Optimal st-PMMA/C₆₀ helical inclusion complexes via tunable energy landscapes for the application of an Ag SERS-active substrate

This article is part of a collection of articles related to the 19th International Small-Angle Scattering Conference (SAS2024) in Taipei, Taiwan.

In bio-inspired systems, the hierarchical structures of biomolecules are mimicked to impart desired functions to self-assembled materials. However, these hierarchical architectures are based on multicomponent systems, which require not only a well defined primary structure of functional molecules but also the programming of self-assembly pathways. In this study, we investigate pathway complexity in the energy landscape of the syndiotactic poly(methyl methacrylate) (st-PMMA)/C₆₀/toluene complex system, where C₆₀ and toluene serve as guests in the st-PMMA helical host. Structural characterization revealed that st-PMMA preferentially wraps around C₆₀, forming a thermodynamically favorable helical inclusion complex (HIC). However, during the preparation of the st-PMMA/C₆₀ HIC, a lengthy guest-exchange pathway was discovered, where the st-PMMA/toluene HIC transformed into the st-PMMA/C₆₀ HIC. This pathway complexity may hinder the formation of the st-PMMA/C₆₀ HIC within a feasible timeframe. Given that the energy landscape can be modulated by temperature, the st-PMMA host can directly wrap around C₆₀ in higher temperature ranges, thereby bypassing the guest-exchange process and increasing the st-PMMA/C₆₀ HIC formation efficiency. Additionally, after self-assembly programming, the st-PMMA/C₆₀ HIC can serve as an excellent photochemical reduction site. The well dispersed nanodomains of the st-PMMA/C₆₀ HICs act as nanoparticle templates for surface-enhanced Raman scattering (SERS) hotspot fabrication. We successfully utilized these HIC templates to synthesize self-assembled SERS-active silver nanoparticle arrays, demonstrating their potential for use in chemical sensing applications. In summary, a clear energy landscape can guide supramolecular engineering to achieve the desired supramolecular architectures by selecting appropriate self-assembly pathways.

1. Introduction

In nature, hierarchical architectures such as ribosomes, DNA and lipid membranes are formed through multicomponent self-assembly processes (Harayama & Riezman, 2018). To ensure the formation of complex structures, it is necessary to maintain the fidelity of molecular architecture such as chain length and sequence in polypeptides and DNA strands (Zaher & Green, 2009). Nature cleverly uses enzymes to realize accurate sequences in polynucleotides or polypeptide chains (Novacek et al., 2024). However, providing the well defined molecular architecture of biomolecules alone does not guarantee the formation of these intricate complexes because self-assembly is pathway dependent (Knowles et al., 2014). The hierarchical self-assembly of biomolecules is encoded in the collective contribution of various non-covalent interactions between bio-entities (e.g. salt bridges, hydro­phobic interactions, hydrogen-bond interactions) (Szilágyi & Závodszky, 2000). Although the competition among these non-covalent interactions further brings out several self-assembled states with local or global energy minima, living systems can navigate pathways within the energy landscape to guide biomolecules toward the desired supramolecular architectures with specific physiological functions (Adamcik & Mezzenga, 2018; Ma et al., 2020; Nishimura & Akiyoshi, 2020).

Learning from nature, bio-inspired self-assembly usually attempts to mimic the definitive primary structure of bio­mol­ecules, for instance, chain length and sequence control. This approach opens the first step toward endowing synthetic molecules with structural complexity in supramolecular chemistry. For example, synthetic helical polymers are usually obtained by controlling stereoregularity, such as syndiotactic polystyrene and poly(acetyl­enes) (Yashima et al., 2016). Among the synthetic stereoregular polymers, syndiotactic poly(methyl methacrylate) (st-PMMA) is a representative example that exhibits helical wrapping behavior in two-/multicomponent systems, similar to natural helical polymers such as DNA strands and polysaccharides (Yashima et al., 2016; Zhang & Seelig, 2011; Fittolani et al., 2020). st-PMMAs can form helical inclusion complexes (HICs) with specific guest molecules through molecular recognition, for example, aromatic solvents, polycyclic aromatic hydro­carbons, fullerenes and isotactic PMMA (it-PMMA) (Yashima et al., 2016; Kawauchi et al., 2010; Kawauchi et al., 2011; Ren et al., 2018). st-PMMA subsequently undergoes a conformational change to the helical structure, providing a cavity ca 1 nm in size, which allows for the inclusion of these guest molecules through the induced-fit mechanism (Kawauchi et al., 2010; Kawauchi et al., 2011; Ren et al., 2018; Bosshard, 2001). Unlike DNA wrapping through complementary hydrogen-bond inter­actions, although st-PMMA complexation is driven by van der Waals forces, its binding selectivity to guest molecules can be still determined by various traits of its chemical structure such as size, chain length and stereoregularity (Kawauchi et al., 2010; Ren et al., 2018; Kajihara et al., 2020). Thus, the unique self-assembly behavior of the st-PMMA multicomponent system extends its applications to sensing, separation and catalysis, in addition to electronic and opto­electronic materials (Qi et al., 2013; Li et al., 2022; Chen et al., 2023).

