Introduction

On the basis of microscopic molecular recognition of a number of biological molecules, living organisms are ubiquitous macroscopic assemblies formed from supramolecular structures1,2,3. In various materials science fields, macroscopic assemblies in artificial systems based on microscopic molecular recognition are promising, for example, self-healing materials4,5,6. However, construction of artificial macroscopic assemblies based on molecular recognition remains a great challenge. Previously, we have reported macroscopic molecular recognition using a simple system of a polyacrylamide-based hydrogel possessing cyclodextrin (CD) and guest moieties7,8,9. According to their binding affinities, gels possessing αCD, βCD and γCD moieties recognize gels possessing their respective counterparts. Recently, we have focussed on macroscopic observations of molecular recognition using aromatic compounds because aromatic compounds absorb light and consequently, may transfer energy and electrons. We have reported that a gel possessing βCD moieties can discriminate the substitution position on a naphthyl group with the appropriate contents of βCD and naphthyl moieties, although these moieties have similar binding constants10. More recently, we have reported selective photoresponsive switching of macromolecular recognition using an azobenzene moiety11.

Pyrene is a popular fluorophore because it emits fluorescence with a relatively high quantum yield in both its monomer and dimer states12. On the other hand, pyrene is highly hydrophobic; its solubility in water is very low (<10−6 M) and easily forms dimers and aggregates in aqueous media13,14. In the present study, we modified a polyacrylamide gel with pyrenyl (Py) moiety (Py-gel) and investigated its interaction with gels possessing CD moieties (αCD-gel, βCD-gel and γCD-gel) (Fig. 1), using a mixed solvent of water and dimethyl sulfoxide (DMSO). The selectivity of Py-gel can be switched by varying the volume fraction of DMSO in the mixed solvent system (xDMSO).

Figure 1: Chemical structures of Py-gel and CD-gels.
figure 1

CD-gels were αCD-gel, βCD-gel and γCD-gel. Mol% are those in monomer feed, which are nearly identical to those in the obtained polymer gel.

Results

Preparation of Py-gel and CD-gels

The gels used in the present study were prepared by conventional radical terpolymerization of acrylamide, N,N′-methylenebis(acrylamide) and a monomer possessing a Py or CD moiety (Fig. 1). During gel preparation, water was used as the solvent for CD-gels, while DMSO was used as the solvent for the Py-gel owing to the solubility of the Py monomer in water. After gel preparation, DMSO was replaced with water for the Py-gel.

Interaction of Py-gel with CD-gels in the mixed solvent.

The interaction of Py-gel with CD-gels was investigated by shaking gel pieces in a mixed solvent of water and DMSO with various volume fractions of DMSO (xDMSO), as can be seen in Fig. 2. At xDMSO=0, Py-gel pieces interact only with γCD-gel pieces to form a gel assembly (Fig. 2a), and the resulted gel assembly is sufficiently stable to be picked up (Supplementary Fig. S1). At xDMSO=0.2, Py-gel pieces interact with both βCD-gel and γCD-gel pieces to form a larger assembly (Fig. 2b). As xDMSO increases up to 0.5, Py-gel pieces interact with βCD-gel pieces, but not the γCD-gel pieces (Fig. 2c). At xDMSO=1, Py-gel pieces and CD-gel pieces do not interact. Furthermore, Py-gel pieces do not interact with αCD-gel pieces regardless of xDMSO because αCD is too small to include the Py moiety. These observations indicate the selectivity of Py-gel strongly depends on the composition of the mixed solvent system. Py-gel prefers γCD-gel at xDMSO=0. Py-gel interacts with both βCD-gel and γCD-gel at xDMSO=0.2. However, Py-gel prefers βCD-gel at xDMSO=0.5. Thus, adjusting the mixed solvent composition can switch the selectivity of Py-gel, allowing macroscopic discrimination of αCD-, βCD- and γCD-gel.

Figure 2: Pictures of the interaction of Py-gel (yellow) with αCD-gel (blue), βCD-gel (red) and γCD-gel (green) in a glass petri dish.
figure 2

(a) Py-gel pieces interact with γCD-gel pieces at xDMSO=0. (b) Py-gel pieces assemble with both βCD-gel and γCD-gel pieces at xDMSO=0.2. (c) Py-gel pieces prefer βCD-gel pieces at xDMSO=0.5. The insets were taken under irradiation with 365 nm light. The interaction between Py-gel and αCD-gel was not observed. Scale bar, 1.0 cm.

