Carbon Nanofiber Composites: Preparation

May 18, 2014 - Nano-biodevice Research Group, Biomedical Research Institute, National ... for sugar (e.g., glucose, fructose, sucrose, and maltose) ox...
1 downloads 0 Views 6MB Size
Download PDF
Article pubs.acs.org/ac

Pd−Ni Alloy Nanoparticle/Carbon Nanofiber Composites: Preparation, Structure, and Superior Electrocatalytic Properties for Sugar Analysis Qiaohui Guo,†,§ Dong Liu,†,§ Xueping Zhang,† Libo Li,† Haoqing Hou,‡ Osamu Niwa,# and Tianyan You*,† †

State Key Laboratory of Electroanalytical Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun, Jilin 130022, China ‡ College of Chemistry and Chemical Engineering, Jiangxi Normal University, Nanchang, Jiangxi 330027, China # Nano-biodevice Research Group, Biomedical Research Institute, National Institute of Advanced Industrial Science and Technology, Central 6, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8566, Japan S Supporting Information *

ABSTRACT: Novel Pd−Ni alloy nanoparticle/carbon nanofiber (Pd−Ni/CNF) composites were successfully prepared by a simple method involving electrospinning of precursor polyacrylonitrile/Pd(acac)2/Ni(acac)2 nanofibers, followed by a thermal process to reduce metals and carbonize polyacrylonitrile. The nanostructures of the resulting Pd−Ni/CNF nanocomposites were carefully examined by a combination of scanning electron microscopy (SEM), transmission electron microscopy (TEM), high resolution transmission electron microscopy (HRTEM), high-angle annular dark field (HAADF)-scanning transmission electron microscopy (STEM), energy dispersive X-ray (EDX), thermogravimetric analysis (TGA), X-ray diffraction (XRD), and X-ray photoelectron spectra (XPS). For all the nanocomposites, the Pd−Ni alloy nanoparticles (NPs) were dispersed uniformly and embedded firmly within the framework or on the surface of CNF. The size, composition, and alloy homogeneity of the Pd−Ni alloy NPs could be readily tailored by controlling the feed ratio of metal precursors and the thermal treatment process. Cyclic voltammetric studies showed enhanced redox properties for Pd−Ni/CNF-based electrodes relative to the Ni-metal electrode and significantly improved electrocatalytic activity for sugar (e.g., glucose, fructose, sucrose, and maltose) oxidation. The application potential of Pd−Ni/CNF-based electrodes in flow systems for sugars detection was explored. A very low limit of detection for sugars (e.g., 7−20 nM), high resistance to surface fouling, excellent signal stability and reproducibility, and a very wide detection linear range (e.g., 0.03−800 μM) were revealed for this new type of Pd−Ni/CNF nanocomposite as the detecting electrode. Such detection performances of Pd−Ni/CNF-based electrodes are superior to those of state-of-the-art nonenzymatic sugar detectors that are commercially available or known in the literature.

T

fouling during sugar analysis and the requirement for electrode surface regeneration limit the wide applications of Pt- or Aubased PAD in flow systems for continuous analysis of sugars. An alternative method for sugar detection is the constant potential amperometry, which is more applicable to flow systems and has the advantage of instrumental simplicity.

he rapid development of food technologies in the past decade, especially in fields of additives and artificial sweeteners, has imposed new challenges in quality control and food safety.1 As a result, quick and accurate determination of additives and sugars are currently of great importance in the food industry, biotechnology, and clinical chemistry and diagnosis.2 For efficient sugar analysis, nonenzymatic electrochemical detection techniques, such as pulsed amperometric detection (PAD) using Pt or Au electrodes,3,4 have been widely used due to their good sensitivity and cost effectiveness. However, the susceptibility of Pt and Au electrodes to surface © 2014 American Chemical Society

Received: February 23, 2014 Accepted: May 17, 2014 Published: May 18, 2014 5898

dx.doi.org/10.1021/ac500811j | Anal. Chem. 2014, 86, 5898−5905

Analytical Chemistry

Article

framework or on the surface of CNF. The nanostructures and atomic Pd−Ni alloy nature were carefully examined by a combination of techniques including scanning electron microscopy (SEM), transmission electron microscopy (TEM), high resolution transmission electron microscopy (HRTEM), high-angle annular dark field (HAADF)-scanning transmission electron microscopy (STEM), energy dispersive X-ray (EDX), thermogravimetric analysis (TGA), X-ray diffraction (XRD), and X-ray photoelectron spectra (XPS). Electrochemical studies showed more enhanced electrocatalytic properties for sugar (e.g., glucose, fructose, sucrose, and maltose) oxidation. Superior sugar electrochemical detection properties, including low limit of detection, high surface fouling resistance, excellent signal stability and reproducibility, and wide detection linear range, were revealed for this new type of Pd−Ni/CNF nanocomposite when used as the detecting electrode in the flow system. The application of Pd−Ni/CNF-based electrode in the liquid chromatography-electrochemical system for sugar analysis in commercial honey products was explored.