In st-PMMA-based applications, the self-assembly process occurs in the multi-component system, where at least two types of guests can bind with the st-PMMA host (Yashima et al., 2016; Kawauchi et al., 2010; Kawauchi et al., 2011). Competition among the guest molecules for binding may lead to multiple self-assembled states in the energy landscape, hindering structural control over the st-PMMA multicomponent system (Kawauchi et al., 2011; Kawauchi et al., 2008). For instance, during the formation of the st-PMMA/C₆₀ HIC where toluene is used as the solvent, st-PMMA/toluene HICs are also formed in solution due to toluene acting as a guest of st-PMMA hosts. To control the complex architecture, conventional supramolecular strategies rely on precise synthesis to tailor the st-PMMA structure, for example, regulation of the rr content and molecular weight, which can improve the binding specificity of a specific guest (Ren et al., 2018). Kajihara et al. (2020) tried to perfect the stereoregularity of the st-PMMA chain, but the amount of C₆₀ encapsulated in the st-PMMA helix increased only marginally, and the encapsulation ratio was considerably lower than the ideal value of 28 wt%. By contrast, bio-systems allow for the formation of complex architectures by precisely controlling bio-entities, which entails guiding them to the correct pathway in the energy landscape (Knowles et al., 2014; Ma et al., 2020). This observation motivated us to explore the thermodynamic stability of each supramolecular species in the st-PMMA multicomponent system and explore the self-assembly pathways in its energy landscape.

Herein, we investigate the complexity of self-assembly pathways in the three-component st-PMMA/C₆₀/toluene system, where C₆₀ and toluene act as guests in the st-PMMA helical host, as illustrated in Scheme (a). The concentration- and temperature-dependent structural characterizations of st-PMMA-based HICs are firstly revealed through simultaneous small- and wide-angle X-ray scattering (SAXS and WAXS). Three self-assembled species are then identified in the st-PMMA/C₆₀/toluene system: helical-rich st-PMMA clusters, st-PMMA/toluene HICs and st-PMMA/C₆₀ HICs [Scheme (b)]. In terms of binding affinity, the tighter binding of st-PMMA and C₆₀ indicates that the st-PMMA/C₆₀ HIC is thermodynamically favorable. Next, we strategically apply temperature modulation to the energy landscape to accelerate the formation of st-PMMA/C₆₀ HIC structures and increase the C₆₀ encapsulation efficiency considerably. We find that programming the self-assembly pathway can lead to a high encapsulation ratio in st-PMMA/C₆₀ HICs without the need for the time-consuming guest-exchange pathway. Furthermore, the resulting st-PMMA/C₆₀ HICs exhibit well dispersed C₆₀ domains and act as effective reduction templates for Ag nanoparticle (Ag-NP) synthesis. These Ag-NPs intensify surface-enhanced Raman scattering (SERS) activity and outperform Ag-decorated C₆₀ crystals in rhodamine 6G detection. By programming self-assembly pathways, this approach optimizes C₆₀ encapsulation and enhances Ag-NP production, thereby demonstrating the potential for synthesizing targeted supramolecular architectures for functional applications.

2. Experimental

2.1. Materials and methods

C₆₀ (purity = 99.5%) and st-PMMA (number-average mol­ecular weight M_n = 31.2 kg mol⁻¹ and rr content = 86%) were purchased from Alfa Aesar and Polymer Source Inc., respectively. All other reagents and solvents were purchased from Sigma–Aldrich and were used without purification. The rr content of the st-PMMA was measured using ¹H nuclear magnetic resonance (NMR) spectrometry. NMR spectra were recorded using an Agilent Unity-400 NMR spectrometer, where CDCl₃ was employed as a deuterated solvent to identify the molecular structures at 25°C. M_n and the polydispersity index of st-PMMA were determined using a gel permeation chromatograph equipped with a JASCO liquid chromatograph, comprising a JASCO PU-4180 pump, JASCO RI-4030 detector and Stragel columns (HR1, HR2 and HR4). Tetra­hydro­furan (THF) was utilized as the eluent at a flow rate of 1.0 ml min⁻¹ and temperature of 30°C. The measurements were performed at 30°C.

