Computational design, synthesis, and mechanochromic properties of new thiophene-based π-conjugated chromophores

Article abstract of DOI:10.1002/chem.201203672

The possibility of exploiting supramolecular architectures for the preparation of innovative mechanochromic devices has been extended by designing novel thienyl-substituted 1,4-bis(ethynyl)benzene dyes, which are characterized by a conjugated, rigid, rodl

Full text of DOI:10.1002/chem.201203672

FULL PAPER
DOI: 10.1002/chem.201203672
Computational Design, Synthesis, and Mechanochromic Properties of New
Thiophene-Based p-Conjugated Chromophores
Giacomo Prampolini,*[a] Fabio Bellina,*[b] Malgorzata Biczysko,[c] Chiara Cappelli,[b, d]
Luciano Carta,[d] Marco Lessi,[b] Andrea Pucci,*[a, b] Giacomo Ruggeri,[a, b] and
Vincenzo Barone[d]
Abstract: The possibility of exploiting
supramolecular architectures for the
preparation of innovative mechano-
chromic devices has been extended by
designing novel thienyl-substituted 1,4-
screening was set up that was able to
select a derivative with optimal spectral
features. The effective combination of
experimental and computational inves-
tigations allowed us to spot those ho-
mologues with already potential aniso-
tropic and aggregachromic features and
characterized by the best spectral prop-
erties and luminescent response. The
best candidate was synthesized and dis-
persed into polyethylene matrix,
a
indeed achieving an “in silico de-
signed” mechanochromic material. Be-
sides the specific applications of this
novel material, the integration of com-
putational and experimental techniques
reported here defines an efficient pro-
tocol that can be applied to make a se-
lection among similar dye candidates,
which constitute the essential respon-
sive part of such supramolecular devi-
ces.
bisACHTUNGTRENNUNG(ethynyl)benzene dyes, which are
characterized by a conjugated, rigid,
rodlike core structure. This new family
of chromophores was synthesized ac-
cording to a simple two-step sequential
cross-coupling reaction, and the optical
properties were investigated in solution
and in a polymeric matrix. To tune the
Keywords: chromophores · compu-
tational design · dyes/pigments · lu-
minescence · mechanochromic ma-
terials
mechanochromic
performances
in
smart polymer materials,
a
virtual
Introduction
“smart” materials, which are endowed with specific optical
properties that can be tuned and controlled by external
stimuli (light, heat, mechanical stress, pH variations, etc.).[1]
“Smart” devices can thus be designed for many different ap-
plications, such as camouflage systems, anti-counterfeiting,
attoreactor sensors, informational displays, or white-emitting
vesicles. Soft intelligent materials, including synthetic poly-
mers among others, are based on the assembly of different
units performing specific functions; for example, compounds
able to respond to a wide variety of stimuli can be inserted
within polymeric matrices (either in interfacial regions or in
more complex supramolecular architectures) to obtain
highly tuneable platforms.[2] Among these mechanorespon-
sive materials, thanks to their possible use for efficient in
situ sensing, particular attention has been recently[3] devoted
to that class of compounds, referred to as mechanochromic,
the color (in absorption and/or fluorescence emission) of
which changes upon mechanical stress. In this case, since
commercial processable polymers (polymer commodities)
display negligible electronic response in the technologically
relevant frequency range (from visible to near-infrared),
color and photoresponse are most conveniently achieved by
introducing properly designed active species.[4] Organic
chromophores, usually based on extended p-conjugated
structures, show large photoresponses in a spectral region
that can be easily tuned from the visible to the near-infrared
by appropriate molecular design. The possible formation of
supramolecular dye architectures adds one more layer of
Impressive progress in material engineering and in the
design of nanohybrid compounds has been inspired over the
past few decades by the growing interest in new soft
[a] Dr. G. Prampolini, Dr. A. Pucci, Prof. G. Ruggeri
Istituto per i Processi Chimico Fisici
Consiglio Nazionale per le Ricerche
Area della Ricerca, via Moruzzi 1
56125 Pisa (Italy)
Fax : (+39)050-3152230
E-mail: giacomo.prampolini@pi.ipcf.cnr.it
[b] Prof. F. Bellina, Dr. C. Cappelli, Dr. M. Lessi, Dr. A. Pucci,
Prof. G. Ruggeri
Dipartimento di Chimica e Chimica Industriale
Universitꢀ di Pisa, via Risorgimento 35
56126 Pisa (Italy)
Fax : (+39)050-2219260
E-mail: bellina@dcci.unipi.it
apucci@dcci.unipi.it
[c] Dr. M. Biczysko
Center for Nanotechnology Innovation@NEST
Istituto Italiano di Tecnologia
Piazza San Silvestro 12, 56127 Pisa (Italy)
[d] Dr. C. Cappelli, L. Carta, Prof. V. Barone
Classe di Scienze
Scuola Normale Superiore
Piazza dei Cavalieri 7, 56126 Pisa (Italy)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201203672.
