Conformation Control of a Conjugated Polymer through Complexation

Jun 21, 2016 - Control of higher-order polymer structures attracts a great deal of interest for many researchers when they lead to the development of ...
0 downloads 0 Views 4MB Size
Download PDF
Article pubs.acs.org/Langmuir

Conformation Control of a Conjugated Polymer through Complexation with Bile Acids Generates Its Novel Spectral and Morphological Properties Youichi Tsuchiya,*,†,‡ Takao Noguchi,§ Daisuke Yoshihara,†,‡ Bappaditya Roy,§ Tatsuhiro Yamamoto,†,‡ and Seiji Shinkai*,†,§,∥ †

Nanotechnology Laboratory, Institute of Systems, Information Technologies and Nanotechnologies (ISIT), Fukuoka Industry-Academia Symphonicity (FiaS), Kyudaishinmachi 4-1, Nishi-ku, Fukuoka 819-0388, Japan ‡ Centre for Future Chemistry and §Institute for Advanced Study, Kyushu University, Moto-oka 744, Nishi-ku, Fukuoka 819-0395, Japan ∥ Department of Nanoscience, Faculty of Engineering, Sojo University, Ikeda 4-22-1, Nishi-ku, Kumamoto 860-0082, Japan S Supporting Information *

ABSTRACT: Control of higher-order polymer structures attracts a great deal of interest for many researchers when they lead to the development of materials having various advanced functions. Among them, conjugated polymers that are useful as starting materials in the design of molecular wires are particularly attractive. However, an equilibrium existing between isolated chains and bundled aggregates is inevitable and has made their physical properties very complicated. As an attempt to simplify this situation, we previously reported that a polymer chain of a water-soluble polythiophene could be isolated through complexation with a helix-forming polysaccharide. More recently, a covalently self-threading polythiophene was reported, the main chain of which was physically protected from self-folding and chain−chain πstacking. In this report, we wish to report a new strategy to isolate a water-soluble polythiophene and to control its higher-order structure by a supramolecular approach: that is, among a few bile acids, lithocholate can form stoichiometric complexes with cationic polythiophene to isolate the polymer chain, and the higher-order structure is changeable by the molar ratio. The optical and morphological studies have been thoroughly performed, and the resultant complex has been applied to the selective recognition of two AMP structural isomers.

■

nature of low-molecular-weight compounds.9−13 For example, when nucleosides were organized through the boronic acid− diol interaction on the boronic acid-appended poly(L-lysine) chain, the resultant nucleotide assembly could be arranged by a pH-induced conformational change in the poly(L-lysine) chain.11−13 In these reports, major interest has been focused on the morphology control of assemblies consisting of small molecules on polymeric templates. We considered that inversely, conformation control of polymers by molecular assemblies as templates would also be interesting for creating new polymeric structures and functions. For example, foldamers are artificial polymers but can form the highly

INTRODUCTION Control of arrangement, orientation, or conformation of molecular assemblies is one of the most effective and important methods of constructing novel functional materials.1 In nature, their inherent functions are often derived from specific higherorder structures. Proteins are typical examples, the functions of which are associated with their tertiary or quaternary structures. When polypeptides acquire such higher-order structures, molecular chaperons often mediate their folding processes through inter- and intramolecular interactions.2,3 In an artificial system, many researchers have been interested in the construction of such higher-order structures that can generate the expected functions. To design such functional materials, cofusion of the polymeric compound as a template axle with the molecular assembly as peripheral decoration compounds seems to be the most promising approach.4−8 In fact, we previously designed several polymer-based molecular organization systems in which conjugated polymers or carbon nanotubes are finely oriented with the aid of the self-assembling © 2016 American Chemical Society

Special Issue: Tribute to Toyoki Kunitake, Pioneer in Molecular Assembly Received: April 29, 2016 Revised: June 20, 2016 Published: June 21, 2016 12403