Transition metals such as Cu and Ni have been explored as the electrochemical detecting electrode for this application.5−11 In particular, Ni-based electrodes showed promising properties for sugar analysis in alkaline media.7−10 Nevertheless, the detection sensitivity, the fouling resistance, and the electrode stability of Ni-based electrochemical detector still need to be improved for practical applications. Efficient preparation methods for nanostructured Ni electrocatalysts with excellent properties also need to be developed. On the other hand, bimetallic alloy nanoparticle (NP) electrocatalysts are currently attracting much attention owing to their demonstrated higher catalytic activity, selectivity, and stability relative to monometallic NPs.11 To support these NPs for practical applications, carbon materials such as carbon nanotubes (CNT),12,13 carbon nanofibers (CNFs),14 and graphene15 have been extensively explored because of their high surface area for dispersing NPs, porous structure for mass transfer, and high electrical conductivity for charge transfer. For example, alloy NPs/CNF nanocomposites have been prepared and showed promising properties in electrocatalysis,16 energy conversion and storage,17 and photocatalysis.18 Typically, the preparation of CNF-supported alloy NPs involves a step of purification and functionalization of presynthesized CNF and the use of chemical stabilizers19 (e.g., polymers, surfactants, capping reagents, and silica) or harsh oxidants20 (e.g., HNO3 and H2SO4) for efficient dispersion of CNF and NPs. This may damage the carbon structure and decrease the electrical conductivity, adversely affecting the performance. The weakly attached alloy NPs to the CNF surface may suffer from detachment during processing and application, which leads to impaired stability of the nanocomposites. For practical applications of carbon-supported alloy NPs, an efficient and simple preparation method without the need of using chemical stabilizers and strong oxidant is highly desired. Electrospinning has recently emerged as a powerful technique for producing one-dimensional nanofibers.21 The intriguing features of electrospun nanofibers, such as high porosity, large surface area, surface functionality, and good mechanical properties, have found wide applications in the fields of biomedicine,22,23 energy,24 the environment,25 etc. Electrospinning can also be used for preparation of metallic NPs/CNF composites for eletrocatalytic applications by spinning metal precursor-containing polyacrylonitrile (PAN) nanofibers, followed by a thermal treatment.26−32 The advantages of this method include the uniform dispersion of metallic NPs within the framework of CNF, the high electrical conductivity of CNF, and the highly porous and mechanically strong network structure of the resulting composite, which contribute to a high electrocatalytic activity by providing a larger active surface area, preventing NPs from detachment and agglomeration, and facilitating electron and mass transfer within the system. Although some monometallic NP/CNF nanocomposites have been successfully prepared and shown great promises in chemical/biological sensing devices,26,29,30 catalysis,31 and renewable energy-related applications,32 preparation of bimetallic alloy NP/CNF nanocomposites using the electrospinning technique is still scarcely known. Herein, we report a new type of bimetallic Pd−Ni alloy NP/ CNF composites that were successfully prepared by electrospinning precursor nanofibers (i.e., PAN/Pd(acac)2/Ni(acac)2), followed by a controlled thermal process to reduce the metals and to carbonize PAN fibers. All the resulting nanocomposites have uniform dispersion of Pd−Ni alloy NPs within the

■

EXPERIMENTAL SECTION Instruments. Scanning electron microscopy (SEM) experiments were performed on PHILIPS XL-30 ESEM with an accelerating voltage of 20 kV. Transmission electron microscopy (TEM), high-angle annular dark field (HAADF)-scanning transmission electron microscopy (STEM), and STEM-energy dispersive X-ray (EDX) images were taken using a JEM-2010 (HR) microscope operated at 200 kV. X-ray diffraction (XRD) spectra were obtained using a Bruker D8 ADVANCE instrument with Cu Kα radiation (40 kV, 40 mA). X-ray photoelectron spectra (XPS) were recorded on an EscalabMKII spectrometer with Mo Kα X-ray as excitation source. Thermogravimetric analysis (TGA) was performed on a PerkinElmer Pyris 1 TGA instrument. Materials. Polyacrylonitrile (PAN, M w = 860,000), palladium(II) acetylacetonate (Pd(acac)2), nickel(II) acetylacetonate (Ni(acac)2), and dimethylformamide (DMF) were purchased from Sigma-Aldrich. Glucose, sucrose, fructose, and maltose were purchased from Alfa Aesar. All aqueous solutions were prepared using NANO pure water. Preparation of Pd−Ni Alloy Nanoparticle/Carbon Nanofiber Composites. A general preparation of Pd−Ni alloy nanoparticle/carbon nanofiber composites (Pd−Ni/CNF) is described as follows: The spinning solution was prepared by dissolving PAN, Pd(acac)2, and Ni(acac)2 in DMF with different Pd(acac)2/Ni(acac)2 molar ratios (2/8 to 8/2). The electrospinning was performed under an electric field of 100 kV m−1 with a flow rate of 1.2 mL h−1. Reduction and carbonization of PAN/Ni(acac)2/Pd(acac)2 nanofibers were performed as below: (1) stabilization of the precursor nanofibers at 250 °C in air for 3 h; (2) reduction of Ni2+ and Pd2+ and formation of Pd− Ni nanoparticles at 500 °C in a H2/N2 mixture (H2/N2 = 1/2, v/v) for 2 h; (3) carbonization of PAN and formation of Pd−Ni alloy nanoparticles at different carbonization temperatures (Tc) from 700 to 1000 °C for 0.5 h in N2 before cooling down to room temperature. Varying annealing time from 0.5 to 2 h at Tc = 850 °C was also investigated.

