Electronic Energy Transfer Modulation in a Dynamic Foldaxane: Proof-of-Principle of a Lifetime-Based Conformation Probe

: Abiotic aromatic oligoamide foldamers are shown to self assemble in solution to form a double helix, which can accommodate a bichromophoric thread in its central void. While in solution reversible electronic energy transfer is instilled between chromophoric termini of the free, flexible thread as evidenced through delayed luminescence, upon rigidification of the rod the chromophores are mutually distanced and effectively decoupled. Consequently, the chro mophores display their individual photophysical character istics. The observed conformation dependent changes of dynamic luminescence properties, which are particularly sensitive to distance, offers a new strategy for lifetime based detection of geometry on the molecular scale as demonstrated through real time luminescence detection of molecular com plexation leading to foldaxane formation.

The study of dynamic supramolecular architectures,i nclud ing stimulus responsive systems,m imicking structural and mechanical properties of large biomolecules and assemblies is ag rowing research area.In this regard, ar ange of foldamer based architectures [1] have been reported, including helical containers, [2] proton [3] or light induced folding, [4] chiroptical switches, [5] and metal/ion responsive systems. [6]Most recently, interpenetration of as ynthetic aromatic oligoamide helix by al inear dumbbell thread has been demonstrated, giving rise to dynamic pseudorotaxane structures termed foldaxanes. [7]erein we report photoactive foldaxane prototypes and the application of reversible electronic energy transfer (REET), and resulting luminescence lifetime increase (see below), as anew method to probe conformational changes/distances,i n this case accompanying the unfolding refolding of ad ouble helical synthetic aromatic oligoamide [8] foldamer around ad umbbell shaped bichromophoric rod (Figure 1a).7b] In the final complex, the distance between terminal chromophoric units is fixed by the long rigid helix, which forms as heath around the alkyl chain, thus altering REET between them.REET gives rise to excited state lifetime elongation resulting from short range bidirectional triplet triplet energy transfer/ shuttling,and as such is fundamentally different from popular Fçrster resonance energy transfer (FRET), which is unidirec tional.Indeed the current approach is much more sensitive to distance r at the nanometer or subnanometer scale,asitrelies on aDexter rather than aFçrster energy transfer mechanism (Figure 1b,eq.1, where J is the integral overlap between the normalized donor emission and acceptor absorption spectra, L is the effective average Bohr radius of the excited and unexcited states of the donor and acceptor (0.7 6 , with an average value of 1.5 [9] ), and K is the experimental factor).Additionally,r esulting long luminescence lifetimes equally offer the possibility of time gated detection, which is not hampered by background fluorescence and higher photon densities can also be obtained with respect to lanthanide complexes. [10]EET can be instilled in bichromophoric molecular systems when specific energetic and kinetic conditions are respected, but examples are relatively few. [11 13]Photoactive thread 1 (Figure 1c)c omprises an inorganic Ru(bpy) 3 2+ like chromophore (bpy = 2,2' bipyridine), whose emissive 3 MLCT state is quasi isoenergetic [12] with an organic triplet on ap yrene ( 3 Pyr) unit that is located at the opposite terminus of along and flexible saturated molecular thread.This energy similarity allows energy to shuttle back and forth between the two chromophores (Supporting Information, Figure S13) when they can approach one another.Considering that deexcitation of an organic triplet is slower than that of a 3 MLCT state,and at short distances intercomponent energy transfer is more rapid, the system can reach an excited state equilibrium and emission would emanate predominantly from the inorganic center, with pyrene acting as an energy storage reservoir. [12]Thus it would be anticipated that alonger luminescence lifetime would be obtained in the free thread (according to eq. 2), while conformation juxtaposition of the helical host on the flexible thread would decouple the chromophores and emission lifetime would decrease.I nt he absence of additional quenching mechanisms,this decrease of emission lifetime should not be accompanied by ac hange in quantum yield.It is noteworthy that the approach to signal interchromophore distance,i ntroduced