DOI: 10.1021/acs.langmuir.6b01639 Langmuir 2016, 32, 12403−12412

Article

Langmuir ordered structures by themselves:14−18 the specific folded structures are constructed owing to the inter- and intramolecular interactions among their main and/or side chains. Polythiophenes known as typical conjugated polymers also behave as foldamers: they consist of the thiophene units connected at the α-position and result in two possible conjugated conformations, the syn-planar form generating a helically folded conformation and the antiplanar form generating an extended zigzag conformation. Polythiophenes generally form a random-coiled structure in solution. When a bulky chiral group is introduced into the side chain, it tends to adopt the helical conformation. For example, Nilsson et al. found that chiral water-soluble polythiophenes form the helical folded structure and the helical motif is governed by the chirality of their side-chain group.19,20 We also reported that a water-soluble achiral polythiophene can be converted to a chiral folded conformation through complexation with a helixforming polysaccharide.21 This polysaccharide called schizophyllan has a triple-stranded right-handed helical structure in nature and forces polythiophenes to adopt the right-handed helical structure, regardless of the right- or left-handed chirality of the original polythiophene structure. The strong helixforming power stems from the coincidence of the helical pitch and the diameter between the polythiophene chain and the polysaccharide chain, which enables these two polymers to form a stable cohelical complex structure. As a result, the absorption maximum and the fluorescence maximum of the complexed polythiophene are red-shifted because of its extended conjugation length. Similar chiral cohelical complexes of polythiophene with DNA or proteins were reported by a few research groups.22,23 Meanwhile, the linearly extended structure of conjugated polymers is regarded as a molecular wire and has attracted the interest of many researchers. So far, the optical and electrochemical properties of linear polythiophenes have been discussed on the basis of their aggregated state or their theoretical calculations.24−32 The properties of the unimolecularly isolated polythiophene were discussed only on the interface or in the suspended state.33−38 The difficulty is related to a tendency inherent to polythiophenes that the more planar the π-system is, the more easily the aggregate is formed. Recently, Sugiyasu, Takeuchi, and others reported a unimolecularly extended polythiophene stably present in solution, the structure of which consists of the unique picket-fenced, selfthreading skeleton.39−42 This molecular design is based on the covalent bonds and the bulky substituents in order to suppress the polythiophene aggregation. A similar molecular design has been often employed to isolate the conjugated polymers.43−49 For example, Aida et al. reported an isolated nanowire in which the covalently bonded dendrons encapsulate the poly(phenyleneethynylene) backbone like an envelope. We considered, from a supramolecular viewpoint, that in order to enjoy the reversible interconversion between the random coil structure and the linearly extended structure, the molecular design based on the noncovalent-bond approach would be more interesting. To suppress the aggregation of conjugated polymers, surfactants often play an effective role. For example, the emission intensity of some polymers can be enhanced by its injection into the surfactant micelle above the critical micelle concentration (cmc) because of the isolation or pseudodilution of polymer chains.50−53 For polyelectrolytes, in general, the introduction of oppositely charged surfactants below the cmc rather induces the aggregation of fluorescent

polymers because of the charge neutralization, and their fluorescence is accordingly quenched.54 However, there are a few reports that the introduction of the surfactant below the cmc rather enhances the fluorescence intensity of polymers. For example, Chen et al. reported that the fluorescence of anionic poly(phenylenevinylene) is enhanced by the addition of the cationic surfactant below the cmc because of the suppression of polymer chain folding.55 Cagnoli et al. also reported that the fluorescence intensity of the cationic polythiophene aggregated in water is recovered by the addition of cholate or deoxycholate below the cmc.56 This fluorescence enhancement is ascribed to the hydrophobic interaction between the polythiophene main chain and added bile acids, making the cationic polythiophene more flexible and more isolated. In this article, we wish to evaluate the properties of cationic polythiophenes in the presence of the higher concentration of bile acids above the cmc. Therein, we unexpectedly found that the further addition of bile acids to the aqueous cationic polythiophene (PT-1, Scheme 1) solution induces a consecutive color change from Scheme 1. Schematic Structures of PT-1 and Bile Acids

yellow, via orange or pink, to red (Figure 1). The violet color originating from the aggregated polythiophene was not