■

RESULTS AND DISCUSSION Morphological Structure and Composition. SEM and TEM were used to characterize the morphologies of Pd−Ni/ CNF nanocomposites. Figure 1A−D shows typical SEM images 5899

dx.doi.org/10.1021/ac500811j | Anal. Chem. 2014, 86, 5898−5905

Analytical Chemistry

Article

Figure 1. Typical SEM images of Pd−Ni/CNF nanocomposites with varying Pd/Ni molar feed ratios (Tc = 850 °C): (A) Pd20Ni80/CNF, (B) Pd30Ni70/CNF, (C) Pd50Ni50/CNF, and (D) Pd80Ni20/CNF. (E) Typical particle size distribution histogram of Pd30Ni70/CNF nanocomposite prepared at Tc = 850 °C, and (F) a typical plot of dependence of the particle size on Pd content in the Pd−Ni alloy nanoparticles.

Figure 2. Typical TEM (A, D, G), HRTEM (B, E, H), and STEM-HAADF (C, F, I) images of Pd30Ni70/CNF prepared at different Tcs: 800 °C (A, B, C), 850 °C (D, E, F), and 900 °C (G, H, I).

of Pd−Ni/CNF with different Pd/Ni molar feed ratios, which were carbonized at Tc = 850 °C, i.e., Pd20Ni80/CNF, Pd30Ni70/

CNF, Pd50Ni50/CNF, and Pd80Ni20/CNF. All nanocomposites have comparable CNF diameters in the range of 300−500 nm 5900

dx.doi.org/10.1021/ac500811j | Anal. Chem. 2014, 86, 5898−5905

Analytical Chemistry

Article

fringes, with lattice spacing being 0.212 nm for Pd30Ni70/CNF (Tc = 800 °C), 0.210 nm for Pd30Ni70/CNF (Tc = 850 °C), and 0.209 nm for Pd30Ni70/CNF (Tc = 900 °C), corresponding to the (111) plane of Pd−Ni alloys. The decrease in lattice spacing with Tc may be attributed to an improved crystallinity of Pd−Ni alloy NPs at higher temperatures. The alloy structure of Pd−Ni NPs was verified by HAADFSTEM and XRD. HAADF-STEM pictures of Pd−Ni NPs in Pd30Ni70/CNF show complete overlapping of Pd and Ni atoms in the NPs (Figure 2C,F,I), demonstrating the atomic mixture of Pd and Ni without phase segregation. This alloy structure was further confirmed by XRD analysis. As shown in Figure S4, Supporting Information, for Pd−Ni/CNF composites with varying Pd/Ni molar ratios (i.e., 20/80 to 80/20) at the same Tc of 850 °C, a single (111) plane diffraction peak was observed at a diffraction angle between those of Ni/CNF (2θ = 45.5°) and Pd/CNF (2θ = 40.3°). This agrees with the Pd/Ni binary phase diagram by Lu et al.,35 which shows that Pd and Ni can form an isomorphous alloy with faced-centered cubic (fcc) structure over the entire composition range. The fcc(111) diffraction peak of Pd−Ni/CNF shifts gradually to larger 2θ angles with increasing Ni content, indicating a decreased lattice constant due to the incorporation of a smaller Ni atom. Figure S4B, Supporting Information, displays the XRD spectra of Pd30Ni70/ CNF prepared at different Tcs. For Pd30Ni70/CNF (Tc = 700 °C), two peaks at 2θ = 40−43° were observed, due probably to fcc(111) diffractions of imperfect Pd−Ni alloy crystals. However, no diffraction peak due to monometallic Pd or Ni crystals was revealed, as compared to the diffraction patterns of Pd/CNF and Ni/CNF. At Tc = 750 °C, the resulting Pd−Ni/ CNF composites showed only one (111) plane diffraction peak, suggesting the formation of atomic Pd−Ni alloys. The diffraction peak became more intense and shifted to larger 2θ angles with increasing Tc, indicating a better crystallinity and larger crystallite sizes at higher Tcs. The crystallite sizes of Pd30Ni70/CNF were estimated by Scherrer’s equation to be 10.2 nm for Tc = 800 °C, 32.8 nm for Tc = 850 °C, and 55.4 nm for Tc = 900 °C. These results agree very well with the particle sizes obtained from previous TEM experiments (Table S2, Supporting Information). To acquire more information about the Pd−Ni NPs in the composites, especially the surface composition and electronic properties, XPS spectroscopy of the resulting Pd−Ni/CNF composites was measured. As shown in the high-resolution spectra for Pd(3d) and Ni(2p) (Figure S5A, Supporting Information), the binding energy of Pd(3d5/2) for Pd−Ni/ CNF composites prepared at Tc = 850 °C increases with an increasing content of Ni in the alloy, i.e., 335.2, 335.5, 335.7, and 336.0 eV for Pd80Ni20/CNF, Pd50Ni50/CNF, Pd30Ni70/ CNF, and Pd20Ni80/CNF, respectively. The binding energy of Pd(3d5/2) for Pd/CNF was found to be 335.0 eV. Similarly, the binding energies of Ni(2p) shift to lower energy with increasing Pd content. The shift of core level binding energies of Pd and Ni in Pd−Ni/CNF composites can be ascribed to differences in valence electron density, as well as to changes in crystal field potential, relaxation energy, and work function.36,37 Since surface segregation may happen in alloy nanostructures, the surface composition of Pd was evaluated by XPS. As expected, the relative Pd/Ni ratios on the surface of Pd−Ni/ CNF composites are different from the feed molar ratios and the EDX composition analysis results (Table S1, Supporting Information). For example, Pd30Ni70/CNF composites prepared at different Tcs were found to have surface Pd contents

