Yoshimitsu Sagara1,2, Marc Karman2, Ester Verde-Sesto2,3, Kazuya Matsuo1, Yuna Kim1, Nobuyuki Tamaoki1, Christoph Weder2. 1. Research Institute for Electronic Science, Hokkaido University , N20, W10, Kita-Ku, Sapporo 001-0020, Japan. 2. Adolphe Merkle Institute, University of Fribourg , Chemin des Verdiers 4, CH-1700 Fribourg, Switzerland. 3. POLYMAT, University of the Basque Country UPV/EHU, Joxe Mari Korta Center , Avda. Tolosa 72, 20018 Donostia-San Sebastián, Spain.
Abstract
The integration of mechanophores, motifs that transduce mechanical forces into chemical reactions, allows creating materials with stress-dependent properties. Typical mechanophores are activated by cleaving weak covalent bonds, but these reactions can also be triggered by other stimuli, and this renders the behavior unspecific. Here we show that this problem can be overcome by extending the molecular-shuttle function of rotaxanes to mechanical activation. A mechanically interlocked mechanophore composed of a fluorophore-carrying macrocycle and a dumbbell-shaped molecule containing a matching quencher was integrated into a polyurethane elastomer. Deformation of this polymer causes a fluorescence turn-on, due to the spatial separation of fluorophore and quencher. This process is specific, efficient, instantly reversible, and elicits an easily detectable optical signal that correlates with the applied force.
The integration of mechanophores, motifs that transduce mechanical forces into chemical reactions, allows creating materials with stress-dependent properties. Typical mechanophores are activated by cleaving weak covalent bonds, but these reactions can also be triggered by other stimuli, and this renders the behavior unspecific. Here we show that this problem can be overcome by extending the molecular-shuttle function of rotaxanes to mechanical activation. A mechanically interlocked mechanophore composed of a fluorophore-carrying macrocycle and a dumbbell-shaped molecule containing a matching quencher was integrated into a polyurethane elastomer. Deformation of this polymer causes a fluorescence turn-on, due to the spatial separation of fluorophore and quencher. This process is specific, efficient, instantly reversible, and elicits an easily detectable optical signal that correlates with the applied force.
The capability to visualize
mechanical forces in polymers by an optical response is useful to
investigate stress transfer and failure mechanisms and valuable for
applications that range from tamper-proof packaging to structural
integrity monitoring.[1−4] Early attempts to create mechanochromic polymers involved the incorporation
of conjugated segments, whose absorption changes due to conformational
rearrangements,[5] fluorophores with assembly
dependent emission characteristics,[6,7] and supramolecular
mechanisms that allow translating mechanical forces into optical signals.[8−10] The use of mechanophores that undergo mechanically induced chemical
reactions is a newer approach toward polymers with mechanoresponsive
behavior.[11−15] Chromogenic responses can be achieved by coupling such mechanophores
with chromophores, affording motifs in which force-induced bond cleavage
leads to optical changes.[1,16−24] Spiropyrans, which undergo electrocyclic ring-opening reactions
to colored merocyanines,[1,4,20,21] are the most-widely investigated
mechanochromophores; other examples include diarylbibenzofuranones[22,23] and hexaarylbiimidazoles,[18] which dissociate
into colored radicals. Though the integration of such motifs into
macromolecules has afforded many mechanochromic materials, the fundamental
operating principle has two important limitations. As the mechanism
involves the force-induced reduction of chemical activation barriers,
the same reactions generally occur upon exposure to heat or light,
which renders the responses unspecific. For example, spiropyrans are
well-known photochromic dyes[25] and not
only change color upon application of force but also exposure to light.[1] Another problem is that most mechanically triggered
chain scission reactions are irreversible, either because the reactants
are spatially separated, the products react further, or the activation
barrier for the reverse reaction is too high.[13,17,19,20] Thus, even
if the reverse reaction is a priori feasible, it is not necessarily
enabled upon releasing the stress, but may require additional activation.
