Phase‐Inversion Method for Incorporation of Metal Nanoparticles into Carbon‐Nanotube/Polymer Composites

A general method for facile incorporation of water-soluble metal nanoparticles in carbon-nanotube (CNT)/polymer composites is described. Gold nanoparticle (AuNP)/CNT/polysulfone (PSf) composites are prepared by the phase-inversion method (see picture), which involves immobilization of the nanoparticles and coagulation of the soft composite at same time. Colloidal metal and semiconductor nanoparticles (NPs) and quantum dots are widely recognized as potential functional elements in carbon nanotube (CNT)/polymeric devices due to the wide range of their interesting optical, electronic, magnetic, and electrochemical properties.1 There is great interest in coupling of CNTs and polymers with metallic or semiconducting NPs and quantum dots.2 There are two categories of method for preparing NP/CNT composites with desirable compositions. The first type is a one-step method, such as electrospinning;3 the second type is a two-step preparation method involving chemical or electrochemical synthesis and deposition of NPs and polymer.4-7 For example, Chen et al. and Kong et al. have reported the synthesis of Fe3O4 NPs on CNTs followed by spontaneous precipitation of polyaniline on the NP/CNT hybrid.4, 5 Wang et al. electrochemically deposited polyaniline on CNTs and subsequently electrochemically deposited gold NPs (AuNPs).6 Carillo et al. incorporated AuNPs in polyethyleneimine/poly(acrylic acid)-coated CNTs by using the layer-by-layer technique.7 The methods described above employ water-soluble polymers or precursors for NP/CNT/polymer hybrid preparation. Herein, we describe a facile and general way of incorporating NPs into CNT/polymer composites by a phase inversion method. In this one-step method, CNTs and water-insoluble polymer are dissolved in an organic solvent that is miscible with water. The suspension of CNTs in such solvent is stabilized by wrapping the CNTs with polymeric chains. After dispensing the CNT/polymer suspension onto the desired surface, an aqueous solution containing water-soluble metal NPs is deposited on the CNT/polymer suspension. Consequently, the solvent phase is exchanged (i.e., N,N-dimethylformamide (DMF) for water), the CNT composite wrapped with water-insoluble polymer precipitates, and metal NPs present in the aqueous phase are incorporated into the composite. We illustrate the composite preparation process in Scheme 1. The CNTs were first functionalized by refluxing in nitric acid (2 M) at room temperature for 24 h, to give surface carboxyl groups at the defect sites of the outer graphene layer of the nanotubes and to remove residual metal-catalyst impurities.8-11 Subsequently, the carboxylic acid-functionalized multi-wall CNTs were washed with distilled water, dried, and dispersed in DMF (2.5 mg mL−1) containing polysulfone (PSf; 85 mg mL−1). The mixture was deposited onto the required substrate. When an aqueous solution of colloidal gold was added to the polymer composite suspension, the AuNP/CNT/PSf composite started to coagulate. The AuNPs were immobilized into the PSf that coated the CNTs. Preparation process of the AuNP/CNT/polysulfone (PSf) composite. A) Deposition of CNT/PSf suspension on the substrate. B) Addition of an aqueous solution of AuNPs, and subsequent coagulation of the PSf and incorporation of AuNPs by the phase inversion method. C) Formation of AuNP/CNT/PSf composite. Figure 1 shows typical scanning electron microscopy (SEM) images of soft AuNP/CNT/PSf composite at lower (A) and higher (B) resolution. The images demonstrate that AuNPs (represented by bright dots in the SEM image) are well dispersed in the polymer matrix. It is also clear from Figure 1B that the PSf coaxially wraps the CNTs and that all CNTs are encapsulated by AuNP-doped polymer. This was further confirmed by transmission electron microscopy (TEM). SEM images of AuNP/CNT/PSf composite at A) 10 000 × and B) 40 000 × magnification. To investigate the presence and localization of AuNPs in the CNT/PSf hybrid material, a Z-contrast scanning TEM (STEM) study of AuNP/CNT/PSf composite was performed with a high-angle annular dark-field (HAADF) detector. Figure 2 shows HAADF-STEM images at different magnifications, in which I ≈ Z2 where I is intensity and Z is atomic number. The higher intensities correspond to heavier elements. In Figure 2, all of the components of the AuNP/CNT/PSf composites can be observed, such as high-contrast gold spherical particles and weakly contrasted CNTs wrapped in PSf. HAADF-STEM confirms that the CNTs are completely wrapped by PSf and that the AuNPs are scattered in the polymer and do not cluster on the CNT surface. HAADF-STEM images of AuNP/CNT/PSf composite. The elemental composition of the AuNP/CNT/PSf composite was analyzed by SEM/energy-dispersive X-ray (EDX) spectroscopy. SEM/EDX analysis of the section of AuNP/CNT/PSf composite that is depicted in Figure 1B is shown in Figure 3. The EDX spectrum reveals strongly the presence of sulfur, which originates from the PSf matrix, and gold from the AuNPs, as indicated by the characteristic elemental emissions of the S Kα and Au Mz, Au Mα, Au Lα, and Au Lβ lines. Note that Ni Kα originates from residual metal-catalyst NPs intercalated in the CNTs.12, 13 Quantitative analysis of the EDX spectra reveals that the atomic percentage of carbon is 90.47% (83.52 wt%), of oxygen is 8.31% (10.22 wt%), of sulfur is 0.96% (2.38 wt%), and of gold is 0.26% (3.88 wt%). The standard deviation of AuNP loading in the composite is 0.022 at% (n = 12). EDX spectrum of AuNP/CNT/PSf composite acquired in the SEM/EDX configuration. To gain insight into the structure of the AuNP/CNT/PSf composite, we employed X-ray photoelectron spectroscopy (XPS) to measure the elemental composition of the CNT/PSf and AuNP/CNT/PSf composites (Figure 4). As expected, strong signals from carbon (C 1s), oxygen (O 1s), sulfur (S 2p), and impurities (silicon, sodium, chlorine) were observed for both CNT/PSf (Figure 4A) and AuNP/CNT/PSf (Figure 4B) composites. The weak signal of the Au 4f peak was observed in the wide spectrum for the AuNP/CNT/PSf composite, as shown in Figure 4B. The atomic concentration of gold was evaluated to be 0.01%. To confirm the peak energies for Au 4f, a detailed scan was performed with X-ray scanning at 100 W to improve the signal-to-noise ratio (see inset of Figure 4B). The peak energies of Au 4f7/2 and Au 4f5/2 were assigned as 84.0 and 87.7 eV, respectively, which corresponded to the gold metallic state. The atomic concentration obtained from XPS measurements is significantly lower than that of the EDX measurements. The inelastic mean free path of Au 4f is less than 2 nm, so the intensity of Au 4f will be decreased if the particles are located underneath the top surface. The intensity can also be decreased by the sample surface topography; a reasonable explanation of the XPS data is that AuNPs are completely integrated and wrapped within PSf. The conclusion from the XPS results is consistent with the HAADF-STEM observations. X-ray photoelectron spectra of A) CNT/PSf composite and B) AuNP/CNT/PSf composite. Expanded spectra in the 200 to −20 eV binding energy region are also shown. The insets in the expanded spectra are detailed scan spectra of Au 4f. Electrochemical impedance spectroscopy was used for measuring the changes in capacitance and resistance to heterogeneous electron transfer of AuNP/CNT/PSf and CNT/PSf composites. The corresponding Nyquist impedance plots are shown in Figure 5A. It is clear from the graphs that AuNP/CNT/PSf composite (gray line, circles) demonstrates lower resistance to heterogeneous electron transfer and also higher capacitance than the CNT/PSf composite (black line, squares), which is beneficial for several electrochemical applications, including sensing/biosensing and battery applications. The corresponding values of resistance to heterogeneous electron transfer (Rp) were found to be 21.6 and 34.9 kΩ for the AuNP/CNT/PSf and CNT/PSf composites, respectively, and the capacitances were 275.2 and 244.0 µF, respectively. A) Nyquist impedance plots of a) CNT/PSf composite electrodes and b) AuNP/CNT/PSf composite electrodes for a 10 mM ferricyanide solution. The supporting electrolyte was 0.1 M KCl. Inset: the Randles equivalent circuit. B) Cyclic voltammograms resulting from the electrochemical properties of 10 mM ferricyanide at a) CNT/PSf composite electrodes and b) AuNP/CNT/PSf composite electrodes. Also shown are the corresponding cyclic voltammograms of the