CO₂-expanded liquid assisted self-assembly between Disperse Red 1 and PS-b-P4VP

Abstract

This work shows that CO₂-expanded liquids (CXLs) facilitate the modulation of morphology and photoluminescence (PL) performance of the self assembled fluorescent composite (SAFC), which was formed between Disperse Red 1 (DR1) and polystyrene-block-poly(4-vinylpyridine) (PS-b-P4VP) in CO₂-expanded ethanol. We find that the anti-solvent effect of CXLs with increasing pressure could effectively regulate the amount and type of the hydrogen bonds with DR1 units as well as the confinement extent of DR1, along with the structure and fluorescence (FL) behavior of the SAFC. The morphology and emission behavior of the SAFC was strongly dependent on the CXL pressure and the molar ratio (R) of 4VP to DR1. The SAFC emission revealed a non-monotonic feature against CXL pressure, which increased with a pressure rise below the threshold pressure (P_T) and then decreased at higher pressures above P_T. The maximum FL of SAFC at 5.58 MPa and R = 2800 was enhanced by 3.4-fold compared with that of the pristine mixture. Confinement and hydrogen bonds are the two major factors responsible for the pressure dependence of the emission. Moreover, the hydrogen bonds formed between the DR1 and P4VP blocks are the principal contribution to SAFC emission than those between DR1 and ethanol.

---

1. Introduction

In recent years, a number of research groups have investigated the photoresponsive behavior of non-covalent azobenzene-containing complexes,¹ and a few works have studied the role of chromophore–polymer intermolecular interactions on the optical performance of polymeric materials.² Polymer–azobenzene complexes are advantageous in the ease of sample fabrication and optimization of the composition of the polymeric probe, and have proven to compete successfully with covalently-functionalized polymers in terms of the magnitude, stability as well as the efficiency.^2c These potential advantages make polymer–azobenzene complexes a promising type of materials for various photonics applications. The well-studied systems are usually poly(4-vinylpyridine)-containing block copolymers,³ and many scientists have examined the self-assembly behavior of poly(styrene-block-4-vinylpyridine) (PS-b-P4VP) and small molecules coupled with the P4VP blocks via hydrogen bonds.^3a,4 Hydrogen bonds usually play a remarkable role in building supramolecular structures through self-assembly due to their appropriate bonding energy,⁵ and it is known that both the strength and extent of hydrogen bonds rely on the specific affinities between the hydrogen bond acceptors and donors.⁶ Zhang and coworkers⁷ performed a combined theoretical study using density functional theory (DFT) for the blends system of PS-b-P4VP and poly(4,4′-oxydiphenylenepyromellitamic acid) (POAA) in both microscopic and mesoscopic scales, and obtained the binding energies, geometric information, and infrared (IR) spectra for the possible hydrogen bonds in the blends. It was clear that the hydrogen bonds between P4VP and POAA, and those among POAA chains were competitive and the major ones could decide the morphology evolution from lamellar to sphere. Clodt and coworkers⁸ reported the self-assembled membranes formed via hydrogen bonds between pyridine units of PS-b-P4VP and hydroxyl groups of the carbohydrates (α-cyclodextrine, α-(D)-glucose, and saccharose)^8a or poly(ethylene glycol),^8b and they found that the addition of carbohydrates preferred to form a type of isoporous composites membranes with higher water flux. Chen and coworkers^4d investigated the complexation-induced phase behaviour of octyl gallate (OG) and PS-b-P4VP system in several solvents, and they concluded that the structure of the complex could be regulated owing to the varying hydrogen bonding strengths between OG and PS-b-P4VP in different conventional solvents. Ali and coworkers^3a examined the emission of Disperse Red 1 (DR1) with PS-b-P4VP at 353.2 K in toluene and ethanol, respectively. They found that the emission of DR1 confined in the hard core of the PS-b-P4VP micelles (in toluene) was more stronger than that in the soft corona (in ethanol) due to the higher degree of confinement within the core, which slowed the trans-to-cis transition of DR1.

As one of the most studied azobenzene derivatives, DR1 has been explored a variety of applications in optical data storage and optical switching devices⁹ for its photoisomerization properties.^3a,10 Toro et al. studied the emission of DR1 in methanol, ethylene glycol, glycerol and phenol at room temperature, they found that alcohols could form stronger hydrogen bonds with nitrogen of the diazo group (–N=N–) of the molecule as well as with the nitrogen atoms of the nitro and the N-ethyl-N-(2-hydroxyethyl) group.¹¹ Besides, relatively weaker hydrogen bonds could be formed with oxygen atoms in the nitro and hydroxyl groups of DR1. And among the hydrogen bonds formed between solvent and DR1, the one between the –OH group of methanol and the nitrogen of diazo group of DR1 was the major contribution to favor emission. And the π → π* transition was regarded as the only contribution to the absorption and fluorescence (FL) although there exists π → π* and n → π* transitions in DR1. Like the majority of chromophores with an extended π-system, DR1 with an azobenzene moiety is sensitive to aggregation effects due to chromophore–chromophore intermolecular interactions. The nature of these interactions and the packing arrangement of the chromophores also depend on their local environment,^2c,12 thus the selection of smart solvent system is an alternative way to promote its emission. Many self assembly studies were carried out in conventional solvent system with the focus on the changes in polymer composition^8a,13 and concentration, solvent composition,¹⁴ pH,¹⁵ the type and amount of the additives (e.g., carbohydrates,^8a acids, bases, or salts),^4b,4f–i as well as temperature. To the best of our knowledge, few works were available using CO₂-expanded liquids (CXLs) as solvent media to prepare the self-assembled aggregates. As we know, CXLs exhibited many beneficial properties in materials processing compared with conventional solvents owing to their tunable solvating power, fluidity, and numerous other properties that can be tuned in response to pressure changes.¹⁶ CXLs belong to a class of ideal solvents for practically any application while synchronously reducing the environmental burden of a given process via considerable replacement of organic solvents with eco-friendly CO₂.^4e And polymer processes in CXLs have been employed to pressure tune the structures and particle sizes of crystalline polymers,¹⁷ and control the morphology and emission of the self assembled fluorescent composite (SAFC) between PS-b-P4VP and the dye molecule 4-(9-(2-(4-hydroxyphenyl)ethynyl)-7,10-diphenylfluoranthen-8-yl)-phenol.^4e