2.2. Infrared spectroscopy

Attenuated total reflection (ATR)/Fourier transform-infrared (FT-IR) spectra of the st-PMMA/toluene and st-PMMA/C₆₀/toluene solution/gel were obtained using a PerkinElmer Spectrum 3 spectrometer equipped with a ZnSe crystal ATR attachment. The IR spectra were recorded over 16 scans in the wavenumber range 920–780 cm⁻¹. Synchrotron FT-IR spectroscopy measurements were performed at the endstation of the TLS14A beamline of the Taiwan Light Source at the National Synchrotron Radiation Research Center (NSRRC), Taiwan.

2.3. Simultaneous small-angle X-ray scattering/wide-angle X-ray scattering

Simultaneous SAXS/WAXS measurements were recorded using the TPS BL13A beamline at NSRRC. The scattering signals were collected using the Eiger X 9M and Eiger X 1M detectors. The wavelength of the X-rays was 0.827 Å. The scattering vector magnitude, q, related to the scattering angle (2θ) and photon wavelength (λ), was calculated using the equation q = 4π sin(θ)/λ. Samples were loaded into quartz capillary tubes and sealed by silicone resin. Finally, the tubes were mounted on the temperature-controlled stage to conduct the SAXS/WAXS measurements.

2.4. X-ray photoelectron spectroscopy

X-ray photoelectron spectroscopy (XPS) was performed at the endstation of the BL24A beamline at NSRRC. The XPS endstation is equipped with a load lock chamber for sample loading, an ultrahigh vacuum preparation chamber for sample preparation and a main analysis chamber for electron spectroscopy measurements, which are carried out using a SPECS NAP 150 electron energy analyzer.

2.5. Scanning electron microscopy

Scanning electron microscopy (SEM) analysis was performed using a JEOL JSM-7610F microscope operating at an accelerating voltage of 5 kV.

2.6. Raman spectroscopy

Raman spectroscopy was performed using a laboratory-built micro-Raman system. A Cobolt 532 samba laser was used to irradiate the samples and scattering signals were collected using a Southport Jademat NM system equipped with a Kymera 328i B1 Andor spectrometer. The laser exposure time was set to 3 s for each spectrum.

3. Result and discussion

3.1. Solvent effect on st-PMMA HICs

Molecular characterization results of the st-PMMA host with M_n = 31 kg mol⁻¹ and rr content = 86% are presented in Figs. S1 and S2 of the supporting information. To investigate the solvent effect on the st-PMMA self-assembly, THF and toluene were used as solvents to prepare the st-PMMA solutions at [st-PMMA] = 0.4 M (80 mg ml⁻¹). The st-PMMA exhibited better solubility in THF than in toluene (Minei et al., 2014). As illustrated in the inset of Fig. 1(a), the st-PMMA/THF system formed a solution, but the st-PMMA/toluene system was in a gel state. Fig. 1(a) presents the WAXS profiles of the st-PMMA/THF and st-PMMA/toluene systems to characterize their HIC structures. In the profile of the st-PMMA/toluene gel, the diffraction peaks at q_tol,helix = 0.38 Å⁻¹, q_helix,pitch = 0.78 Å⁻¹ and q_helix,intra = 0.94 Å⁻¹ result from the st-PMMA/toluene HICs, corresponding to the interhelical packing distance (d_tol,helix = 16.5 Å), helical pitch (d_helix,pitch = 8.0 Å) and intramolecular distance (d_helix,intra = 6.7 Å) along the helical axis, respectively (Kawauchi et al., 2011). By contrast, the WAXS profile of the st-PMMA/THF solution exhibits two broad amorphous halos at q_hc,inter = 0.48 Å⁻¹ and q_hc,intra = 0.89 Å⁻¹, which correspond to the interchain (d_hc,inter = 13.1 Å) and intrachain (d_hc,intra = 7.0 Å) distances in the amorphous helical clusters of st-PMMA chains. This result suggests that the HICs were not formed in THF because the weak interaction between st-PMMA and THF induced spontaneous aggregation of the st-PMMA chains into small, helical-rich amorphous clusters [bottom model in Fig. 1(b)]. Interestingly, in the st-PMMA/toluene gel, although a portion of the helical st-PMMA chains were wrapped around toluene, forming HICs, the scattering peak at q_hc,inter indicated that a fraction of amorphous helical chains remained [Fig. 1(b)].

Figure 1 (a) Inverted vial test photographs and WAXS profiles of the st-PMMA/toluene gel and st-PMMA/THF solution at [st-PMMA] = 0.4 M. (b) Structural illustration of the st-PMMA/toluene HIC and helical-rich st-PMMA cluster. (c) IR spectra of the st-PMMA/toluene gel and st-PMMA/THF solution at [st-PMMA] = 0.4 M.