and length of tens of micrometers. Pd−Ni alloy NPs dispersed uniformly within and on the surface of CNF with a narrow size distribution (Figure 1E). It is interesting to note that the size of Pd−Ni alloy NPs increases almost linearly with Pd content from 26.2 nm for Pd20Ni80 to 60.5 nm for Pd80Ni20 (Figure 1F), which may be accounted for by the larger atomic size of Pd as well as different crystal nucleation and growth mechanisms of Pd−Ni alloys with different Pd contents.33 Carbonization temperature was also found to have substantial effect on the Pd−Ni particle sizes. For Pd−Ni/CNF with a specific molar Pd/Ni feed ratio and with the same annealing time, the particle size increases steadily with Tc. For example, Pd30Ni70/CNF has a particle size of ca. 9.5 nm at Tc = 800 °C, ca. 52.5 nm at Tc = 900 °C, and ca. 120.5 nm at Tc = 1000 °C (Figures 2 and S1, Supporting Information). At all temperatures, the Pd−Ni alloy NPs were found uniformly dispersed within and on the surface of CNF. Similar to the temperature effect, extended annealing also led to an increase in the particle size (Figure S2, Supporting Information). For Pd30Ni70/CNF composites that were carbonized at Tc = 850 °C and annealed for 0.5, 1, and 2 h, the Pd−Ni particle sizes were found to be 32.4 ± 2.6, 49.2 ± 3.4, 75.4 ± 5.3 nm, respectively, which grew almost linearly with annealing time. The uniform dispersion of Pd−Ni NPs was not affected by the prolonged annealing. Interparticle distances appear to be lengthened with increased annealing time; no obvious agglomeration of NPs was found. The effect of Tc and annealing time on NP size may be explained by the thermodynamic size effect.31 Thus, the above studies have shown that the size of Pd−Ni alloy NPs in Pd−Ni/CNF composites could be readily tuned by controlling the feed ratio of metal precursors, Tc, and annealing time. The resulting Pd−Ni NPs dispersed uniformly and embedded firmly within the CNF framework, which contribute to the good stability against high-temperature sintering.34 The compositions of Pd−Ni/CNF composites were characterized by EDX and TGA. A typical EDX spectrum of Pd30Ni70/CNF is given in Figure S3A, Supporting Information, which shows peaks corresponding to C, O, Pd, and Ni elements. The atomic ratio of Pd/Ni was estimated to be 31.4/68.6, which agrees very well with the feed molar ratio of Pd/Ni, i.e., 3/7. Similarly, EDX spectra of all other Pd−Ni/CNF composites showed comparable Pd/Ni atomic ratios to their corresponding feed ratios (Table S1, Supporting Information), indicating a good controllability of composite compositions through the feed ratio of precursors. The total metal content in the Pd−Ni/ CNF composites was evaluated by TGA in air at a heating rate of 10 °C/min. Typical TGA curves of Pd30Ni70/CNF prepared at Tc = 750, 850, and 900 °C are shown in Figure S3B, Supporting Information. An early weight loss of ca. 5% was found for all the composites at around 100 °C, due to evaporation of absorbed water. Major weight loss of Pd30Ni70/ CNF occurred at around 450 °C, which corresponds to the oxidation of carbon fibers. In comparison, Pd30Ni70/CNF (Tc = 750 °C) appears to have lower decomposition temperature (ca. 300 °C) than Pd30Ni70/CNF obtained at higher Tcs. All the Pd30Ni70/CNF composites gave comparable metal residues in the range of 19.6−20.5 wt %. Alloy and Crystal Structures. The alloy and crystal structures of Pd−Ni/CNF composites were studied using HRTEM, HAADF-STEM, and XRD. Figure 2B,E,H displays typical HRTEM pictures of Pd30Ni70/CNF composites that were carbonized and annealed at different temperatures. All composites show good crystallinity and well-resolved lattice 5901