The interaction strength of the gels was investigated by stress−strain measurements using a tensile meter. A gel assembly of Py-gel and CD-gel cubes was set on a tensile meter, and the stress was monitored as a strain was applied (Supplementary Fig. S2). Using the stress−strain data, Fig. 3 plots the stress at rupture as a function of xDMSO. For Py-gel/γCD-gel assemblies, the stress at rupture decreases as xDMSO increases with a shoulder at xDMSO~0.4. In contrast, the stress at rupture increases for Py-gel/βCD-gel assemblies as xDMSO increases, reaches a maximum at xDMSO~ 0.4, and then decreases. These observations confirm selectivity of Py-gel switches; Py-gel prefers γCD-gel at xDMSO <0.2. At xDMSO~0.2, Py-gel interacts with both βCD-gel and γCD-gel. At 0.2 <xDMSO≤0.6, Py-gel prefers βCD-gel. However, at xDMSO=1, Py-gel does not interact with CD-gels because inclusion complexes are not formed in DMSO.

Figure 3: Stress at rupture as a function of xDMSO for the Py-gel/βCD-gel (red) and Py-gel/γCD-gel assemblies (green).
figure 3

Py-gel prefers γCD-gel at xDMSO <0.2. At xDMSO~0.2, Py-gel interacts with both βCD-gel and γCD-gel. At 0.2 <xDMSO≤0.6, Py-gel prefers βCD-gel. At xDMSO=1, Py-gel does not interact with CD-gels. Stress-strain measurements were repeated at least three times, and the error bars denote the standard deviations.

Fluorescent study on the model systems

To understand the selectivity switching mechanism, fluorescence measurements were carried out using a model polymer, that is, polyacrylamide modified with a 1-mol% Py moiety (pAAm/Py, Supplementary Fig. S3). Supplementary Fig. S4a shows the steady-state fluorescence spectra of pAAm/Py at various xDMSO in the absence of CD. At xDMSO=0, the spectrum exhibits monomer fluorescence in the region of 360–380 nm and an excimer emission around 485 nm. The Py content of pAAm/Py is relatively low, but Py moieties easily form dimers within the same polymer chain owing to hydrophobicity. As xDMSO increases, the intensity of the excimer emission decreases until almost disappears at xDMSO >0.8. The ratio (IE/IM) of the fluorescence intensities around 376 and 485 nm, which are ascribed predominantly to monomer and excimer emissions, respectively, were calculated and plotted against xDMSO (Supplementary Fig. S4b). IE/IM decreases as xDMSO increases and reaches ca. 0.02 at xDMSO >0.8. The same xDMSO dependency of IE/IM occurs in the fluorescence spectra of Py-gel (Supplementary Fig. S5), indicating that the addition of DMSO makes the Py moiety more solvophilic and inhibits the formation of the excimer. In addition, insets in Fig. 2 visibly indicate the formation of the dimer in Py-gel at xDMSO=0 (bluish green light emission) and the monomer at xDMSO=0.5 (blue light emission) under photoirradiation.

Fluorescence spectra were also measured for pAAm/Py at xDMSO=0, 0.2, 0.5 and 1 in the presence of various concentrations of βCD or γCD to understand the effects of xDMSO on the interaction of the Py moiety with βCD or γCD (Supplementary Fig. S6). At xDMSO=0, the fluorescence spectra for pAAm/Py are nearly independent of the βCD concentration. On the other hand, as γCD concentration increases, the intensity of the monomer fluorescence decreases while the intensity of excimer emission increases, demonstrating that the Py moiety interacts with γCD in its dimer state but not βCD at xDMSO=0. At xDMSO=0.2, the intensity of the monomer fluorescence increases as βCD increases. On the other hand, the intensity of the excimer fluorescence slightly increases as γCD increases. These observations indicate that the Py moiety interacts with βCD in its monomer state, but with γCD in its dimer state at xDMSO=0.2. At xDMSO=0.5, the intensity of monomer fluorescence increases as βCD increases, but the fluorescence spectra are almost independent of the γCD concentration, indicating that the Py moiety interacts only with βCD. In addition, at xDMSO=1, that is, in DMSO, the spectra do not change on the further addition of βCD or γCD. Consistent with previous findings15,16, these observations lead to the conclusion that the Py moiety interacts with βCD in its monomer state to form a 1:1 complex, whereas the Py moiety interacts with γCD in its dimer state to form a 2:1 complex.