Meanwhile, the confinement or aggregation of the fluorophore^3a,18 enhances FL emission of the DR1, which might be an effective way of improving its FL intensity. Under such guidance, the present work focuses on exploring the connection between polymer aggregation and the optical response of the SAFC via the modular tunability offered by the hydrogen bonds formed between PS-b-P4VP and DR1 with the assistance of CXLs technology (Scheme 1). The confinement effect or the aggregation degree of DR1 on the emission performance of the SAFC was investigated by regulating the solvated state of P4VP chains which was easily tuned by the CXLs pressure. Furthermore, we try to distinguish the emission contributions from hydrogen bonds between H-bond (a) and H-bond (b), herein H-bond (a) represents the hydrogen bonds formed between the –OH group of DR1 and the nitrogen atoms of the pyridine ring of P4VP blocks, and H-bond (b) between the –OH group of ethanol and the nitrogen of diazo group (–N=N–) of DR1. We also clarify such questions as follows. Which type of hydrogen bonds is predominating in quantity? Are there any confinement difference between H-bond (a) and H-bond (b)? Do they enhance the emission of SAFC in equal or different weight? Will the amount and type of hydrogen bonds are pressure dependent? We found that the changes in the solubilities of PS-b-P4VP with increasing CXLs pressure will affect the amount of the hydrogen bonding between the P4VP block and DR1, further influence the self-assembly behavior of the SAFC in CXLs, along with the luminescent performance of the resultant SAFC. This work shows that CXLs method facilitates the modulation of the morphology and the emission performance of the SAFC, which provides a promising pathway of developing optical materials. We believe that such studies will not only contribute to the basic understanding of the role of intermolecular interactions on the optical response of polymeric materials, but also help investigate the underlying luminescent mechanisms of the azobenzene-containing polymers.

Scheme 1 Structure and emission control of the PS-b-P4VP–DR1 composites using CO₂ expanded ethanol (left) schematic representation of the CXLs-assisted self assembly between PS-b-P4VP and DR1; (right) the evolution of morphology and emission intensity (circles) of PS-b-P4VP–DR1 composites with CXLs pressure.

Following the introduction, Section 2 describes mainly the preparation of copolymer PS-b-P4VP and the SAFC between PS-b-P4VP and DR1 in CO₂-expanded ethanol. Section 3 discusses the morphological transitions and the mechanism of pressure and concentration dependence of SAFC emission, focusing on major factors affecting the FL performance of SAFC. The final section presents a summary of this work.

2. Experimental section

2.1 General information

The average molecular weight (M_w and M_n) and polydispersity index (PDI, M_w/M_n) were obtained by Gel Permeation Chromatography (GPC) (Waters 515 HPLC pump, Waters 2414 Refractive index detector, Waters Corporation, USA). The morphologies of the SAFC were obtained by transmission electron microscopy (TEM), which was performed by JEOL JEM-2100 microscope with an accelerating voltage of 200 kV (JEOL Ltd, Tokyo, Japan). Fourier transform infrared (FT-IR) spectra were recorded on a NEXUS-470 spectrometer (Nicolet, USA). The fluorescence measurements were carried out at 298.2 K on a Hitachi FL-7000 spectrofluorometer with an excitation wavelength at 490 nm.

2.2 Materials

The monomer 4-vinylpyridine (4VP, J&K Scientific Ltd, Beijing, 95%) was distilled under reduced pressure before used. And styrene (St, Tianjin Chemical Reagent, 98%) was washed with an aqueous solution of sodium hydroxide (5 wt%) three times and then with water until neutralization. After being dried with anhydrous magnesium sulfate for 24 h, the monomer was distilled at reduced pressure. 2,2-Azobis(isobutyronitrile) (AIBN, Shanghai Chemical Reagent, 98%) was purified by recrystallization from methanol. Tetrahydrofuran (THF, Tianjin Chemical Reagent, 98%) was distilled in the presence of sodium. Dimethylformamide (DMF, Tianjin Chemical Reagents Co.) was distilled from CaH₂ under reduced pressure. Benzyl dithiobenzoate (BDTB) was prepared according to the procedure reported.¹⁹ CO₂ with a purity of 99.95% was provided by Henan Keyi Gas Company (Zhengzhou, China). DR1 (J&K Scientific Ltd, Beijing, ≥95%) and ethanol (HPLC grade, Sigma-Aldrich, ≥99.9%) were used as received.

2.3 Preparation of PS and PS-b-P4VP

2.3.1 Preparation of PS. In a typical polymerization procedure, an appropriate amount of St, BDTB, and AIBN (with a molar ratio of 100/1/0.2) were added into a 25 mL dry glass tube, followed by three freeze–vacuum–thaw cycles. The tube was sealed under vacuum and then immersed into an oil bath at 353.2 K with magnetic stirring. After reaction for 9 h, the tube was cooled to room temperature immediately. The polymer was dissolved with certain amount of THF, and then precipitated by dropping the solution into anhydrous methanol, followed by filtration to obtain the pink polystyrene (PS). Repeat the dissolving-precipitation procedure three times, afterwards the product was dried in a vacuum oven for 24 h to achieve the polystyrene (PS) (yield: 47.0%).

2.3.2 Preparation of PS-b-P4VP. A prescribed amount of AIBN, PS (as chains transfer agent, CTA), 4VP and DMF were successively added into a dry glass tube with a magnetic bar. After three freeze–vacuum–thaw cycles, the tube was sealed under vacuum and then the sealed tube was immersed in an oil bath at 353.2 K. After prescribed time, the tube was rapidly cooled down to room temperature. The copolymer was dissolved with certain amount of DMF, and then the copolymer was precipitated by adding the polymer solution into a mixture of petroleum ether and absolute ether (1 : 1, volume ratio). After filtration, the precipitation was dried in a vacuum oven for 24 h to obtain the light pink block copolymer PS-b-P4VP (yield: 52.5%).