dx.doi.org/10.1021/ac500811j | Anal. Chem. 2014, 86, 5898−5905

Analytical Chemistry

Article

(relative to the total molar amount of Pd and Ni) of 35.2% for Tc = 750 °C, 42.4% for Tc = 800 °C, 54.5% for Tc = 850 °C, and 94.2% for Tc = 900 °C (Table S2, Supporting Information). The higher the Tc, the more is the enrichment of Pd at the Pd−Ni alloy NPs surface. This Pd-enriched surface of Pd−Ni NPs from high-temperature treatment may be attributed to the lower surface energy and larger atomic size of Pd relative to Ni. Electrochemical Properties and Electrocatalytic Activity for Glucose Oxidation. To understand the electrochemical properties of the Pd−Ni/CNF composites, cyclic voltammetry (CV) analyses of paste electrodes (PdNiCFP) prepared from Pd−Ni/CNF with varying Pd/Ni ratios and different Tcs in 0.1 M NaOH aq. were performed. All PdNiCFP electrodes, independent of Pd/Ni ratios and Tc, showed a pair of redox peaks in the potential range of 0.3−0.5 V, corresponding to the Ni(II)/Ni(III) redox couple in alkaline medium (Figures S6 and S7, Supporting Information). In comparison with Ni/CNF-based paste electrode (NiCFP), PdNiCFP exhibited lower anodic peak potentials and enhanced redox current (Figure S6, Supporting Information), indicating facilitated Ni(II)/Ni(III) redox processes in Pd−Ni/CNF composites. Among PdNiCFP electrodes that were prepared from Pd−Ni/CNF with varying Pd/Ni ratios (i.e., 2/8 to 8/2) but the same Tc of 850 °C, paste electrode basing on Pd30Ni70/ CNF showed the lowest anodic peak potential of ca. 0.4 V and the highest peak currents, suggesting a high redox activity. The CV curves of PdNiCFP electrodes from Pd30Ni70/CNF having different Tc were shown in Figure S7, Supporting Information. For Pd30Ni70CFP (Tc = 750 °C), the electrochemical behavior was found to be comparable to that of NiCFP. As the Tc increased to 800 °C or above, however, the reversibility of Pd30Ni70CFP improved dramatically, accompanied by a shift of anodic peak potential toward lower values and higher anodic current. The anodic peak current reached the maximum at Tc = 850 °C and decreased at higher Tc (i.e., 900 °C). The increase in anodic peak current with Tc from 750 to 850 °C may be attributed to a higher electrical conductivity of Pd−Ni/CNF due to a more complete carbonization of CNF fibers, while the decrease in anodic peak current from 850 to 900 °C could be understood by the surface enrichment of Pd and a low content of electrochemically active Ni at higher Tc. To explore the potential of Pd−Ni/CNF composites as efficient electrocatalysts, glucose electro-oxidation on asprepared PdNiCFP was studied. The electron transfer resistance (Rct), which is known to dictate electron transfer kinetics of redox probe at the electrode interface,38 was first measured using electrochemical impedance spectroscopy (EIS) (Figure 3A). Rct of Pd30Ni70CFP (107.2 Ω) is smaller than those of PdCFP (153.3 Ω) and NiCFP (200.5 Ω), suggesting a synergistic effect of Pd and Ni in Pd−Ni alloy NPs, which facilitates the electron transfer during redox processes. This agrees with previous results from the CV study, where lower anodic peak potentials and higher redox current were observed for PdNiCFP electrodes relative to NiCFP. The electrochemical response of Pd−Ni/CNF-based electrodes to sugar was probed in 0.1 M NaOH aq. Upon addition of 1 mM of glucose, PdCFP showed no change in the featureless CV curve in the potential range of 0.1−0.6 V (Figure S8A, Supporting Information), whereas Pd30Ni70CFP (Tc = 850 °C) exhibited substantial increase in both anodic and cathodic peak currents at ca. 0.40 and 0.32 V, respectively, as well as a positive shift of anodic peak potential by ca. 0.02 V (Figure 3B). In comparison with NiCFP electrode (Figure S8B, Supporting

Figure 3. (A) Nyquist plots of PdCFP, Pd30Ni70CFP, and NiCFP electrodes in 0.1 M KCl solution containing 5 mM [Fe(CN)6]3−/4−. (B) CV curves of Pd30Ni70CFP electrode in 0.1 M NaOH aq. with (red line) and without (black line) 1 mM glucose. Scan rate: 50 mV s−1. (C) Current responses of Ni-metal, NiCFP, and Pd-NiCFP electrodes based on Pd−Ni/CNF with different Pd/Ni ratios (Tc = 850 °C) upon successive addition of 0.2 mM glucose at a potential of 0.40 V vs Ag/ AgCl. (D) Current responses of Pd30Ni70CFP electrode (Tc = 850 °C) upon successive addition of glucose solutions with varying concentrations; applied potential: 0.40 V vs Ag/AgCl.

Information), Pd30Ni70CFP has a lower potential (by ca. 0.05 V) and much higher current for the anodic oxidation peak in the presence of glucose, indicating an enhanced electrocatalytic activity for Pd−Ni/CNF composites. Similar electrocatalytic phenomena were observed for other sugars such as fructose, sucrose, and maltose (Figure S9, Supporting Information). No electrode fouling was revealed during 50 cycles of voltammetric measurements. Given that many glucose sensors using metal or alloy electrodes are susceptible to chloride ion poisoning,39 the antifouling ability of Pd30Ni70CFP electrode was probed by performing CV measurements in 0.1 M NaOH aq. containing 0.2 M NaCl. No change in the current response of Pd30Ni70CFP to glucose due to the presence of a high concentration of Cl− was discernible (Figure S10, Supporting Information), confirming the good resistance to surface fouling. Amperometric Detection of Glucose. Nonenzymatic glucose sensors taking advantage of electrocatalytic oxidation of glucose are currently of intense research interest. Given the significant changes in the redox current intensity of PdNiCFP upon addition of glucose, amperometric detection of glucose using Pd−Ni/CNF-based electrodes appear promising. Thus, Pd−Ni/CNF composite paste electrodes having different Pd/ Ni ratios (Tc = 850 °C) were further evaluated for their amperometric responses to glucose at a potential of 0.4 V in 0.1 M NaOH aq. As shown in Figure 3C,D, upon successive addition of glucose solutions with constant (i.e., 0.2 mM) or varying concentrations, significant and quick responses were observed for all the PdNiCFP electrodes with 95% steady-state current achievable within 2 s. In comparison, the Ni-metal electrode showed the lowest sensitivity among all the tested electrodes. The NiCFP electrode exhibited much improved 5902

