-

crease is obtained for the nitro derivatives 3 and 3a, for which

the values of nearly 30 D are calculated for the Franck–Condon

(FC) geometry corresponding to S₀. Because the dipole mo-

ments calculated for the optimized excited cis-keto forms are

lower, one can expect that the enol form will be stabilized

Figure 13. Black: difference of the calculated (B3LYP/6-31+G(d,p)) steady-

state IR spectra for the enol and keto forms of 1. Grey: evolution associated

difference spectra (EADS) of species A obtained from target analysis.

Time-resolved IR measurements were also performed on the

nitro substituted anil 3 (Figure S14 in the Supporting Informa-

tion). All the spectral changes end after a few picoseconds.

The decay of the excited state is slightly faster than the recov-

ery of the ground state, probably because of the involvement

of a hot ground state, as testified by the positive feature on

the low frequency side of the bleach. The intensity of the sig-

nals is lower than for the other compounds and the quality of

the experimental data did not allow carrying out a target anal-

ysis using all the regions.

&

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

10

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

with respect to the keto species in more polar solvents. We

checked the dependence of the S₁dipole moment in the enol

form of 2 on the proton position by increasing the proton in

steps of 0.1 ꢃ from the optimized location (OH=0.998 ꢃ). For

the OH distances of 0.998, 1.1, 1.2 1.3, and 1.4 ꢃ the following

values of dipole moments were predicted: 16.96, 16.44, 15.41,

14.95, and 15.91 D. In the S₁-optimized cis-keto form, the pre-

dicted dipole moment is 15.85 D. These result suggest that the

barrier to proton transfer should increase in polar environment,

in excellent agreement with the experimental results, which

reveal slower phototautomerization in acetonitrile than in

toluene.

the electron promotion from a highest occupied molecular or-

bital (HOMO) to the lowest unoccupied one (LUMO). For the

ground state geometry these orbitals are delocalized over the

whole molecule, even though the HOMO is preferentially locat-

ed on the dimethylaniline part, whereas LUMO, on the nitro-

phenyl moiety. For the twisted S₁geometry (aꢂ908) HOMO

and LUMO become localized on the donor and acceptor parts,

respectively. This leads to a TICT (twisted intramolecular charge

transfer) state nature of S₁, with the dipole moment approach-

ing 40 D. Actually, the dipole moment is already considerable

for the FC geometry of S₁. Its increase upon twisting in 3 and

3a is opposite to the behaviour predicted for 2 and 2a, of

which the structure becomes more planar in S₁(a=08 and

178, respectively) and the dipole moment decreases.

Table 2 shows the calculated ground- and excited-state

values of the lengths of bonds that may be involved in torsion-

al dynamics described by a, b, and g angles (Scheme 2). We

also included the C11N2 bond, to check the possibility of twist-

ing of the diethylamino group. Important differences are ob-

tained for different derivatives, which may explain the photo-

physical behaviour. Therefore, we discuss these results in detail

in the next section.

Optimization of the S₁geometry for model molecules, as

well as for 1 was either not convergent or was leading to

twisted structures (bꢂ908) with energies very close to that of

the ground state. This is in agreement with previous findings

that showed twisting towards non-planar geometries, corre-

sponding to the regions of S₀/S₁conical intersec-

tion.^{[20,21,41,42,50,53]}However, S₁geometry could be optimized for

other model compounds, anils, and boranils. For 2, a complete-

ly planar structure was found. Also for the boranil derivative

2a a more planar structure is predicted in S₁, with the angle a

decreasing to 178 from 408 in S₀. The predicted Stokes shifts

compare satisfactorily with the experiment. These results sug-

gest that upon passing from SA to push–pull substituted anils,

the initially excited cis-enol form should become more stable

with respect to twisting. Comparison of the electronic charges

on the oxygen and nitrogen atoms involved in intramolecular

H bonds in the S₀, S₁(FC) and optimized S₁states revealed that

the driving force for tautomerization is coupled to the twisting

about the C7N1 bond. For 1, the respective values on the

oxygen are ꢀ0.32, ꢀ0.31. and ꢀ0.19; for nitrogen, the calcula-

tions yield 0.03, 0.03 and ꢀ0.07. The corresponding values for

2 are ꢀ0.57, ꢀ0.56, and ꢀ0.56 (oxygen) and ꢀ0.62, ꢀ0.59, and

ꢀ0.62. These results indicate that blocking or slowing down

the twisting around C7N1 should lead to a decrease in the ex-

cited state proton transfer rate. This prediction is confirmed by

the decay times of the initially excited forms: about 50 fs in

SA,^[31]250–500 fs in 1, and 0.8–2.5 ps in 2.

The different behaviour of the nitro derivatives observed in

the experiments was also reproduced by excited state geome-

try optimizations. In contrast to the planarized geometries of 2

and 2b, a charge transfer S₁state with a twisted structure (a

ꢂ908) was obtained for both 3 and 3a (and also for the nitro

model compound, SANO₂). One should note that the twisted

form is now characterized by the angle a close to 90 degrees

and so it is different from the CI1 and CI2 structures shown in

Scheme 2, for which the angles b and g were involved. The cal-

culated S₁!S₀transition energies are somewhat higher than

the positions of the F₂emission in 3 and 3a. It should be

noted, however, that, because of a huge S₁dipole moment,

the emission in acetonitrile should be strongly red-shifted with

respect to the position predicted for an isolated molecule. The

origin of charge transfer character is clearly seen upon inspec-

tion of molecular orbitals (Figure S16 in the Supporting Infor-

mation). The S₁state is represented, both in absorption and

emission by a practically (98%) single configuration, involving

Table 2. Calculated (B3LYP/6-31G(d,p)) lengths [pm] of the bonds rele-

vant for the torsional dynamics.