The number-average molecular weight (M_n) and polydispersity index (M_w/M_n) of PS and PS-b-P4VP are tabulated in Table 1.

Table 1 Polymerization of PS and PS-b-P4VP at 353.2 K and different reaction conditions

Sample	t (h)	Initiator	BDTB/AIBN	Mn^a	Mw/Mn^a	Yield (%)
a The number-average molecular weight (Mn, g mol^−1) and polydispersity index (Mw/Mn) of PS and PS-b-P4VP were determined by GPC in DMF at 313.2 K, flow rate 1 mL min^−1.
PS	9	AIBN	10 : 2	5130	1.36	47.0
PS-b-P4VP	20	AIBN		13 960	1.38	52.5

---

2.4 Preparation of SAFC in CO₂-expanded ethanol

The experimental apparatus was described in our previous work.^17a Typically, appropriate amount of DR1 were dissolved in ethanol. A certain amount of PS-b-P4VP were dispersed in ethanol at a suitable temperature, the mixture was then ultrasonicated for 5 min in order to disperse the block copolymer homogeneously in the solution. The DR1/ethanol solution was then added into the dispersion in a tube. After equilibration, the tube containing the mixture was quickly transferred into a stainless-steel autoclave at the temperature of 353.2 K. CO₂ was then charged into the autoclave to achieve the desired pressure within a short time. After 2 h treatment in CO₂-expanded ethanol, the system was cooled down and slowly depressurized, afterwards the SAFC sample was collected and labeled.

A droplet of the micellar solution was dropped on a carbon-coated copper grid placed on a filter paper, followed by air-drying. The carbon-coated copper grids were then exposed in staining agent iodine for 6 h to make the P4VP block visible so that the contrast is enhanced.

To understand the molar ratio of 4VP/DR1 (the ratio of 4VP/DR1 is named as R) dependence on morphology change and emission intensity of the SAFC, the reaction mixtures were examined with R selected to be 1400 (i.e., PS-b-P4VP and DR1 in ethanol at concentration of 5 mg mL⁻¹ and 1.8 × 10⁻⁵ mol L⁻¹, respectively) and 2800 (PS-b-P4VP and DR1 at concentration of 5 mg mL⁻¹ and 9.0 × 10⁻⁶ mol L⁻¹), respectively. The CXLs pressures were selected to range from 3.0 to 8.3 MPa, and the SAFC samples free of CO₂ (0.1 MPa) were prepared for comparison.

3. Results and discussion

3.1 FTIR analyses

Fig. 1 provides the FTIR spectra of the pristine PS-b-P4VP and the SAFC samples with R = 1400 at different conditions. The samples of SAFC were dried from a 5 mg mL⁻¹ ethanol solution in KBr. Firstly, the PS-b-P4VP (Fig. 1a) exhibited a characteristic and symmetric stretching of the free pyridine ring at 1025 cm⁻¹. The intensities at 1025 cm⁻¹ for the PS-b-P4VP–DR1 composite (Fig. 1b and c) decreased considerably, suggesting a significant reduction in the amount of free pyridine in the P4VP block due to the formation of the complex with DR1.^3d,20 Secondly, the C–N stretching of the pyridine ring of the PS-b-P4VP–DR1 composite at 1221 cm⁻¹ (ref. 21) decreased with increasing CXLs pressure compared to that of the PS-b-P4VP after the formation of hydrogen bonds. Thirdly, the new pyridinium salt peak^3a observed at 1670 cm⁻¹ for the PS-b-P4VP–DR1 composite (Fig. 1b and c) further supported the formation of hydrogen bonds with the DR1 unit, and its intensity increased with increasing CXLs pressure in comparison to the pure PS-b-P4VP. Thus, it is convincing from the FTIR data that the self-assembled fluorescent composite (SAFC) of DR1 and PS-b-P4VP were successfully made due to the hydrogen-bonding interactions between the pyridine groups of P4VP block and the –OH groups of DR1.

Fig. 1 FTIR spectra of PS-b-P4VP and SAFC. (a) Pure PS-b-P4VP; SAFC (with R = 1400) were obtained from the mixture of DR1 (1.8 × 10⁻⁵ mol L⁻¹) and PS-b-P4VP (5 mg mL⁻¹) in CO₂-expanded ethanol (353.2 K/varied pressure for 2 h) (b) SAFC of 0.1 MPa; (c) SAFC of 6.28 MPa.

3.2 The dispersibility and volume expansion of the reaction mixtures at 353.2 K

When CO₂ dissolves into ethanol, the volume of the mixture expands, forming a CO₂ expanded ethanol mixture. The extent of miscibility of CO₂ and ethanol is usually reflected in the volume expansion of the liquid phase after CO₂ injection. The volume expansion of the liquid phase, ΔV/V₀, at a certain temperature T and pressure P can be expressed as^17c

ΔV/V₀ = [V(T, P) − V₀(T, P⁰)]/V₀(T, P⁰) (1)

where V is the liquid volume, V₀ is the liquid volume at the reference condition, and P⁰ represents the reference pressure. The reference condition represents the liquid phase at the atmospheric pressure. As a class II liquids,²² ethanol could dissolve large amount of CO₂ and expand considerably, and consequently may experience significant changes in many physical properties, such as the polarity and solubility or dispersibility of the mixture. Fig. 2 displays the isothermal volumetric expansion and dispersibility images of the reaction mixtures under study. At 353.2 K and R = 1400 (Fig. 2), it is found that the dispersibility state of the reaction mixtures transits from transparent and homogeneous state (Fig. 2a, 5.50 MPa, 29% volume expansion) through emulsion (Fig. 2b, 6.75 MPa, 39% volume expansion) to solution with visible precipitation (Fig. 2c, 8.33 MPa, 48% volume expansion). We could find reasonable explanation from the pressure-dependent anti-solvent effect of CO₂. As illustrated in Fig. 2, the volumes of the mixture became increased with increasing pressure at 353.2 K. It is evident that the miscibility of CO₂ is enhanced with increasing pressure, that is, CO₂ will exert more anti-solvent effect as pressure increases, which may lead to a worse solubility and more folded chains of P4VP block, and thus the aggregated amount of PS-b-P4VP is supposed to increase significantly with increasing CXLs pressure, as reflected in the pressure-dependent dispersibility state of the reaction system (Fig. 2).