dx.doi.org/10.1021/ac500811j | Anal. Chem. 2014, 86, 5898−5905

Analytical Chemistry

Article

sensitivity relative to Ni-metal attributable to the increased electrocatalytically active surface area as a result of the good dispersion of Ni NPs in supporting CNF. In comparison with NiCFP, all PdNiCFP electrodes showed even better sensitivity to glucose, except for Pd80Ni20CFP that had lower current responses than NiCFP, due probably to the low surface content of electrocatalytically active Ni. In general, the amperometric sensitivity of the tested electrodes follow the order of Ni-metal < Pd80Ni20CFP < NiCFP < Pd50Ni50CFP < Pd20Ni80CFP < Pd30Ni70CFP electrode (Figure 3C). The Pd30Ni70CFP electrode showed the highest sensitivity to glucose, which is more than 2-fold of that for NiCFP. The calibration plot of Pd30Ni70CFP electrode for glucose detection could be obtained in a wide range from 0.1 μM to 5.4 mM (R = 0.9992) with the limit of detection (LOD) of ca. 60 nM (S/N = 3). Such a glucose detection sensitivity is superior to most state-of-the-art nonenzymatic glucose sensors such as Pd-SWNT40 and PtNiERGO/GCE electrodes,38 which can be attributed to the unique electronic structure of atomic Pd−Ni alloy as revealed from the XPS study (Figure S5, Supporting Information). The modified electronic structure in Pd−Ni alloy should alter the surface active sites and its local reactivity,41 contributing to the enhanced sensitivity to glucose. Meanwhile, the geometry of bimetallic alloy,38 the appropriate surface Ni content, and the uniform dispersion of Pd−Ni NPs in conductive CNF framework should all facilitate the electron transport and mass transfer in Pd−Ni/CNF-based electrodes and contribute to high detection sensitivity. Additionally, the Pd30Ni70/CNF modified electrode also exhibited desirable reproducibility. A relative standard deviation of 4.8% was obtained toward 0.1 mM glucose for 6 different Pd30Ni70/CNF modified electrodes. Application in a Liquid Chromatography-Electrochemical System for Sugar Detection. To further explore the potential of Pd−Ni/CNF composites for sugar detection, the application of Pd30Ni70CFP electrode as an amperometric detector in a flow system, i.e., liquid chromatography-electrochemical system (LC-EC) and flow-injection analysis (FIA), was investigated. A sugar mixture of monosaccharides (i.e., glucose and fructose) and disaccharides (i.e., sucrose and maltose) was prepared and used to evaluate the detection sensitivity. The sugars were subject to separation by an anionexchange column (Hamilton RCX-10) and detection at the Pd30Ni70CFP electrode. As shown by the chromatographic curves (a) and (c) in Figure 4A, for sugar concentrations of 10 and 0.02 μM, respectively, in a mobile phase of 0.04 M NaOH aq., excellent signal strength and separation of the four sugars were obtained at the Pd30Ni70CFP electrode with a low application potential of 0.4 V. In comparison with Ni-metal electrode (Figure 4A, curve b), Pd30Ni70CFP electrode exhibited much higher sensitivity to the sugar mixture at a concentration of 10 μM. The LODs of Pd30Ni70CFP for glucose, fructose, sucrose, and maltose were found to be 7, 9, 15, and 20 nM, respectively. The linear ranges with correlation coefficient (R) larger than 0.9990 are typically from ca. 0.03 μM up to a few hundred (e.g., 500−800) μM (Table S3, Supporting Information). In comparison with previously reported LOD values for electrocatalytic electrodes, for example, Ni alloy electrodes (LOD = 0.1 μM)8,9 and Ni-based CME electrodes (LOD = 0.33−0.55 μM),7 Pd30Ni70CFP electrode exhibited superior detection performances. Pd30Ni70CFP electrode was also found to have excellent signal reproducibility and electrocatalytic stability in the flow system. FIA experiments showed little signal loss in peak

Figure 4. Sugar detection at Pd30Ni70CFP electrode in a LC-EC system for (A) a mixture of glucose (1), fructose (2), sucrose (3), and maltose (4) and (B) a commercial honey sample (curves a and c for detection signals from Pd30Ni70CFP electrode, and curve b for signals from Nimetal electrode). Mobile phase was 0.04 M NaOH aq.; separation column was Hamilton RCX-10 anion-exchange column (250 × 4 mm i.d.) plus CarboPac PA 100 guard column (50 × 4 mm i.d.); injection volume, 10 μL; concentration of sugars, 10 μM for curves a and b and 0.02 μM for curve c; flow rate, 0.6 mL min−1; applied potential, 0.40 V vs Ag/AgCl.