Herein we show that these problems can be overcome by a new mechanophore
type that exploits the molecular-shuttle function of rotaxanes[26−28] and allows mechanically accessing and trapping an entropically unfavorable
state, which cannot be stabilized by other stimuli. As a first example,
we studied a mechanoluminophore formed by the interlocked assembly
of a fluorophore-carrying cycle and a dumbbell-shaped molecule containing
a matching quencher (Figure ). The activation barrier for shuttling the cycle along the
shaft is of the order of 10–20 kcal·mol–1 and can readily be overcome by thermal activation.[26] Due to charge-transfer interactions, the cycle is preferably
located around the quencher and the fluorophore is switched off. The
force-induced displacement of the two elements, concomitant with a
fluorescence turn-on, is possible by incorporating the rotaxane into
a polymer, and macroscopic deformation of this material.
Figure 1
Operating principle
of rotaxane-based mechanoluminophores. The
interlocked assembly contains a fluorophore-carrying cycle (gray/green),
a dumbbell-shaped molecule with a matching quencher (brown), and two
stoppers (blue). Reactive groups (red) permit the integration into
a polymer chain (gray). In the idle state, the cycle is preferably
located around the quencher (top) and the fluorescence is quenched.
The force-induced displacement of the two elements (bottom) causes
a fluorescence turn-on.
Operating principle
of rotaxane-based mechanoluminophores. The
interlocked assembly contains a fluorophore-carrying cycle (gray/green),
a dumbbell-shaped molecule with a matching quencher (brown), and two
stoppers (blue). Reactive groups (red) permit the integration into
a polymer chain (gray). In the idle state, the cycle is preferably
located around the quencher (top) and the fluorescence is quenched.
The force-induced displacement of the two elements (bottom) causes
a fluorescence turn-on.The rotaxane-based mechanoluminophore 1 (Figure a) was inspired by
works of Sanders and Stoddart, who pioneered neutral donor–acceptor
catenanes[29] and rotaxanes[30] based on electron-rich 1,5-disubstituted naphthalene crown
ethers and the electron-poor 1,4,5,8-naphthalenetetracarboxylic diimide
(NpI) motif. The latter was expected to quench the emission of 4,7-bis(phenylethynyl)-2,1,3-benzothiadiazole
(BTH). Thus, the widely used 1,5-dinaphtho[38]crown-10 cycle was modified
to include this emitter. The [2]rotaxane 1 was prepared
in a template-directed synthesis that relied on the self-assembly
of this motif and NpI[30] via a modified
Huisgen 1,3-dipolar cycloaddition between an azide and an alkyne.[31] Tris(p-tert-butylphenyl)phenylmethane stoppers[32] were
used to lock-in the structure and two hydroxyl groups were strategically
placed on the cycle and the dumbbell to enable the covalent integration
into polymer chains. The nonreactive [2]rotaxane 2, the
free luminophore 3, and the free dumbbell 4 were prepared as reference materials (Figure a).
Figure 2
(a) Molecular structures of rotaxanes 1 and 2, luminophore 3, and dumbbell 4. (b) Synthesis and molecular structure of the mechanophore-containing
polyurethane 1-PU. This polymer was prepared from poly(tetrahydrofuran)
(PTHF), 4,4′-methylenebis(phenylisocyanate) (MDI), rotaxane 1, and 1,4-butanediol (BDO) in a polymerization catalyzed
by dibutyltin dilaurate (DBTDL). The x:y and x:z ratios were determined
by 1H NMR spectroscopy and the fraction of 1 in the monomer feed.