Fig. 2 The dispersibility image and volume expansion of the reaction mixtures of DR1 (1.8 × 10⁻⁵ mol L⁻¹) and PS-b-P4VP (5 mg mL⁻¹) in CO₂-expanded ethanol as a function of pressure at 353.2 K (a) 5.50 MPa; (b) 6.75 MPa; (c) 8.33 MPa.

The dispersibility images at different pressures could help us find the threshold pressure (P_T) to homogeneously disperse the copolymers, estimate the aggregation or folding degree of P4VP chains, and the further confinement effect on the DR1, and then the accompanying emission behavior of the SAFC. Herein, P_T refers to the pressure where the copolymers have the maximum degree of aggregation, and the copolymers start precipitation when pressure is above P_T. Moreover, the volume expansion and the accompanying dispersibility transition behavior (Fig. 2) would influence the degree of intramolecular rotation of DR1 and help us to regulate the FL intensity of SAFC at suitable experimental conditions. Thus, we can expect that the pressure-dependent anti-solvent effect may open a promising opportunity for the morphology and luminescence tuning of PS-b-P4VP/DR1 composite, as illustrated in Section 3.3.

3.3 The regulation of morphology and emission of SAFC in CO₂-expanded ethanol

Firstly, as ethanol is P4VP selective, the SAFC of PS-b-P4VP–DR1 in ethanol (0.1 MPa, free of CO₂) shows spherical micelles (Fig. 3a and 4a) with PS cores surrounded by P4VP coronal chains, and then DR1 could be located in the spherical coronas of the micelles through hydrogen bonding with P4VP. We think the confinement of DR1 comes from two major contributions, one is from hydrogen bonds with P4VP blocks, and the other from the aggregation or clustering of copolymers resulted from the CXLs-induced changes in solvated state and mobility of copolymers. As will be demonstrated in the following section, the CXLs anti-solvent effect is responsible for the regulation of the amount and type of hydrogen bonds in the reaction mixture (Section 3.3.1). Besides, the solubility change in copolymer induced by the anti-solvent effect of CXLs leads to the transformation in folding degree of P4VP chains, followed by the confinement difference of the DR1 unit (Section 3.3.2).

Fig. 3 TEM images of SAFC. SAFC (with R = 1400) were obtained from the mixture of DR1 (1.8 × 10⁻⁵ mol L⁻¹) and PS-b-P4VP (5 mg mL⁻¹) in CO₂-expanded ethanol (353.2 K/varied pressure for 2 h) (a) 0.1 MPa; (b) 4.15 MPa; (c) 5.50 MPa; (d) 6.28 MPa; (e) 8.33 MPa.

Fig. 4 TEM images of SAFC. SAFC (with R = 2800) were obtained from the mixture of DR1 (9.0 × 10⁻⁶ mol L⁻¹) and PS-b-P4VP (5 mg mL⁻¹) in CO₂-expanded ethanol (353.2 K/varied pressure for 2 h) (a) 0.1 MPa; (b) 3.05 MPa; (c) 4.78 MPa; (d) 5.58 MPa; (e) 8.32 MPa.

3.3.1 The CXLs-induced evolution in the type and amount of hydrogen bonds, and the influence on the SAFC emission. We find that the SAFC formed in ethanol free of CO₂ (0.1 MPa) is more emissive than the corresponding pristine mixture (Fig. 5A and 6A), which may be resulted from the formation of hydrogen bonds with DR1. It is known that the confinement of the dye unit through intermolecular hydrogen bonding favors the FL emission via the restriction of the trans-to-cis isomerization of the azo compound.¹¹ Herein, the hydrogen bonds could be formed either between DR1 and ethanol,¹¹ or between DR1 and P4VP blocks of PS-b-P4VP.^3a And among the hydrogen bonds formed between solvent and DR1, the one between the –OH group of ethanol and the nitrogen of diazo group (–N=N–) of DR1 was the major contribution to promote emission based on the work reported by Toro and coworkers.¹¹ Meanwhile, the FTIR data shown in Section 3.1 confirmed the existence of another important type of hydrogen bonds between DR1 and P4VP blocks. If H-bonds (a) denotes the hydrogen bonds formed between the –OH group of DR1 and the nitrogen atoms of the pyridine ring of P4VP blocks, and H-bonds (b) represents those between the –OH group of ethanol and the nitrogen of diazo group (–N=N–) of DR1, we believe that the combinational contributions from H-bonds (a) and H-bonds (b) make the SAFC more emissive than its pristine mixture. But which type of hydrogen bonds is predominating in quantity? Are there any confinement difference between H-bonds (a) and H-bonds (b)? Do they enhance the emission of SAFC in equal or different weight? Will the amount and type of hydrogen bonds are pressure dependent?

Fig. 5 (A) FL spectra of the SAFC (with R = 1400), and the pristine mix represents the mixture of DR1 (1.8 × 10⁻⁵ mol L⁻¹) and PS-b-P4VP (5 mg mL⁻¹) in ethanol without formation of the SAFC. (B) Pressure dependence of the maximum FL intensity (I) of SAFC. SAFC were obtained from the mixture of DR1 (1.8 × 10⁻⁵ mol L⁻¹) and PS-b-P4VP (5 mg mL⁻¹) in CO₂-expanded ethanol (353.2 K/varied pressure for 2 h). λ_ex = 490 nm.