The helical conformation of st-PMMA primarily consists of trans–trans (tt) conformations (Spéváček, 1978; Berghmans et al., 1994). As depicted in Fig. 1(c), the IR spectra of both st-PMMA/THF and st-PMMA/toluene exhibit CH₂ rocking bands in the 840–870 cm⁻¹ range, corresponding to the trans–trans (ν_tt ≃ 860 cm⁻¹) and trans–gauche (ν_tg ≃ 841 cm⁻¹) CH₂ vibration modes. The I₈₆₀/I₈₄₁ ratio was used to assess the extent of helical conformation in st-PMMA (Berghmans et al., 1994). In THF, helical st-PMMA exhibited a lower I₈₆₀/I₈₄₁ ratio of 0.69, which increased to 0.91 in the toluene solution. This result indicated that HIC formation promoted a more ordered helical conformation of the st-PMMA chains.

3.2. Gelation behavior of st-PMMA/toluene system

Studies have demonstrated that st-PMMA readily forms HICs with aromatic solvents (Spéváček, 1978; Berghmans et al., 1994). Therefore, it is essential to explore the thermodynamically favored structure during gelation of the st-PMMA/toluene system. An inverted vial test with varying [st-PMMA] concentrations in toluene solution (Fig. S3) indicated that the sol-to-gel transition occurred when the [st-PMMA] concentration reached 0.4 M, highlighting a strong correlation between the gelation behavior and self-assembly of HIC structures. To probe the structural evolution under varying [st-PMMA] concentrations, simultaneous SAXS/WAXS measurements were conducted, as illustrated in Figs. 2(a) and 2(b) (Liu et al., 2019). To analyze the hierarchical structures within this two-component system, the SAXS profiles of the sol and gel states were fitted using the Beaucage and gel-like models, respectively (details provided in the supporting information) (Mallam et al., 1991; Shibayama et al., 1992; Beaucage, 1995). Fig. 2(c) presents the structural parameters derived from the SAXS analyses, and Fig. 2(d) illustrates the structural evolution during a reversible sol-to-gel transition in the st-PMMA/toluene system.

Figure 2 (a) SAXS and (b) WAXS profiles of st-PMMA/toluene solution for [st-PMMA] from 0.05 to 0.4 M. (c) Concentration-dependent variation of structural parameters (ξ, Rg,hc and Rg,HIC) derived from SAXS model fitting in (a). (d) Illustration of structural evolution in the reversible sol-to-gel transition of the st-PMMA/toluene system. Temperature-dependent (e) SAXS and (f) WAXS profiles of st-PMMA/toluene at [st-PMMA] = 0.4 M. (g) Variation of ξ, Rg,hc and Rg,HIC derived from the SAXS model fitting in (e) during the gel-to-sol transition process.

At the lowest [st-PMMA] of 0.05 M, only weak scattering halos corresponding to the disordered st-PMMA chains within the helical-rich clusters appeared in the WAXS region (q = 0.2–1.1 Å⁻¹). In the SAXS region (q = 0.003–0.2 Å⁻¹), the SAXS profiles in Fig. 2(a) exhibit a single scattering knee, which was attributed to the mesomorphic helical-rich st-PMMA clusters with a gyration radius (R_g,hc) of 32 nm. As [st-PMMA] increased from 0.05 to 0.2 M, the intensity of q_hc,inter increased gradually [Fig. 2(b)], indicating an increase in the quantity of amorphous st-PMMA clusters in the solution. When [st-PMMA] exceeded 0.1 M, the diffraction peaks (q_tol,helix, q_helix,pitch and q_helix,intra) became evident, indicating gradual crystallization of the st-PMMA/toluene HICs within the helical-rich clusters. Moreover, the SAXS profiles [Fig. 2(a)] exhibited two scattering knees: one in the low-q region (q < 0.02 Å⁻¹) and the other in the higher-q region (0.02 < q < 0.2 Å⁻¹). These knees corresponded to the larger helical-rich clusters (R_g,hc) and smaller helical bundles of st-PMMA/toluene HICs (R_g,HIC), respectively, as fitted using the Beaucage model. Shifting of these scattering knees toward lower q values with increasing [st-PMMA] reflected the growth of R_g,hc and R_g,HIC, as depicted in Fig. 2(c).

At [st-PMMA] = 0.4 M, gel formation occurred as mesomorphic helical-rich clusters collided and formed a larger-scale 3D network structure [Fig. 2(d)]. The diffraction peaks of st-PMMA/toluene HICs became prominent, while the peak intensity of q_hc,inter decreased, indicating a higher HIC content within the gel network [Fig. 2(b)]. On the basis of the gel-like model fitting, the correlation length (ξ) of the network structure and R_g,HIC were determined to be 185 and 5.9 nm, respectively. Fig. 2(c) summarizes the hierarchical sizes across the solution and gel states for various [st-PMMA]. From sol to gel, the larger domain (R_g,hc and ξ) grew from 55 to 185 nm, while R_g,HIC increased from 5 to 5.9 nm. Crystallization of the st-PMMA/toluene HICs occurred hierarchically within the pre-existing large helical-rich clusters, which lowered the nucleation barrier considerably. This process resembles a two-step crystallization (Chuang et al., 2011), where conformational and concentration fluctuations induce phase-separated domains, facilitating subsequent crystallization. Thus, the helical-rich st-PMMA clusters act as intermediates for forming thermodynamically favored st-PMMA/toluene HICs.