C6N1

C7N1

C7C8

C11N2

SA

S₀

S₁

S₀

S₁

S₀

S₁

S₀

S₁

S₀

S₁

S₀

S₁

S₀

S₁

S₀

S₁

S₀

S₁

S₀

S₁

S₀

S₁

141

134

141

132

141

134

140

139

140

147

140

132

143

137

140

137

142

137

140

146

142

147

128

139

130

142

128

139

130

133

130

128

129

142

131

139

130

132

137

130

127

132

127

145

142

144

139

145

141

145

141

144

145

144

139

141

140

143

146

140

141

143

148

140

146

–

138

139

–

SADMA

SAF

SACN

SANO₂

1

[a]

–

138

139

137

138

137

138

137

138

137

136

137

1a

2

2a

3

3a

[a] Values for the S₁cis-keto form, as the optimization converged to this

structure.

Discussion

Both experimental and theoretical evidence show that the

dual emission, observed for the anils 1 and 2 and absent in

their boranil analogues 1a, 1b, and 2a is due to photoinduced

proton transfer (ESIPT) from the OH group to the N1 nitrogen

atom. FU, TA, and time-resolved IR experiments demonstrate

that tautomerization occurs at 293 K in hundreds of femtosec-

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

11

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

onds to a few picoseconds. The reaction is thus much slower

than in the parent SA. The rates strongly depend on solvent

polarity and the nature of the substituent. The sensitivity to

the substituent can be explained by the gradual increase of

the barrier to proton transfer in S₁upon switching from SA to

the anils derivatives 1 and 2. As this barrier reflects the

strength of the intramolecular OH···N hydrogen bond, it may

be instructive to compare the OH stretching frequencies calcu-

lated for different molecules. For the ground state, the calcula-

tions for model compounds predict that SA substitution by the

dimethylamino group leads to the decrease of this frequency,

that is, to a stronger hydrogen bond, whereas, the opposite

effect is observed for ꢀF, ꢀCN, and ꢀNO₂substitution. This was

also confirmed by the calculated OH stretching frequencies of

anils: 3191, 3215, and 3218 cm^ꢀ1for 1, 2, and 3, respectively.

Since the excited-state geometries could not be optimized for

all molecules, such systematic comparison was not possible for

S₁. However, a huge decrease in the OH frequency calculated

for 2 in S₁, to 2794 cm^ꢀ1can be noted, accompanied by the

decrease of the OꢀN distance from 263 to 258 pm, which

shows a drastic increase of driving force for proton transfer.

The lower proton transfer rates in a polar environment can

be explained by the preferential stabilization of the excited,

more polar cis-enol form. We note that the calculated dipole

moment values of the excited cis-enol are significantly higher

than those of cis-keto species only for the non-relaxed Franck–

Condon geometries, and therefore this conclusion is valid for

the fast reaction regime. However, we checked that calcula-

tions yield a similar trend, that is, decrease of the S₁dipole

moment upon elongation of the OH bond for both FC and op-

timized S₁geometries of cis-enol.

value is obtained if one uses the value of 260 ps as the fluores-

cence decay of the excited cis-enol form. The discrepancy dis-

appears if the 260 ps decay is associated with the proton-trans-

ferred species, for which both FU experiments and calculations

predict a lower radiative rate; one should note that, for this

reason, the 260 ps component is detected by SPC and TA, but

not FU. All these findings indicate that the low energy emis-

sion of 3 could correspond to the twisted tautomeric species.

Still, we note that the long-lived transient, a characteristic sign

of ESIPT, is observed only for toluene solutions of 3. It may be

that the rotation around a slows down the twisting around g.

Another explanation is a much shorter excited state lifetime

due to quenching of the S₁state in polar solvents, which we

discuss below.

A spectacular effect is that the fluorescence quantum yields

and lifetimes of boranils in solvents with low dielectric con-

stant are higher than those of the corresponding anils, by as

much as one or two orders of magnitude. This makes a signifi-

cant contribution of twisting involving a and b angles in the

excited anils rather improbable as major non-radiative depopu-

lation channels, since such processes are allowed for both anils

and boranils. On the other hand, this result emphasizes the im-

portance of relaxation which changes the value of g, because

this process is blocked in boranils. Such relaxation is possible

in the cis-keto forms, and should lead to the photochromic

trans-keto structure. Indeed, long-lived transients are observed

for anils, but not for boranils.

The boranils, devoid of the hydroxyl group, show very

strong fluorescence in nonpolar environment, but the quan-

tum yield and lifetime drastically decrease for polar solvents. It

seems that the emitting species have the same structure in sol-

vents of different polarity. This is indicated by the values of the

radiative constants, practically the same in different solvents:

(4ꢁ1)ꢂ10⁸s^ꢀ1for 1a, (1.2ꢁ0.3)ꢂ10⁸s^ꢀ1for 1b, and (3ꢁ1)ꢂ

10⁸s^ꢀ1for 2a. The observed fourfold decrease of the radiative

constant in 1b versus 1a is in excellent agreement with theo-

retical predictions (Table S1 in the Supporting Information).

The large dependence of fluorescence intensity of boranils

on temperature and solvent viscosity in polar environments in-

dicates that the quenching involves a large amplitude motion,

most probably twisting. Due to the rigidifying character of BX₂,

relaxation involving the angle g is rather improbable; also the

calculations do not predict such a possibility. The calculated

C11N2 bond lengths (Scheme 1) remain the same in S₀and S₁,

which rules out the twisting of the diethylamino group. Thus,

two angles, a and b have to be taken into account (Scheme 2).

Twisting around the C7N1 bond, involving b, is predicted by

calculations for SA and SADMA to guide the excited molecule

towards the region of CI with the ground state. For twisting in-

volving a, such behaviour is not predicted. We therefore con-

clude that the quenching channel is related to the increase of

b. Since the effect strongly increases with solvent polarity, such

twisting most probably involves an electronic state with

a large dipole moment. Since no transient intermediate was

detected, and the rates of S₁decay and S₀repopulation seem

to be equal, such a state only provides a doorway to efficient

deactivation, but is not populated itself.