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

complexity to the problem, providing at the same time a
wealth of exciting opportunities. With respect to the isolated
chromophores, molecular aggregates can display entirely
new features, and are extremely sensitive to the medium
properties, so that aggregachromic materials have been de-
veloped exploiting precisely this sensitivity.[3,5] In any case,
the crucial and responsive part of a mechanochromic mate-
rial is the organic chromophore, either chemically linked
onto the polymeric support or blended into it. In the latter
case, some general criteria that should be satisfied by the
dye can be defined:[3b] as a rigid rodlike shape or the pres-
ence of electron-withdrawing or -donating groups on the ar-
omatic core of the dye.
thienyl-substituted 1,4-bisACTHNGUTERNNU(G ethynyl)benzenes 1 bearing donor
and acceptor groups on the two terminal thiophene rings.
This class of dyes is characterized by a rigid, rodlike, highly
conjugated molecular structure, which is also easily functio-
nalizable by addition of push–pull peripheral groups.
Moreover, the high electron delocalization of such com-
pounds, favored by the presence of the central benzene and
the two lateral conjugated thiophene rings, should in princi-
ple give rise to bright absorption/emission in the UV/visible
range, coupled with a strong sensitivity to the nature of the
push/pull substituents. This is consistent with the hypothesis
that different push/pull pairs could be able to sensitively
shift the absorption/emission maximum wavelength towards
different spectral regions. In our design protocol, quantum
mechanical methods are employed for the in silico screening
of UV/Vis absorption and emission spectra characterizing a
series of derivatives 1, based on the same aromatic skeleton
and bearing substituent groups of different electron-donat-
ing and electron-accepting capability. Once the most promis-
ing compound is selected, it is effectively synthesized and
characterized by measuring its absorption and emission
spectra in solution. Next, the experimental spectra can be
compared with the computed ones to assess the quality of
the predictions. Eventually, the target chromophore is dis-
However, the choice of the best candidate, among a set of
similar target chromophores, may be rather difficult, since
an accurate prediction of the spectral features (including
Stokes shift or line shapes) can become very cumbersome if
solely based on chemical intuition. Based on this, the availa-
bility of reliable and predictive computational methods[2,6]
can be very beneficial, since the results of an in silico
screening can effectively guide the synthetic work, with
clear benefits in terms of cost and time. Evidently, a strong
interplay between theoreticians and experimentalists is re-
quired. For instance, when dealing with mechanochromic
materials, it is crucial to select those chromophoric units
that have some of the following characteristics: a well-ori-
ented transition dipole, a large Stokes shift for high-emission
efficiency, optical behavior depending on the aggregation
extent, and, in case of anti-counterfeiting needs, strong ab-
sorption in the near-UV region and emission in the visible
range.[3b] In particular, once a basic molecular candidate pos-
sessing the required features has been spotted through ex-
periments, an in silico screening can be performed on a
larger set of homologues that differ from the parent chro-
mophore by a small number of substituent groups. Once one
or more targets have been identified by calculations, the ex-
perimental efforts may be concentrated on them. Such a
conceptual procedure can be successful only if the computa-
tional predictions are reliable, and appropriate approaches
are therefore required, which allow the direct simulation of
spectroscopic signals, while retaining a reasonable computa-
tional cost. This request appears to be satisfied by DFT-
based methods, the reliability of which in simulating the
spectral properties of molecular organic systems of medium
up to large dimensions has been widely validated.[7] More-
over, recent methodological advancements have demonstrat-
ed that not only vertical absorption and emission transitions
can be reliably obtained, but that spectral shapes can be di-
rectly simulated also.[8]
persed in
a linear low-density polyethylene (LLDPE)
matrix, and optical properties of the resulting mechanochro-
mic material are investigated. Finally it might be worth men-
tioning that the film aggregachromic behavior, at the base
of the mechanochromic response of dye/polymer blends, is
beyond the scope of the current study and will be investigat-
ed later.
Results and Discussion
Preparation and characterization of thienyl-substituted 1,4-
bis
ACHTUNGTRENNUNG
a small set of thienyl-substituted 1,4-bisACHTUNGTRENNUNG
1a–e, bearing substituents of different natures in terms of
the acceptor/donor capability, was synthesized. Compounds
1a–d were prepared according to a simple and effective
two-step reaction sequence involving a) a Cassar–Heck reac-
tion between the trimethylsilyl-protected p bridge 2 and a 5-
substituted 2-bromothiophene 3, and b) a novel Pd/Cu-cata-
lyzed Sila–Sonogashira cross-coupling between the trime-
thylsilyl-protected alkynes 4, resulting from step (a), and
bromides 3 (Scheme 1). In detail, when [(4-ethynylphenyl)-
ethynyl]trimethylsilane (2)[10] was treated with bromides
Organic fluorophores featuring heteroaromatics as the
main p-conjugated backbones usually display increased po-
larizability, stability, and thermal and chemical robustness
required for fabrication processes of plastic materials. In
particular, thiophene-based p-conjugated molecules have at-
tracted great interest owing to their remarkable electrical
and optical features.[9] Starting from these observations, we
focused our synthetic efforts on the preparation of novel
3a–c in the presence of [PdCl2ACHTNUTRGNEUNG(PhCN)2] (5 mol%), tBu3P
(10 mol%), and DABCO (2.0 equiv) as the base in MeCN
at room temperature,[11] the required alkynes 4a–c were iso-
&
2
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Thiophene-Based Chromophores
FULL PAPER
Dispersion into a polymeric matrix and properties of a se-
lected functional dye: When dispersed in a LLDPE matrix
at a concentration of 0.02 wt%, 1d displays optical features
mostly similar to that in solution, thus indicating a good
compatibility between the componentsꢂ mixture. The dichro-
ic behavior of the 1d/LLDPE mixture was evaluated by
UV/Vis experiments in linearly polarized light (Figure 1) on
Scheme 1. Synthesis of compounds 1a–d.