Figure 1. Photographs of the PT-1 aqueous solution with various concentrations of bile acids: (a) cholate, (b) deoxycholate, and (c) lithocholate; top, under visible light; bottom, under UV365 irradiation; [PT-1] = 0.20 mM. 12404

DOI: 10.1021/acs.langmuir.6b01639 Langmuir 2016, 32, 12403−12412

Article

Langmuir observed at all.32 We wish to discuss the origin of this unique red polythiophene and the role of bile acids interacting with the cationic polythiophene through the electrostatic force. Furthermore, we have demonstrated a potential application of the fluorescence color change arising from the conformational change of this conjugated polymer to the recognition of the nucleotide structural isomers.

1. In Figure 2, the solutions were clear in the solid line region whereas they were somewhat cloudy in the dotted line region. It is seen from Figure 2 that the red-shift by added lithocholate was induced at less than 0.10 mM. The similar spectral change by added deoxycholate and cholate was induced at around 0.20 and 2.0 mM, respectively. In addition, the CD signals of PT-1 were observed in the red-shifted π−π* excitation region in the presence of deoxycholate (0.40 mM) and lithocholate (0.10 mM) (Figure S2). These results support the view that the PT-1 main-chain conformation is regularly maintained through complexation with these bile acids.57−61 It is possible to explain the complexation mode between PT1 and bile acids below the cmc if one can assume that charge neutralization and hydrophobic interaction are operating predominantly in aqueous solution.62−64 Bile acids have a hydrophilic α-face bearing several hydroxyl groups and a hydrophobic β-face bearing three methyl groups located on the steroid plane. In the first point, their facial amphiphilic nature should suppress the self-association of the PT-1 main chain because of the cooperative action of the electrostatic interaction and the hydrophobic interaction.56 In the second point, the redshifted peaks appear, from the lower concentration, in the order of lithocholate < deoxycholate < cholate. This order is consistent with the number of hydroxyl groups on the α-face: that is, lithocholate bearing only one hydroxyl group is most hydrophobic among the three bile acids and can interact with PT-1 most strongly. As these complexes were formed by charge neutralization, they resulted in the precipitates after 15−60 min at 20 °C. Therefore, the measurements were carried out before this precipitate formation. In the TEM and AFM images of PT-1/lithocholate (2:1 molar ratio) complex, the formation of the extended wire structure was confirmed (Figure S3). The AFM image shows that the diameter of the single wire can be estimated to be ca. 0.8 nm (Figure S3d). These results consistently support the view that the spectral changes in the absorption and fluorescence spectra are rationalized in terms of isolation and extension of the PT-1 main chain. Because of the solution turbidity, however, it is difficult to further discuss the interaction mode between PT-1 and lithocholate more quantitatively in aqueous solution. Complexation of Polythiophene with Lithocholate in DMSO/Water Mixed Solvent. Taking advantage of lithocholate that can interact with PT-1 more strongly than other bile acids, we tried to clarify the detailed interaction mechanisms in a homogeneous medium. When PT-1 and lithocholate were mixed in DMSO/water mixed solvents, no precipitation was observed in the relatively low water percentage region, and therein PT-1 showed spectral changes similar to those in water (Figure S4). We confirmed that the emission peak at around 526 nm ascribable to the random-coiled PT-1 disappeared in the Vw0.4 (DMSO/water = 6:4 v/v) solution (Figure S5), indicating that PT-1 is fully complexed with lithocholate. Hereafter, this Vw0.4 composition was used as a standard medium. Figure 3a shows the absorption spectral titration of PT-1 (0.20 mM) with lithocholate. With the increase in the lithocholate concentration, the absorption peak at 465 nm arising from the π−π* transition of PT-1 was shifted to 540 nm with a tight isosbestic point. Because PT-1 itself is optically inactive and adopts an achiral random-coiled conformation, no CD peak appeared in the π−π* transition region. When lithocholate was added, a few CD peaks appeared in the redshifted π−π* transition region (Figure 3b): the positive peak