The behaviour of 3 can be interpreted in terms of an elec-

tron transfer coupled with conformational change and, possi-

bly, proton transfer. The dual pattern of the emission is now

observed also for the boranil analogues 3a and 3b. Time evo-

lution of the stimulated fluorescence in the TA spectra record-

ed in solvents of different polarity indicates a large value of

the dipole moment of the emitting species. Moreover, huge

dependence of the F₂emission on solvent polarity and viscosi-

ty, as revealed by temperature dependence of the emission,

points not only to the large dipole moment of the emitting

species, but also to large amplitude motion in the relaxation

process involved in its creation. We therefore assign the low

energy emission to the twisted (aꢂ908) polar structure, calcu-

lated for both 3 and 3a. Since such structure can be achieved

without proton transfer, the question is whether the twisting is

accompanied by tautomerization in the anil 3. The experiment

provides several arguments in favour of this hypothesis. The

first is the observation of somewhat different TA patterns in 3

and 3a. Also, the time-resolved fluorescence spectra look dif-

ferent for 3 and 3a at early delays (Figure 6) before they

evolve into a low energy band, similar in the two compounds

and characterized by a small value of the radiative constant. Fi-

nally, it is instructive to compare the radiative rates estimated

for the initially excited form. Using the data in Table 1, one ob-

tains for 3a similar values, (1–3)ꢂ10⁸s^ꢀ1, for both polar and

nonpolar solvents. However, for 3 in toluene, a much lower

&

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

12

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

In order to understand the role of substituents in the excited

state torsional dynamics, we analyse the results of calculations

regarding changes of bond lengths upon excitation (Table 2).

Practically the same pattern is observed for all the anils and

boranils in their ground electronic state. In contrast, the excit-

ed-state values of the C6N1 and C7N1 bond lengths, responsi-

ble for a and b torsional motions, drastically change with the

nature of substituent. For molecules without an electron ac-

cepting group (SA, SADMA) or those with a weak acceptor

(SAF, 1 and 1a) the C6N1 and C7N1 bond lengths decrease

and increase in S₁, respectively. The opposite is true for the

compounds with strong acceptors (SANO₂, 3, 3a): now the

C6N1 bond becomes longer and the C7N1 bond shorter. The

cyano-substituted molecules (SACN, 2, 2a) are on a borderline

between these two patterns, the two bonds being of similar

lengths in S₀and S₁. The influence of the dimethylamino group

is rather minor, as can be deduced from the comparison of

corresponding pairs (SA vs. SADMA, SAF vs. 1 and 1a, SACN

vs. 2 and 2a, and SANO₂vs. 3 and 3a).

It now becomes clear why differently substituted anils and

boranils should exhibit different photophysics. Excitation of

nitro-substituted chromophores weakens the C6N1 and

strengthens the C7N1 bond. Naturally, this facilitates the a tor-

sional relaxation. On the contrary, in molecules without strong

electron acceptors, b-relaxation is preferred in the excited

state. In chromophores with moderate electron-accepting

properties, both types of torsional dynamics should be slowed

down. As a result, planar S₁conformation is more favourable.

BF₂or BPh₂, the rate of S₁deactivation is drastically reduced.

On the other hand, intramolecular charge transfer accompa-

nied by twisting about C6N1 bond, as in 3 or 3a, leads to

a state of which the lifetime is considerably longer than in the

case when rotational dynamics involves C7N1 and C7C8

bonds. In other words, effective relaxation increasing the value

of angle a suppresses torsional dynamics involving b and g.

The complicated photophysics of Schiff bases is the result of

many possible processes that can occur with similar rates in

the excited molecule: proton transfer, electron transfer, torsion-

al dynamics engaging several bonds, cis–trans isomerization.

Our results allow suppressing or enhancing these processes by

a judicious choice of substituent or solvent. Thus, excited state

proton transfer rate can be slowed down by increasing the po-

larity of the surrounding medium. The intramolecular electron

transfer can be enhanced by placing a stronger electron ac-

ceptor at C3; in contrast, the role of the donor substituent at

C11 seems to be minor. Most important, a general model has

been proposed which allows predicting the trends in photo-

physical changes caused by substitution. It is based on the

changes in length of important bonds responsible for torsional

dynamics: C6N1, C7N1 and C7C8.

We hope that the present results will contribute to the un-

derstanding of intricacies in the photophysics of Schiff bases.

Currently, we extend the investigations to structures contain-

ing two and three identical proton-transfer functionalities in

one chromophore. Finally, one should not overlook the appli-

cation potential of molecules of which the emission properties

are very strongly affected by the environment.

Summary

Experimental Section

Anils and boranils exhibit photophysical behaviour that is

strongly sensitive to both substitution pattern and environ-

ment. Dual emission in anils is due to intramolecular excited

state proton transfer. Boranils 1a, 1b, and 2a, devoid of

proton-transfer functionality, exhibit single fluorescence from

a state of which the electronic structure is similar to that of

the corresponding anil. On the contrary, in the nitro deriva-

tives, dual emission is observed for both anils and boranils. Its

origin is the intramolecular electron transfer accompanied by

twisting about the C6N1 bond.

Synthetic procedures

All reactions were performed under a dry atmosphere of argon. All

chemicals were used as received from commercial sources without

further purification. Dichloroethane was distilled over P₂O₅under

an argon atmosphere. Thin layer chromatography (TLC) was per-

formed on silica gel or aluminium oxide plates coated with fluores-

cent indicator. Chromatographic purifications were conducted

using 40–63 mm silica gel.

The primary emission of both anils and boranils is quenched

in polar solvents. Transient-absorption experiments for boranils

indicate that the quenching occurs directly to the ground

state, without formation of a “dark”, long-lived intermediate.

On the other hand, long-lived transient species are detected

for anils and can be assigned to trans-keto forms.