current intensity for Pd30Ni70CFP electrode during 60 consecutive glucose injections (Figure S11A, Supporting Information). In contrast, the peak current intensity dropped by ca. 38% of its initial value for Ni-metal electrode in the same experiment (Figure S11B, Supporting Information). The signal reproducibility of Pd30Ni70CFP electrode was also found to be much better than Ni-metal electrode, as evidenced by the smaller relative standard deviation (RSD) of 1.45% for Pd30Ni70CFP and 10.3% for Ni-metal. To probe the longterm stability of Pd30Ni70CFP in the flow system, the day-to-day stability was evaluated for a period of one month by performing five injections of glucose a day. As expected, excellent stability of the electrode and signal reproducibility were found, with the RSD less than 4.7%. The good stability of Pd−Ni/CNF as electrochemical detecting electrode for sugars in a flow system can be accounted for by the Pd−Ni NPs that are uniformly dispersed and embedded within the CNF framework, which prevented the electrocatalytically active NPs in the composites from detachment and agglomeration and also depressed the Ni corrosion in alkaline solution. Meanwhile, the chemically inert CNF matrix may absorb to a certain degree impurities and help to prevent surface fouling,42 contributing to the stability of PdNiCFP electrodes. Application in Honey Analysis. To further evaluate the application potential of Pd−Ni/CNF as electrochemical detector in flow systems (e.g., LC-EC), analysis of commercial honey for their sugar compositions using Pd30Ni70CFP electrode was carried out. It is known that typical honeys contain up to 27 sugars, with most of them being monosaccharides, i.e., glucose (ca. 30−35 wt %) and fructose (ca. 36−42 wt %). Sucrose has been commonly used as additive in honey; however, its content is regulated by government. For 5903

dx.doi.org/10.1021/ac500811j | Anal. Chem. 2014, 86, 5898−5905

Analytical Chemistry

Article

(5) Wang, C. G.; Yin, L. W.; Zhang, L. Y.; Gao, R. J. Phys. Chem. C 2010, 114, 4408−4413. (6) Zadeii, J. M.; Marioli, J.; Kuwana, T. Anal. Chem. 1991, 63, 649− 653. (7) Fermier, A. M.; Colón, L. A. J. High Resolut. Chromatogr. 1996, 19, 613−616. (8) Marioli, J. M.; Luo, P. F.; Kuwana, T. Anal. Chim. Acta 1993, 282, 571−580. (9) Mu, Y.; Jia, D. L.; He, Y. Y.; Miao, Y. Q.; Wu, H. L. Biosens. Bioelectron. 2011, 26, 2948−2952. (10) You, T. Y.; Niwa, O.; Chen, Z. L.; Hayashi, K.; Tomita, M.; Hirono, S. Anal. Chem. 2003, 75, 5191−5196. (11) Shin, Y.; Bae, I. T.; Arey, B. W.; Exarhos, G. J. J. Phys. Chem. C 2008, 112, 4844−4848. (12) Wu, B. H.; Hu, D.; Kuang, Y. J.; Liu, B.; Zhang, X. H.; Chen, J. H. Angew. Chem., Int. Ed. 2009, 48, 4751−4754. (13) Wu, B. H.; Kuang, Y. J.; Zhang, X. H.; Chen, J. H. Nano Today 2011, 6, 75−90. (14) Hao, C.; Ding, L.; Zhang, X. J.; Ju, H. X. Anal. Chem. 2007, 79, 4442−4447. (15) Song, W.; Li, D. W.; Li, Y. T.; Li, Y.; Long, Y. T. Biosens. Bioelectron. 2011, 26, 3181−3186. (16) Hsin, Y. L.; Hwang, K. C.; Yeh, C. T. J. Am. Chem. Soc. 2007, 129, 9999−10010. (17) Guo, C. X.; Guai, G. H.; Li, C. M. Adv. Energy Mater. 2011, 1, 448−454. (18) Guo, C. X.; Yang, H. B.; Sheng, Z. M.; Lu, Z. S.; Song, Q. L.; Li, C. M. Angew. Chem., Int. Ed. 2010, 49, 3014−3017. (19) Wang, D.; Li, Z. C.; Chen, L. W. J. Am. Chem. Soc. 2006, 128, 15078−15079. (20) Kim, Y. T.; Ohshima, K.; Higashimine, K.; Uruga, T.; Takata, M.; Suematsu, H.; Mitani, T. Angew. Chem., Int. Ed. 2006, 45, 421−425. (21) Zhang, F.; Yuan, C. Z.; Zhu, J. J.; Wang, J.; Zhang, X. G.; Lou, X. W. Adv. Funct. Mater. 2013, 23, 3909−3915. (22) Nie, H. R.; Shen, X. X.; Zhou, Z. H.; Jiang, Q. S.; Chen, Y. W.; Xie, A.; Wang, Y.; Han, C. C. Carbohydr. Polym. 2011, 85, 681−686. (23) Yang, D. Y.; Niu, X.; Liu, Y. Y.; Wang, Y.; Gu, X.; Song, L. S.; Zhao, R.; Ma, L. Y.; Shao, Y. M.; Jiang, X. Y. Adv. Mater. 2008, 20, 4770−4775. (24) Miao, Y. E.; Zhu, G. N.; Hou, H. Q.; Xia, Y. Y.; Liu, T. X. J. Power Sources 2013, 226, 82−86. (25) Miao, Y. E.; Wang, R. Y.; Chen, D.; Liu, Z. Y.; Liu, T. X. ACS Appl. Mater. Interfaces 2012, 4, 5353−5359. (26) Huang, J. S.; Wang, D. W.; Hou, H. Q.; You, T. Y. Adv. Funct. Mater. 2008, 18, 441−446. (27) Guo, Q. H.; Zhou, X. P.; Li, X. Y.; Chen, S. L.; Seema, A.; Greiner, A.; Hou, H. Q. J. Mater. Chem. 2009, 19, 2810−2816. (28) Hu, X. B.; Zhao, B. Y.; Song, Y. P.; Huang, B.; Cheng, G.; Hu, K. A. Carbon 2005, 43, 79−85. (29) Hu, G. Z.; Zhou, Z. P.; Guo, Y.; Hou, H. Q.; Shao, S. J. Electrochem. Commun. 2010, 12, 422−426. (30) Zhang, J.; Zhu, X. L.; Dong, H. F.; Zhang, X. J.; Wang, W. C.; Chen, Z. D. Electrochim. Acta 2013, 105, 433−438. (31) Hou, H. Q.; Deneker, D. H. Adv. Mater. 2004, 16, 69−73. (32) Nam, S. H.; Shim, H. S.; Kim, Y. S.; Dar, M. A.; Kim, J. G.; Kim, W. B. ACS Appl. Mater. Interfaces 2010, 2, 2046−2052. (33) Watson, K. D.; Tatsinkou Nguelo, S. E.; Desgranges, C.; Delhommelle, J. CrystEngComm 2011, 13, 1132−1140. (34) Hou, H. Q.; Schaper, A. K.; Weller, F.; Greiner, A. Chem. Mater. 2002, 14, 3990−3994. (35) Lu, P.; Teranishi, T.; Asakura, K.; Mikio, M.; Toshima, N. J. Phys. Chem. B 1999, 103, 9673−9682. (36) Wang, K. W.; Chung, S. R.; Liu, C. W. J. Phys. Chem. C 2008, 112, 10242−10246. (37) Kim, K. S.; Winograd, N. Chem. Phys. Lett. 1975, 30, 91−95. (38) Gao, H. C.; Xiao, F.; Ching, C. B.; Duan, H. W. ACS Appl. Mater. Interfaces 2011, 3, 3049−3057. (39) Sun, Y.; Buck, H.; Mallouk, T. E. Anal. Chem. 2001, 73, 1599− 1604.