Discussion

On the basis of the characterization data, the selectivity switching mechanism for the Py-gel dependence on xDMSO is considered (Fig. 4). At xDMSO=0, a significant fraction of the Py moieties on the Py-gel surface form dimers and interact with γCD moieties on the γCD-gel. In contrast, at xDMSO=0, the fraction of the monomer Py moiety on the Py-gel surface is too low to interact with the βCD moieties on the βCD-gel surface. At xDMSO=0.2, the Py moiety is slightly solvophilic, and the fractions of monomer and dimer of the Py moieties on the gel surface are sufficient to interact with both βCD-gel and γCD-gel to form larger assemblies containing all three types of pieces. At xDMSO=0.5, the formation of dimer Py moieties on the Py-gel surface is insignificant, but the monomer Py moieties still interact with βCD moieties on the βCD-gel surface. Thus, Py-gel interacts only with the βCD-gel.

Figure 4: Conceptual illustrations for selectivity switching of Py-gel as a function of xDMSO.
figure 4

(a) At xDMSO=0, Py dimers are formed and included by γCD on the interface. (b) At xDMSO=0.2, Py monomers and dimers coexist in the gel that interact with both βCD and γCD. (c) At xDMSO=0.5, Py dimers are separated into monomers that included by βCD on the interface of gel assembly. Gels in yellow, red and green colour represent Py-gel, βCD-gel and γCD-gel, respectively.

In conclusion, the selectivity of Py-gel can be switched by changing the composition of the mixed solvent system composed of water and DMSO, because the monomer and dimer fractions of the Py moieties on the gel surface depend on xDMSO; the monomer and dimer of Py moieties prefer the βCD and γCD moieties, respectively. These observations allow the macroscopic discrimination of αCD-gel, βCD-gel and γCD-gel using Py-gel.

Methods

Measurements

1H NMR spectra were measured on a JEOL JNM-ECA500 spectrometer, and chemical shifts were referenced to the solvent values (2.49 and 4.79 p.p.m. for DMSO-d6 and D2O, respectively). Positive-ion matrix-assisted laser desorption/ionization time-of-flight mass spectrometry experiments were performed using a Shimadzu/KRATOS Axima CFR Ver.2.2.3 mass spectrometer calibrated by α-cyano-4-hydroxycinnamic acid and insulin. Fluorescence spectra were obtained on a HITACHI F-2500 spectrophotometer, using a 1-cm quartz cuvette. The slit widths for both excitation and emission sides were kept at 2.5 nm during measurement. All fluorescence spectra were reported as averages of three scans with an error less than 5%. The stress-strain curves for gel assemblies were recorded using a Yamaden RE-33005 Rheoner creep meter. Each sample (5×10 mm2 in size) was measured at a rate of 0.1 mm s−1 at room temperature.

Materials

1-Pyrenemethylamine hydrochloride was purchased from Sigma-Aldrich. Acryloyl chloride was obtained from Tokyo Chemical Industry. Triethylamine, acrylamide (AAm), ammonium peroxodisulfate (APS), N,N,N′,N′-tetramethylethylenediamine (TMEDA), N,N′-methylenebis(acrylamide) (MBA), acetone, methanol, DMSO (spectra grade), NaHCO3 and NaOH were purchased from Nacalai Tesque. N,N-Dimethylformamide and dichloromethane (DCM) were purified by utilizing a Glass Contour solvent dispensing system. Water was purified by a Millipore Milli-Q system. Cyclodextrins (αCD, βCD and γCD) were purchased from Junsei Chemical. and recrystallized twice from water before using. Mono(6-deoxyacrylamido)-α-cyclodextrin (A6αCD), mono(6-deoxyacrylamido)-β-cyclodextrin (A6βCD) and mono(6-deoxyacrylamido)-γ-cyclodextrin (A6γCD) were prepared according to the procedure reported previously10,17. Other reagents were reagent grade and used without further purification.