General procedure I

To a solution of aniline B (1 equiv) and p-TsOH (trace amounts) in

dry ethanol was added aldehyde A (1 equiv). The resulting solution

was refluxed overnight. After cooling, the desired anil precipitated.

It was then filtered, washed with ethanol, pentane and dried under

vacuum.

The investigation of anils and boranil shows the importance

of all three angles (a, b, g in Schemes 1 and 2) relevant for tor-

sional dynamics and, in consequence, rapid excited-state deac-

tivation. Moreover, our study provides examples of chromo-

phores in which efficient channels of excited-state deactivation

either reinforce or compete against each other. For instance,

photoinduced proton transfer in anils 1 and 2 in nonpolar en-

vironments induces rapid radiationless deactivation, most

probably through twisting motion. Once the possibility of tau-

tomerization is removed by replacement of the OH proton by

General procedure II

To a stirred solution of anil (1 equiv) in anhydrous dichloroethane

at 858C was added BF₃·OEt₂(2.5 equiv). A colour change occurred

or a precipitate appeared revealing the formation of an intermedi-

ate. Then DIEA (2.5 equiv) was added to form the desired boranil.

The course of reaction was monitored by TLC analysis. After cool-

ing, the reaction mixture was washed with a saturated solution of

NaHCO₃and extracted with dichloromethane. Organic layers were

dried over MgSO₄and evaporated under vacuum. The product was

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

13

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

purified by silica gel chromatography and eluted with a mixture of

dichloromethane and petroleum ether.

Synthesis of 1b

General procedure III; yellow powder; yield: 78%. ¹H NMR

(200 MHz, [D₆]acetone): d=8.44 (1H, s, CH Imine), 7.35 (5H, m, CH

Arom), 7.12 (8H, m, CH Arom), 6.94 (2H, m, CH Arom), 6.32 (1H,

dd, CH, J₁=8.8 Hz, J₂=2.2 Hz, CH Arom), 6.02 (1H, d, CH, J=

2.2 Hz, CH Arom), 3.47 (4H, q, J=6,9 Hz, CH₂NEt₂), 1.17 ppm (6H,

t, J=6,9 Hz, CH₃NEt₂); ¹³C NMR (100 MHz, [D₆]acetone): d=164.3,

159.9, 156.2, 143.0, 134.7, 133.7, 126.5, 126.4, 126.3, 125.6, 115.0,

114.7, 109.6, 105.2, 97.8, 44.5, 12.1 ppm; ¹¹B NMR (128 MHz,

[D₆]acetone): d=5.17 ppm (s, BPh₂); elemental analysis calcd (%)

for C₂₉H₂₈BFN₂O: C 77.34, H 6.27, N 6.22; found: C 77.04, H 5.90, N

5.94; EI-MS (m/z): 450.1 (100).

General procedure III

To a solution of anil (1 equiv) in distilled toluene was added BPh₃

(1.5 equiv). The resulting mixture was stirred at 908C. The course

of reaction was monitored by TLC analysis. After cooling, the sol-

vent was removed under vacuum and the residue was purified by

silica gel chromatography eluted with a mixture of dichlorome-

thane and petroleum ether.

Synthesis of 1

1

Synthesis of 2a

General procedure I; yellow powder; yield: 85%. H NMR (200 MHz,

[D₆]acetone): d=13.30 (1H, s, OH), 8.61 (1H, s, CH Imine), 7.24 (5H,

m, CH Arom), 6.36 (1H, m, CH Arom), 6.13 (1H, s, CH Arom), 3.46

(4H, q, CH₂NEt₂), 1.19 ppm (6H, t, CH₃NEt₂); ¹³C NMR (75 MHz,

acetone): d=163.2, 162.8, 135.8, 135.1, 123.3, 123.2, 116.8, 116.5,

109.9, 104.7, 98.1, 45.1, 12.9 ppm; elemental analysis calcd (%) for

C₁₇H₁₉N₂OF: C 71.33, H 6.69, N 9.78; found: C 71.24, H 6.57, N 9.68;

EI-MS (m/z): 286.1 (100).

General procedure II; yellow powder; yield: 84%. ¹H NMR

(200 MHz, CDCl₃): d=8.05 (1H, brs, CH Imine), 7.67 (4H, ABsys,

J

_AB=8.8 Hz, n₀d=20.0 Hz, CH Arom), 7.24 (1H, d, ³J=8.8 Hz, CH

3

4

Arom), 6.40 (1H, dd, J=8.8 Hz, J=2.4 Hz, CH Arom), 6.23 (1H, d,

⁴J=2.4 Hz, CH Arom), 3.48 (4H, q, J=6,8 Hz, CH₂NEt₂), 1.26 ppm

3

(6H, t, ³J=6,8 Hz, CH₃NEt₂); ¹³C NMR (75 MHz, [D₆]acetone): d=

163.4, 160.2, 158.1, 148.1, 136.1, 134.2, 124.8, 119.1, 111.2, 108.4,

98.0, 45.9, 13.0 ppm; ¹¹B NMR (128 MHz, [D₆]acetone): d=0.90 ppm

(t, J=16.6 MHz, BF₂); elemental analysis calcd (%) for C₁₈H₁₈BF₂N₃O:

C 63.37, H 5.32, N 12.32; found: C 63.18, H 5.17, N 12.09; EI-MS (m/

z): 341.1 (100), 322.1 (35).