background electrolyte (c and d for CNT/PSf and AuNP/CNT/PSf composite electrodes, respectively). Conditions: scan rate, 100 mV s−1; phosphate-buffered saline (PBS), pH 7.0. Cyclic voltammetry of ferricyanide (10 mM; Figure 5B) shows that the oxidation/reduction signal of the ferrous/ferric couple is about two times higher at the AuNP/CNT/PSf composite electrode than on the CNT/PSf composite electrode. The magnitude of the anodic current is 1.40 mA (at Epa = 250 mV) and 0.66 mA (at Epa = 215 mV) for AuNP/CNT/PSf and CNT/PSf composites, respectively; the cathodic current is 1.57 mA (at Epc = 80 mV) and 0.82 mA (at Epc = 111 mV) for AuNP/CNT/PSf and CNT/PSf composite electrodes, respectively (all potentials are uncorrected from ohmic drop). It is also possible to note a higher background current of the AuNP/CNT/PSf composite electrode than of the CNT/PSf composite electrode, which is due to the higher capacitance of the AuNP/CNT/PSf composite and is consistent with the electrochemical impedance measurements discussed above. The relative standard deviations (RSDs) of the voltammetric experiments using different electrodes (n = 12) are ΔEpp = 7.1% and Epa = 9.3%. Such RSDs are comparable with those of screen-printed electrodes fabricated by different methods.14, 15 To evaluate the practical suitability of AuNP/CNT/PSf composites for biosensing applications, we coupled the enzyme horseradish peroxidase (HRP) within the AuNP/CNT/PSf composite. Figure 6 shows the chronoamperometric response of HRP/CNT/PSf- and HRP/AuNP/CNT/PSf-based biosensors in the electrocatalytic reduction of H2O2 to H2O at −0.2 V using hydroquinone as mediator. The amperometric response of the HRP/AuNP/CNT/PSf biosensor to H2O2 is very fast (within 1 s) and it exhibits a current nine times higher than that of the HRP/CNT/PSf biosensor. The higher response of the AuNP-modified composites is expected to result from the larger conductive area provided by the AuNPs, which improves the electron transfer and increases the reduction current for hydrogen peroxide. In the context of electrochemical experiments, it is important to mention that even a small number of electroactive NPs at very small (under 1%) random array coverage may exhibit behavior analogous to that of the corresponding macroelectrode due to heavily overlapping diffusion layers, as was shown theoretically and experimentally for voltammetry at micro/nanoelectrode arrays.16-19 Importantly, the typical thickness of the diffusion layer (in the framework of conditions used in our voltammetric experiments) is about 54 µm. According to the TEM and SEM images, the distance between AuNPs (which behave as gold nanoelectrodes) is on the order of tens to hundreds of nanometers and their diffusion layers are indeed heavily overlapped. Therefore, AuNPs embedded in CNT/PSf composite are responsible for the enhanced electrochemical properties of AuNP/CNT/PSf composites. Amperometric response of a) HRP/CNT/PSf and b) HRP/AuNP/CNT/PSf composite-based biosensors to 2.4 mM hydrogen peroxide at a potential of −0.2 V. Conditions: PBS buffer, pH 7.0; concentration of hydroquinone, 1.8 mM. The electrodes are chemically stable in aqueous media. However, note that at potentials of 1.1 V and higher (vs. Ag/AgCl) in strongly acidic media the AuNPs undergo electrochemical oxidation and dissolution.20 In conclusion, we have developed a simple, fast, and general method for facile incorporation of water-soluble metal NPs in CNT/polymer composites. The AuNP/CNT/PSf composites were successfully prepared by the phase inversion method, which involved immobilization of the NPs and coagulation of the soft composite at the same time. The AuNPs were dispersed in a polymer coating of the CNTs. The phase inversion incorporation technique is easy, very fast (about 5 min for the whole process), and versatile. It also allows incorporation of biomolecules in the composite. The electrochemical impedance spectroscopic and voltammetric data clearly prove that AuNPs enhance the CNT/PSf composite and that they bring advantageous properties, which should be greatly beneficial in the areas of highly sensitive sensors and biosensors and for battery/energy-storage applications. The phase inversion method can incorporate water-soluble metallic and semiconductor