(a) Molecular structures of rotaxanes 1 and 2, luminophore 3, and dumbbell 4. (b) Synthesis and molecular structure of the mechanophore-containing
polyurethane 1-PU. This polymer was prepared from poly(tetrahydrofuran)
(PTHF), 4,4′-methylenebis(phenylisocyanate) (MDI), rotaxane 1, and 1,4-butanediol (BDO) in a polymerization catalyzed
by dibutyltin dilaurate (DBTDL). The x:y and x:z ratios were determined
by 1H NMR spectroscopy and the fraction of 1 in the monomer feed.Upon excitation at 365 nm, solutions (c =
1 ×
10–5 M) of the free luminophore 3 fluorescence
brightly. The luminophore exhibits solvatochromic behavior (Figures S1 and S2) with emission maxima between
515 (hexane) and 601 nm (acetonitrile). Adding the free dumbbell 4 does not impact the fluorescence of a solution of 3 (Figure S3). Solutions of rotaxanes 1 and 2 show no emission, and the fluorophore’s
absorption is slightly red-shifted (absorption maximum = 441 nm) relative
to that of 3 (438 nm, Figures S4
and S5). Along with the characteristic shifts[30] of the 1H NMR signals of the aromatic protons
caused by the interaction of the NpI and the cyclic part in 1 and 2 (Figure S6), these data indicate the formation of [2]rotaxanes and demonstrate
that in their idle state the BTH fluorescence is indeed quenched.Adapting an established strategy,[33−35] we incorporated rotaxane 1 in a statistical manner into a linear, segmented polyurethane
elastomer through a low-temperature polyaddition reaction (1-PU, Figure b). To avoid
interactions among multiple mechanophores, the rotaxane content was
limited to ca. 0.4 wt % via the content of 1 in the monomer
feed. The incorporation of 1 in the polymer was confirmed
by UV–vis spectroscopy (Figure S7), but not detectable by NMR spectroscopy (Figure
S8). We also synthesized a control polyurethane without mechanophore
(PU) and formulated physical blends of this polymer with
rotaxane 2 (2inPU) or a mixture of luminophore 3 and dumbbell 4 (3,4inPU). All
compositions were processed into 60–130 μm thin films
by solvent casting. Differential scanning calorimetry, thermogravimetric,
and dynamic mechanical analyses show that the thermomechanical properties
of 1-PU and PU are identical and match those
of similar polyurethanes[33,34] (Figures S9–S11). Uniaxial strain–stress measurements
of 1-PU, PU, 2inPU, and 3,4inPU films reveal the expected elastic deformation behavior
and show that the mechanical characteristics of the different compositions
are within experimental error identical (Figure
S12 and Table S1).As intended, as-prepared 1-PU films do not exhibit
any appreciable emission upon UV excitation, but upon deformation,
the characteristic fluorescence of the BTH motif instantly appears
(Figure a, Supporting Movie S1, Figure
S13a). This behavior is reversible; as soon as the stress is
released, the sample relaxes and ceases to emit light. By contrast,
neither of the reference materials shows any mechanically induced
fluorescence changes (Figure S13b,c, Supporting Movies S2 and S3). Films of 2inPU do not fluoresce in either unstretched or stretched
state (Figure S14). On the other hand, 3,4inPU films fluoresce brightly, regardless of the strain
applied (Figure S15). Thus, these results
show that the mechanoluminophore 1 functions as intended
and depicted in Figure ; macroscopic forces are transferred to the rotaxanes and cause a
displacement of the emitter-carrying cycle to the periphery of the
dumbbell, which persists until the stress is released. The reference
experiments confirm that the force transduction from the macroscopic
to the molecular level requires a covalent integration of the mechanophore
in the polymer.
Figure 3
(a) Images of a 1-PU film show that the fluorescence
is turned on upon deformation and switched off when the stress is
removed. Top and bottom images were taken under irradiation with 365
nm light and ambient illumination, respectively. (b) Emission spectra
of a pristine 1-PU film in the initial state and upon
uniaxial deformation to the indicated strains. (c) Emission spectra
of the sample used in (b) recorded upon gradually releasing the stress
and thereby reducing the strain; the stress-free state had a residual
elongation of 190%, due to hysteresis. All spectra were recorded with
λex = 385 nm.