example, most commercially available honeys contain 0.5−1.5 wt % sucrose, and the top limit is no more than 5 wt % for many European countries.43 Thus, quantitative analysis of sugars, especially sucrose, is an important test for evaluating the quality of honey. As shown in Figure 4B, the Pd30Ni70CFP electrode was found to be an effective detector for assaying the major sugar content of honey, i.e., glucose, fructose, sucrose, and maltose. In contrast, the Ni-metal electrode showed no signal to sucrose and maltose in honey (Figure 4B-b). Given the positive results, five different honey samples were further analyzed for their sugar contents using the Pd30Ni70CFP electrode in the LCEC system (Table S4, Supporting Information). All the chromatographic analyses showed effective detection and quantification of the major sugar contents, with sucrose contents being in the range of 0.5−5%, except for one of the samples (i.e., #1 in Table S4, Supporting Information) that contains 8.65% of sucrose, suggesting adulterated honey.

■

CONCLUSIONS Novel bimetallic Pd−Ni alloy NP/CNF nanocomposites were successfully prepared by electrospinning the precursor nanofibers, i.e., PAN/Pd(acac)2/Ni(acac)2, followed by a controlled thermal treatment for reduction of metals and carbonization of PAN fibers. The resulting Pd−Ni alloy NPs had particle sizes in the range of 9−120 nm, depending on the Pd/Ni feed ratio, carbonization temperature, and annealing time. A uniform dispersion of Pd−Ni NPs within the framework or on the surface of CNF was observed. The atomic alloy structure of Pd−Ni NPs in the composites was verified. Due to the formation of bimetallic alloy, the electronic properties of Pd and Ni were modified, as evidenced by the XPS study, which contribute to the enhanced redox properties of Pd−Ni/CNFbased electrodes. Electrochemical studies on Pd−Ni/CNFbased electrodes showed excellent sensitivity and stability toward sugar oxidation, as well as good application potential in flow systems for continuous analysis of sugars.

■

ASSOCIATED CONTENT

* Supporting Information S

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Fax: (+86) 431-85262850. Author Contributions §

Q.G. and D.L. contributed equally to this work.

Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS We are grateful for the financial support from the National Natural Science Foundation of China (No. 21222505). REFERENCES

(1) Godfray, H. C. J.; Beddington, J. R.; Crute, I. R.; Haddad, L.; Lawrence, D.; Muir, J. F.; Pretty, J.; Robinson, S.; Tomas, S. M.; Toulmin, C. Science 2010, 327, 812−818. (2) Su, L.; Jia, W. Z.; Hou, C. J.; Lei, Y. Biosens. Bioelectron. 2011, 26, 1788−1799. (3) Johnson, D. C.; LaCourse, W. R. Anal. Chem. 1990, 62, 589A− 597A. (4) LaCourse, W. R.; Johnson, D. C. Anal. Chem. 1993, 65, 50−55. 5904

dx.doi.org/10.1021/ac500811j | Anal. Chem. 2014, 86, 5898−5905

Analytical Chemistry

Article

(40) Meng, L.; Jin, J.; Yang, G. X.; Lu, T. H.; Zhang, H.; Cai, C. X. Anal. Chem. 2009, 81, 7271−7280. (41) Lu, Y. Z.; Chen, W. ACS Catal. 2012, 2, 84−90. (42) McCreery, R. L. Chem. Rev. 2008, 108, 2646−2687. (43) Sue, J. W.; Hung, C. J.; Chen, W. C.; Zen, J. M. Electroanalysis 2008, 20, 1647−1654.

5905

dx.doi.org/10.1021/ac500811j | Anal. Chem. 2014, 86, 5898−5905