Synthesis of 2

1

General procedure I; yellow crystals; yield: 38%. H NMR (300 MHz,

CDCl₃): d=13.25 (1H, s, OH), 8.40 (1H, s, CH Imine), 7.47 (4H,

ABsys, J_AB=8.7 Hz, n₀d=113.7 Hz, CH Arom), 7.18 (1H, d, ³J=

8,7 Hz, CH Arom), 6.28 (1H, dd, ³J=8.7 Hz, ⁴J=2.4 Hz, CH Arom),

6.19 (1H, d, ⁴J=2.4 Hz, CH Arom), 3.42 (4H, q, ³J=7.2 Hz, CH₂

NEt₂), 1.23 ppm (6H, t, ³J=7.2 Hz, CH₃NEt₂); ¹³C NMR (75 MHz,

CDCl₃): d=164.4, 162.3, 153.2, 152.8, 134.7, 133.6, 121.8, 119.3,

109.1, 108.5, 104.6, 97.8, 44.9, 12.9 ppm; elemental analysis calcd

(%) for C₁₈H₁₉N₃O: C 73.69, H 6.53, N 14.32; found: C 73.45, H 6.28,

N 14.09; EI-MS (m/z): 293.1 (100).

Synthesis of 3a

General procedure II; orange powder; yield: 72%. ¹H NMR

(200 MHz, [D₆]acetone): d=8.67 (1H, s, CH Imine), 8.13 (4H, ABsyst,

J

_AB=9.1 Hz, n₀d=99.4 Hz), 7.52 (1H, d, CH Arom), 6.63 (1H, dd, J=

9.1 Hz, CH Arom), 6.19 (1H, d, CH, J₁=9.1 Hz, J₂=2.4 Hz, CH Arom),

3.62 (4H, q, J=7.3 Hz, CH₂NEt₂), 1.27 ppm (6H, t, J=7.3 Hz, CH₃

NEt₂); ¹³C NMR (75 MHz, [D₆]acetone): d=162.5, 157.3, 157.2, 146.2,

134.6, 125.0, 123.5, 107.7, 98.0, 45.4, 12.7 ppm; ¹¹B NMR (128 MHz,

[D₆]acetone): d=0.88 ppm (t, J=16.7 Hz, BF₂); elemental analysis

calcd (%) for C₁₇H₁₈BF₂N₃O₃: C 56.54, H 5.02, N 11.64; found: C

56.32, H 4.72, N 11.42; EI-MS (m/z): 361.1 (100), 313.1 (25).

Synthesis of 3

General procedure I; red powder; yield: 76%. ¹H NMR (200 MHz,

CDCl₃): d=13.23 (1H, s, OH), 8.43 (1H, s, CH Imine), 7.77 (4H,

ABsyst, J_AB=8.8 Hz, n₀d=189.8 Hz, CH Arom), 7.18 (1H, d, J=

8.8 Hz, CH Arom), 6.28 (1H, dd, J₁=8.8 Hz, J₂=1.9 Hz, CH Arom),

6.19 (1H, d, J=1.9 Hz, CH Arom), 3.42 (4H, q, J=6,8 Hz, CH₂NEt₂),

1.20 ppm (6H, t, J=6,8 Hz, CH₃NEt₂); ¹³C NMR (75 MHz, CDCl₃): d=

164.3, 162.3, 154.8, 152.8, 144.9, 134.7, 125.2, 121.3, 109.0, 104.6,

97.5, 44.8, 12.7 ppm; elemental analysis calcd (%) for C₁₇H₁₉N₃O₃: C

65.16, H 6.11, N 13.41; found: C 64.82, H 5.82, N 13.19; EI-MS (m/z):

313.1 (100).

Synthesis of 3b

1

General procedure III; yellow orange powder; yield: 85%. H NMR

(200 MHz, [D₆]acetone): d=8.65 (1H, s, CH Imine), 8.05 (2H, m, CH

Arom), 7.30 (7H, m, CH Arom), 7.17 (6H, m, CH Arom), 6.38 (1H,

dd, CH, J₁=8.8 Hz, J₂=2.2 Hz, CH Arom), 6.02 (1H, d, CH, J=

2.2 Hz, CH Arom), 3.47 (4H, q, J=7,3 Hz, CH₂NEt₂), 1.17 ppm (6H,

t, J=7,3 Hz, CH₃NEt₂); ¹³C NMR (75 MHz, acetone): d=165.9, 160.5,

158.0, 152.7, 136.3, 135.0, 134.7, 131.1, 128.3, 127.5, 126.7, 125.9,

124.6, 111.2, 107.1, 98.6, 45.6, 13.0 ppm; ¹¹B NMR (128 MHz, ace-

tone): d=5.59 ppm (s, BPh₂); elemental analysis calcd (%) for

C₂₉H₂₈BN₃O₃: C 72.97, H 5.91, N 8.80; found: C 72.74, H 5.77, N

8.62; EI-MS (m/z): 477.1 (100).

Synthesis of 1a

General procedure II; yellow powder; yield: 43%. ¹H NMR

(200 MHz, [D₆]acetone): d=8.49 (1H, s, CH Imine), 7.62 (2H, m, CH

Arom), 7.47 (1H, d, J=9.2 Hz, CH Arom), 7.25 (2H, m, CH Arom),

6.55 (1H, dd, J₁=9.2 Hz, J₂=2.4 Hz, CH Arom), 6.16 (1H, d, J=

2.4 Hz, CH Arom), 3.57 (4H, q, J=7,2 Hz, CH₂NEt₂), 1.24 ppm (6H,

t, J=7,2 Hz, CH₃NEt₂); ¹³C NMR (75 MHz, CDCl₃): d=160.6, 157.3,

135.5, 126.1, 126.0, 116.8, 116.5, 107.5, 98.1, 45.6, 12.9 ppm;

¹¹B NMR (128 MHz, [D₆]acetone): d=0.85 ppm (t, J=16.7 MHz, BF₂);

elemental analysis calcd (%) for C₁₇H₁₈BF₃N₂O: C 61.11, H 5.43, N

8.38; found: C 60.84, H 5.19, N 8.33; EI-MS (m/z): 334.1 (100), 285.1

(30).

NMR characterization

The 300 (¹H), 400 (¹H), 75 (¹³C), 100 (¹³C) 128 (¹¹B) MHz NMR spectra

were recorded at room temperature with perdeuterated solvents

with residual protonated solvent signals as internal references. The

128 (¹¹B) MHz NMR spectra were recorded at room temperature

with borosilicate glass as reference.