Preparation of N-(1-pyrenylmethyl)acrylamide (A1Py)

To a solution of 1-pyrenemethylamine hydrochloride (1.07 g, 4.00 mmol) and acryloyl chloride (0.95 ml, 12 mmol) in DCM (55 ml), a solution of triethylamine (1.62 ml, 16.0 mmol) in DCM (5.0 ml) solution was added dropwise at 0 °C. The reaction was allowed to run overnight at room temperature. After evaporation of the solvent, A1Py was purified by silica gel column chromatography using a mixed solvent of ethyl acetate and hexane (1/3, v/v) and obtained as a pale yellow powder: yield 390 mg, 34 %; mp 215–216 °C; 1H NMR (500 MHz, DMSO-d6, 30 °C) δ 8.759 (brs, 1H, NH), 8.382 (d, J=9.2 Hz, 1H, pyrene), 8.320–8.231 (m, 4H, pyrene), 8.156 (s, 2H, pyrene), 8.073 (t, J=7.6 Hz, 1H, pyrene), 8.038 (d, J=7.7 Hz, 1H, pyrene), 6.312 (dd, J=17.1 and 10.1 Hz, 1H, vinyl), 6.185 (dd, J=17.1 and 2.3 Hz, 1H, vinyl), 5.630 (dd, J=10.1 and 2.4 Hz, 1H, vinyl), 5.093 (d, J=5.8 Hz, 2H, CH2). 13C NMR (125 MHz, DMSO-d6, 30 °C) δ 164.44 (C=O), 132.58 (PyC1), 131.56 (PyC11), 130.74 (C=C), 130.26 (PyC14), 130.17 (C=C), 128.16 (PyC10), 127.59 (PyC13), 127.31 (PyC12), 127.04 (PyC6), 126.88 (PyC5), 126.21 (PyC9), 125.54 (PyC8), 125.23 (PyC3), 125.14 (PyC7), 124.69 (PyC2), 124.04 (PyC16), 123.88 (PyC15), 123.15 (PyC4), 40.46 (C-N). Anal. Calcd for C20H15NO(H2O)0.13: C, 83.50; H, 5.35; N, 4.87. Found: C, 83.50; H, 5.34; N 4.89. Positive ion matrix-assisted laser desorption/ionization time-of-flight mass spectrometry: m/z=285.1, [M+H+] 285.2.

Preparation of a model polymer (pAAm/Py)

A predetermined amount of AAm and A1Py were dissolved in dimethylformamide. After purging with dry argon for 30 min, APS (3 mg, 1×10−5 mol) was added to the monomer solution. The reaction mixture was placed into a cuvette equipped with a stirrer and sealed. The cuvette was warmed with an oil bath thermostated at 60 °C overnight. The reaction mixture was poured into excess of methanol to give a precipitate. The polymer obtained was recovered by filtration and dried under vacuum.

Preparation of Py-gel

A predetermined amount of AAm (405.2 mg, 5.70 mmol), MBA (27.8 mg, 0.18 mmol) and A1Py (34.2 mg, 0.12 mmol) were dissolved in DMSO (3.0 ml). After purging with dry argon for 30 min, APS (6 mg, 3×10−5 mol) was added to the monomer solution. The reaction mixture was placed into a 1-cm quartz cuvette and sealed. The cuvette was set into a 60-°C oven for 8 h. The gel was washed with DMSO and soaked in excess of the solvent used.

Preparation of CD-gels

A predetermined amount of AAm, MBA and a CD monomer (A6αCD, A6βCD or A6γCD) were dissolved in water (2.0 ml) in a 1-cm quartz cuvette. After addition of APS (3 mg, 1×10−5 mol) and TMEDA (2.5 μl, 2×10−5 mol), the reaction mixture immediately turned into a gel, which was washed with water to remove unreacted monomers and initiators. αCD-gel, βCD-gel and γCD-gel were dyed with blue (Coomassie brilliant blue R-250), red (food pigment red NO.102) and green (cyan NO.1 and yellow NO.4) dyes, respectively, by immersing into a solution of the corresponding dye. The CD-gels were then immersed in excess of the solvent used.

Additional information

How to cite this article: Zheng, Y. et al. Switching of macroscopic molecular recognition selectivity using a mixed solvent system. Nat. Commun. 3:831 doi: 10.1038/ncomms1841 (2012).