lated in 60, 65, and 37% yield, respectively (Scheme 1). The
target chromophores 1a–d were successfully obtained from
a Sila–Sonogashira-type coupling involving the reaction of
trimethylsilylalkynes 4a–c with thienyl bromides 3c–e in the
Figure 1. UV/Vis absorption spectra as a function of the polarization
angle q of a LLDPE-oriented film (Dr=5) containing 1d (0.02 wt%).
presence of [PdCl2ACHTNUGTRNE(UGN PhCN)2] (5 mol%), tBu3PHBF4
(10 mol%), CuI (10 mol%), and BnBu3NCl (20 mol%), as
the phase-transfer catalyst (PTC) in toluene and aqueous
NaOH at 408C, according to a recent synthetic protocol de-
veloped by us.[12] In this way, thiophene-based derivatives
1a–d were isolated in 50–95% yields. Similarly, the symmet-
rically substituted derivative 1e was obtained with 68% iso-
lated yield by reacting a molar excess amount of bromide
the uniaxially oriented film (drawing ratio Dr=5). When
the polarization of the incident light is parallel to the
stretching direction (q=08), the film shows a broad absorp-
tion band that decreases in the perpendicular configuration
(q=908) but provides only a moderate dichroic ratio R (A08/
A908) of 5. This behavior suggests that the fairly extended
conjugated structure of 1d is likely responsible for the limit-
ed anisotropic properties of the oriented film. This is in
agreement with the fact that a slightly pronounced rod-
shaped chromophore is often characterized by a transition
dipole moment barely oriented along its molecular axis.[13]
The luminescence of the LLDPE film containing 0.1 wt%
of 1d was recorded in the pristine form and after uniaxial
deformation (Figure 2). The emission band is pointed at
about 460 nm and with a Stokes shift of about 90 nm, simi-
larly to what was observed in THF (Table 1). No evident dif-
ferences in luminescence occurred after film deformation,
except for the signal decreasing due to film thinning. This
behavior indicates that 1d was molecularly dispersed within
3a
with
1,4-bis[(trimethylsilyl)ethynyl]benzene
(5)
(Scheme 2).
Scheme 2. Synthesis of compound 1e.
Table 1. Experimental maximum absorption/emission wavelengths (lmax
and quantum yields of compounds 1a–e in THF (10ꢀ6 m).
)
lmax (abs) [nm]
lmax (em) [nm]
Ff[a]
The absorption and emission spectra of chromophores
1a–e were experimentally measured as solutions in THF.
All of the chromophores 1a–e show an absorption maxi-
mum in the near-UV region, in the range between 350 and
370 nm. Looking at the emissions, compound 1d displays
both the largest Stokes shift (96 nm) and one of the highest
quantum yields, thereby making it suitable as a possible can-
didate for the preparation of mechanochromic materials.
1a
1b
1c
1d
1e
348
368
352
372
348
403
bb[b]
bb[c]
468
0.26
0.13
0.02
0.22
0.29
395
[a] Fluorescence quantum yield was determined relative to quinine sul-
fate in 0.1m H2SO4. [b] Broad band (bb) from 420 to 570 nm. [c] Broad
band from 395 to 460 nm.
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

G. Prampolini, F. Bellina et al.
confidently applied to simulate broad spectral features. All
details of this first validation can be found in the Supporting
Information; here we only point out that adiabatic ap-
proaches should be always considered for a detailed analysis
of well-resolved experimental spectra.
A final validation of the reliability of the computational
protocol in the specific context of thiophene-based chromo-
phores was performed for compounds 1a–e, the measured
absorption and emission spectra of which (Table 1) were
compared to their VE, VG, and AS counterparts with re-
spect to both transition energies (Table 2) and whole spectra
(Figure 3).
Table 2. Computed and experimental (exptl) maximum absorption/emis-
sion wavelengths (lmax) of compounds 1a–e.
Figure 2. Emission of a LLDPE film containing 1d (0.1 wt%) before and
after orientation at Dr=5.
l
max (abs) [nm]
lmax (em) [nm]
VE
VG
AS
exptl
VE
VG
AS
exptl
1a
1b
1c
1d
1e
354
376
359
377
353
350
374
362
374
350
350
373
364
373
349
348
368
352
372
348
427
462
433
463
426
392
420
395
430
392
395
424
398
435
395
403
bb[a]
bb[b]
468
the LLDPE matrix and it did not generate significant p–p
stacked chromophore structures even at the highest concen-
tration investigated. No mechanochromism was therefore
achieved.[13]
395
[a] Broad band (bb) from 420 to 570 nm. [b] Broad band from 395 to
460 nm.
