Functionalization of single-walled carbon nanotubes with thermo-responsive poly(N-isopropylacrylamide): effect of the polymer architecture

Abstract

The functionalization of single-walled carbon nanotubes (SWNTs) using non-covalently interacting stimuli-responsive polymers in aqueous media has opened up a new area of study for SWNTs in biology and medicine. We prepared thermo-responsive pyrene-labelled four-arm star-shaped poly(N-isopropylacrylamide) (p-SPNIPAM) and linear p-PNIPAM by atom transfer radical polymerization and then used these two polymers to modify SWNTs. The two polymers could be attached onto the sidewall of nanotubes by π–π stacking between the pyrene group and the SWNTs, but p-SPNIPAM imparted a better solubility to the SWNTs than p-PNIPAM. In addition, when the temperature of the dispersion was above the cloudy point of the polymer solution, the p-SPNIPAM/SWNT hybrids formed compact large diameter bundles, whereas the p-PNIPAM/SWNTs formed small loose bundles floating in water. The aggregated SWNTs hybrids were re-dispersed below the cloudy point. Thus SWNT hybrids are not only dispersed in water, but also undergo reversible dispersion and aggregation by tuning temperature, making them attractive in biomedical or sensory applications. The architecture of the thermo-sensitive polymers also affects the dispersibility of SWNTS and their “smart” behaviour.

---

Introduction

Single-wall carbon nanotubes (SWNTs)^1,2 are promising functional nanomaterials in a wide range of applications, including biocompatible transportation,³ drug delivery⁴ and cancer therapy,⁵ as a result of their extraordinary electrical,⁶ optical⁷ and mechanical properties.⁸ The strong bundling/aggregation behaviour of pristine SWNTs, however,^9,10 not only causes poor solubility, but also impedes their application in situations where individual tubes or small-sized bundles are required to obtain the best performance.^11–13 When used as switchable sensors in biological or medicinal chemistry, it is necessary to control the dispersion or aggregation of SWNTs in the solvent using external stimuli.^14–17 The non-covalent functionalization of SWNTs with stimuli-responsive polymers can be used to address these issues, as this approach not only tunes the solubility, but, more importantly, preserves the intrinsic structure and properties of the SWNTs.^18–20

A number of stimuli-responsive polymers triggered by heat,^21–24 pH,²⁴ redox,²⁵ light^26–29 and CO₂ ³⁰ have been used to non-covalently wrap nanotubes for exfoliation to produce “intelligent” SWNTs. Among these methods, thermo-responsive polymers, the physicochemical properties of which change abruptly on minor stimuli from changes in temperature, have been widely investigated for the modification of SWNTs. The nanohybrids obtained show promising applications in biomedicine and intelligent materials. For example, poly(N-isopropylacrylamide) (PNIPAM), a classic thermo-responsive homopolymer with a lower critical solution temperature (LCST) of 32 °C (close to that of the human body temperature),^22,34 has been directly applied to the dispersion and tuning of the macrostructure of SWNTs.²¹ Unfortunately, PNIPAM has only a modest physical affinity for SWNTs, which is not sufficient for the effective stabilization of aqueous SWNT suspensions.^22,23 To improve the physical interaction of PNIPAM with SWNTs without affecting its thermo-responsive behaviour, Grunlan and coworkers²² incorporated pyrene groups into PNIPAM, which significantly improved its affinity for SWNTs because of the formation of π–π stacking between pyrene and the nanotubes.^18,20 Aqueous SWNT suspensions stabilized with pyrene-containing PNIPAM are stable for weeks with no visible sedimentation. These SWNT hybrids also show a temperature-sensitive response in water. To develop new temperature-responsive polymeric dispersants, the same researchers²³ used pyrene-containing poly(N-cyclopropyl-acrylamide), an analogue of PNIPAM, to functionalize SWNTs to tailor the dispersibility of SWNTs through a thermal trigger.

However, the overwhelming majority of thermo-responsive polymers reported so far for modifying SWNTs are confined to linear polymers. To the best of our knowledge, there have been few reports of the effect of the molecular structure of thermo-responsive polymers on the dispersion of SWNTs. In the past few years, as a result of developments in polymer synthesis, polymers with various architectures have been produced and been investigated.^31,32 For example, star-shaped polymers, with a 3D, hyper-branched structure with linear arms of the same or different molecular weight (MW) emanating from a central core, show a smaller hydrodynamic radius and lower solution viscosity than linear polymers of the same MW and composition.^32,33 It is anticipated that the star-shaped thermo-responsive polymer architecture may affect the properties of the polymer-functionalized SWNTs, including the dispersibility and responsiveness. Thus discerning the difference caused by the polymer architecture is particularly helpful for the design of polymeric dispersants.

In this paper, ​we used an atom transfer radical polymerization (ATRP) technique to synthesize a thermo-responsive four-armed star architecture poly(N-isopropylacrylamide) (p-SPNIPAM) using a pyrene-containing tetrafunctional compound as the initiator (Scheme 1a). PNIPAM is a widely studied polymer for use in biological applications³⁴ and the pyrene group can be adsorbed onto the sidewall of nanotubes via π–π interactions.^35,36 To unravel the effect of the polymer architecture on the dispersal of SWNTs, the linear p-PNIPAM counterpart prepared by the same method was also examined (Scheme 1b). Both polymers were used to functionalize SWNTs and then the dispersibility and responsiveness of the SWNT hybrids in water were studied. This work may help in the understanding of the effect of polymer structure on the dispersion and responsiveness of SWNTS in solvents, which may offer a reference for designing dispersants for nanomaterials. The SWNT hybrids underwent dispersion and aggregation with a change in temperature, making them attractive in biomedical or sensory applications.

Scheme 1 Synthetic pathways for preparing (a) p-SPNIPAM and (b) p-PNIPAM.

Experimental

Materials

SWNTs (purity > 90%, OD 1–2 nm, length 5–30 μm) prepared by a chemical vapour deposition procedure were kindly provided by Timesnano (Chengdu, China) and used without further treatment. N-Isopropylacrylamide (NIPAM, 97%, Tokyo Kasei Kagyo Co.) was purified by recrystallization from a mixture of benzene and n-hexane (2 : 3 v/v). Glycidol (96%), 2-bromoisobutyryl bromide (98%), copper(I) bromide (CuBr) (98%), 1-pyrenemethylamine hydrochloride (95%) and 1-pyrenemethanol (98%) were purchased from Aldrich and used without further purification. Tris(2-(dimethylamino)ethyl)amine (Me₆TREN) was prepared using a previously reported procedure.^37,38 Triethylamine (Aldrich, 99%) and dichloromethane (DCM) (Guanghua Chemicals Co. Ltd, 99.9%) were stirred overnight over CaH₂ and distilled under reduced pressure prior to use. All other reagents were purchased from Shanghai Chemical Reagent Co. Ltd and were used as received unless specified otherwise.