(a) Images of a 1-PU film show that the fluorescence
is turned on upon deformation and switched off when the stress is
removed. Top and bottom images were taken under irradiation with 365
nm light and ambient illumination, respectively. (b) Emission spectra
of a pristine 1-PU film in the initial state and upon
uniaxial deformation to the indicated strains. (c) Emission spectra
of the sample used in (b) recorded upon gradually releasing the stress
and thereby reducing the strain; the stress-free state had a residual
elongation of 190%, due to hysteresis. All spectra were recorded with
λex = 385 nm.Several experiments were conducted to investigate the mechanochromic
transduction in 1-PU in greater detail and relate optical
response to macroscopic deformation. Figure b shows that the emission intensity increases
with the applied strain λ = (L – L0)/L0 and the behavior
is reversible (Figure c), even though 1-PU displays the typical hysteresis
and cyclic softening (Figure a) associated with the evolution of the domain structure of
segmented polyurethanes upon deformation.[36] The stress and the emission intensity are clearly correlated (Figures b,c and Figure S16), as apparent from the overlay of
the stress–strain curves and the emission intensities measured
in a separate experiment (Figure a). A comparison of the emission intensities of fully
stretched films (λ = 600%) of 1-PU and 3,4inPU suggests that deformation activates at least 29% of the mechanophores
in 1-PU (Figure S17). The
fluorescence intensities recorded over 20 loading–unloading
cycles with strains of 190 and 600% (Figure b) reveal reversible and rapid switching
between constant on/off states. A gradual decrease of the emission
intensity is observed for a 1-PU film that was kept in
the stretched state (Figure c, and Figure S18), mirroring the
dynamic stress relaxation function.[37] This
suggests that the molecular relaxation processes of the polyurethane
are intimately related to the stresses exerted on the covalently incorporated
rotaxane. No optical changes were observed in a similar stress relaxation
experiment with a 3,4inPU reference film (Figure S18).
Figure 4
(a) Overlay of strain (lines) and emission
intensity (symbols)
measured separately upon uniaxially deforming pristine 1-PU films (black solid line and squares), releasing the stress (black
dotted line and triangles), deforming the sample again (red solid
line and squares), and releasing the stress again (red dotted line
and triangles). (b) Emission intensities recorded upon deforming a 1-PU film for 20 cycles (first and second cycles omitted).
The emission intensities were determined at strains of 190 (relaxed
state) and 600%. (c) Nominal stress (line) and emission intensity
(symbols) measured as a function of time after deforming a pristine 1-PU film to a strain of 600% and keeping the strain constant.
(d) Emission decay profiles of a pristine 1-PU film recorded
upon uniaxial deformation to nominal strains of 400 and 600% (green
and blue lines) and after releasing the stress and stretching the
sample again to 400 and 600% (orange and red lines). The emission
decay profile of a 3,4inPU reference film (black line)
is also shown. All emission intensities were measured at 530 nm, with
λex = 385 (a–c) or 405 nm (d). The intensities
in panels a–c were extracted from uncorrected emission spectra.
(a) Overlay of strain (lines) and emission
intensity (symbols)
measured separately upon uniaxially deforming pristine 1-PU films (black solid line and squares), releasing the stress (black
dotted line and triangles), deforming the sample again (red solid
line and squares), and releasing the stress again (red dotted line
and triangles). (b) Emission intensities recorded upon deforming a 1-PU film for 20 cycles (first and second cycles omitted).
The emission intensities were determined at strains of 190 (relaxed
state) and 600%. (c) Nominal stress (line) and emission intensity
(symbols) measured as a function of time after deforming a pristine 1-PU film to a strain of 600% and keeping the strain constant.