Computational screening of push–pull thienyl-substituted
1,4-bisACHTUNGTRENNUNG(ethynyl)benzenes dyes
The VE approach is able to reproduce absorption maxima
well, with an average error between calculated and experi-
mental values of about 6 nm. Even more relevant is the fact
that the calculations are able to correctly reproduce the ex-
perimental trend moving from one chromophore to the
other (i.e., 1aꢁ1e<1c<1b<1d). In addition, in all cases
the most intense transition is due to a HOMO–LUMO exci-
tation, involving the frontier orbitals of p character, so that
the electronic transition of interest can be classified as p!
p*. Moving to emission, as already found in the preliminary
tests (see the Supporting Information), the VE approach
gives less accurate results and is therefore less effective for
reliable predictions. For instance, in compounds 1a and 1e,
a larger average error between experimental results and cal-
culations is obtained (around 20 nm). Furthermore, the ab-
sence in the VE approach of any reliable treatment of the
line shape makes the comparison and assignment of the ex-
perimental and calculated absorption maxima difficult. This
is most evident for the emission band of compounds 1b and
1c, in which the broad shape of the experimental band
makes a precise assignment of the maximum emission wave-
length rather questionable. For such reasons, a simple check
on vertical excitation energies has been followed by more
refined VG and AS approaches. Absorption and emission
spectra, computed with different approaches, are compared
to their experimental counterparts in Figure 3 for com-
pounds 1a, 1c, and 1e. Similar results were obtained for
other homologues. Inspection of Figure 3 confirms the abili-
ty of both VG and AS approaches to give a maximum ab-
sorption wavelength in excellent agreement with experi-
ment, but also to accurately describe the experimental line
shapes. More importantly, the improvement with respect to
Validation of the protocol: As the effect on the spectral fea-
tures of the dye of selected combination push/pull substitu-
ents is not easily predictable based only on chemical intu-
ition, we decided to perform a virtual screening over several
candidate pairs. Among the available computational proto-
cols, the most popular vertical energy (VE) approach relies
on the computation of vertical excitation energies, which are
further convoluted to simulate line shapes. Such a treatment
completely neglects the influence of nuclear motions, de-
spite the well-recognized notion that a proper account of vi-
brational effects is often mandatory to correctly interpret
the experimental findings.[7] In this respect, the calculation
of reliable linear electronic spectra, including vibronic con-
tributions (vertical gradient (VG), adiabatic shift (AS), or
adiabatic Hessian (AH) approaches; see the Supporting In-
formation for details), has recently become feasible for
medium-to-large systems,[8,14] and in several cases the inclu-
sion of vibrational contributions is a key factor for a correct
interpretation of the experimental outcomes.[6j,15] Finally, the
surrounding environment can be taken into account in a
time-independent approach by employing continuum solva-
tion models,[6e,f,16] which provide rather accurate results at a
relatively low computational cost and can be extended to
several spectroscopic properties,[6c,17] with further considera-
tion of spectral line shapes.[17c,18] A first assessment of the
adequacy of the specific computational model (i.e., choice
of the DFT functional/basis set, VG and AS approximations,
and solvation model) has been performed on both phenyl-
and thiophene-based p-conjugated chromophores, by com-
paring the computed spectra with their experimental coun-
terparts. It was found that both VG and AS models can be
&
4
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Thiophene-Based Chromophores
FULL PAPER
AS approaches are also in agreement with the trend experi-
mentally obtained in luminescence, namely, 1e<1c<1a<
1b<1d. These results suggest that the virtual screening of
the absorption properties of novel chromophore candidates
can be performed at the VE level. Similarly, although the
computational cost of VG/AS fluorescence spectra is cer-
tainly balanced by the gain in accuracy and reliability, the
large number of targets to be investigated suggested that we
should perform an initial preliminary screening at the VE
level also with respect to emission. Thereafter, a restricted
set of homologues will be selected based on VE results and
considered for AS calculations. The final target compound
to be effectively synthesized will be thus chosen within this
second restricted set.
Virtual screening: Beside some substituents already consid-
ered in the previous set 1a–1e (as the CHO group), several
new chromophores 1, characterized by a strong electron-
withdrawing 2-methylenemalonic unit, were included in the
new screening set. As a matter of fact, this latter type of
substituent should be characterized by both a large Stokes
shift, generally associated with highly luminescent dyes, and
an increased capability to aggregate, owing to the augment-
ed polarity and the pronounced rodlike shape.[13] The pres-
ence of push–pull substituents may also drive the transition
dipole moment along the molecular axis, thus conferring to
the molecule anisotropic potentialities.[19] The VE screening
results are summarized in Table 3, by specifying, for each
homologue of the target set, the position of the absorption
and emission peaks and the computed Stokes shifts.
Table 3. VE screening results: predicted maximum absorption/emission
wavelengths and Stokes shifts in nm.
X
Y
lmax (abs)
lmax (em)
Stokes shift
Figure 3. Computed and experimental spectra are shown for compounds
1a, 1c, and 1e. Experimental spectra are reported with a black solid line.
Dashed, solid, and dotted lines are employed for spectra computed in the
VE, VG, and AS approximations, respectively. Absorption intensity:
molar absorption coefficient in dm3 molꢀ1 cmꢀ1; emission intensity in
H
OH
H
CHO
COOH
347
377
370
434
418
426
370
364
433
412
419
388
382
442
428
436
418
463
454
534
520
527
461
454
529
516
520
491
479
562
558
567
71
86
84
100
102
101
91
CH=
G
CH=C(CN)AHCUTNGTRENNUNG
mJmolꢀ1 sꢀ1
.