&

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

14

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

Stationary FTIR spectra were recorded on a Perkin–Elmer Spectrum

One FT-IR spectrometer (ATR, Golden Gate), with solid samples.

Spectral measurements

The spectral grade solvents used for spectral measurements: cyclo-

hexane, tetrahydrofuran, diethyl ether, methanol, n-propanol, valer-

onitrile, butyronitrile, acetonitrile (Merck), and toluene (Roth) were

checked for the presence of fluorescing impurities. Butyronitrile

was distilled just prior to experiments.

Calculations

The calculations of electronic and vibrational structure in the

ground electronic state were performed using density functional

theory (DFT). For excited state geometry optimization and for the

calculation of electronic transition energies, time-dependent DFT

model was applied. Two software packages were used, Gaussi-

an09^[62]and ADF.^[63–65]

Electronic absorption spectra were recorded on a Shimadzu UV

3100 spectrophotometer. Stationary luminescence spectra were

obtained using an Edinburgh FS900 CDT, Jasny^[54]or compact

Jasny^[55]spectrofluorimeters, equipped with sample temperature

control systems.

Transient absorption spectra were recorded using a setup based

on 1 kHz amplified Ti:Saphire system (Spectra-Physics) as described

in detail elsewhere.^[56,57]The second harmonic (l=400 nm) was

used to excite the sample. The energy per pulse at the sample was

around 1 mJ. A white-light continuum generated by focusing

pulses at 800 nm into a CaF₂plate was used for probing. The po-

larization of the probe beam was at the magic angle relative to

that of the pump pulses. All spectra were corrected for the chirp

of the probe pulses. The full width at half maximum of the instru-

ment response function was around 200 fs.

Acknowledgements

This work has been financed by the funds from the Polish Na-

tional Science Centre (DEC-2011/01/B/ST4/01130 and 2011/01/

B/ST2/02053). We acknowledge a computing grant from the In-

terdisciplinary Centre for Mathematical and Computational

Modeling. This research was also supported in part by PL-Grid

Infrastructure. We acknowledge the CNRS and ECPM for pro-

viding research facilities and financial support (MRT fellowship

for D.F.). J.D. thanks NanOtechnology, Biomaterials and aLterna-

tive Energy Source for the ERA Integration project from the Eu-

ropean Union (FP7-REGPOT-CT-2011-285949-NOBLESSE). Sup-

port from the Swiss National Foundation (Nr. 200020-147098)

and the University of Geneva is also acknowledged.

Time-resolved fluorescence spectra were measured using femto-

second up-conversion setup, consisting of femtosecond oscillator

followed by a home-built regenerative amplifier.^[58]The amplified

beam was split into two parts: one was used to generate second

harmonic beam at 400 nm, while the rest of the beam was used to

drive a noncollinear optical parametric amplifier. Second harmonic

beam was used for excitation of a sample enclosed in a 200 mm

thick quartz cuvette. The fluorescence was then collected by an all-

reflective Schwarzschild objective, and then imaged onto the sur-

face of a BBO nonlinear crystal. The fluorescence was temporally

gated by an IR beam (1020 nm) generated by a parametric amplifi-

er. The BBO crystal was wobbling, which allowed gating the fluo-

rescence beam in a broad spectral range at once. Such upconvert-

ed spectrum was then imaged onto an input slit of a spectrometer.

Time-resolved spectra (uncorrected for spectral sensitivity) were

measured for different time delays between excitation and gating

pulses, which was controlled by the use of a translation stage. The

temporal resolution was 150 fs, which was mostly determined by

the duration of the excitation pulse. The time constants of the

measured fluorescence maps were retrieved using global analysis,

as described elsewhere.^[59]

Keywords: anils · boranils · photochemistry · photophysics ·

ultrafast spectroscopy

[1] M. H. V. Huynh, T. J. Meyer, Chem. Rev. 2007 , 107 , 500 4 –5 064.

[2] S. Hammes-Schiffer, Acc. Chem. Res. 2009 , 42 , 188 1 –1 889.

[3] C.-C. Hsieh, C.-M. Jiang, P.-T. Chou, Acc. Chem. Res. 2010 , 43 , 136 4 –

1374.

[4] H. Petek, J. Zhao, Chem. Rev. 2010 , 110 , 708 2 –7 099.

[5] M. Barroso, L. G. Arnaut, S. J. Formosinho, RSC Catal. Ser. 2012, 8, 126–

151.

[6] D. R. Weinberg, C. J. Gagliardi, J. F. Hull, C. F. Murphy, C. A. Kent, B. C.

Westlake, A. Paul, D. H. Ess, D. G. McCafferty, T. J. Meyer, Chem. Rev.

2012 , 112 , 401 6 –4 093.

[7] O. S. Wenger, Acc. Chem. Res. 2013 , 46 , 151 7 –1 526.

[8] J. Bonin, M. Robert, Photochem. Photobiol. 2011 , 87 , 119 0 –1 203.

[9] Z. R. Grabowski, K. Rotkiewicz, W. Rettig, Chem. Rev. 2003 , 103 , 389 9 –

4031.

[10] T. Scherer, I. H. M. van Stokkum, A. M. Brouwer, J. W. Verhoeven, J. Phys.

Chem. 1994 , 98 , 1053 9 –1 0549.

[11] W. Rettig, M. Maus, in Conformational Changes Accompanying Intramo-

lecular Excited State Electron Transfer (Ed.: J. Waluk), Wiley-VCH, Wein-

heim, 2000, pp. 1–55.

[12] Z. R. Grabowski, Pure Appl. Chem. 1992, 64, 1249–1255.

[13] J. Herbich, B. Brutschy, in TICT Molecules, Vol. 4 (Ed.: V. Balzani), Wiley-

VCH, Weinheim, 2001, pp. 697–741.