NPs and quantum dots into the CNT/polymer matrix in a one-step process. In principle, this process is applicable to any polymer that is soluble in water-miscible organic solvent. The simplicity of the method will contribute widely to various fields of application. An S-4800 field-emission scanning electron microscope (Hitachi) was employed to obtain SEM images. A JEM 2100F field-emission transmission electron microscope (JEOL) was used to obtain TEM images in the STEM mode (spot size, 0.4 nm; 200 kV; HAADF). All voltammetric and electrochemical impedance spectroscopy experiments were performed using an electrochemical analyzer Autolab 302 (Ecochemie, The Netherlands). Electrochemical experiments were carried out in a 5-mL voltammetric cell at room temperature (25 °C) using a three-electrode configuration. A platinum electrode served as an auxiliary electrode and an Ag/AgCl electrode as reference electrode. X-ray photoelectron spectra were obtained with a PHI Quantera SXM (ULVAC-PHI) spectrometer, which was equipped with monochromatic Al Kα X-rays. The analysis area was defined by the X-ray beam. The beam was 100 µm in diameter at a beam energy of 15 kV/25 W for wide-scan spectra and 1.5 × 0.1 mm2 scanned beam at a beam energy of 20 kV/100 W for detailed scan spectra; the photoelectron take-off angle was 45°. The binding energy was calibrated to the Au 4f7/2 line of 83.94 eV and a pass energy of 26 eV was used. Wide-scan spectra were obtained at a pass energy of 280 eV with 0.5 eV steps, and detailed scan spectra for Au 4f were obtained at pass energy 26 eV with 0.1 eV steps. CNTs (outer diameter 30–50 nm, length 0.5–200 µm) were first purified by stirring in 2 M nitric acid for 24 h.21 Subsequently, purified CNTs were dispersed in DMF at a concentration of 5 mg mL−1, and the suspension was then placed in an ultrasonic bath for 10 min. PSf (BASF Ultrasons S 3010 natur) was dissolved separately in DMF (87 mg PSf mL−1) and both solutions were mixed together in a ratio of 1:1 and subsequently sonicated for 30 min to achieve a stable suspension. The final CNT concentration was 2.5 mg mL−1. Immediately after deposition of CNT/PSf (15 µL) on the required substrate, the colloidal gold solution (15 µL; 20 nm in diameter, citrate stabilized) was dropped onto the CNT/PSf. The CNT/PSf suspension started instantly to precipitate due to the insolubility of PSf in water. To ensure the complete coagulation of the composite and completely remove the DMF, distilled water (30 µL) was dropped onto the modified nanocomposite. For control experiments (preparation of CNT/PSf composites), distilled water was used instead of colloidal gold solution. The working electrodes were fabricated by screen printing using a Dek248 semiautomatic system. The squeegees used were of the soft polymer type, and the pressure applied during the printing process was set to 7 kg cm−2. A double-sweep process was programmed at a speed of 20 mm s−1. AuNP/CNT/PSf composite was deposited on the surface of the screen-printed electrode by the above-described procedure. Acheson carbon ink (Electrodag 400B, Acheson Colloids Co.), conductive silver ink (Electrodag 6037 SS), and insulating ink (Minico M 7000) were used for the preparation of screen-printed electrodes. AuNP colloidal solution was mixed with HRP aqueous solution (25 IU mL−1) for 10 min under stirring in a 1:5 ratio. After preparation of the CNT/PSf suspension and its deposition onto the screen-printed electrodes as described above, the electrode was immersed in the HRP/AuNP solution. In this step, the polymer coagulated at the same time as the AuNPs and HRP were incorporated into the matrix. After 5 min of phase inversion, the HRP/AuNP/CNT/PSf biosensor was rinsed with distilled water for 1 min. Cyclic voltammetric experiments were carried out at a scan rate of 100 mV s−1 with PBS (50 mM phosphate, 100 mM KCl, pH 7.0). Electrochemical impedance spectroscopy was performed in 0.1 M KCl. Fe(CN)63−/4− was used as electrochemical probe at a concentration of 10 mM. For TEM measurements, AuNP/CNT/PSf composites were cut by microtome and then transferred to a TEM grid. For SEM and XPS measurements, the AuNP/CNT/polymer composite was placed onto a polycarbonate substrate. EDX spectra were obtained in the SEM/EDX configuration.