CH=C(CN)2
CHO
COOH
SH
90
96
CH=
E
CH=C(CN)
N
104
101
103
97
120
130
131
VE is most significant for emission than for absorption. In
fact, the increased accuracy of the vibronic approaches is ap-
parent for the 1a and 1e chromophores, for which the sepa-
ration between the two most intense peaks is now evident
and very well reproduced, and the accuracy of the predicted
maximum wavelength peak is increased, with respect to VE
computed values, by roughly a factor of 2.
CH=C(CN)2
CHO
COOH
NH2
CH=
CH=C(CN)
CH=C(CN)2
A
AHCTUNGTRENNUNG
As already noted, the emission bands obtained experi-
mentally for compounds 1b and 1c are rather broad and do
not present any clear pattern of vibrational structure. This
makes the comparison with their simulated counterparts
rather difficult. Broadly speaking, as far as 1c is concerned,
VG and AS (but also VE) spectra fall within the wide spec-
tral region (395–460 nm) of the experimental emission.
More importantly, emission maxima obtained with VG and
It appears that different pairs of X and Y substituents
highly influence the absorption/emission peak positions. The
homologue characterized by X=Y=H (labeled 1j, see the
Supporting Information) is reported in the first row and can
be taken as a reference to evaluate the direction of the com-
puted shifts. In this respect, all X/Y pairs shift toward larger
values, that is, lower energy regions of the spectrum, both in
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

G. Prampolini, F. Bellina et al.
absorption and in emission. In the former case, as far as
ꢃpullꢂ substituents are concerned, the maximum shift
(around 90 nm with respect to 1j) is observed with Y=
(CH=ACHTUNGTRENNUNG(C9H4O2)), whereas the minimum one (about 30 nm)
is registered when aldehydic or carboxylic groups are em-
ployed. Also the redshifts due to the X ꢃpushꢂ substitutions
appear, for a given Y, rather regular, with NH2 >OHꢁSH.
The predicted redshift in emission with respect to 1j is even
more marked, being as large as 100 nm. In other words, the
substitution of hydrogen atoms with the selected X/Y pairs
systematically increases the Stokes shift. It may be worth
noting that, at variance with the absorption, no regular
trend can be spotted on Y substitution for a given X. For in-
stance, the redshift caused by the CH=ACTHUNRGTNEUNG(C9H4O2) and CH=
C(CN)2 Y substituents is 111 and 102 nm, respectively, when
combined with X=SH, and 144 and 149 nm, respectively,
when combined with X=NH2. These results indicate the ab-
sence of a clear trend and rule out the possibility of a
simple rationalization of the results, based only on ꢃpush/
pullꢂ effects, thereby highlighting the important role that an
in silico screening can play toward improved rational molec-
ular design.
Figure 4. Absorption (dashed line) and emission (solid lines) spectra of
the restricted screening set 1 f–1h, reported along with the reference
compound 1j. All spectra have been simulated within the AS approach
and convoluted with a 750 cmꢀ1 half width at half-maximum (HWHM)
Gaussian function. Absorption intensity: molar absorption coefficient in
dm3 molꢀ1 cmꢀ1; emission intensity in mJmolꢀ1 sꢀ1
.
Based on the former VE screening, a new restricted
screening set was selected by considering that the largest
Stokes shift is registered for chromophores substituted with
Y=CH=C(CN)2. Absorption and emission spectra have
been therefore simulated at the AS level for this “pull”
group in combination with all three considered “push” ones,
giving rise to target compounds 1 f, 1g, and 1h (X=OH,
SH, and NH2, respectively).
SH group upon electronic excitation. Therefore, 1 f could be
suggested as the best candidate for further development.
However, because we had to take into account the necessity
of modulating the interaction between the dye molecule and
its miscibility within a polymer matrix, compound 1i (in
which the hydroxyl “push” substituent was replaced by a
methoxyl group) was finally chosen as the target chromo-
phore to be prepared.
Absorption and emission bands were re-computed accord-
ing to this latter substitution: the resulting spectra are re-
ported in Figure 5 together with the ones obtained for ho-
The AS computed spectra, presented in Figure 4, clearly
show that the final target cannot be identified among the
considered homologues solely based on VE results. In fact,
despite the fact that 1h seems to be the most promising
target (Table 2) owing to the largest Stokes shift computed
at the VE level and the similar transition intensities found
for all three compounds, the direct simulation of the absorp-
tion and emission spectra reported in Figure 4 shows that al-
though 1h exhibits the largest spectral window between ab-
sorption and emission maxima, at the same time much
lower absorption and, even more importantly, emission in-
tensities are to be expected.
For such reasons, the 1 f and 1g chromophores, showing
large Stokes shift and at the same time good absorption and
emission features, are particularly suitable for the applica-
tions considered here. Between the latter two systems, 1g
(i.e., the one with X=SH) may show some disadvantages
related to the significant change of the steric position of the
Figure 5. Experimental (black solid lines) and computed (gray dashed
and solid lines for absorption and emission, respectively) spectra of the
final target compound 1i. Computed spectra for 1 f are reported for com-
parison. All spectra simulated within the AS approach and convoluted
with a Gaussian function of 750 cmꢀ1 HWHM. Absorption intensity in
dm3 molꢀ1 cmꢀ1; emission intensity in mJmolꢀ1 sꢀ1
.