■

RESULTS AND DISCUSSION Complexation of Polythiophene with Bile Acids in Aqueous Solution. As seen from the photographs in Figure 1, the aqueous solutions were clear enough and suitable for spectral measurements as long as the solution color was yellow. On the other hand, the aqueous red solutions that appeared at the high deoxycholate and lithocholate concentrations were somewhat cloudy, so the subsequent spectral data collected in this concentration region should be used only qualitatively. It is seen from Figure 1 that the yellow-to-red transition of the lithocholate complex occurred at the lowest concentration and that of the deoxycholate complex was next. The solution color of the cholate complex was yellow up to 1.0 mM, and the color transition was induced above 2.0 mM (Figure 2 and Figure S1).

Figure 2. (a) Absorbance ratio changes (590 nm vs 415 nm) of PT-1 as a function of bile acid concentrations: cholate (△), deoxycholate (○), and lithocholate ( × ). [PT-1] = 0.20 mM. (b) Enlarged view of the bile acid concentration range of (a). The broken line means that the solution is somewhat cloudy.

When cholate or deoxycholate, which has three or two hydroxyl groups, respectively (Scheme 1), was added to the aqueous PT-1 solution, the absorption spectra (Figure S1) changed in a two-step manner: that is, the first absorbance decrease at 415 nm followed by a wavelength shift to a few peaks at around 590 nm. The fluorescence spectral change was similar: that is, the first abrupt fluorescence increase in the emission peak (526 nm) followed by a shift in the emission maximum to the longer wavelength (around 605 nm, Figure S1). The first fluorescence increase is attributed to the inhibition of polymer-chain folding induced by the association with the bulky bile acids.55,56 In this concentration region, the circular dichroism (CD) signal from PT-1 was not observed. The results suggest that added cholate or deoxycholate mainly plays a role in inhibiting the polymer-chain folding and cannot yet enforce PT-1 to adopt some regular chiral conformation. In the second step, PT-1 showed the red-shifted absorption spectra by the further addition of bile acids, with the absorbance of the 590 nm peak being increased as accompanied by the decrease in the 415 nm peak. Figure 2 shows plots of the absorbance ratio (590 nm/415 nm) of PT-1 against the concentration of three different bile acids in water. These spectral changes were reflected by the color changes in Figure 12405

DOI: 10.1021/acs.langmuir.6b01639 Langmuir 2016, 32, 12403−12412

Article

Langmuir

solution was held at 80 °C and then cooled to 20 °C, the CD pattern was reinverted to the original CD sign that was observed in the titration at 0.20 mM. In their absorption spectra, on the other hand, there was no significant difference regardless of the preparation methods. Previously, we observed a similar CD inversion of PT-1.65 When ATP was successively added to the PT-1 solution, the same CD pattern was observed initially: however, above the charge-equivalent ATP concentration, it was totally inverted to the symmetrical CD pattern. The CD inversion implies that the chiral structure of supramolecular assemblies is dramatically changed. We found a few other examples of the polythiophene CD inversion. Bidan et al. reported that the CD band of polythiophene is inverted by the addition of poor solvent,66 indicating that the aggregation mode of polythiophene is subjected to the solvatochromic effect. Similarly, Bouman and Meijer reported the CD inversion of polythiophene in the film state:57 this polythiophene film shows a negative Cotton effect at 25 °C and no CD signal at 170 °C, and slow cooling regenerates the original CD signal whereas fast cooling gives the inverted CD signal. They concluded that the CD inversion is ascribable to the difference between the thermodynamic control and the kinetic control. In the present system, the CD sign of PT-1 is affected by the concentration and the heating temperature of the PT-1/lithocholate complex. What is the origin of this unique CD inversion phenomenon? The treatment processes suggest that the origin is related to the difference between the thermodynamic control and the kinetic control. One clue to this proposal is obtained from XRD measurements (Figure 5). The samples without the heat-and-