herein, may be applied to any matched chromophore pair, allowing variation of timescales and colors,a dding to the versatility of the approach. [13]onsidering the free bichromophoric rod 1,w hose syn thesis is detailed in the Supporting Information, the electronic absorption spectrum (Figure S8) shows the distinct additive contributions of the pyrene and ruthenium based chromo phores,showing that the long insulating bridge electronically isolates the chromophores.Characteristic red emission (l max = 620 nm) is observed irrespective of excitation wavelength as excitation energy is ultimately efficiently funneled to the emissive ruthenium moiety.H owever,e nhanced oxygen sensitivity is observed (F deoxy /F oxy = 9.6) with respect to the parent chromophore,w hich is diagnostic of an enhanced triplet lifetime that was confirmed through analysis of time resolved luminescence decays.I ndeed, while parent Ru (bpy) 3 2+ is characterized by as ub microsecond monoexpo nential decay in degassed CHCl 3 (ca.700 ns), the situation is different for bichromophore 1,where abiexponential decay is observed (t = 28 ns,2 0% and 2.5 ms, 80 %: Figure 2b).This behavior can be rationalized by considering ad ynamic excited state equilibrium involving both chromophores, where the short component corresponds to the time required for energy to be redistributed between the chromophores to attain an excited state equilibrium (at ar ate k eq = 1/t = 3.5 10 7 s À1 )a nd the microsecond decay corresponds to the deexcitation of the equilibrated excited states.T he relatively slow approach to equilibrium for 1,c ompared to sub ns processes in bichromophoric species with shorter linkers, [12,13] may be explained by the fact that the Ru II center of the thread is attached to apyrene energy reservoir through along chain, requiring reorganization to allow the transfer to occur, consequently slowing it down.Owing to Brownian motion and bond rotation, the distance between the inorganic (Ru 2+ ) core and pyrene is constantly changing.Thed istribution of distances between the Ru core and pyrene is described by aG aussian function centered at 1.0 nm with FWHM of 0.5 nm, using the method described by Lakowicz (see the Supporting Information, Figure S14). [14]n the dynamic equilibrated excited state,energy shuttles forwards and backwards between the 3 MLCT state to the 3 Pyr,and the observed long luminescence lifetime component is aw eighted average of kinetic deexcitation parameters according to the distribution of excitation energy between the two chromophores (Figure 1b,eq.2, where a corresponds to the fraction of populated ruthenium like triplets and 1 a corresponds to the fraction of pyrene like triplets).Time resolved transient absorption spectroscopy was conducted to evidence the REET process and evaluate energy distribution (see the Supporting Information).Synchronous changes in absorption of at ransient signal in the visible spectral region (namely ground state bleaching,G SB,r ecovery) and time resolved emission decay are shown in Figure 2. From analysis of the initial recovery of GSB to am etastable state,a n equilibrium constant of K eq = 5 AE 1could be obtained (see the Supporting Information).From this K eq value the proportion of energy stored on 3 MLCT and 3 Pyr states at agiven moment, was estimated at 16 %(a) and 84 %(1 a), respectively.T he rate of backwards ( 3 Pyr to 3 MLCT) energy transfer k b = k eq /(K eq + 1) = 6.0 10 6 s À1 ,w hile forward ( 3 MLCT to 3 Pyr) energy transfer rate k f = k eq k b = 3.6 10 7 s À1 .
To assess the winding of af oldamer around the bichro mophoric dumbbell guest, the new aromatic oligoamide sequence 4 was prepared.7c] Oligomer 4 was obtained by elongation of the Q F (7 amino 8 fluoro 2 quino linecarboxylic acid) segment of 3 (Figure 1c), which after Boc deprotection was coupled with the acid chloride of a Q F 2 dimer (Supporting Information, Scheme S2).Elongated strand 4 was likewise shown to form parallel and antiparallel (4) 2 duplexes but with enhanced stabilities compared to (3) 2 (Figure 3a).Decreasing the concentration of 4 in CDCl 3 from 0.5 mm to 0.125 mm caused some duplex dissociation (Sup porting Information, Figure S3).Integration of the 1 HNMR signals of the single versus the double helices allowed calculation of ad imerization constant (K dim )o f5 .9 10 4 Lmol À1 at 298 K( Supporting Information, Figure S3), avalue permitting titrations at 4mm with aproportion of (4) 2 higher than 95 %.Thed ouble helical nature of (4) 2 was also evidenced in the solid state by X ray crystallography (Fig ure 3d).