&
6
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Thiophene-Based Chromophores
FULL PAPER
mologue 1 f. It is worth mentioning that in this case, both
absorption and emission spectra were computed at the AS
level to enforce the prediction and improve the comparison
with the experimental band.
Synthesis and experimental characterization of compound
1i: Compound 1i was synthesized according to the reaction
sequence summarized in Scheme 3, and its absorption/emis-
sion spectra were recorded on solutions in THF.
Figure 6. UV/Vis absorption (left) and emission (right) spectra of
a
LLDPE film containing 1i (0.02 wt%).
Scheme 3. Synthesis of compound 1i.
The resulting peaks were found at 422 nm in absorption
and at 512 nm in emission. The a posteriori experimental
spectra are also reported in Figure 5. The agreement with
the theoretical predictions is very good, both for absorption
and emission peaks and intensities. Moreover, use of the AS
method allowed us to predict a line shape, which is also in
good agreement with the experimental bands. Finally, in line
with the validation results, the overall reproduction of the
absorption spectrum is better than the one found for emis-
sion, which results in a slightly underestimated Stokes shift.
However, all in all, the computed screening succeeded in in-
dicating a target chromophore with the chosen characteris-
tics.
Figure 7. UV/Vis absorption spectra as a function of the polarization
angle q of a LLDPE-oriented film (Dr=5) containing 1i (0.02 wt%).
Inset: digital images of the film under near-UV light (366 nm), observed
through a linear polarizer
Dispersion into a polymeric matrix and properties of the
ꢀdesignedꢁ compound: Once the predicted spectroscopic be-
havior of target chromophore 1i was confirmed, this dye
was coupled with a polymeric matrix to assemble the smart
material, the final aim of this work. First, 1i (0.02 wt%) dis-
persed within a LLDPE matrix showed absorption bands at
450 and 480 nm, respectively, redshifted about 25 nm with
respect to the solution in THF. Conversely, the emission of
the film was at 498 nm, that is, blueshifted 14 nm compared
with 1i in solution (Figure 6).
Moreover, the overall vibronic pattern of the optical
bands is also different, possibly owing to increased rigidity
provided by the polymeric environment. These characteris-
tics can be also attributed to the poor molecular interaction
between 1i chromophores and the aliphatic repeating units
of the supporting LLDPE matrix, even though no spectral
bands attributed to the formation of dye aggregates are
found at this concentration. The dichroic behavior of the 1i/
LLDPE mixture was confirmed by UV/Vis experiments in
linearly polarized light (Figure 7) on the uniaxially oriented
film (drawing ratio Dr=5). When the polarization of the in-
cident light is parallel to the stretching direction (q=08), the
film showed an absorption maximum peak at 450 nm. On
the contrary, in the perpendicular configuration (q=908) the
absorption band is mostly suppressed, thus clearly indicating
a pronounced anisotropic behavior. The dichroic ratio R
(A08/A908) is as high as 25, which is 5 times higher than that
recorded for the 1d/LLDPE-oriented film. This result is as-
cribed both to its rodlike shape, which allows the transition
dipole moment to be well aligned with the molecular
axis,[13a,19] and to its dispersion within PE at a molecular
level. The same behavior was recorded for the emission as
evidenced by images of the film recorded under the illumi-
nation of a 366 nm UV lamp with the superimposition of a
linear polarizer oriented parallel and perpendicular to the
drawing direction (Figure 7, inset).
In analogy with previous findings for perylene-based
dyes,[20] by increasing the 1i concentration within LLDPE
(from 0.02 to 0.1 wt%) the fluorescence of the film was
mostly suppressed at all wavelengths, which suggests very
strong p–p packing among 1i chromophores (Figure 8). In-
terestingly, application of a mechanical stress produced a
partial but effective disruption of 1i aggregates, thus restor-
ing the typical green emission of the film coming from iso-
lated non-interacting dyes (500–550 nm). Molecular disper-
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

G. Prampolini, F. Bellina et al.
indeed a fundamental issue, especially in the presence of a
rather large set of chromophores. The effective combination
of experimental design and the computational screening al-
lowed the selection of the best chromophore in terms of
highly anisotropic features, aggregachromic properties, and
bright luminescence. Thereafter, a smart material was creat-
ed by assembling the chromophore with a polymeric matrix.
In fact, the selected dye when dispersed into linear low-den-
sity polyethylene showed emission features depending on
the dye-aggregation extent, that is, a remarkable quenching
of luminescence occurs at the highest concentration. Once a
mechanical stress was applied, the dye promptly restored its
bright green emission thanks to the rupture of its supramo-
lecular structure provided by the polymer-matrix orientation
displaying distinct mechanochromic behavior.
From a computational point of view, the present study
also demonstrates that experimental effects (e.g., solvation)
are well reproduced by simulations, which underlies the abil-
ity of such an approach to dissect the role of several, con-
comitant effects in tuning optical properties. A further step
forward in the computational screening could be to take
into account, at a reliable level of accuracy, the interactions
of the polymeric matrix with the chromophore, thus extend-
ing the predictions to the whole mechanochromic material.
Indeed, work in this direction is in progress in our laborato-
ry.