Figure 3. Spectral changes in the PT-1 DMSO solution (Vw0.4) by the addition of lithocholate: (a) absorption spectra and (b) CD spectra. [PT-1] = 0.20 mM. (c) Absorbance changes at 465 nm (broken line) and 540 nm (solid line). (d) CD intensity changes at 490 nm (broken line) and 590 nm (solid line).

and the negative peak were observed at 590 and 500 nm, respectively. The finding proves that the PT-1 main chain is chirally oriented through the complexation with lithocholate. Figure 3c,d shows that the plots are saturated at 0.20 mM, which is the concentration equivalent to the incubated PT-1 unit concentration. This supports the view that lithocholate interacts with one PT-1 unit in a 1:1 molar ratio. In addition, the sigmoidal curvatures suggest that the complexation proceeds according to the cooperative action of the electrostatic interaction and the van der Waals interaction among lithocholate molecules on the PT-1 backbone. Interestingly, we found that the CD spectra of the 1:1 PT-1/ lithocholate complex are strongly affected by the concentration and the heat-and-cool treatment. First, we noticed that when the complex concentration is enhanced up to 2.0 mM, the solution produces the inverted CD pattern. Even when this solution was diluted to 0.20 mM, this inverted CD pattern was maintained (Figure 4). When this diluted 0.20 mM solution was held at 60 °C and then cooled to 20 °C, the CD intensity was somewhat weakened. Surprisingly, when this 0.20 mM

Figure 5. XRD spectra of (a) the PT-1/lithocholate complex with heat-and-cool treatment, (b) the PT-1/lithocholate complex without heat-and-cool treatment, and (c) lithocholic acid.

cool treatment showed a featureless pattern indicative of the amorphous structure. In contrast, the samples with the heatand-cool treatment showed a crystal-like XRD pattern that is more or less similar to that of lithocholate. In differential scanning calorimetry (DSC), lithocholate itself gave a melting point peak at around 190 °C, whereas the PT-1/lithocholate complex did not give any peak regardless of the heat-and-cool treatment (Figure S6). These results consistently support the view that the CD sign is inverted by the lithocholate environment wrapping the PT-1 main-chain, i.e., depending upon whether it is amorphous or crystal-like. Thus, the original CD sign is due to the thermodynamic control facilitating the crystal growth whereas the inverted CD sign is due to the

Figure 4. CD spectra of PT-1/lithocholate complex in DMSO solution (Vw0.4) with and without heat-and-cool treatment; [PT-1] = [lithocholate] = 0.20 mM; heating, 60 or 80 °C for 30 min; cooling, 20 °C for 10 min. Samples were prepared as for the 2.0 mM solution at 20 °C except for the reference sample, which was prepared the same as the 0.20 mM solution at 20 °C without the heat-and-cool treatment. 12406

DOI: 10.1021/acs.langmuir.6b01639 Langmuir 2016, 32, 12403−12412

Article

Langmuir

is, the shorter the fluorescence lifetime becomes.68,69 In addition, the decrease in the fluorescence lifetime is induced by the formation of the aggregate. One can understand, therefore, that the conjugation of the PT-1/lithocholate complex is elongated by 3 days of incubation according to the thermodynamic control. The long-time-span spectral change in polythiophene was also reported by Danesh et al.70 When a cationic surfactant was mixed with an anionic polythiophene under high-concentration conditions, the complex formed the stable gels and the random structure of polythiophene remained. After the dilution of the complex solution, the polythiophene main chain was elongated with the red shift of the absorption and fluorescence maxima over several hours. This polythiophene/surfactant complex was finally precipitated. On the other hand, our PT-1/lithocholate complex seems more stable because it did not give any precipitate even for 6 months after the heat-and-cool treatment. Taking these lines of information into consideration, we carried out AFM measurements. Figure 7 shows the AFM