To confirm the thermal stability of the st-PMMA/toluene HICs, temperature-dependent SAXS/WAXS profiles of the st-PMMA/toluene gel were obtained at [st-PMMA] = 0.4 M, as illustrated in Figs. 2(e) and 2(f). During heating, the diffraction peaks q_tol,helix, q_helix,pitch and q_helix,intra diminished gradually and they disappeared at temperatures exceeding 45°C, leaving only the q_hc,inter scattering peak. Meanwhile, the SAXS profiles were reduced to a single scattering knee with reduced intensity at high q. Fig. 2(g) presents the size changes (ξ, R_g,hc and R_g,HIC) derived from model fitting. The gel-to-sol transition occurred at around 35°C, while the disassembly temperature of the st-PMMA/toluene HICs (T_m,tolHIC) is approximately 50°C. Additionally, these phase transitions were confirmed by performing inverted vial tests (Fig. S4) and recording temperature-dependent IR spectra (Fig. S5). The results indicated that the st-PMMA/toluene HICs played a critical role in the reversible sol-to-gel transition, as illustrated in Fig. 2(d).

3.3. Formation of st-PMMA/C₆₀ HICs along temperature-controlled pathways

In supramolecular host–guest systems involving multiple guest molecules (Zwaag et al., 2015; Valera et al., 2018), the formation of various self-assembled structures is influenced heavily by the competition between the guest components. Consequently, in the st-PMMA/C₆₀/toluene system, the st-PMMA host can form HICs with both C₆₀ and toluene molecules (Yashima et al., 2016; Berghmans et al., 1994). Therefore, identifying the pathways to the thermodynamically favorable HICs is crucial for developing applications of the st-PMMA/C₆₀/toluene system. Herein, we first prepared st-PMMA/C₆₀/toluene samples with varying [st-PMMA] from 0.05 to 0.4 M while maintaining a constant C₆₀ mixing ratio of 7 wt% relative to [st-PMMA]. This setup allowed us to create a competitive environment for comparing the binding affinities of both C₆₀ and toluene to the st-PMMA host.

Figs. 3(a) and 3(b) depict the SAXS and WAXS profiles of the st-PMMA/C₆₀/toluene system with various [st-PMMA]. Similarly to the st-PMMA/toluene system (Fig. 2), the two-level structure was characterized by ξ, R_g,hc and R_g,HIC determined through fittings using the Beaucage and gel-like models for the st-PMMA/C₆₀/toluene system, as depicted in Fig. 3(c). The diffraction peak (q_C60,helix) at q ≃ 0.30 Å⁻¹, depicted in Fig. 3(b), grew gradually as [st-PMMA] increased, together with the above-mentioned peaks of the helical-rich st-PMMA clusters and st-PMMA/toluene HICs. This q_C60,helix peak, indicative of the packing distance between the st-PMMA/C₆₀ helices [as illustrated in Fig. 3(d)], was absent in the st-PMMA/toluene system [Fig. 2(b)], and it emerged only after the addition of C₆₀. The packing distances (d_C60,helix = 20.9 Å) of the st-PMMA/C₆₀ HICs were greater than those of the st-PMMA/toluene HICs (d_tol,helix = 16.5 Å) owing to the larger molecular size of C₆₀. Furthermore, the q_C60,helix peak was observed at [st-PMMA] = 0.05 M, while the q_tol,helix peak of the st-PMMA/toluene HICs was observed at 0.2 M. Guest molecules with stronger binding affinity generally facilitate the formation of guest–host complexes at lower concentrations (Matulis et al., 2005). Consequently, under these competitive conditions in the st-PMMA/C₆₀/toluene system, the st-PMMA hosts preferentially assembled with C₆₀ owing to their stronger binding affinity.

Figure 3 (a) SAXS and (b) WAXS profiles of the st-PMMA/C60/toluene system for [st-PMMA] of 0.05–0.4 M. (c) Variation of ξ, Rg,hc and Rg,HIC derived from the SAXS model fitting in (a). (d) Illustration of structural evolution in the st-PMMA/C60/toluene complex system. Temperature-dependent (e) SAXS and (f) WAXS profiles of the st-PMMA/C60/toluene system at [st-PMMA] = 0.4 M. (g) Variation of ξ, Rg,hc and Rg,HIC derived from SAXS model fitting in (e) during the gel-to-sol transition.