Fig. 6 (A) FL spectra of the SAFC (with R = 2800), and the pristine mix represents the mixture of DR1 (9.0 × 10⁻⁶ mol L⁻¹) and PS-b-P4VP (5 mg mL⁻¹) in ethanol without formation of the SAFC. (B) Pressure dependence of the maximum FL intensity (I) of SAFC. SAFC were obtained from the mixture of DR1 (9.0 × 10⁻⁶ mol L⁻¹) and PS-b-P4VP (5 mg mL⁻¹) in CO₂-expanded ethanol (353.2 K/varied pressure for 2 h). λ_ex = 490 nm.

In ethanol free of CO₂ (0.1 MPa), the SAFC (R = 2800, 0.1 MPa) is more emissive (Fig. 5B and 6B) with more dense spherical micelles (Fig. 4a) than that with R = 1400 and 0.1 MPa (Fig. 3a), which may imply that there are more amount of H-bonds (a) in R = 2800 series since the hydrogen bonds between DR1 and ethanol cannot lead to any aggregation of copolymers, and the emission contribution from H-bonds (a) seems greater than that from H-bonds (b). We think that the confinement difference on the DR1 unit resulted from different types of hydrogen bonds would help explain the observed SAFC fluorescence in this work. Perhaps for the hydrogen bonds between solvent and DR1, the ethanol molecules may not considerably hamper the dihedral twist of DR1 along –N=N– bond, which makes the H-bonds (b) less restrictive and thus contributes less to the SAFC emission. In contrast, for the hydrogen bonds (H-bond (a)) between PS-b-P4VP and DR1, the P4VP chains attached to the terminal –OH groups of DR1 could appreciably restrain the motion of longer arms in the dye unit and therefore offers greater contribution to the FL enhancement of SAFC. In other words, the H-bond (a) might be the determinant, the more amount of H-bond (a), the more emissive the corresponding SAFC.

According to the results observed in dispersibility (Fig. 2) and the pressure dependence of maximum FL intensity of SAFC (Fig. 7B), it is reasonable to suppose the threshold pressure (P_T) around 5.60 MPa (R = 2800) or 6.30 MPa (R = 1400) in our reaction mixture, which corresponds to the maximum degree of copolymers aggregation. As pressure increases, more CO₂ can be dissolved in ethanol, and the amount of H-bond (b) decreases owing to the reduction in the number of ethanol molecules surrounding DR1. It might be that the H-bond (b) helps stabilize the excited state of DR1, the less H-bond (b) in the system, the relatively higher energy of the excited state of DR1, and then the more blue shift the corresponding emission spectra. Consequently, at constant R value, the reduction in the amount of H-bond (b) results in a continuous emission spectral blue shift with increasing pressure below P_T (Fig. 7A). Whereas at pressures above P_T, the amount of H-bond (b) keeps constant, and the emission spectra show no further blue shift thereafter (Fig. 7A). In contrast, at comparable pressures, the pressure dependence of emission spectra shows smaller blue shift in R = 1400 series (Fig. 7A), which may be owing to the relatively more amount of H-bond (b) in this series compared with that in the high R value series. That is, there might be more H-bond (a) and less H-bond (b) in the R = 2800 series.

Fig. 7 (A) Pressure dependence of the maximum emission wavelength (λ_max) of the SAFC. (B) Pressure dependence of the maximum FL intensity (I) of SAFC. I₀ = the maximum FL intensity of the pristine mixture. SAFC (with R = 1400 or R = 2800) were obtained from the mixture of DR1 (1.8 × 10⁻⁵ or 9.0 × 10⁻⁶ mol L⁻¹) and PS-b-P4VP (5 mg mL⁻¹) in CO₂-expanded ethanol (353.2 K/varied pressure for 2 h). λ_ex = 490 nm.

At R = 1400, the FTIR intensity of pyridinium salt peak (1670 cm⁻¹) of the SAFC at 6.28 MPa (Fig. 1c) is stronger than that at 0.1 MPa (Fig. 1b), which means that the amount of H-bond (a) increases with pressure rise below P_T. Although the amount of H-bond (b) keeps decreasing with increasing pressure below P_T, the SAFC still exhibits emission enhancement (Fig. 7B), which also indicates that in CXLs, the amount of H-bonds (a) is increasing with pressure rise at the expense of those H-bond (b). Simultaneously, as CXLs pressure increases, the P4VP chains become more folded due to the worse solubility resulted from the anti-solvent effect of CO₂, and thus the continuous increase in the SAFC emission is partially owing to the more restrained trans-to-cis isomerization of azo group of DR1 coupled with P4VP. While at pressures above P_T, the diblock copolymers start partially precipitated (Fig. 2c), leading to a gradual reduction in the effective amount of 4VP units coupled with DR1, and therefore the corresponding SAFC is less aggregated even if the P4VP chains become even more coiled (Fig. 3e and 4e). Accordingly, the number of hydrogen bonds between DR1 and P4VP (H-bond (a)) decreases with increasing pressure when pressure is above P_T, and the FL intensity keeps declined consequently (Fig. 7B). In brief, the H-bond (a) offers greater contribution to SAFC emission than that of H-bond (b), and confinement effect and H-bond (a) might be the major factors responsible for the pressure dependence of emission.