kinetic control to give the amorphous structure (Figure 5). These results clearly indicate that the complexation between PT-1 and lithocholate is formed under thermodynamic control at lower concentration (0.20 mM) but under kinetic control at higher concentration (2.0 mM). The kinetically metastable state can be changed to the thermodynamically stable state by heating. These findings imply that the present complex system is useful as a heat-triggered memory utilizing the CD signal. The temperature-induced change in the PT-1 main-chain conformation should be reflected by the fluorescence spectra. As shown in Figure S7a, the emission peak at 550 nm arising from the random-coil structure of PT-1 appeared at the higher temperature (80 °C), whereas the emission peak at 605 nm arising from the more extended structure appeared at the lower temperature (20 °C). These peaks could be reversibly regenerated several times (Figure S7b). It is seen from Figure S7b that the peak intensity at 550 nm gradually increases with the recycle number. This is probably a sort of aging effect in growing the fully random-coil conformation. This state transformation of the PT-1/lithocholate complex is probably related to the critical micellar temperature (65 °C) inherent to lithocholate:62,67 above 65 °C, the lithocholate assemblies disappear and their template effect is almost effaced. Structural Aspect of the Polythiophene−Lithocholate Complex. Here, we noticed that the PT-1/lithocholate complex with the heat-and-cool treatment shows a decrease in fluorescence at 605 nm and an increase in CD signals over several hours (Figure S8). After the heat-and-cool treatment, the fluorescence maximum of the PT-1/lithocholate complex was already shifted to longer wavelength (605 nm), which was often rationalized in terms of the polythiophene aggregation. However, the complex maintained a relatively strong fluorescence, with the intensity being about half of that for PT-1 itself (Figure 6). The 3 days of incubation at 20 °C

Figure 7. AFM images of the PT-1/lithocholate complex on HOPG, scale bars 1.0 μm: (a) 0.20 mM sample (Vw0.4, with heat-and-cool treatment); (b) 200-fold-diluted sample of the 0.20 mM sample (Vw0.4, with heat-and-cool treatment) with water, incubated at 20 °C for 10 min after dilution, and (c) 200-fold-diluted sample of the 0.20 mM sample (Vw0.4, with heat-and-cool treatment) with water, without incubation. (d) Height profiles of (b) and (c).

images of the PT-1/lithocholate complex on the HOPG substrate. The extended fibrous structures were observed for the sample prepared from the 200-fold-diluted solution (1.0 μM, 10 min of incubation after dilution with water). The height of the fibers was estimated to ca. 1.6 nm. This value is only a little larger than the lithocholate molecular size (1.5 nm),71,72 suggesting that the linear PT-1 main chain is interacting with lithocholate molecules. Here, we confirmed that the 200-folddiluted solution gives absorption and fluorescence spectra that are nearly the same as the original spectra (0.20 mM) (Figure S10). This result suggests that the red-shifted spectra obtained from the PT-1/lithocholate complex are not due to the stacked PT-1 main chain but are due to the extended one. As observed for the spectral changes, the TEM images also showed the timeand temperature-dependent morphological changes (Figure S11). The sample solutions of the PT-1/lithocholate complex (0.20 mM) after the heat treatment (95 °C for 30 min) were incubated for 10 min, 1 h, or 3 days at 20 °C and then subjected to TEM measurements. The sample obtained from 10 min of incubation gave the featureless random structure. In the sample incubated for 1 h, both the fiberlike structure and

Figure 6. Excitation and emission spectra of PT-1 (broken line), the PT-1/lithocholate complex after heating to 80 °C for 30 min and cooling to 20 °C for 10 min (solid line), and the PT-1/lithocholate complex after heating to 95 °C for 30 min and cooling to 20 °C for 3 days (dotted line). [PT-1] = [lithocholate] = 0.20 mM.

induced the decrease in the fluorescence intensity with a slight peak wavelength shift. We compared the fluorescence lifetime of the PT-1/lithocholate complex without incubation and after 3 days of incubation. Without incubation, the fluorescence lifetime of the PT-1/lithocholate complex was 0.6 ns, whereas the fluorescence lifetime became much shorter (