[14] E. Nosenko, G. Wiosna-Sałyga, M. Kunitski, I. Petkova, A. Singh, W. J.

Buma, R. P. Thummel, J. Waluk, Angew. Chem. Int. Ed. 2008 , 47 , 603 7 –

6040; Angew. Chem. 2008 , 120 , 612 6 –6 129.

Time-correlated single photon counting (TC-SPC) measurements

were carried out using a home-built setup. The excitation was pro-

vided by a PicoQuant LDH-D-C-375 laser, emitting at 379 nm, with

pulse width of about 60 ps and pulse energy in the range of 30–

50 pJ. The emission, after passing through a Digikrçm CM112

double grating monochromator working in a subtractive mode,

was detected by an HPM-100-40 hybrid detector (Becker & Hickl)

coupled to a Becker & Hickl SPC-830 TR-SPC module. The decay

profiles were analysed using commercially available packages:

Fluofit version 3.3 (Picoquant) and FAST version 3.3 (Edinburgh

Instruments).

[15] E. Hadjoudis, I. M. Mavridis, Chem. Soc. Rev. 2004, 33, 579–588.

[16] E. Hadjoudis, S. D. Chatziefthimiou, I. M. Mavridis, Curr. Org. Chem. 2009,

13, 269–286.

[17] R. S. Becker, C. Lenoble, A. Zein, J. Phys. Chem. 1987 , 91 , 351 7 –3 524.

[18] R. S. Becker, C. Lenoble, A. Zein, J. Phys. Chem. 1987 , 91 , 350 9 –3 517.

[19] V. C. Vargas, J. Phys. Chem. A 2004 , 108 , 28 1 –2 88.

[20] M. I. Knyazhansky, N. I. Makarova, A. V. Metelitsa, A. I. Shif, V. P. Pichko,

E. P. Ivakhnenko, Mol. Cryst. Liq. Cryst. Sci. Technol. Sect. A 1997, 298,

391–396.

Femtosecond transient IR spectra were recorded upon 400 nm ex-

citation using the pump-probe set up described elsewhere,^[60,61]in

three different probe windows, from 1500 to 1730 cm^ꢀ1. The spec-

tra were scaled and merged using the corresponding factor calcu-

lated for the first window recorded. Further windows until

1980 cm^ꢀ1were checked, but no transient signal was present. At

least six measurements were recorded in each window and

averaged.

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

15

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Paper

[21] M. I. Knyazhansky, A. V. Metelitsa, A. J. Bushkov, S. M. Aldoshin, J. Photo-

chem. Photobiol. A 1996 , 97 , 12 1 –1 26.

[22] K. Kownacki, L. Kaczmarek, A. Grabowska, Chem. Phys. Lett. 1993 , 210 ,

37 3 –3 79.

[23] A. Grabowska, K. Kownacki, L. Kaczmarek, J. Lumin. 1994 , 6 0 –6 1 , 88 6 –

890.

[24] K. Kownacki, A. Mordzinski, R. Wilbrandt, A. Grabowska, Chem. Phys.

Lett. 1994 , 227 , 27 0 –2 76.

[47] D. Frath, S. Azizi, G. Ulrich, R. Ziessel, Org. Lett. 2012 , 14 , 477 4 –4 777.

[48] T. Sekikawa, O. Schalk, G. Wu, A. E. Boguslavskiy, A. Stolow, Springer

Proc. Phys. 2012 , 125 , 31 3 –3 15.

[49] M. L. Horng, J. A. Gardecki, A. Papazyan, M. Maroncelli, J. Phys. Chem.

1995 , 99 , 1731 1 –1 7337.

[50] J. M. Ortiz-S ꢃnchez, R. Gelabert, M. Moreno, J. M. Lluch, J. Chem. Phys.

2008, 129, 214308.

[51] R. Destro, A. Gavezzotti, M. Simonetta, Acta Crystallogr. Sect. B 1978 , 34 ,

286 7 –2 869.

´

[25] A. Mordzinski, K. Kownacki, R. Wilbrandt, C. Rios, J. Inf. Rec. Mater. 1994,

21, 601–604.

[52] S. Chibani, A. Charaf-Eddin, B. Le Guennic, D. Jacquemin, J. Chem.

Theory Comput. 2013 , 9 , 312 7 –3 135.

[53] J. M. Ortiz-S ꢃnchez, D. Bucher, L. C. T. Pierce, P. R. L. Markwick, J. A.

McCammon, J. Chem. Theory Comput. 2012, 8, 2752–2761.

[54] J. Jasny, J. Lumin. 1978 , 17 , 14 9 –1 73.

[26] A. Grabowska, K. Kownacki, J. Karpiuk, S. Dobrin, L. Kaczmarek, Chem.

Phys. Lett. 1997 , 267 , 13 2 –1 40.

[27] A. Grabowska, K. Kownacki, L. Karczmarek, Acta Phys. Pol. A 1995, 88,

1081–1088.

[28] C. Okabe, T. Nakabayashi, Y. Inokuchi, N. Nishi, H. Sekiya, J. Chem. Phys.

2004 , 121 , 943 6 –9 442.

[29] N. Otsubo, C. Okabe, H. Mori, K. Sakota, K. Amimoto, T. Kawato, H.

Sekiya, J. Photochem. Photobiol. A 2002 , 154 , 3 3 –3 9.

[30] W. M. F. Fabian, L. Antonov, D. Nedeltcheva, F. S. Kamounah, P. J. Taylor,

J. Phys. Chem. A 2004 , 108 , 760 3 –7 612.

[31] M. Ziꢄłek, J. Kubicki, A. Maciejewski, R. Naskre˛cki, A. Grabowska, Phys.

Chem. Chem. Phys. 2004, 6, 4682–4689.

[32] M. Ziꢄłek, J. Kubicki, A. Maciejewski, R. Naskre˛cki, A. Grabowska, J.