Characterization

¹H NMR spectra were recorded at 25 °C on a Bruker AV300 NMR spectrometer at 300 MHz. Chemical shifts (δ) are reported in parts per million (ppm) with reference to the internal standard protons of tetramethylsilane (TMS).

The MW and MW distribution of the polymers were determined by gel permeation chromatography (GPC) using a Waters Model 1515 pump, a Model 2410 refractive index detector and an OH-pak KB-803 column operated at an oven temperature of 25 °C. THF was used as the mobile phase at a flow rate of 0.8 mL min⁻¹ and the column was calibrated using monodispersed polystyrene as a standard.

UV-visible-NIR measurements were recorded out on a computer-manipulated dual-beam spectrometer (UV-visible 4100, Hitashi, Japan) operated at a resolution of 1 nm at 25 °C. Based on previously reported procedures,²⁸ the absorbance was recorded in the wavelength range 1600–400 nm. However, the wavelength range 400–1300 nm was finally selected because the absorbance intensity of the UV-visible-NIR spectra of our suspension at wavelengths > 1300 nm was beyond the measurement range of the instrument. The SWNT hybrid suspension was diluted 10 times prior to the UV-visible-NIR measurements.

Fluorescence spectrometry was performed on a Cary Eclipse spectrometer (Varian, USA). The fluorescence emission spectra were scanned between 350 and 600 nm. The excitation wavelength was set at 335 nm and the excitation and emission slit widths were set at 10 and 1.5 nm, respectively. The suspension of SWNT/polymer hybrids was diluted 10 times prior to the fluorescence experiments. The ratio of excimer emission intensity (I_E) to monomer emission intensity (I_M) was estimated from the ratio of the peak heights at 460.06 and 375.23 nm.

Transmission electron microscopy (TEM) images were obtained using a Hitachi H600 electron microscope operated at an acceleration voltage of 75 kV and the samples for TEM measurements were prepared by placing one drop of sample on a copper grid coated with carbon. To observe the dispersed SWNTs more clearly, the samples were stained with phosphotungstic acid for negative staining TEM observations.

Thermogravimetry analysis (TGA) was conducted on a 299-F1 thermal analysis system (NETZSCH, Germany). Samples were heated in a flow of N₂ (50 mL min⁻¹) from room temperature to 600 °C at a heating rate of 10 °C min⁻¹. To prepare the samples for TGA analysis, 10 mL of a SWNT/polymer hybrid suspension were filtered through a PTFE microporous membrane (220 nm). The hybrids left on the membrane were washed repeatedly with water and then freeze-dried for 24 h. The original SWNTs, p-SPNIPAM and p-PNIPAM were used without further treatment in the TGA measurements.

UV-visible spectra were obtained at 25 °C on a UNICO UV-Vis 4802 double-beam spectrophotometer (Shanghai, China).

Preparation of tetrafunctional initiator

1-Pyrenemethylamine hydrochloride (1.00 g) was washed with DCM (100 mL) and 2 M NaOH (100 mL) to extract 1-pyrenemethylamine into the organic layer, which was then dried under anhydrous Na₂SO₄. After the removal of the solvent, 1-pyrenemethylamine (0.85 g, yield 98.8%) was obtained as a white solid for used in further reactions.

A solution of 1-pyrenemethylamine (0.80 g, 0.45 mmol) in 30 mL of THF was added dropwise to a solution of glycidol (4.08 g, 3.60 mmol) in 10 mL of THF over a period of 30 min under N₂ protection at 0 °C. The reaction mixture was then stirred for 24 h at 28 °C. After removal of the solvent, the crude product was purified by silica chromatography using CH₂Cl₂/CH₃OH (12 : 1) as the eluent to give compound 1 at a yield of 77.9%. ¹H NMR (DMSO-d₆ + D₂O) (Fig. S1, ESI†), δ/ppm: 8.42–7.92 (m, 9H), 4.37 (m, 2H), 3.63 (m, 2H), 3.27–3.13 (m, 4H), 2.66–2.52 (m, 4H). ESI-HRMS (Fig. S2, ESI†): calcd: 308.1862 (1·H⁺). Found: m/z = 308.1861.

Compound 1 (0.90 g, 2.37 mmol) and triethylamine (3.84 g, 0.38 mol) were dissolved in 20 mL of THF and cooled in an ice bath. After 15 min, 2-bromoisobutyryl bromide (8.70 g, 0.38 mol) dissolved in 20 mL of THF was added dropwise over 30 min. The reaction mixture was stirred at 0 °C for 2 h and then at room temperature for 24 h. The insoluble triethylamine hydrobromide salt was filtered, THF was evaporated by rotary distillation and the crude product obtained was dissolved in DCM followed by washing with 2 M NaOH (50 mL). After removing the DCM, the product was further purified by silica chromatography using petroleum ether/ethyl acetate (8 : 1) as the eluent to give the compound for use as the final tetrafunctional initiator 2 at a yield of 52.3%. ¹H NMR (CDCl₃) (Fig. S3, ESI†), δ/ppm: 8.42–7.92 (m, 9H), 5.30 (m, 2H), 4.51–4.39 (m, 4H), 4.11–4.07 (m, 2H), 3.02–2.95 (m, 4H), 2.66–1.96 (m, 24H). ESI-HRMS (Fig. S4, ESI†): calcd: 975.9916 (2·H⁺). Found: m/z = 975.9928.

Preparation of monofunctional initiator

The monofunctional initiator 3 was synthesized by esterification between 1-pyrenemethanol and 2-bromoisobutyryl bromide. The synthesis procedure was similar to that for compound 2 and the yield was 66.7%. ¹H NMR (CDCl₃) (Fig. S5, ESI†), δ/ppm: 8.42–7.92 (m, 9H), 5.92 (m, 2H), 1.93 (s, 6H). ESI-HRMS (Fig. S6, ESI†): calcd: 403.0310 (3·Na⁺). Found: m/z = 403.0313.

Preparation of star p-SPNIPAM

The tetrafunctional initiator 2 (0.400 g, 0.408 mmol), NIPAM (4.608 g, 40.8 mmol), Me₆TREN (0.376 g, 1.632 mmol) and solvent (8 mL, DMF/H₂O = 5 : 1) were added to a reaction tube and the solution was degassed by three freeze–thaw cycles. Then CuBr (0.233 g, 1.632 mmol) was introduced into the tube under N₂ protection. The reaction system was again degassed by one freeze–thaw cycle. The reaction was then carried out at 0 °C under a slightly positive pressure of N₂. The reaction solution became green and more viscous as polymerization proceeded. The polymerization was stopped after about 24 h by immersing the reaction tube into liquid nitrogen for about 5 min. The mixture was then diluted with THF and passed through a neutral Al₂O₃ column to remove the ATRP catalyst. After the solvent had been removed, the product was precipitated from excess ether and dried under vacuum. To remove impurities such as the residual monomer and low molecular weight product, the dried solid product was dissolved in distilled water and transferred into a dialysis tube (MW cut-off 3500 Da) for dialysis against distilled water for 1 week. Next, the product was concentrated in a rotary evaporator and dried under vacuum. A white solid product of p-SPNIPAM was obtained (4.2 g) at a yield of 84%. ¹H NMR (D₂O) (Fig. S7, ESI†), δ/ppm: 8.42–7.92 (br m), 7.92–7.4 (br s), 3.90 (br s), 2.20–1.40 (br m), 1.19 (br s). FTIR (in KBr) (Fig. S8, ESI†): 3298 cm⁻¹ (NH, amide valence), 2970 cm⁻¹ (C–H, valence), 1642 cm⁻¹ (C=O amide band I), 1540 cm⁻¹ (C=O amide band II).