Foldaxane formation was first assessed using 1 HNMR titrations.Upon adding rod 1 to asolution of (4) 2 ,the signals corresponding to the parallel and antiparallel dimers were progressively replaced by an ew single set of peaks corre sponding to 1&(4) 2 (Figure 3a c).After 24 hours,t he reso nances in the 1 HNMR spectra were unchanging and specific shifts in NH amide resonances could be assigned to hydrogen bond interactions between the two carbamate groups of the rod and the double helix.At thermodynamic equilibrium the association constant (K ass )w as found to be 1700 Lmol À1 .Additionally,c rystallographic evidence was sought to unam biguously show the interpenetrated nature of 1&(4) 2 .W hile single crystals of suitable quality of the 1:1host guest complex could not be grown, undoubtedly hampered by the presence of L and D stereoisomers of the ruthenium moiety as well as the P and M isomers of the helix, single crystals of the analogous complex 2&(4) 2 were successfully obtained. [15,16]As shown in Figure 3e and 3f,t he antiparallel double helix envelops the alkyl segment of the rod, thereby distancing the opposite terminus of sequestered 1 with the presence of bifurcated hydrogen bonds between the NH of the pyridine trimer (P 3 )ateach extremity of the double helix and the two carbonyl oxygen atoms of the rod (d NH•••OC % 2.95 ).Impor tantly,the X ray structure allows interchromophore distances to be estimated at ! 1.8 nm.This is noteworthy as it is generally considered that triplet triplet energy transfer, which depends on distance according to eq. 1i si nefficient beyond 1nm, which would mean the REET process would no longer be operational in the host guest complex.Conse quently,ashortening of the luminescence lifetime with respect to the free thread would be anticipated.Figure 4 shows the luminescence decays of both the dilute free thread 1 and amixture of 1 and (4) 2 (1:32, respectively;2mm of (4) 2 ), slowly leading to 77.5 %f ormation of complex 1&(4) 2 in solution as demonstrated by 1 HNMR (see above).A luminescence lifetime of 1.2 msi st hus obtained, half that of the free thread allowing discrimination of free and bound species.
Thed ata analysis merits consideration.During the com plexation experiments there are different emissive species in solution, principally free 1 and the 1&(4) 2 complex, which could be at different stages of binding between 1 and (4) 2 .
Concerning free 1,i ts emission lifetime is fixed and is about 2.5 msindegassed CDCl 3 ,depending on concentration, while the lifetime of bound 1 is anticipated to have an unknown distribution which can complicate data treatment of such systems.A st he primary goal in these experiments was to directly observe the effect of complexation on the rod elongation process and determine the rate of formation of 1&(4) 2 ,this was fulfilled by establishing the fraction of bound 1,t racked using emission lifetime dependence versus post mixing time in the complexation experiment.Rather than using distributions for emission lifetime of bound 1,emission decays were treated using as um of two exponents A exp( t/ t short ) + B exp( t/t long ), where t long stands for lifetime of unbound 1 and t short for the 1&(4) 2 complex.This allows direct observation of the growth of the fraction of 1&(4) 2 during the experiment.As imilar trend was obtained if the emission decay is fitted using one exponent (Supporting Information, Figure S11).Indeed, tail fitting of luminescence decays is often employed for facility of imaging longer components in complex mixtures.D uring the complexation process,a st he fraction of 1&(4) 2 grows,t he rate of emission increases,s ob yt reating emission decays with one exponent, ageneral trend could be observed in the obtained dependency t = f(t), allowing to ascertain whether the complexation process has several characteristic time constants or only one and determine values for them.
Figure 5i llustrates luminescence lifetime dependence as afunction of time during the complexation process leading to 1&(4) 2 as an averaged value t = (A•t short + B•t long )/(A + B), where t long was fixed during the fitting procedure.The changes of fraction of t long during the experiment represent the decrease of the fraction of unbound 1.F igure 5s hows that the higher the relative concentration of 1,t he longer the emission lifetime after the complexation process is com pleted.Forthe systems with thread:host ratios 1:32, 1:12, 1:6, the corresponding lifetime values of 1.1 ms, 1.3 ms, and 1.5 ms were obtained.As the binding constant for 1&(4) 2 equals 1700 m À1 ,7 7.5 %, 76 %, and 75 %o f1 is bound in the 1&(4) 2 complex at stoichiometries 1:32, 1:16, and 1:6, respectively.Differences in lifetimes could be understood if the intermo lecular interactions between pyrene and Ru II center of differ ent 1 molecules were taken into account.At higher concen tration of 1 there are more possibilities for REET to occur between two different 1 molecules.