(d) Emission decay profiles of a pristine 1-PU film recorded
upon uniaxial deformation to nominal strains of 400 and 600% (green
and blue lines) and after releasing the stress and stretching the
sample again to 400 and 600% (orange and red lines). The emission
decay profile of a 3,4inPU reference film (black line)
is also shown. All emission intensities were measured at 530 nm, with
λex = 385 (a–c) or 405 nm (d). The intensities
in panels a–c were extracted from uncorrected emission spectra.Emission lifetime measurements
provide further support of the molecular
mechanism at play (Figure d, Figures S19 and S20). In a moderately
deformed 1-PU film (λ = 400%), the luminophores
are closer to the quencher groups and the emission lifetime is shorter
than in a maximally deformed film (λ = 600%), whose emission
lifetime approaches that of the 3,4inPU reference sample,
which does not change upon deformation (Figure
S20). To demonstrate that the mechanical switching in 1-PU is indeed a highly specific process, we monitored the
fluorescence of this material at elevated temperature. Gratifyingly,
no obvious changes of the emission color and intensity were observed
(Figure S21). Even at 100 °C, the
luminescence of an unstretched film remains switched off, whereas
stretched samples fluoresce brightly. However, annealing 1-PU at 150 °C caused an increase of the emission intensity at 530
nm over time, which suggests dethreading and/or decomposition of the
rotaxane at this temperature (Figure S22). Chain cleavage could also be achieved by ultrasonication of a
THF solution of 1-PU, which led to a gradual and significant
decrease of the molecular weight and an increase of the emission intensity
(Figure S23). The fact that no emission
turn-on but a similar molecular-weight decrease was observed when
a corresponding solution of PU and rotaxane 2 was similarly treated (Figure S24) suggests
that at least some of the mechanophores are irreversibly cleaved upon
sonication of 1-PU.In summary, a new mechanophore
type was accessed by incorporating
a weakly interacting fluorophore/quencher pair that displays conformation-dependent
optical characteristics in a mechanically interlocked rotaxane. The
deformation of a rubbery polyurethane containing the first mechanoluminophore
based on this design leads to rapid and reversible fluorescence switching,
and the optical response correlates with the macroscopic deformation.
The concept should be applicable to other interlocked molecules, including
catenanes and knots, various chromophore systems, notably charge-transfer
pairs that change their absorption color upon dissociation, and different
polymer types. In addition to the attractive features already mentioned,
it appears straightforward to tailor the optical properties via the
choice of the chromophores, and the forces required to dissociate
the fluorophores and quenchers are very small. Thus, such supramolecular
mechanophores may also be useful to visualize ultrasmall mechanical
forces, such as those generated by living cells.
Authors: Zhixing Chen; Jaron A M Mercer; Xiaolei Zhu; Joseph A H Romaniuk; Raphael Pfattner; Lynette Cegelski; Todd J Martinez; Noah Z Burns; Yan Xia Journal: Science Date: 2017-08-04 Impact factor: 47.728
Authors: Yulan Chen; A J H Spiering; S Karthikeyan; Gerrit W M Peters; E W Meijer; Rint P Sijbesma Journal: Nat Chem Date: 2012-06-03 Impact factor: 24.427
Authors: Ashley L Black Ramirez; Zachary S Kean; Joshua A Orlicki; Mangesh Champhekar; Sarah M Elsakr; Wendy E Krause; Stephen L Craig Journal: Nat Chem Date: 2013-08-04 Impact factor: 24.427
Authors: Douglas A Davis; Andrew Hamilton; Jinglei Yang; Lee D Cremar; Dara Van Gough; Stephanie L Potisek; Mitchell T Ong; Paul V Braun; Todd J Martínez; Scott R White; Jeffrey S Moore; Nancy R Sottos Journal: Nature Date: 2009-05-07 Impact factor: 49.962
Authors: Diederik W R Balkenende; Souleymane Coulibaly; Sandor Balog; Yoan C Simon; Gina L Fiore; Christoph Weder Journal: J Am Chem Soc Date: 2014-07-10 Impact factor: 15.419
Authors: Shilin Yu; Arkady Kupryakov; James E M Lewis; Vicente Martí-Centelles; Stephen M Goldup; Jean-Luc Pozzo; Gediminas Jonusauskas; Nathan D McClenaghan Journal: Chem Sci Date: 2021-06-04 Impact factor: 9.825
Authors: Seth A Sharber; Kuo-Chih Shih; Arielle Mann; Fanny Frausto; Terry E Haas; Mu-Ping Nieh; Samuel W Thomas Journal: Chem Sci Date: 2018-05-24 Impact factor: 9.825