Preparation of linear p-PNIPAM

The synthesis process of p-PNIPAM was similar to that of p-SPNIPAM, except for the ratio of the mixed solvent (DMF : H₂O 9 : 1). The yield was 1.1 g (64.2%). ¹H NMR (D₂O) (Fig. S9†) δ/ppm: 8.42–7.92 (br m), 7.92–7.4 (br s), 3.90 (br s), 2.20–1.40, (br m), 1.19 (br s). FTIR (in KBr) (Fig. S8, ESI†): 3298 cm⁻¹ (NH, amide valence), 2970 cm⁻¹ (C–H, valence), 1642 cm⁻¹ (C=O amide band I), 1540 cm⁻¹ (C=O amide band II).

Preparation of linear PNIPAM without a pyrene group

The synthesis of linear PNIPAM without a pyrene group was similar to that of p-SPNIPAM, except the initiator was ethyl 2-bromo-2-methylpropanoate. The yield was 1.1 g (64.2%). ¹H NMR (D₂O) (Fig. S10†), δ/ppm: 7.92–7.4 (br s), 3.90 (br s), 2.20–1.40 (br m), 1.19 (br s).

Preparation of SWNT/polymer hybrid dispersion

The dispersion of SWNTs was prepared following a previously reported procedure.^22,28 A 20 mg mass of SWNTs was added into 20 mL of the prepared polymer solution (0.83 mM), followed by sonication (KQ-100, Kunshan Ultrasound Instrument Co., China) for 30 min (100 W, 40 kHz) at room temperature. The resultant suspension was then centrifuged at 3500 rpm (Shanghai Medical Instruments Co., Ltd, China) for 10 min to give a homogeneous dispersion of SWNTs.

Results and discussion

Synthesis and characterization

ATRP, a powerful and versatile controlled radical polymerization technique, enables precise control over the MW, MW distribution and functionality of polymers.^39–41 In addition, it can be carried out under benign conditions and is tolerant of most functional groups,^39–41 which is desirable for attaching pyrene groups onto the end of polymer chains. Thus the ATRP process was used here to prepare both star-shaped and linear NIPAM-based polymers.

Scheme 1a shows the general synthesis route used for the preparation of p-SPNIPAM star-shaped polymers. The first step was the preparation of the tetrafunctional ATRP initiator using 1-pyrenemethylamine as a precursor. The primary amine group of 1-pyrenemethylamine was reacted with excess glycidol ([glycidol]/[NH₂] = 8.0) to give the tetrahydroxyl compound 1. Further reaction between the terminal primary amine group and glycidol was evidenced by the ¹H NMR spectra. Fig. 1a shows that the signals at 3.1–3.8 ppm (e + d) are the methylene protons next to the four terminal hydroxyl groups, which showed that the starting material was transformed into tetrahydroxyl compound 1. Following the esterification of the hydroxyl groups of 1 with excess 2-bromoisobutyryl bromide, the ATRP tetrafunctional initiator 2 was obtained. By comparing its ¹H NMR spectrum with that of compound 1, the signals at 3.1–3.8 ppm were seen to be downshifted to 4.2–4.6 ppm (e + d, Fig. 1b) and were attributed to the ester methylene protons. New signals were observed at 1.6–1.9 ppm (f, Fig. 1b) that resulted from the methyl groups of the four terminal bromoisobutyryl groups of 2, which confirmed the successful synthesis of the tetrafunctional initiator 2.

Fig. 1 ¹H NMR spectra of (a) compound 1, (b) compound 2 and (c) p-SPNIPAM, as shown in Scheme 1a.

The four-armed star p-SPNIPAM was obtained by polymerizing NIPAM via ATRP using 2 as an initiator. To avoid deactivation of the copper catalyst through complexation with the amide groups in PNIPAM, Me₆TREN was selected as the ligand. Fig. 1c shows the ¹H NMR spectrum of p-SPNIPAM. The characteristic signals of both NIPAM and pyrene can be clearly identified; the signals at 3.9 (c) and 1.1–2.2 ppm (d + b) were assigned to the –NHCH(CH₃)₂ and –CH₂–CH and –NHCH(CH₃)₂ groups in PNIPAM and that at 7.90–8.40 ppm (a) was ascribed to –CH in pyrene, suggesting that the pyrene group was introduced into PNIPAM. Based on the integration of the signals at 7.90–8.40 and 3.91 ppm, the polymerization degree (DP) and M_n,nmr of p-SPNIPAM can be calculated from eqn (1) and (2), respectively.

M_n = DP × M_NIPAM + M_initiator (2)

where A_{3.91 ppm} and A_{7.90–8.40 ppm} are the integrated areas of the protons in pyrene and the –CH– of NIPAM; M_NIPAM and M_initiator are the MWs of NIPAM and the initiator, respectively. DP and M_n,nmr were close to the theoretical values (Table 1). The DP of each PNIPAM branch was calculated to be 24.5.

Table 1 Characteristics of p-SPNIPAM, p-PNIPAM and NIPAM

Polymer	[M]o/[I]o^a	DPn^b	Mn,theo^c (g mol^−1)	Mn,nmr^b (g mol^−1)	Mn,GPC^d (g mol^−1)	Mw/Mn^d	Cloudy point^e (°C)
a [M]o/[I]o refers to the molar feed ratio of NIPAM : initiator.b Estimated from ^1H NMR spectra.c Expected number-average molecular weight from polymerization stoichiometry.d Determined by GPC in THF at 25 °C (PSt calibration).e Determined by UV-visible spectrophotometry in 10 mg mL^−1 aqueous solutions.
p-SPNIPAM	100	98	12 270	12 049	9520	1.25	29.8
p-PNIPAM	100	95	11 680	11 115	7930	1.26	31.0
PNIPAM	100	—	10 700	—	7480	1.43	33.1

---

The preparation of linear p-PNIPAM was similar to that of p-SPNIPAM (Scheme 1b). Based on the ¹H NMR data, the DP and M_n,nmr values of p-PNIPAM were also calculated from eqn (1) and (2) and the results, which are close to the theoretical values, are given in Table 1.