This interpretation could also explain the observation of such dependencies in fraction change of t long (Figure 5b).The time constant for foldaxane 1&(4) 2 formation (t)atdifferent ratios between 1 and (4) 2 were determined by fitting the experimental curves (Figure 5) with am ono exponential function.Ther esulting values determined from changes in averaged luminescence lifetime t and from fraction of the long lifetime of 1&(4) 2 are very similar and equal (t 1:32 = 195 AE 25 min, t 1:12 = 250 AE 25 min, t 1:6 = 290 AE 50 min). 1HNMR folding kinetic experiments were also carried out on solutions of 1&(4) 2 (Supporting Information, Figures S4 and S5) and proved to be in good agreement with luminescence lifetime measurements.I ndeed, the folding of (4) 2 around 1 could be calculated to take place at ar ate of k = 1.8 min À1 Lmol À1 .From this value alifetime of 406 min at 2mm ([1] Concerning mechanistic aspects of foldaxane winding, slow kinetics and solid state data (crystallographic structures of (4) 2 and 2&(4) 2 show the cavity aperture to be smaller than the terminal chromophores of the thread) are consistent with an unfolding refolding mechanism of the helix around the  dumbbell shaped rod.This would suggest partial (and stepwise) or alternatively complete double helix unzipping to accommodate the guest.As the luminescence lifetime is gradually shortening over the timeframe of the experiment to reach af ixed value,t his is consistent with the progressive extension of the rod induced by the winding of the helix starting from one end, slowly advancing and ultimately expulsion of the lead chromophore from the opposite end of the host.
In conclusion, we demonstrated herein that as ensitive luminescence lifetime probe for distance or conformational change can be developed by exploiting the modulation of reversible electronic energy transfer processes.T his was illustrated using afoldaxane architecture,the slow formation of which could be followed in real time.I np rinciple the reported strategy should prove general for aw ealth of energetically and kinetically matched chromophore pairs across awide spectral and temporal window.Additionally,in the current case experiments were carried out in the submilli to micromolar concentration ranges,b ut more or less concentrated samples could be used depending on chromo phores used (inherent lifetimes,a bsorption, and emission quantum yield) and detection sensitivity,respectively.Indeed, qualitative and to ac ertain degree quantitative prediction of the photophysical properties may be possible using this strategy.T hese systems could, for example,p otentially find use in studying ar ange of biological systems where nano metric conformational changes and/or rigidification accom panying interaction is present.Studies along these lines are currently in progress.

Figure 1 .
Figure 1.a) Representation of the parallel antiparallel equilibriumof adouble helix and the assembly modulated photophysical properties.b) Equations describing Dexter type electronic energy transfer (eq. 1) and the observed luminescence lifetime when REET is operational (eq.2).c) Structural formula of bichromophoric rod 1,i ts precursor 2, and the aromatic oligoamidesequences 3 and 4.H exafluorophosphate counter anions are omitted in 1 for clarity.

Figure 2 .
Figure 2. a) Transient absorption spectra of 1 in CDCl 3 (l exc 440 nm) at varying time domains.The gray box corresponds to excitation diffusion.b) Time resolved emission decay of 1 in CDCl 3 (l ex 465 nm, l obs 620 nm) and ground state bleaching (GSB, l obs 450 460 nm) recovery of excited 1.

Figure 3 .
Figure 3. Partial 1 HNMR spectra (300 MHz, 298 K) showing the amide region of (4) 2 (2 mm)inC DCl 3 :a)inthe absence of guest, b), c) in the presence of 1 b) 0.5 equiv and c) 2equiv guest.The signals of the empty parallel and antiparallel double helix (4) 2 are marked with * and &,r espectively.The signals of 1&(4) 2 are denoted with *.d ) f)Solid state structures[16]  of d) the free antiparallel double helix (4) 2 ,e), f) the foldaxane 2&(4) 2 .The antiparallel strands of (4) 2 are shown in gray and white tube (d,e) and CPK (f)r epresentations.The rod like guest 2 is shown in CPK representation.The bipyridine and the pyrene stoppers are shown in green and blue, respectively.The carbamatef unctions and the linear alkyl chain of the guest are shown in gold and red, respectively.The gray arrows illustrate the unscrewing motion of (4) 2 to accommodate the guest.[7c]

Figure 5 .
Figure 5. Real time complexation (time constant t)o f1 within (4) 2 obtained by fitting a) temporal evolution of the fraction of long luminescence lifetime component and b) changes in observed average luminescence lifetime consideringabiexponential decay.