Figure 8. Emission of a LLDPE film containing 0.02 and 0.1 wt% 1i
before and after orientation at Dr=5. Inset: digital image of the oriented
film under irradiation at 366 nm.
sion flanked by the high luminescent response due to the 1i
optical characteristics give to the film a distinct contrast in
emission between the oriented portion and the pristine
sample (Figure 8, inset). This mechanochromism is substan-
tially different from that illustrated in the literature.[3b,21] In
these cases, the film changes its emission upon drawing
owing to the disruption of chromophore aggregates, which
also emit but at higher wavelengths than the fluorescence of
the isolated dyes. In our case, the mechanism is equally ef-
fective since the film passes after drawing from the OFF
state to the ON state, such as that occurring in light
switches.
Acknowledgements
This work has been supported by the Fondazione Cassa di Risparmio di
Pisa under “POLOPTEL” project no. 167/09.
These results are a clear demonstration of the agreement
between the computational screening and the experimental
results. A dye useful for the development of mechanochro-
mic polymer devices must have an extended conjugation
flanked by a rod-shaped structure, which is able to promote
both the dichroic properties, when molecularly dispersed,
and the formation of stacked assemblies. Stokes shift and
quantum efficiency are the most relevant features provided
by the computational analysis, since dyes that emit a bright-
er luminescence in the visible region allow larger optical
variations at modest solicitations to be obtained.
[1] a) X. Zhang, S. Rehm, M. M. Safont-Sempere, F. Wꢄrthner, Nat.
Chem. 2009, 1, 623–629; b) Y. Sagara, T. Kato, Nat. Chem. 2009, 1,
605–610; c) Y. Sagara, T. Kato, Angew. Chem. 2011, 123, 9294–
9298; Angew. Chem. Int. Ed. 2011, 50, 9128–9132; d) P. Anzenbach-
er, F. Li, M. A. Palacios, Angew. Chem. 2012, 124, 2395; Angew.
Chem. Int. Ed. 2012, 51, 2345.
[2] J.-L. Bredas, S. R. Marder, E. Reichmanis, Chem. Mater. 2011, 23,
309–309.
[3] a) A. Pucci, R. Bizzarri, G. Ruggeri, Soft Matter 2011, 7, 3689–3700;
b) A. Pucci, G. Ruggeri, J. Mater. Chem. 2011, 21, 8282–8291.
[4] a) D. A. Davis, A. Hamilton, J. L. Yang, L. D. Cremar, D. Van
Gough, S. L. Potisek, M. T. Ong, P. V. Braun, T. J. Martinez, S. R.
White, J. S. Moore, N. R. Sottos, Nature 2009, 459, 68–72; b) Y.
Chen, A. J. H. Spiering, S. Karthikeyan, G. W. M. Peters, E. W.
Meijer, R. P. Sijbesma, Nat. Chem. 2012, 4, 559–562.
Conclusion
[5] C. Weder, J. Mater. Chem. 2011, 21, 8235–8236.
The efficient design of innovative dye-doped polymeric ma-
terials with smart attributes requires the fine tuning of sev-
eral parameters, among which the modulation of the optical
characteristics of chromophores as a function of their aggre-
gation behavior appears to be the most intriguing. Within
this framework, the synthetic approach allowed the prepara-
[6] a) C. J. Cramer, Essentials of Computational Chemistry, 2nd ed.,
Wiley, Chichester, 2004; b) P. Kirkpatrick, Nat. Rev. Drug Discovery
2005, 4, 959; c) M. Cifelli, L. De Gaetani, G. Prampolini, A. Tani, J.
Phys. Chem. B 2008, 112, 9777–9786; d) M. McCullagh, T. Prytkova,
S. Tonzani, N. D. Winter, G. C. Schatz, J. Phys. Chem. B 2008, 112,
10388–10398; e) L. R. Dalton, S. J. Benight, L. E. Johnson, D. B.
Knorr, I. Kosilkin, B. E. Eichinger, B. H. Robinson, A. K. Y. Jen,
R. M. Overney, Chem. Mater. 2011, 23, 430–445; f) S. Difley, L.-P.
Wang, S. Yeganeh, S. R. Yost, T. V. Voorhis, Acc. Chem. Res. 2010,
43, 995–1004; g) D. Beljonne, J. Cornil, L. Muccioli, C. Zannoni, J.-
L. Brꢅdas, F. Castet, Chem. Mater. 2011, 23, 591–609; h) A. Pedone,
G. Prampolini, S. Monti, V. Barone, Phys. Chem. Chem. Phys. 2011,
13, 16689–16697; i) R. J. Macfarlane, B. Lee, M. R. Jones, N. Harris,
tion of novel thienyl-substituted 1,4-bisACHTNUTRGNEUNG(ethynyl)benzene
dyes characterized by a well-defined rodlike conjugated
core with optical properties remarkably predicted by relia-
ble computational methods. The selection of the best candi-
date for the preparation of mechanochromic devices was
&
8
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Thiophene-Based Chromophores
FULL PAPER
G. C. Schatz, C. A. Mirkin, Science 2011, 334, 204–208; j) S. Fantac-
ci, F. De Angelis, Coord. Chem. Rev. 2011, 255, 2704–2726; k) G. C.
Schatz, J. Phys. Chem. Lett. 2011, 2, 125–126; l) L. Zhu, Y. Yi, Y. Li,
E.-G. Kim, V. Coropceanu, J.-L. Brꢅdas, J. Am. Chem. Soc. 2012,
134, 2340–2347.