GPC was used to determine the M_n,GPC and polydispersity index (PDI) of the polymers. The PDI was low at around 1.25, which is consistent with the characteristics of ATRP polymerization.^39,40 In addition, the GPC elution peak of the polymers was monomodal and fairly symmetrical (Fig. 2), indicating that well-defined polymers were successfully obtained.⁴¹ The GPC curve of p-SPNIPAM suggested that the major products were the desired four-arm star polymer because it does not show a tail at the lower MW side ascribed to polymer chains containing four, three or two PNIPAM branches caused by intramolecular irreversible termination reactions or inefficient initiation during polymerization.⁴² The deviations that were seen between M_n,GPC, M_n,nmr and M_n,theo, which may be attributed to the GPC values, are the MWs relative to the narrow polydispersity polystyrene standards. Thus M_n,nmr is used in further discussions.

Fig. 2 GPC traces of p-SPNIPAM and p-PNIPAM.

Thermo-responsive behaviour of polymers

PNIPAM is a typical thermo-responsive polymer, i.e. it is hydrophilic and is soluble in water below its LCST, but it becomes hydrophobic and insoluble at higher temperatures. To investigate whether the two polymers in which the pyrene groups were incorporated retained their LCST behaviour, the aqueous polymer solutions were first observed visually at different temperatures. Fig. 3 (inset images) shows that the polymer solutions were transparent at 25 °C, but became milky white at 36 °C, indicating that the two polymers have a phase transition temperature.

Fig. 3 Temperature dependence of optical transmittance at 600 nm obtained for aqueous solutions of p-SPNIPAM (10 mg mL⁻¹), p-PNIPAM (10 mg mL⁻¹) and PNIPAM (10 mg mL⁻¹). The inset images are p-SPNIPAM (a and d), p-PNIPAM (b and e) and PNIPAM (c and f) at 25 and 36 °C, respectively.

To determine the cloudy point accurately, the variation in transmittance with temperature was monitored using a UV-visible spectrophotometer. Fig. 3 shows that the polymer solutions had a sharp transition when the temperature was increased to a certain value (ca. 30 °C), reflecting the decrease in solubility of the polymers. The cloudy point is usually defined as the midpoint of the temperature–transmission curve.^43–46 The cloudy point of PNIPAM without a pyrene group was 33.1 °C, close to previously reported values.⁴⁷ However, the cloudy points of p-SPNIPAM and p-PNIPAM were 29.8 and 31.0 °C (Fig. 3), respectively, which were lower than that of PNIPAM without a pyrene group (33.1 °C). This can be ascribed to the incorporation of the hydrophobic pyrene group, which enhanced the attractive hydrophobic interactions of the polymers.^48,49 In addition, the cloudy point of star p-SPNIPAM was lower than that of linear p-PNIPAN, which was attributed to the hydrophobic core and the end-groups of the branches.⁴⁹

Dispersibility of SWNTs in water

To investigate the effect of the stimuli-responsive polymer architecture on the dispersibility of the nanotubes, two polymer solutions of equal concentration (0.83 mM) were used to disperse the SWNTs. The SWNT dispersions were obtained by sonication to mix the SWNTs with the polymers, followed by centrifugation. Fig. 4 shows the appearance of the aqueous dispersions of pristine and modified SWNTs. The original SWNTs were insoluble in water and sedimented at the bottom of the vial. However, the SWNTs modified by either p-SPNIPAM or p-PNIPAM were dispersed in water, forming a stable homogeneous dispersion with no observable settling over a period of 2 months.

Fig. 4 Appearance of SWNT/polymer hybrids in water after two months of storage: (a) original SWNTs; (b) p-SPNIPAM/SWNTs; and (c) p-PNIPAM/SWNTs.

To reveal the state of dispersion of the hybrids, UV-visible-NIR spectrometry, a technique common used to characterize dispersions of SWNTs,^50–52 was carried out. Fig. 5a shows that a strong absorbance was observed at 600–1300 nm, which is in the main range for observing the absorption of a favourable dispersion.²⁸ However, it should be noted that only the average spectroscopic characteristics of dispersed nanotubes were obtained because the SWNTs used in this work were always a mixture, in which the SWNT varied in both tube diameter and helicity, as evidenced in our previous report.²⁹ When the 750–870 nm region was enlarged (inset, Fig. 2b), sharp peaks attributable to the absorption features of the van Hove transition semiconducting tubes⁵² were found, indicative of the exfoliated SWNTs. It also should be noted that the absorbance of the p-SPNIPAM/SWNTs was higher than that of p-PNIPAM/SWNTs, which was attributed to the fact that there were more SWNTs in the p-SPNIPAM/SWNTs hybrids (detailed discussion given in the next section).

Fig. 5 (a) UV-visible-NIR spectra of p-SPNIPAM/SWNTs (0.083 mM p-SPNIPAM, 0.052 mg mL⁻¹ SWNTs) and p-PNIPAM/SWNTs (0.083 mM p-PNIPAM, 0.040 mg mL⁻¹ SWNTs) in water (25 °C). These samples were diluted ten times relative to their original dispersion. TEM images of (b) p-SPNIPAM/SWNTs and (c) p-PNIPAM/SWNTs dispersed in water.

In parallel with the spectral analysis, TEM observations were used to observe the SWNT hybrids suspended in water. To visualize these more clearly, a negative staining technique was used on the SWNT samples, which provides reverse-contrast negative electron optical images for the unstained component.^53–55 The white lines surrounded by grey areas in Fig. 5b and c are the SWNTs because the stain did not permeate the cavity of the tubes. The observed diameter of the tubes ranged from 3 to 6 nm, indicating that the SWNTs were converted into individual tubes and small bundles. A polymer layer was also observed (labelled in Fig. 5b and c), indicating that the polymers were attached to the tubes.

As mentioned above, the solution of p-SPNIPAM/SWNTs contained more SWNTs than the solution of p-PNIPAM/SWNTs. To confirm this, we diluted the two dispersions 10 times and found a much darker solution for the p-SPNIPAM/SWNTs (Fig. 6, inset), implying that the solution of p-SPNIPAM/SWNTs contained more SWNTs than the solution of p-PNIPAM/SWNTs.^56,57 To confirm this, the UV-visible absorption was measured because the absorbance is proportional to the concentration of exfoliated SWNTs.^56,57 Based on previously reported procedures,⁵⁷ the absorbance at 400 nm was compared with the concentration of the soluble SWNTs in these suspensions. Fig. 6 shows that the absorbance of the p-SPNIPAM/SWNTs (1.88) at 400 nm was much higher than that of the p-PNIPAM/SWNTs (0.98), indicating a higher concentration of soluble SWNTs in the p-SPNIPAM/SWNTs. After removing the insoluble products by centrifugation, it was found that the concentration of SWNTs in the final p-SPNIPAM and p-PNIPAM solutions were ca. 0.52 and 0.40 mg mL⁻¹, respectively. All these observations suggest that p-SPNIPAM has a higher dispersion efficiency.