3.3.2 The CXLs-induced changes in solvated state and mobility of P4VP chains, and their influences on the SAFC morphology and emission. Firstly we will focus on the evolution of the SAFC structure with CXLs pressure, suppose the major factors that may influence the morphology of the SAFC aggregates, and present the possible reasons for their effectiveness. At R = 1400, the structure of SAFC micelles (Fig. 3) evolves from smaller-sized and less compact spheres (0.1 MPa), through larger and compact spheres (4.15 MPa) and medium-extent aggregates of interconnected rods (5.50 MPa), to higher-extent clusters of bicontinuous rods (6.28 MPa), and then to less aggregated rods plus spheres (8.33 MPa) with increasing CXLs pressure. While at R = 2800, the morphology of SAFC aggregates (Fig. 4) transits from spheres (0.1 MPa), through short branched rods (3.05 MPa) and moderate-degree aggregates of bicontinuous rods (4.78 MPa), to more aggregated bicontinuous rods (5.58 MPa), and then to less clustered spheres plus rods (8.33 MPa) with pressure rise. So it is obvious that the SAFC structure is strongly pressure tunable from the TEM images shown in Fig. 3 and 4. The observed pressure dependence of morphological transition of the SAFC might be determined by two major factors, the CXLs-induced changes in solvated state of copolymers, as well as the amount of the hydrogen bonds between DR1 and P4VP block. The reasons could be explained as follows. As ethanol is P4VP selective, the observed SAFC spherical micelles in ethanol (Fig. 3a and 4a) are composed of a spherical PS cores surrounded by P4VP coronal chains, along with DR1 units being located in the coronas via hydrogen bonding with P4VP blocks. The spherical aggregates are usually the first clusters to form, which are generally regarded as the starting morphology for other upcoming micelles.²³ Similarly, rods consist of cylindrical PS cores and P4VP–DR1 coronas surrounding the cores, which stands for the next step in the arrangement of copolymer chains in response to the change in block or solvent media parameters. Furthermore, the bicontinuous rods at higher CXLs pressures are composed of three-dimensional networks of interconnected branched rods. In a mean-field theoretical approach, Izzo and Marques demonstrated that a morphological evolution from spheres to rods might occur as the corona-forming block length decreases.²⁴ And it is exactly the case that the corona-forming P4VP chains are getting more folded (or the length of corona portion is becoming decreased) as the CXLs pressure increases. Based on the dispersibility behavior shown in Fig. 2, it appears that the solubility of copolymers becomes worse with increasing CXLs pressure, and the solvated state of P4VP chains is getting more folded or clustered (Fig. 3 and 4) with pressure rise at pressures lower than P_T. Accordingly, for the R = 1400 series, initially we obtained the smaller-sized and less compact SAFC spheres in ethanol (Fig. 3a), then the SAFC spheres at 4.15 MPa become larger and compact (Fig. 3b) due to the enhanced folding degree of P4VP chains induced by the anti-solvent effect of CXLs. As the further shrinking of the corona-forming P4VP chains at 5.50 MPa, the SAFC formed a medium-extent aggregates of interconnected rods (Fig. 3c). Then the bicontinuous SAFC rods with the highest-extent clusters (Fig. 3d) were obtained at the threshold pressure of 6.28 MPa, along with a continuously reducing contact area between P4VP and the CXLs. Yet, the SAFC rod-like micelles obtained at 8.33 MPa become less aggregated (Fig. 3e) despite the steep further shrinking of P4VP coronal chains. Here the morphological transition with pressure shows a non-monotonic trend, which reaches the highest-extent aggregates at the threshold pressure (P_T), and then follows by a less aggregated structure when pressure is higher than P_T. A more plausible explanation would be, the copolymers start precipitated at pressures above P_T (Fig. 2c), along with a drop in the effective amount of 4VP units bonded with DR1, and consequently the SAFC becomes less clustered despite the even more coiled P4VP chains (Fig. 3e). Besides, the morphological transition with increasing CXLs pressure also reduces the total interfacial free energy due to the decreasing contact area between P4VP and the solvent.²³ And the above analysis on the pressure dependence of morphological transition also works for the R = 2800 series. However at comparable pressures and the same copolymer concentrations for the two series, which guarantees the identical solvated state of copolymer chains, the SAFC aggregates become more dense or higher clustering degree in the higher R value series (Fig. 3 and 4), highly likely owing to the more amount of H-bond (a) in the R = 2800 series as stated in Section 3.3.1. That is, the more amount of H-bond (a), the higher aggregation degree the corresponding SAFC. Briefly, the CXLs-induced changes in solvated state of copolymers may offer major control on both the SAFC morphology and aggregation degree of micelles at constant R value, whereas the amount of H-bond (a) helps effectively regulate the clustering degree of the SAFC micelles at comparable pressures.

It seems the more folded or clustered the P4VP chains, the stronger the confinement effect of DR1, and the more emissive of the corresponding SAFC. At pressures below P_T, as shown in Fig. 5 (R = 1400) and Fig. 6 (R = 2800), the FL of SAFC is getting enhanced with increasing CXLs pressure at R values of 1400 and 2800, respectively, whereas the emission keeps decreased at pressures above P_T. For the R = 2800 series, the structure of SAFC micelles (Fig. 4) changes from spheres (0.1 MPa), through moderately aggregated bicontinuous rods (4.78 MPa), to more aggregated bicontinuous rods (5.58 MPa), and then to less clustered spheres plus rods (8.33 MPa) with increasing pressure, accompanied by the SAFC emission being 1.1, 2.7, 3.4 and 2.0-fold enhanced (Fig. 7B) than the pristine mixture at 0.1, 4.78, 5.58 and 8.32 MPa, respectively. And at R = 1400 series, the SAFC morphology (Fig. 3) transits from less compact spheres (0.1 MPa), through mildly aggregated interconnected rods (5.50 MPa), to more densely clustered bicontinuous rods (6.28 MPa), and then to less aggregated rods coexisting with spheres (8.33 MPa) with increasing CXLs pressure, along with the FL enhancement of SAFC being 1.1, 2.3, 2.6 and 1.7-fold higher (Fig. 7B) than the pristine mixture at 0.1, 5.50, 6.28 and 8.33 MPa, respectively. On the basis of the aforementioned results, we are able to deduce that the confinement of DR1 becomes the most effective in the highly aggregated bicontinuous rods (5.58 MPa) compared with the moderately (4.78 MPa) and less aggregated rods (3.05 MPa), as illustrated in Fig. 4 and 7B for R = 2800 series, that is, the higher the degree of aggregation of micelles, the more emissive the corresponding SAFC. Similarly the more compact the spherical micelles, the more stronger the confinement effect and then the greater enhancement of the SAFC emission, as was reflected in Fig. 3 (R = 1400) and Fig. 7B. Moreover, the confinement in the bicontinuous rods with more folded or clustered corona chains is more efficient in the emission improvement of SAFC than that in the spheres with longer corona chains, as shown in Fig. 3, 4 and 7B. Meanwhile, the decreased mobility of P4VP coronal chains with increasing pressure strengthens the confinement degree of DR1 and favors the SAFC emission effectively owing to the weakening non-radiative decay. When the pressure reached 5.58 MPa (R = 2800), the P4VP chains become further worse in mobility and even greater in folding degree, which makes the corresponding SAFC the most emissive due to the strongest confinement effect of DR1 in this case.