[16] a) S. Miertus, E. Scrocco, J. Tomasi, Chem. Phys. Lett. 1981, 55, 117;
b) C. J. Cramer, D. G. Truhlar, Chem. Rev. 1999, 99, 2161; c) J.
Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 2005, 105, 2999–3093;
d) J. Mosnacek, M. Bertoldo, C. Kosa, C. Cappelli, G. Ruggeri, I.
Lukac, F. Ciardelli, Polym. Degrad. Stab. 2007, 92, 849–858; e) A.
Klamt, WIREs Comput. Mol. Sci. 2011, 1, 699–709; f) J. Tomasi,
WIREs Comput. Mol. Sci. 2011, 1, 855–867.
[7] V. Barone, Computational Strategies for Spectroscopy: From Small
Molecules to Nano Systems, Wiley, Hoboken, 2011.
[17] a) J. L. Rivail, D. Rinaldi, V. Dillet, Mol. Phys. 1996, 89, 1521–1529;
b) C. Cappelli in Continuum Solvation Models in Chemical Physics:
Theory and Applications (Eds.: B. Mennucci, R. Cammi), Wiley,
Chichester, 2007; c) B. Mennucci, C. Cappelli, C. A. Guido, R.
Cammi, J. Tomasi, J. Phys. Chem. A 2009, 113, 3009–3020; d) A. V.
Marenich, C. J. Cramer, D. G. Truhlar, C. A. Guido, B. Mennucci, G.
Scalmani, M. Frisch, J. Chem. Sci. 2011, 2, 2143–2161.
[18] a) Y. H. Wang, M. Halik, C. K. Wang, S. R. Marder, Y. Luo, J.
Chem. Phys. 2005, 123, 194311; b) R. Improta, V. Barone, F. Santoro,
Angew. Chem. 2007, 119, 409–412; Angew. Chem. Int. Ed. 2007, 46,
405–408; c) F. Avila Ferrer, R. Improta, F. Santoro, V. Barone, Phys.
Chem. Chem. Phys. 2011, 13, 17007–17012.
[19] A. Pucci, C. Cappelli, S. Bronco, G. Ruggeri, J. Phys. Chem. B 2006,
110, 3127–3134.
[20] F. Donati, A. Pucci, C. Cappelli, B. Mennucci, G. Ruggeri, J. Phys.
Chem. B 2008, 112, 3668–3679.
[21] a) B. R. Crenshaw, C. Weder, Chem. Mater. 2003, 15, 4717–4724;
b) C. Lçwe, C. Weder, Adv. Mater. 2002, 14, 1625–1629.
[8] a) V. Barone, J. Bloino, M. Biczysko, F. Santoro, J. Chem. Theory
Comput. 2009, 5, 540–554; b) J. Bloino, M. Biczysko, F. Santoro, V.
Barone, J. Chem. Theory Comput. 2010, 6, 1256–1274.
[9] M. Usui, H. Fukumoto, T. Yamamoto, Bull. Chem. Soc. Jpn. 2010,
83, 1397.
[10] J. Fortage, J. Boixel, E. Blart, L. Hammarstrçm, H. C. Becker, F.
Odobel, Chem. Eur. J. 2008, 14, 3467–3480.
[11] A. Soheili, J. Albaneze-Walker, J. A. Murry, P. G. Dormer, D. L.
Hughes, Org. Lett. 2003, 5, 4191–4194.
[12] F. Bellina, M. Lessi, Synlett 2012, 23, 773–777.
[13] a) N. Tirelli, S. Amabile, C. Cellai, A. Pucci, L. Regoli, G. Ruggeri,
F. Ciardelli, Macromolecules 2001, 34, 2129–2137; b) A. Pucci, N.
Tirelli, G. Ruggeri, F. Ciardelli, Macromol. Chem. Phys. 2005, 206,
102.
[14] F. Santoro, R. Improta, A. Lami, J. Bloino, V. Barone, J. Chem.
Phys. 2007, 126, 084509–084513.
[15] a) G. Pietraperzia, M. Pasquini, F. Mazzoni, G. Piani, M. Becucci,
M. Biczysko, D. Michalski, J. Bloino, V. Barone, J. Phys. Chem. A
2011, 115, 9603–9611; b) D. Jacquemin, E. Bremond, I. Ciofini, C.
Adamo, J. Phys. Chem. Lett. 2012, 3, 468–471.
Received: October 15, 2012
Published online: && &&, 0000
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!

G. Prampolini, F. Bellina et al.
Get smart! A new family of lumines-
cent chromophores has been designed.
The procedure used allowed the selec-
tion of the best candidate for the prep-
aration of a novel mechanoresponsive
“smart” device. The integration of the
computational and experimental tech-
niques defines an efficient protocol
that can be applied when a selection
among very similar dye candidates is
required (see figure).
Chromophores
G. Prampolini,* F. Bellina,*
M. Biczysko, C. Cappelli, L. Carta,
M. Lessi, A. Pucci,* G. Ruggeri,
V. Barone ...................... &&&& — &&&&
Computational Design, Synthesis, and
Mechanochromic Properties of New
Thiophene-Based p-Conjugated Chro-
mophores
&
10
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!