Fig. 6 Absorption spectra of soluble SWNTs in water from p-SPNIPAM/SWNTs and p-PNIPAM/SWNTs. The inset images are (a) p-SPNIPAM/SWNTs and (b) p-PNIPAM/SWNTs.

To gain further insights into the interactive mechanism of polymers on the SWNTs, fluorescence spectrometry was used to compare the emission from the suspensions of the two polymers and their SWNT dispersions because the fluorescence spectrum is known to be very sensitive to the mutual interactions between the SWNTs and the dispersant.^34–36 Fig. 7 shows that p-SPNIPAM had three sharp bands and one broad band around 350–450 nm and 460 nm, respectively, which are the characteristic monomer and excimer emissions of pyrene.^58–60 In contrast, p-PNIPAM only showed emission from the monomer. After the SWNTs had dispersed in the polymer solution, the fluorescence emission was almost completely quenched as a result of electron transfer and/or energy transfer,⁶¹ indicative of π–π interactions between the pyrene group and the SWNTs.^34–36 Thus the spectroscopic evidence supports the proposal that p-SPNIPAM or p-PNIPAM is wrapped around the SWNTs by π–π stacking; in other words, the improved dispersibility of the SWNT/polymer hybrids in water is attributed to the increased physical interactions caused by the adsorbed polymer barrier layer, which overcomes the van der Waals forces responsible for the bundling of the nanotubes.

Fig. 7 Fluorescence emission spectrum of (a) p-SPNIPAM/SWNTs (0.083 mM p-SPNIPAM, 0.052 mg mL⁻¹ SWNTs) and (b) p-PNIPAM/SWNTs (0.083 mM p-PNIPAM, 0.040 mg mL⁻¹ SWNTs) in water at 25 °C.

The quenching efficiency can be calculated from fluorescence spectrometry. The quenching efficiency of the SWNTs towards p-SPNIPAM was 97.35% at 375.23 nm (Fig. 7a), clearly indicating a strong molecular interaction between the polymer and the SWNTs. In contrast, the p-PNIPAM/SWNT dispersion with the same molar ratio between the polymer and the SWNTs showed that only 85.17% of the fluorescence emission was quenched (Fig. 7b), suggesting a much stronger interaction between p-SPNIPAM and the SWNTs than that between p-PNIPAM and the SWNTs,⁶² i.e. p-SPNIPAM is inclined to be absorbed onto the nanotubes. The adsorption ability of the polymer is also evidenced by the TGA results, as the weight loss ratio increased with increasing amounts of polymer chains attached to the sidewalls of the nanotubes. From the TGA curve in Fig. 8, the weight percentages of p-SPNIPAM and p-PNIPAM in the SWNT hybrids are ca. 56.7 and 42.5 wt%, respectively. Based on the calculation described in our previous paper,³⁰ the number of p-SPNIPAM molecules attached to one nanotube was 1863, higher than that of p-PNIPAM (1268), which is consistent with the fluorescence spectra. The polymer content in the hybrids measured by TGA was lower than the actual content (polymer : SWNTs ca. 10 : 1 w/w), implying that there was an equilibrium between the bound and unbound polymers, or that unbound polymers are somehow needed for the SWNTs to be dispersed in water.⁶³ The results from the quenching of the fluorescence emission and TGA suggest that star p-SPNIPAM had higher adsorption ability than linear p-PNIPAM, leading to a higher coverage of p-SPNIPAM on the SWNTs. The higher coverage of the polymer on the nanotube sidewalls will increase the steric hindrance and prevent agglomeration, resulting in a greater solubility of the SWNTs in water.^56,64 With this in mind, it is reasonable to conclude that the higher concentration of SWNTs in p-SPNIPAM/SWNTs should be attributed to the higher adsorption ability of p-SPNIPAM. Pan et al.⁷⁰ also found that star polymers have a strong dispersibility for carbon nanotubes when they used six-armed star poly(L-lactic acid) with a triphenylene core to modify multiwalled carbon nanotubes.

Fig. 8 TGA traces for original untreated SWNTs, p-SPNIPAM/SWNT hybrids, p-PNIPAM/SWNT hybrids and the corresponding polymers p-SPNIPAM and p-PNIPAM.

How can the different dispersion capabilities of the two polymers for SWNTs be explained? From the fluorescence emission peaks, it was found that, apart from the monomer pyrene, the pyrenes in star-shaped p-SPNIPAM form excimers, strongly suggesting that microstructural differences exist between these polymers. To further reveal these differences, polymer solutions with different concentrations were examined by fluorescence spectrometry. Fig. S11 (ESI)† shows that p-SPNIPAM had monomer and excimer emission peaks in the measured concentration range, whereas the emission peaks of p-PNIPAM were mainly monomers. The plots of excimer-to-monomer ratios, I_E/I_M, versus polymer concentration provide more direct information about the status of pyrene, reflecting the chain conformation. Fig. 9 clearly shows that the excimer in p-SPNIPAM is much higher than that in p-PNIPAM over the whole concentration range. Remarkably, 13.2 times I_E/I_M of p-SPNIPAM with respect to that of p-PNIPAM occurs at 0.83 mM, which is exactly the polymer concentration used in this work, indicating that the pyrene in star-shaped p-SPNIPAM is likely to form excimers.

Fig. 9 I_E/I_M as a function of polymer concentration in deionized H₂O.

As they have the same components, in addition to only one labelled pyrene molecule at the end of the polymer chain, the differences should arise as a result of the architecture of the polymers. Star-shaped polymers usually exhibit a smaller hydrodynamic radius than linear polymers.^32,33 To confirm this, the rheological properties of the two polymers were examined. Fig. S12† shows the viscosity as a function of the shear rate of p-SPNIPAM (0.83 mM) and p-SPNIPAM (0.83 mM) at 25 °C. Both polymers have a low viscosity; however, the viscosity of p-PNIPAM is higher than that of p-SPNIPAM, implying that the hydrodynamic radius of p-PNIPAM higher than that of p-SPNIPAM.³³ The results of viscosity testing confronts the GPC observation that p-SPNIPAM had a larger M_n,GPC value than p-PNIPAM. This may be because the absolute size of p-SPNIPAM is larger than that of p-PNIPAM calculated from the ¹H NMR results; furthermore, both polymers showed extended conformation in THF (used as the GPC eluent) because THF is a good solvent for the various components of the polymer, including the NIPAM units, the hydrophobic pyrene core and the end-groups of the chain. However, when both polymers were dissolved in water (such as during viscosity testing), p-SPNIPAM may show a more contracted conformation, i.e. a small hydrodynamic radius, when compared with p-PNIPAM because p-SPNIPAM has more hydrophobic groups, especially in the core and at the end of branches. Pyrene is a highly hydrophobic group, and is inclined to interact with relatively hydrophobic groups, such as the polymer main chain or other pyrene groups.⁶⁵ The presence of the excimer in the star p-SPNIPAM solution implies that the pyrene group in p-SPNIPAM is exposed to the solvent and then contacts with other pyrene groups (Scheme 2a) because the smaller hydrodynamic radius of p-SPNIPAM leads to poor interactions between pyrene and the polymer backbone. The excimer fluorescence emission is modest, indicating the poor stacking between the pyrene groups.^58,59,66,67 Under sonication, the p-SPNIPAM chains are easily attached to the sidewalls of the nanotubes. However, the pyrene group in linear p-PNIPAM exists as a monomer, from which we can speculate that pyrene is protected by the loosely coiled chain because of the larger hydrodynamic radius (Scheme 2b), which may cause a barrier for interactions between p-PNIPAM and the SWNTs. More importantly, compared with p-PNIPAM, the smaller hydrodynamic radius of p-SPNIPAM will lead to a lower steric hindrance between the polymers attached to the nanotubes, resulting in more polymer molecules adsorbing onto the SWNTs. Both characteristics give a higher dispersion efficiency to star-shaped p-SPNIPAM.