Furthermore, Fig. 3(b) shows that the diffraction intensities increased noticeably, which highlighted the enhanced crystallinity of both the HICs in the st-PMMA/C₆₀/toluene system. This result aligned with the macroscopic results obtained in the inverted vial test (Fig. S6), where the critical gelation concentration for the st-PMMA/C₆₀/toluene system was only [st-PMMA] = 0.2 M, which was half the concentration required for the st-PMMA/toluene system. However, R_g,HIC remained approximately 2–5 nm [Fig. 3(c)], similar to that of the st-PMMA/toluene system [Fig. 2(c)]. This implies that the nucleation density of the HICs dominated the gelation behavior in the st-PMMA/C₆₀/toluene system. As [st-PMMA] further increased to 0.4 M, the corresponding length (ξ) of the gel network became smaller than that at 0.2 M [Fig. 3(c)], indicating increased network density. This phenomenon was also reflected in the rheological measurement results (Fig. S7), where the gel modulus (G′) of the st-PMMA/C₆₀/toluene gel (G′ = 80 Pa) was higher than that of the st-PMMA/toluene gel (G′ = 20 Pa).

Figs. 3(e)–3(g) depict the temperature-dependent SAXS and WAXS profiles along with the fitted parameters. These profiles demonstrate the thermal stabilities of the st-PMMA/toluene and st-PMMA/C₆₀ HICs during the gel-to-sol transition. As illustrated in Fig. 3(g), the correlation length of the gel network increased slightly with temperature owing to thermal expansion. Above 55°C, the gel network disintegrated, accompanied by a continuous decrease in the size of the HICs (R_g,HIC). This gel-to-sol transition was further confirmed by the inverted vial test (Fig. S8). Interestingly, as the gel collapsed, the q_tol,helix peak disappeared, whereas the q_C60,helix peak remained visible until 75°C. This finding suggests that the melting temperature (T_m,C60HIC) of the st-PMMA/C₆₀ HICs was approximately 75°C, considerably higher than T_m,tolHIC of the metastable st-PMMA/toluene HICs (around 50°C). Therefore, the st-PMMA/C₆₀ HICs emerged as the thermodynamically favorable species compared with the metastable st-PMMA/toluene HICs in the st-PMMA/C₆₀/toluene system.

To further evaluate the maximum C₆₀ encapsulation content in the st-PMMA host, Fig. 4(a) illustrates the procedure for preparing the st-PMMA/C₆₀ HICs from the toluene solution at 25°C. On addition of C₆₀ powder into the st-PMMA/toluene solution (0.4 M, 1 ml), the st-PMMA/C₆₀ inclusion complexation drove the C₆₀ powders to dissolve in the solution. We further used the WAXS tool to probe the pathways toward the st-PMMA/C₆₀ HICs and determined the maximum C₆₀ encapsulation ratio in the st-PMMA HICs at 25°C below T_m,tolHIC. According to the WAXS profiles in Fig. 4(b), before adding C₆₀, both the st-PMMA/toluene and the st-PMMA/C₆₀ HICs coexist in the st-PMMA/toluene solution at 25°C below T_m,tolHIC. As C₆₀ was gradually added from 3 to 22 wt% (2.5–23 mg ml⁻¹), the intensities of the q_tol,helix and q_hc,inter peaks decreased simultaneously. Meanwhile, the intensity of the q_C60,helix peak in the spectrum of the st-PMMA/C₆₀ HIC increased. This observation shows that the st-PMMA/C₆₀ HIC formation follows two self-assembly pathways at 25°C. In one of the pathways, the disordered st-PMMAs in the helical-rich cluster directly wrap around C₆₀s to form the st-PMMA/C₆₀ HICs. The other pathway is determined by competing encapsulation, where C₆₀s undergoes a guest-exchange process with the st-PMMA/toluene HICs. As the C₆₀ con­centration increases to 20 wt%, the disappearance of the q_tol,helix peak confirms that only the st-PMMA/C₆₀ HICs remain in the system, as illustrated in Fig. 4(c). It is well known that C₆₀ only has a limited solubility of approximately 1.5 mg ml⁻¹ in toluene solvent (Guo et al., 2016). Thus, st-PMMA acts not only as the helical host but also as the solubilizing agent to make C₆₀ significantly more soluble in the st-PMMA/toluene solution through inclusion complexation. At 22 wt%, the sharp diffraction at q₍₁₁₁₎ = 0.77 Å⁻¹, attributed to the C₆₀ crystallites, indicates that the C₆₀ fills up the st-PMMA HICs, thereby leading to the precipitation of excess C₆₀. From the thermodynamic standpoint, the maximum C₆₀ encapsulation ratio in the st-PMMA helix with an rr content of 86% is approximately 20 wt%, lower than the ideal encapsulation ratio of 28 wt% in the defect-free st-PMMA helix (rr content = 100%) (Kajihara et al., 2020). This decrease in encapsulation might be ascribed to a few chain defects in the st-PMMAs.