Chem. Phys. 2006, 124, 124518.

[55] J. Jasny, J. Waluk, Rev. Sci. Instrum. 1998 , 69 , 2242.

[56] G. Duvanel, N. Banerji, E. Vauthey, J. Phys. Chem. A 2007 , 111 , 536 1 –

5369.

[57] N. Banerji, G. Duvanel, A. Perez-Velasco, S. Maity, N. Sakai, S. Matile, E.

Vauthey, J. Phys. Chem. A 2009 , 113 , 820 2 –8 212.

[58] B. Białkowski, Y. Stepanenko, M. Nejbauer, C. Radzewicz, J. Waluk, J. Pho-

tochem. Photobiol. A 2012, 234, 100–106.

[59] P. Fita, E. Luzina, T. Dziembowska, C. Radzewicz, A. Grabowska, J. Chem.

Phys. 2006 , 125 , 184508.

[60] M. Koch, A. Rosspeintner, K. Adamczyk, B. Lang, J. Dreyer, E. T. J. Nibber-

ing, E. Vauthey, J. Am. Chem. Soc. 2013 , 135 , 984 3 –9 848.

[61] M. Koch, R. Letrun, E. Vauthey, J. Am. Chem. Soc. 2014 , 136 , 406 6 –4 074.

[62] Gaussian 09, Revision B.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.

Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-

nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,

A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,

K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.

Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,

M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, T. Keith,

R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S.

Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox,

J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Strat-

mann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L.

Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dan-

nenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz,

J. Cioslowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2010.

[63] C. Fonseca Guerra, J. Snijders, G. te Velde, E. Baerends, Theor. Chem. Acc.

1998, 99, 391–403.

[33] M. Ziꢄłek, J. Kubicki, A. Maciejewski, R. Naskre˛cki, W. Łuniewski, A. Gra-

bowska, J. Photochem. Photobiol. A 2006, 180, 101–108.

[34] M. Ziꢄłek, K. Filipczak, A. Maciejewski, Chem. Phys. Lett. 2008, 464, 181–

186.

´

[35] M. Ziꢄłek, G. Burdzinski, K. Filipczak, J. Karolczak, A. Maciejewski, Phys.

Chem. Chem. Phys. 2008, 10, 1304–1318.

´

[36] M. Ziꢄłek, G. Burdzinski, J. Karolczak, J. Phys. Chem. A 2009, 113, 2854–

2864.

[37] M. Ziꢄłek, M. Gil, J. A. Organero, A. Douhal, Phys. Chem. Chem. Phys.

2010, 12, 2107–2115.

[38] M. Ziꢄłek, C. Martin, M. T. Navarro, H. Garcia, A. Douhal, J. Phys. Chem. C

2011, 115, 8858–8867.

´

[39] M. Ziꢄłek, G. Burdzinski, A. Douhal, Photochem. Photobiol. Sci. 2012, 11,

1389–1400.

[40] M. Ziꢄłek, B. Cohen, X. Yang, L. Sun, M. Paulose, O. K. Varghese, C. A.

Grimes, A. Douhal, Phys. Chem. Chem. Phys. 2012, 14, 2816–2831.

[41] T. Sekikawa, O. Schalk, G. Wu, A. E. Boguslavskiy, A. Stolow, J. Phys.

Chem. A 2013 , 117 , 297 1 –2 979.

[42] L. Spçrkel, G. Cui, W. Thiel, J. Phys. Chem. A 2013 , 117 , 457 4 –4 583.

[43] M. H. Luo, H. Y. Tsai, H. Y. Lin, S. K. Fang, K. Y. Chen, Chin. Chem. Lett.

2012 , 23 , 127 9 –1 282.

[44] T.-C. Fang, H.-Y. Tsai, M.-H. Luo, C.-W. Chang, K.-Y. Chen, Chin. Chem. Lett.

2013 , 24 , 14 5 –1 48.

[64] G. te Velde, F. Bickelhaupt, S. van Gisbergen, C. Fonseca Guerra, E. Baer-

ends, J. Snijders, T. Ziegler, J. Comput. Chem. 2001 , 22 , 931.

[65] ADF2013, SCM, Theoretical Chemistry, Vrije Universiteit, Amsterdam,

The Netherlands, http://www.scm.com.

[45] E. Hadjoudis, Mol. Eng. 1995 , 5 , 30 1 –3 37.

[46] D. Frath, S. Azizi, G. Ulrich, P. Retailleau, R. Ziessel, Org. Lett. 2011 , 13 ,

341 4 –3 417.

Received: July 31, 2014

Published online on && &&, 0000

&

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

16

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

FULL PAPER

Simple twist of fate: Dual fluorescence

observed in anils is attributed to excited

state intramolecular proton transfer

(ESIPT). Nitro-substituted anil and bora-

nil exhibit intramolecular electron trans-

fer coupled with twisting of the nitro-

phenyl group. A general model is pro-

posed that explains photophysical re-

sponses to different substitution pat-

terns in anils and boranils.

&

Photochemistry

´

J. Dobkowski, P. Wnuk, J. Buczynska,

M. Pszona, G. Orzanowska, D. Frath,

G. Ulrich, J. Massue,

S. Mosquera-Vꢃzquez, E. Vauthey,*

C. Radzewicz, R. Ziessel,* J. Waluk*

&& – &&

Substituent and Solvent Effects on the

Excited State Deactivation Channels in

Anils and Boranils

Chem. Eur. J. 2014, 20, 1 – 17

www.chemeurj.org

17

ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Article Doi

Substituent and solvent effects on the excited state deactivation channels in anils and boranils

DOI: 10.1002/chem.201404669

Source and publish data:

Authors:

Article abstract of DOI:10.1002/chem.201404669

Full text of DOI:10.1002/chem.201404669

Products guided by the article

R&D Labs maybe for 1311090-52-4

Relevant to this article

Hot Product