Scheme 2 Schematic illustration of the different dispersibility of SWNTs for (a) star-shaped p-SPNIPAM and (b) linear p-PNIPAM and their thermo-responsive behaviour in water.

Thermo-responsive behaviour of SWNT hybrids

Having established that polymer/SWNTs are dispersed in an aqueous environment, we investigated whether the dispersibility of the SWNT-based hybrids could be reversibly controlled by tuning the temperature, as p-SPNIPAM and p-PNIPAM are known to be thermo-sensitive. The dispersions of polymer/SWNTs were placed at temperatures both above and below the cloudy point of the polymers, respectively. Before heating, the polymer/SWNT hybrids showed good dispersion in water. However, the hybrids gradually precipitated from the aqueous solution at 40 °C (inset, Fig. 10a). When the suspension was cooled to room temperature, the aqueous hybrid solution became homogeneous (inset, Fig. 10a). This reversibility was effective beyond three cycles of alternate heating and cooling, suggesting that the dispersion/aggregation state of both the polymer/SWNTs hybrids could be reversibly tuned. During the cycling process, UV-visible spectrophotometry was used to monitor the change in transmittance of the dispersion. A wavelength of 600 nm in the visible region was selected to determine the transmittance to exclude the influence of aromatic rings in the polymer.⁶⁸ Fig. 10 shows that the transmittance of the SWNT/polymer dispersion was close to zero at 25 °C, whereas, after heating, the transmittance of the p-SPNIPAM/SWNTs and p-PNIPAM/SWNTs reached 98.0 and 90.1%, respectively, indicating that the dispersion/aggregation of the p-SPNIPAM/SWNTs and p-PNIPAM/SWNTs hybrids can be reversibly switched by temperature. The upper layer of the solution of p-PNIPAM/SWNTs at 40 °C was slightly turbid and the transmittance of the p-PNIPAM/SWNTs was only 90.1% as a result of the aggregation of free p-PNIPAM.⁶⁹ These findings also suggest a lower adsorption of p-PNIPAM on the nanotubes, agreeing with the earlier conclusions.

Fig. 10 Reversible change of the transmittance of (a) the original SWNT dispersions and (b) suspensions diluted ten times at 600 nm on alternate heating and cooling. The inset images in (a) of the (1) p-SPNIPAM/SWNT (0.83 mM p-SPNIPAM, 0.52 mg mL⁻¹ SWNTs) and (2) p-PNIPAM/SWNT (0.83 mM p-PNIPAM, 0.40 mg mL⁻¹ of SWNTs) hybrids dispersed in water refer to the corresponding situations. The concentration of the suspension images in the inset of (b) is one-tenth of that for the suspensions in (a).

SWNT hybrids or hybrid/polymers can form network structures at high concentrations, particularly at temperatures above the cloudy point, which was also proposed by Theato et al.²² in a study in which the polymer concentration was close to that of this work. Networking may affect the responsive behaviour. Thus the responses of the diluted solutions were examined. It was found that p-SPNIPAM/SWNTs aggregated at 40 °C, whereas the p-PNIPAM/SWNTs became darker and more turbid (inset, Fig. 10b). UV-visible spectrophotometry was used to monitor the change in transmittance of the dispersion. Fig. 10b shows that the transmittance of the p-SPNIPAM/SWNT dispersion switched between 4.6 and 98.2% during cooling and heating. On the other hand, although the transmittance of the p-PNIPAM/SWNTs was switchable, the transmittance decreased from 20.2 to 11.5 as a result of free p-PNIPAM molecules. These results imply that the responsive behaviours of the two SWNT hybrids were different.

UV-visible-NIR spectrometry was used to monitor the variation in the dispersibility of the hybrids in dilute solutions on alternating treatments of heating and cooling. Fig. 11a shows that when the temperature increased to 40 °C, the absorbance of the p-SPNIPAM/SWNTs decreased and concomitantly the sharp peaks corresponding to the absorption features of the van Hove transition semiconducting tubes in the region 750–870 nm disappeared at 40 °C, indicating that bundled SWNTs were reformed. When the suspension was cooled to 25 °C, the characteristic sharp peaks appeared again, indicating that the SWNTs were re-dispersed. As for the p-PNIPAM/SWNTs, the variation in the sharp peaks of the dispersed SWNTs was the same as that of the p-SPNIPAM/SWNTs, although an enhanced absorbance was observed, implying that the SWNTs also undergo a reversible dispersion/aggregation transition. However, the absorbance of p-PNIPAM/SWNTs at 40 °C increased as a result of the high turbidity, which is consistent with the observations in the UV-visible spectrum.

Fig. 11 (a) UV-visible-NIR spectra of p-SPNIPAM/SWNTs (0.083 mM p-SPNIPAM, 0.052 mg mL⁻¹ SWNTs) and p-PNIPAM/SWNTs (0.083 mM p-PNIPAM, 0.040 mg mL⁻¹ SWNTs) in water at 25 °C and 40 °C. TEM images of (b) p-SPNIPAM/SWNT and (c) p-PNIPAM/SWNT hybrids in water at 40 °C.

Although the two SWNTs hybrids were bundled at 40 °C, there were differences in their appearance. To obtain more direct information on the aggregation state of the SWNTs and to unravel the discrepancy, TEM was used to visualize the polymer/SWNT hybrid suspension after heating (Fig. 11b and c). Both SWNTs hybrids were bundled, but the p-SPNIPAM/SWNT hybrids formed compact large diameter bundles (50–400 nm), whereas the p-PNIPAM/SWNTs only formed small loose bundles (50–100 nm) floating in water.