Figure 4 (a) Preparation of st-PMMA/C60 HICs in the st-PMMA/C60/toluene solution. (b) WAXS profiles of the st-PMMA/C60/toluene solution with various C60 wt% values at 25°C. (c) Free-energy landscape of the st-PMMA(0.4 M)/C60/toluene system at T < Tm,tolHIC. (d) Free-energy landscape of the st-PMMA/C60/toluene system at Tm,tolHIC < T < Tm,C60HIC. (e) WAXS profiles of the st-PMMA(0.4 M)/C60/toluene solution with various C60 wt% values at 50°C.

Nonetheless, as we follow the energy landscape at 25°C below T_m,tolHIC [Fig. 4(c)] to prepare the st-PMMA/C₆₀ HICs, a lengthy guest-exchange process lasting about 7.5 h is required to achieve an encapsulation ratio of 20 wt% (Mukhopadhyay et al., 2006). Given that the free energy of the system is temperature dependent, temperature can modulate the energy landscape to bypass the guest-exchange process, as illustrated in Fig. 4(d). The WAXS analysis in Fig. 4(e) reveals that, at T = 50°C, within the range T_m,tolHIC < T < T_m,C60HIC, the st-PMMA/toluene HICs disassemble as evidenced by the absence of their diffraction peaks (q_tol,helix, q_helix,pitch and q_helix,intra). As we added more C₆₀s (from 4 to 20 wt%), the increased I_C60,helix at q = 0.30 Å⁻¹ was accompanied only by the decreased I_hc,inter at q = 0.48 Å⁻¹. This result clearly indicates that, at T_m,tolHIC < T < T_m,C60HIC, the complex system directly chose the disordered st-PMMA/C₆₀ co-assembly pathway to form the st-PMMA/C₆₀ HICs, without undergoing the time-consuming guest-exchange process. This accelerated the formation of the st-PMMA/C₆₀ HICs at the encapsulation ratio of 20 wt% in a shorter time of 4.5 h. This adjustment in temperature successfully tuned the energy landscape of the st-PMMA multicomponent system, thereby programming the efficient self-assembly route toward the thermodynamically favorable st-PMMA/C₆₀ HICs.

3.4. st-PMMA/C₆₀ HICs as redox sites for preparing the Ag SERS-active substrate

In sensing applications, numerous studies have demonstrated that the SERS effect in Raman spectroscopy depends on surface plasmons to enhance the light's electric field, which is influenced by the morphology of metallic hotspots, including their size and density (Lee et al., 2008; Zhu et al., 2016; Solis et al., 2017). C₆₀ exhibits redox activity with specific metal ions, such as Ag and Au, through electron-transfer processes (Shin et al., 2010; Shrestha et al., 2013). However, the crystallization behavior of C₆₀ often leads to micrometre- to millimetre-sized crystal morphologies, limiting the surface area available for metal ion adsorption (Wu et al., 2015). Conversely, according to the above structural analysis, st-PMMA/C₆₀ HIC bundles are nanometre sized, thereby providing a considerably larger surface area. Furthermore, use of the efficient co-assembly pathway [Fig. 4(d)] allows for rapid control over the C₆₀ encapsulation ratio within the st-PMMA complex structure. Consequently, the st-PMMA/C₆₀ HICs hold substantial promise as redox templates for fabricating Ag SERS-active substrates.

The preparation procedure of the Ag SERS-active substrate is illustrated in Fig. 5(a). By using the liquid–liquid interface diffusion method (Shrestha et al., 2013), a solution of AgNO₃ in ethanol/H₂O (2 ml) was gradually added to the st-PMMA/C₆₀/toluene complex gel (1 ml) with various C₆₀ encapsulation ratios (7, 14 and 20 wt%). This diffusion process yielded an Ag⁺-loaded st-PMMA complex gel that settled to the bottom of the vessel. After removing the supernatant, the Ag⁺-loaded st-PMMA complex gel was dissolved in a toluene solution and cast onto silicon wafers. The resulting Ag⁺-containing st-PMMA complex film was then subjected to a redox reaction under UV light exposure (λ = 365 nm) to obtain the Ag SERS-active substrate.

Figure 5 (a) Preparation of SERS-active Ag-NP substrates reduced from the st-PMMA/C60 complex film, and SERS measurement of R6G analytes on the Ag SERS-active substrate. (b) XPS spectra of the Ag^+-containing st-PMMA/C60 HIC film and C60 crystals before/after visible light exposure. (c)–(e) SEM images of SERS-active Ag-NP substrates prepared with different C60 encapsulation ratios: (c) 7 wt%, (d) 14 wt% and (e) 20 wt%. (f) Ag-decorated C60 crystals. (g) Raman spectra (SERS) collected from various SERS-active Ag-NP substrates with different [R6G] (10 and 10^−2 µM).