Finally, we may conclude that the hydrogen bond is the key factor to determine the SAFC emission in ethanol free of CO₂. However in CXLs, the CXLs anti-solvent induced solubility change in P4VP chains results in aggregation of copolymers, and in this case, both confinement and H-bond (a) play important roles in tuning the morphology and emission performance of the SAFC. And therefore at comparable pressures, the pressure dependence of SAFC emission is more striking for R = 2800 series owing to the more intensive confinement and more amount of H-bond (a) in this case.

4. Conclusions

The SAFC of PS-b-P4VP and DR1 was successfully prepared in CO₂ expanded ethanol, and the hydrogen bonds between PS-b-P4VP and DR1 were confirmed by FTIR data. TEM and emission results indicate that the structure of micelles and emission performance of SAFC could be easily controlled by the regulation of CXLs pressure. The solubility change in PS-b-P4VP induced by the anti-solvent effect of CXLs could effectively control the amount and type of the hydrogen bonds with DR1 units as well as the confinement extent of DR1, along with the morphology and FL performance of the SAFC. The conformation, aggregation degree and emission behaviors of the SAFC are strongly dependent on the CXLs pressure and the molar ratio of 4VP to DR1. The emission of SAFC reveals a non-monotonic feature against CXLs pressure, with emission increased with pressure rise below P_T and then decreased with increasing pressures above P_T. The maximum FL of SAFC at 5.58 MPa and R = 2800 is enhanced by 3.4-fold compared with that of the pristine mixture. In contrast, the maximum SAFC emission is increased by 2.6-fold at 6.28 MPa and R = 1400. At pressures below P_T, the amount of H-bond (a) is increasing with pressure rise at the expense of those H-bond (b), whereas the amount of both type of hydrogen bonds becomes decreased at pressures above P_T. Confinement effect and hydrogen bonds are two crucial factors responsible for the pressure dependence of emission. Moreover, the H-bond (a) offers more remarkable contribution to SAFC emission than the H-bond (b). This work explores the connection between polymer conformation and the optical response of the SAFC through the hydrogen bond interactions with the assistance of CXLs technology, which shows that CXLs method could indeed manipulate the morphology and emission performance of the SAFC, and at the meantime provides a potential pathway for the development of smart optical materials. For further understanding of the presented strategy in this work, copolymers and dye units with different steric effects such as PS-b-P2VP and DR1 derivatives will be employed in our future plan since the steric effects either from copolymers or dye units may have potential influence not only on the morphology of SAFC but also on the hydrogen bonding ability between PVP and the dye unit. In addition, our future work also involves other solvent candidates such as methanol or tertiary butanol with varying PVP selectivity and hydrogen bonding ability with DR1 derivatives, focusing on their possible influences on the SAFC structure and emission. Such studies will not only contribute to the basic knowledge of the role of intermolecular interactions on the optical response of polymeric materials, but also help uncover the possible emission mechanisms of the azobenzene-containing polymers.

Acknowledgements

This work was supported by funds from the National Natural Science Foundation of China (No. 21543009, 21073167, J1210060), and the Innovative Research Grant for Undergraduate Students of Zhengzhou University (2013–2016).