On the basis of these results, it is reasonable to speculate that the SWNTs were dispersed in water by adsorbing p-SPNIPAM or p-PNIPAM as a result of π–π stacking (Scheme 2). As p-SPNIPAM has a smaller hydrodynamic radius than linear p-PNIPAM, the pyrene group is exposed in the solvent and forms an excimer; the static hindrance between the polymer attached to the tube is smaller, which means that it has a higher dispersion efficiency and interactive capability than linear p-PNIPAM. The temperature-controlled dispersion of the SWNTs in water is ascribed to the polymer shell (Scheme 2). At room temperature, both polymers anchored on the SWNTs showed an entangled conformation as the temperature was lower than their cloudy point,^22,23 leading to good dispersion of the SWNTs. When the temperature was raised above the cloudy point, the amide bonds formed intramolecular H-bonds, which resulted in a compact chain conformation with a hydrophobic surface.^22,23 Thus the polymer–polymer interactions increased, leading to more nanotube bundles. As the star-shaped p-SPNIPAM was densely adsorbed on the tube, the SWNT hybrids formed compact large diameter bundles (Scheme 2a). As for the p-PNIPAM/SWNTs, free p-PNIPAM existed in the dispersion and hindered the intertube interaction, resulting in the formation of small loose SWNTs bundles floating in water (Scheme 2b).

Conclusions

We prepared well-defined thermo-responsive pyrene-labelled 4-arm star-shaped p-SPNIPAM and linear p-PNIPAM by an ATRP technique and then used the two polymers to modify SWNTs. The two polymers could be attached onto the sidewall of nanotubes by π–π stacking between the pyrene group and the SWNTs, but the star-shaped p-SPNIPAM imparted a higher solubility to the SWNTs than linear p-PNIPAM because p-SPNIPAM had a smaller hydrodynamic radius than linear p-PNIPAM. In addition, the dispersion/aggregation state of the p-SPNIPAM/SWNTs and the p-PNIPAM/SWNTs hybrids could be reversibly switched by tuning the temperature as a result of the H-bond transition. However, when the temperature was higher than the cloudy point of the polymers, the p-SPNIPAM/SWNT hybrids formed compact large diameter bundles, whereas the p-PNIPAM/SWNTs formed loose small bundles as a result of free p-PNIPAM. These findings may help to understand the effect of molecular structure of stimuli-responsive polymers on the dispersion of SWNTs and its smart behaviour. SWNT hybrids can undergo reversible dispersion and aggregation by tuning temperature, rendering them attractive in biomedical or sensory applications.

Acknowledgements

This work was financially supported by the Distinguished Youth Fund from the Science and Technology Department of Sichuan Province (2010JQ0029) and the National Natural Science Foundation of China (21273223).