The redox process (Ag⁺ → Ag) within the st-PMMA/C₆₀ HIC template was characterized using XPS. As depicted in Fig. S9, no XPS signals corresponding to Ag⁺/Ag were detected in the st-PMMA substrate, indicating minimal interaction between Ag⁺ and the st-PMMA host. However, Fig. 5(b) shows two distinct peaks at 368.7 and 374.8 eV, which correspond to the 3d_5/2 and 3d_3/2 binding energies of Ag⁺, respectively, in the st-PMMA/C₆₀ HIC film and pure C₆₀ crystal substrates. This suggests that Ag⁺ preferentially binds to C₆₀ through charge-transfer interactions. Upon exposure to UV light, the binding energies of 3d_5/2 and 3d_3/2 shifted to lower values (368.2 and 374.3 eV, respectively), confirming successful redox conversion of the Ag⁺ into metallic Ag-NPs.

Moreover, SEM analysis revealed the distribution of metallic Ag-NPs on the redox templates. As depicted in Figs. 5(c)–5(f), Ag-NPs with an average size of approximately 20 nm were distributed evenly across the st-PMMA/C₆₀ HIC films. As the C₆₀ encapsulation ratio was increased from 7 to 20 wt%, the highest Ag-NP density was achieved on the st-PMMA/C₆₀ (20 wt%) HIC film. By contrast, only a small number of Ag-NPs were found on the C₆₀ crystal substrate. This difference was attributed to the uniformly dispersed C₆₀ on the surface of the st-PMMA/C₆₀ film, which provided numerous active sites that facilitated efficient Ag-NP reduction.

To evaluate the SERS performance of the Ag SERS-active substrates, aqueous solutions of rhodamine 6G (R6G, 10 µl) with varying concentrations (10⁻²–1 µM) were applied to the substrates, as depicted in Fig. 5(a). The Raman spectra of R6G on the Ag-NP substrate, displayed in Fig. 5(g), clearly exhibit the vibrational bands of R6G in the 500–1700 cm⁻¹ range (Jensen & Schatz, 2006). These bands were assigned as follows: the C—C—C in-plane bending band at ν = 610 cm⁻¹, the C—H out-of-plane bending band at ν = 770 cm⁻¹ and the C—H in-plane bending band at ν = 1188 cm⁻¹. Additionally, the aromatic ring stretching bands of R6G were observed between ν = 1310 and 1649 cm⁻¹. By contrast, no distinct Raman peaks of R6G (1 µM) were detected on the Ag-decorated C₆₀ crystal substrate [Fig. 5(g)].

The SERS enhancement factor (EF) of R6G on the Ag-NP substrates was calculated (see the supporting information). The substrates prepared from st-PMMA/C₆₀ templates with encapsulation ratios of 7, 14 and 20 wt% exhibited EF₁₅₁₂ values of 3.1 × 10⁴, 1.5 × 10⁵ and 4.1 × 10⁵ at ν = 1512 cm⁻¹, respectively. The increasing EF correlated to a higher density of Ag-NP formation on the templates with the greater C₆₀ encapsulation. Furthermore, even at an [R6G] as low as 10x⁻² µM, clear Raman peaks were observed on the Ag-NP SERS substrate. These findings demonstrated that following an optimized self-assembly pathway facilitated the rapid formation of st-PMMA/C₆₀ HIC films with the desired C₆₀ encapsulation ratio. The well dispersed st-PMMA/C₆₀ nanodomains allowed for the production of a higher number of Ag-NPs, leading to significant SERS enhancement.

4. Conclusions

This study elucidated the self-assembly pathways of the st-PMMA/C₆₀/toluene system. According to our structural characterization results, st-PMMA exhibited stronger binding affinity for C₆₀ than for toluene, making st-PMMA/C₆₀ HICs the thermodynamically favored structures. Temperature control helped to modulate the energy landscapes of the self-assembled HICs: at T < T_m,tolHIC, the complex guest-exchange pathway delayed st-PMMA/C₆₀ HIC formation, whereas at T_m,tolHIC < T < T_m,C60HIC, the st-PMMA/toluene HICs were suppressed, allowing for rapid st-PMMA/C₆₀ HIC formation. The st-PMMA/C₆₀ HIC film effectively acted as a redox template, generating abundant Ag-NPs for SERS applications. These substrates deliver a high EF of 10⁵, outperforming Ag-decorated fullerene crystals in R6G detection. This work provides a framework for programming self-assembly pathways to design advanced supramolecular materials.

5. Related literature

The following references are cited only in the supporting information: He et al. (2017); Wei & Hore (2021).