Notes and references

1. (a) J. del Barrio, E. Blasco, C. Toprakcioglu, A. Koutsioubas, O. A. Scherman, L. Oriol and C. Sanchez-Somolinos, Macromolecules, 2014, 47, 897–906; (b) J. del Barrio, E. Blasco, L. Oriol, R. Alcala and C. Sanchez-Somolinos, J. Polym. Sci., Part A: Polym. Chem., 2013, 51, 1716–1725.
2. (a) A. Priimagi, J. Vapaavuori, F. J. Rodriguez, C. F. J. Faul, M. T. Heino, O. Ikkala, M. Kauranen and M. Kaivola, Chem. Mater., 2008, 20, 6358–6363; (b) A. Priimagi, K. Lindfors, M. Kaivola and P. Rochon, ACS Appl. Mater. Interfaces, 2009, 1, 1183–1189; (c) A. Priimagi, Doctoral dissertation, Helsinki University of Technology, Finland, 2009.
3. (a) N. Ali and S. Y. Park, Langmuir, 2009, 25, 13426–13431; (b) J. Ruokolainen, R. Makinen, M. Torkkeli, T. Makela, R. Serimaa, G. Ten Brinke and O. Ikkala, Science, 1998, 280, 557–560; (c) J. Ruokolainen, G. Ten Brinke and O. Ikkala, Adv. Mater., 1999, 11, 777–780; (d) J. Ruokolainen, M. Saariaho, O. Ikkala, G. ten Brinke, E. L. Thomas, M. Torkkeli and R. Serimaa, Macromolecules, 1999, 32, 1152–1158.
4. (a) A. Sidorenko, I. Tokarev, S. Minko and M. Stamm, J. Am. Chem. Soc., 2003, 125, 12211–12216; (b) X. Yao, D. Chen and M. Jiang, Macromolecules, 2004, 37, 4211–4217; (c) H. Peng, D. Chen and M. Jiang, Langmuir, 2003, 19, 10989–10992; (d) S. C. Chen, S. W. Kuo and F. C. Chang, Langmuir, 2011, 27, 10197–10205; (e) X. Cheng, S. Huang, H. Li, N. An, Q. Wang and Y. Li, RSC Adv., 2016, 6, 4545–4551; (f) L. Zhang, H. Shen and A. Eisenberg, Macromolecules, 1997, 30, 1001–1011; (g) A. Harada and K. Kataoka, Macromolecules, 1995, 28, 5294–5299; (h) H. Duan, D. Chen, M. Jiang, W. Gan, S. Li, M. Wang and J. Gong, J. Am. Chem. Soc., 2001, 123, 12097–12098; (i) J. Zhu, H. Yu and W. Jiang, Macromolecules, 2005, 38, 7492–7501.
5. (a) S. W. Kuo and F. C. Chang, Macromolecules, 2001, 34, 4089–4097; (b) Y. S. Lu, C. Y. Yu, Y. C. Lin and S. W. Kuo, Soft Matter, 2016, 12, 2288–2300.
6. (a) S. W. Kuo and F. C. Chang, Macromolecules, 2001, 34, 5224–5228; (b) S.-C. Chen, S.-W. Kuo, U. S. Jeng, C.-J. Su and F.-C. Chang, Macromolecules, 2010, 43, 1083–1092; (c) Y. He, B. Zhu and Y. Inoue, Prog. Polym. Sci., 2004, 29, 1021–1051; (d) S.-W. Kuo, P.-H. Tung, C.-L. Lai, K.-U. Jeong and F.-C. Chang, Macromol. Rapid Commun., 2008, 29, 229–233; (e) C. H. Hsu, S. W. Kuo, J. K. Chen, F. H. Ko, C. S. Liao and F. C. Chang, Langmuir, 2008, 24, 7727–7734.
7. C. L. Zhang, J. L. Hu, D. Chen, Y. Zhu, Y. Fan and Y. Liu, Polymer, 2012, 53, 4718–4726.
8. (a) J. I. Clodt, S. Rangou, A. Schroder, K. Buhr, J. Hahn, A. Jung, V. Filiz and V. Abetz, Macromol. Rapid Commun., 2013, 34, 190–194; (b) J. I. Clodt, B. Bajer, K. Buhr, J. Hahn, V. Filiz and V. Abetz, J. Membr. Sci., 2015, 495, 334–340.
9. A. Natansohn and P. Rochon, Chem. Rev., 2002, 102, 4139–4175.
10. (a) C. Pietsch, R. Hoogenboom and U. S. Schubert, Angew. Chem., 2009, 121, 5763–5766; (b) L. Liu, W. Li, K. Liu, J. Yan, G. Hu and A. Zhang, Macromolecules, 2011, 44, 8614–8621.
11. C. Toro, A. Thibert, L. De Boni, A. E. Masunov and F. E. Hernandez, J. Phys. Chem. B, 2008, 112, 929–937.
12. (a) Z. Tang, M. S. Johal, P. Scudder, N. Caculitan, R. J. Magyar, S. Tretiak and H. Wang, Thin Solid Films, 2007, 516, 58–66; (b) C. A. Hunter and J. K. M. Sanders, J. Am. Chem. Soc., 1990, 112, 5525–5534; (c) T. Kunitake, Colloids Surf., 1986, 19, 225–236.
13. N. Hameed, N. V. Salim, J. Parameswaranpillai and B. L. Fox, Mater. Lett., 2015, 147, 92–96.
14. H. Liu, C. Q. Gao, Z. L. Ding and W. Q. Zhang, Macromol. Chem. Phys., 2016, 217, 467–476.
15. (a) F. Liu and A. Eisenberg, J. Am. Chem. Soc., 2003, 125, 15059–15064; (b) Z. Geng, Z. K. Cheng, Y. T. Zhu and W. Jiang, J. Phys. Chem. B, 2016, 120, 5527–5533; (c) Z. Yi, P. B. Zhang, C. J. Liu and L. P. Zhu, Macromolecules, 2016, 49, 3343–3351.
16. (a) M. Anand, S. You, K. M. Hurst, S. R. Saunders, C. L. Kitchens, W. R. Ashurst and C. B. Roberts, Ind. Eng. Chem. Res., 2008, 47, 553–559; (b) F. Dehghani and N. R. Foster, Curr. Opin. Solid State Mater. Sci., 2003, 7, 363–369.
17. (a) Y. Liu, Y. Wei, R. Qin, L. Zhang, L. Yang, H. Li and X. Li, eXPRESS Polym. Lett., 2011, 5, 60–72; (b) M. Su, Y. Wei, L. Qi, Y. Pang, Y. Guo, P. Li, J. Li and H. Li, eXPRESS Polym. Lett., 2011, 5, 1102–1112; (c) W. Xue, L. Qi, X. Li, S. Huang, H. Li, X. Guan, G. Bai and L.-E. Liu, Chem. Eng. J., 2012, 209, 118–125.
18. M. Sutter, S. Oliveira, N. N. Sanders, B. Lucas, A. Hoek, M. A. Hink, A. J. W. G. Visser, S. C. Smedt, W. E. Hennink and W. Jiskoot, J. Fluoresc., 2007, 17, 181–192.
19. J. Chiefari, Y. K. B. Chong, F. Ercole, J. Krstina, J. Jeffery, T. P. T. Le, R. T. A. Mayadunne, G. F. Meijs, C. L. Moad, G. Moad, E. Rizzardo and S. H. Thang, Macromolecules, 1998, 31, 5559–5562.
20. M. J. Hangauer and C. R. Bertozzi, Angew. Chem., Int. Ed., 2008, 47, 2394–2397.
21. Y. Senay and B. Serdar, Struct. Chem., 2009, 20, 423–434.
22. P. G. Jessop and B. Subramaniam, Chem. Rev., 2007, 107, 2666–2694.
23. Y. Mai and A. Eisenberg, Chem. Soc. Rev., 2012, 41, 5969–5985.
24. D. Izzo and C. M. Marques, Macromolecules, 1997, 30, 6544–6549.

---