Notes and references

1. S. Iijima and T. Ichihashi, Nature, 1993, 363, 603–605.
2. P. M. Ajayan, J. M. Lambert, P. Bernier, L. Barbedette, C. Colliex and J. M. Planeix, Chem. Phys. Lett., 1993, 215, 509–517.
3. N. W. S. Kam, M. O'Connell, J. A. Wisdom and H. J. Dai, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 11600–11605.
4. N. W. S. Kam and H. J. Dai, Phys. Status Solidi B, 2006, 243, 3561–3566.
5. B. Kang, D. C. Yu, Y. D. Dai, S. Q. Chang, D. Chen and Y. T. Ding, Small, 2009, 5, 1292–1301.
6. M. Ouyang, J. L. Huang and C. M. Lieber, Acc. Chem. Res., 2002, 35, 1018–1025.
7. S. Nanot, E. H. Haroz, J. H. Kim, R. H. Hauge and J. Kono, Adv. Mater., 2012, 24, 4977–4994.
8. J. P. Salvetat, G. A. D. Briggs, J. M. Bonard, R. R. Bacsa, A. J. Kulik, T. Stockli, N. A. Burnham and L. Forro, Phys. Rev. Lett., 1999, 82, 944–947.
9. L. A. Girifalco, M. Hodak and R. S. Lee, Phys. Rev. B: Condens. Matter Mater. Phys., 2000, 62, 13104–13110.
10. S. Nuriel, L. Liu, A. H. Barber and H. D. Wagner, Chem. Phys. Lett., 2005, 404, 263–266.
11. S. J. Tans and C. Dekker, Nature, 2000, 404, 834–835.
12. P. C. Collins, M. S. Arnold and P. Avouris, Science, 2001, 292, 706–709.
13. T. Rueckes, K. Kim, E. Joselevich, G. Y. Tseng, C. L. Cheung and C. M. Lieber, Science, 2000, 289, 94–97.
14. M. S. Strano, P. W. Barone, S. Baik and D. A. Heller, Nat. Mater., 2005, 4, 86–92.
15. J. H. Kim, D. A. Heller, H. Jin, P. W. Barone, C. Song, J. Zhang, L. J. Trudel, G. N. Wogan, S. R. Tannenbaum and M. S. Strano, Nat. Chem., 2009, 1, 473–481.
16. P. W. Barone and M. S. Strano, Angew. Chem., Int. Ed., 2006, 45, 8138–8141.
17. C. Y. Hong, Y. Z. You and C. Y. Pan, Chem. Mater., 2005, 17, 2247–2254.
18. P. M. Ajayan and J. M. Tour, Nature, 2007, 447, 1066–1068.
19. N. Karousis, N. Tagmatarchis and D. Tasis, Chem. Rev., 2010, 110, 5366–5397.
20. Y. L. Zhao and J. F. Stoddart, Acc. Chem. Res., 2009, 42, 1161–1171.
21. L. W. Chen and D. Wang, Nano Lett., 2007, 7, 1480–1484.
22. P. Theato, K. C. Etika, F. D. Jochum, M. A. Cox, P. Schattling and J. C. Grunlan, Macromol. Rapid Commun., 2010, 31, 1368–1372.
23. (a) K. C. Etika, F. D. Jochum, P. Theato and J. C. Grunlan, J. Am. Chem. Soc., 2009, 131, 13598–13598; (b) K. C. Etika, F. D. Jochum, M. A. Cox, P. Schattling, P. Theato and J. C. Grunlan, Macromolecules, 2010, 43, 9447–9453.
24. J. C. Grunlan, L. Liu and Y. S. Kim, Nano Lett., 2006, 6, 911–915.
25. S. Liang, G. Chen, J. Peddle and Y. M. Zhao, Chem. Commun., 2012, 48, 3100–3102.
26. S. L. Chen, Y. G. Jiang, Z. Q. Wang, X. Zhang, L. M. Dai and M. Smet, Langmuir, 2008, 24, 9233–9236.
27. W. H. Yi, A. Malkovskiy, Y. Q. Xu, X. Q. Wang, A. P. Sokolov, M. Lebron–Colon, M. A. Meador and Y. Pang, Polymer, 2010, 51, 475–481.
28. H. Imahori, T. Umeyama, K. Kawabata, N. Tezuka, Y. Matano, Y. Miyato, K. Matsushige, M. Tsujimoto, S. Isoda and M. Takano, Chem. Commun., 2010, 46, 5969–5971.
29. Z. Guo, Y. Feng, D. Zhu, S. He, H. Liu, X. Shi, J. Sun and M. Qu, Adv. Funct. Mater., 2013, 2013(23), 5010–5018.
30. Z. R. Guo, Y. J. Feng, S. He, M. Z. Qu, H. L. Chen, H. B. Liu, Y. F. Wu and Y. Wang, Adv. Mater., 2013, 25, 584–590.
31. G. Mountrichas, M. Mpiri and S. Pispas, Macromolecules, 2005, 38, 940–947.
32. G. Lapienis, Prog. Polym. Sci., 2009, 34, 852–892.
33. L. Y. Qiu, R. J. Wang, C. Zheng, Y. Jin and L. Q. Jin, Nanomedicine, 2010, 5, 193–208.
34. Z. M. O. Rzaev, S. Dincer and E. Piskin, Prog. Polym. Sci., 2007, 32, 534–595.
35. D. Wang, W. X. Ji, Z. C. Li and L. W. Chen, J. Am. Chem. Soc., 2006, 128, 6556–6557.
36. D. M. Guldi, G. M. A. Rahman, N. Jux, D. Balbinot, U. Hartnagel, N. Tagmatarchis and M. Prato, J. Am. Chem. Soc., 2005, 127, 9830–9838.
37. J. Queffelec, S. G. Gaynor and K. Matyjaszewski, Macromolecules, 2000, 33, 8629–8639.
38. M. Ciampoli and N. Nardi, Inorg. Chem., 1966, 5, 41–44.
39. W. A. Braunecker and K. Matyjaszewski, Prog. Polym. Sci., 2007, 32, 93–146.
40. N. V. Tsarevsky and K. Matyjaszewski, Chem. Rev., 2007, 107, 2270–2299.
41. K. Matyjaszewski and N. V. Tsarevsky, Nat. Chem., 2009, 1, 276–288.
42. Z. S. Ge, Y. L. Cai, J. Yin, Z. Y. Zhu, J. Y. Rao and S. Y. Liu, Langmuir, 2007, 23, 1114–1122.
43. S. B. Lee, S. C. Song, J. I. Jin and Y. S. Sohn, Macromolecules, 2001, 34, 7565–7569.
44. X. M. Liu and L. S. Wang, Biomaterials, 2004, 25, 1929–1936.
45. Y. C. Wang, Y. Li, X. Z. Yang, Y. Y. Yuan, L. F. Yan and J. Wang, Macromolecules, 2009, 42, 3026–3032.
46. A. Khan, J. Colloid Interface Sci., 2007, 313, 697–704.
47. R. Yoshida, K. Uchida, Y. Kaneko, K. Sakai, A. Kikuchi, Y. Sakurai and T. Okano, Nature, 1995, 374, 240–242.
48. Y. F. Wu, X. L. Liu, Y. Wang, Z. R. Guo and Y. J. Feng, Macromol. Chem. Phys., 2012, 213, 1489–1498.
49. R. Plummer, D. J. T. Hill and A. K. Whittaker, Macromolecules, 2006, 39, 8379–8388.
50. S. Brahmachari, D. Das, A. Shome and P. K. Das, Angew. Chem., Int. Ed., 2011, 50, 11243–11247.
51. X. Xin, G. Y. Xu, T. T. Zhao, Y. Y. Zhu, X. F. Shi, H. J. Gong and Z. Q. Zhang, J. Phys. Chem. C, 2008, 112, 16377–16384.
52. M. J. O'Connell, S. M. Bachilo, C. B. Huffman, V. C. Moore, M. S. Strano, E. H. Haroz, K. L. Rialon, P. J. Boul, W. H. Noon, C. Kittrell, J. P. Ma, R. H. Hauge, R. B. Weisman and R. E. Smalley, Science, 2002, 297, 593–596.
53. H. Chakraborty and M. Sarkar, Langmuir, 2004, 20, 3551–3558.
54. L. Zhai, M. Zhao, D. Sun, J. Hao and L. Zhang, J. Phys. Chem. B, 2005, 109, 5627–5630.
55. J. D. Hartgerink, E. Beniash and S. I. Stupp, Science, 2001, 294, 1684–1688.
56. G. J. Bahun, C. Wang and A. Adronov, J. Polym. Sci., Part A: Polym. Chem., 2006, 44, 1941–1951.
57. W. B. Liu, S. F. Pei, J. H. Du, B. L. Liu, L. B. Gao, Y. Su, C. Liu and H. M. Cheng, Adv. Funct. Mater., 2011, 21, 2330–2337.
58. D. F. Anghel, J. L. Toca–Herrera, F. M. Winnik, W. Rettig and R. von Klitzing, Langmuir, 2002, 18, 5600–5606.
59. M. Ingratta and J. Duhamel, Macromolecules, 2007, 40, 6647–6657.
60. J. Yip, J. Duhamel, G. J. Bahun and A. Adronov, J. Phys. Chem. B, 2010, 114, 10254–10265.
61. Y. Tomonari, H. Murakami and N. Nakashima, Chem.–Eur. J., 2006, 12, 4027–4034.
62. J. H. Zou, L. W. Liu, H. Chen, S. I. Khondaker, R. D. McCullough, Q. Huo and L. Zhai, Adv. Mater., 2008, 20, 2055–2060.
63. A. Star, J. F. Stoddart, D. Steuerman, M. Diehl, A. Boukai, E. W. Wong, X. Yang, S. W. Chung, H. Choi and J. R. Heath, Angew. Chem., Int. Ed., 2001, 40, 1721–1725.
64. K. C. Etika, F. D. Jochum, M. A. Cox, P. Schattling, P. Theato and J. C. Grunlan, Macromolecules, 2010, 43, 9447–9453.
65. F. M. Winnik, Macromolecules, 1990, 23, 233–242.
66. F. M. Winnik, S. T. A. Regismond and E. D. Goddard, Langmuir, 1997, 13, 111–114.
67. H. Siu and J. Duhamel, Macromolecules, 2005, 38, 7184–7186.
68. K. Yamamoto, T. Serizawa, Y. Muraoka and M. Akashi, Macromolecules, 2001, 34, 8014–8020.
69. L. W. Chen and D. Wang, Nano Lett., 2007, 7, 1480–1484.
70. L. P. Yang and C. Y. Pan, Macromol. Chem. Phys., 2008, 209, 783–793.

---

Footnote

† Electronic supplementary information (ESI) available: ¹H NMR and ESI-HRMS of compounds, and fluorescence emission spectrum of polymers. See DOI: 10.1039/c6ra00998k

---