Facile Preparation of Graphene-Copper Nanoparticle Composite by in

Nov 19, 2011 - School of Pharmacy, Fudan University, Shanghai 201203, China ... Mussel-Inspired Architecture of High-Flux Loose Nanofiltration Membran...
0 downloads 0 Views 904KB Size
Download PDF
ARTICLE pubs.acs.org/ac

Facile Preparation of Graphene-Copper Nanoparticle Composite by in Situ Chemical Reduction for Electrochemical Sensing of Carbohydrates Qiwen Chen, Luyan Zhang, and Gang Chen* School of Pharmacy, Fudan University, Shanghai 201203, China

bS Supporting Information ABSTRACT: A novel graphene-copper nanoparticle composite was prepared by the in situ chemical reduction of a mixture containing graphene oxide and copper(II) ions using potassium borohydride as a reductant. It was mixed with paraffin oil and packed into one end of a fused capillary to fabricate microdisc electrodes for sensing carbohydrates. The morphology and structure of the graphene-copper nanoparticle composite were investigated by scanning electron microscopy, X-ray diffraction, and Fourier transform-infrared spectroscopy. The results indicated that copper nanoparticles with an average diameter of 20.8 nm were successfully deposited on graphene nanosheets to form a well interconnected hybrid network. The analytical performance of these unique graphene-copper nanoparticle composite paste electrodes was demonstrated by sensing five carbohydrates in combination with cyclic voltammetry and capillary electrophoresis (CE). The advantages of the composite detectors include higher sensitivity, satisfactory stability, surface renewability, bulk modification, and low expense of fabrication. They should find applications in microchip CE, flowing-injection analysis, and other microfluidic analysis systems.

G

raphene, as an important two-dimension nanomaterial, has attracted tremendous scientific and technological attention because of its unique nanostructure and properties since it was first reported in 2004.1,2 It holds great promise for potential applications in many technological fields such as sensors, electronics, supercapacitors, batteries, fuel cells, solar cells, nanocomposites, and hydrogen storage because of its high specific surface area, excellent thermal and electrical conductivity, and strong mechanical strength.3 5 The existing approaches to producing graphene include mechanical exfoliation, chemical vapor deposition, and chemical or thermal reduction of graphite oxide.2 Among them, the last one is considered to be the most economical approach to preparing graphene. Usually, graphite powder is oxidized with some strong oxidants and exfoliated to form graphene oxide (GO) that can be reduced to prepare graphene. As a chemical precursor of graphene, GO is basically a single atomic layer of carbon covered with epoxy, hydroxyl, carbonyl, and carboxyl groups.6 It has been employed to prepare nanocomposites,7 antibacterial paper,8 chemically modified graphene,9 conjugate with proteins,10 etc. Because GO bears abundant oxygen-containing functional groups, it can be well dispersed in aqueous solution and should find a wide range of applications in the preparation of graphenebased composites. Graphene has been employed to prepare electrochemical sensors and biosensors because of its excellent electrical conductivity and electrocatalytic activity.11 14 A variety of approaches have been developed for the fabrication of graphene-based r 2011 American Chemical Society

electrodes. Surface modification was the commonly used method. Graphene was usually dispersed in solvent15 or polymer solutions (such as Nafion16 and chitosan17 solutions) that were subsequently modified on the surface of electrodes. In addition, electrochemical polymerization18 has also been employed to modify graphene on electrodes. Recently, a method based on in situ polymerization was developed for the facile fabrication of graphene-polymer composition electrodes as the end-column amperometric detectors of capillary electrophoresis (CE). The mixtures of graphene with the prepolymers of urea and formaldehyde,19 styrene,20 or methyl acrylate21 were filled into electrode tubes. The filled graphenecontaining mixtures would cure to form graphene-polymer composite electrodes. It has been demonstrated by many groups that electrically conductive graphene showed strong electrocatalytic activity when it was employed to improve the electrochemical response of some bioactive substances.22 24 The high electrical conductivity of graphene and its ability to promote the electrontransfer reactions suggest great promise for electrochemical sensing. As the products of photosynthesis, carbohydrates are the most abundant class of organic compounds in living organisms. A variety of methods based on liquid chromatography,25 CE,26 and microchip CE27 have been developed for their analysis. Prior to ultraviolent or fluorescence detection, carbohydrates need to Received: August 28, 2011 Accepted: November 18, 2011 Published: November 19, 2011 171

dx.doi.org/10.1021/ac2022772 | Anal. Chem. 2012, 84, 171–178

Analytical Chemistry

ARTICLE

be derivatized because there is no chromophore or fluorophore in them. Underivatized carbohydrates can be directly detected at platinum or gold electrodes using pulsed amperometric detection (AD).28,29 The applied potential on these electrodes must be continuously pulsed to reduce electrode poisoning. Alternatively, copper electrodes have been widely applied in the AD of carbohydrates in a strongly alkaline medium without any electrode fouling.30,31 Recently, a variety of nanomaterials have been employed to fabricate electrodes for sensing carbohydrates. Escarpa et al. prepared nickel and nickel-copper nanowires by electroplating using alumina templates.32,33 The magnetic nanowires were loaded on screen-printed carbon electrodes with the aid of magnets for the determination of several carbohydrates. In addition, NiO nanoparticles34 and CuO nanorod bundles35 have been used in the direct AD of carbohydrates. It was been demonstrated that these nanomaterials showed strong electrocatalytic activity toward the oxidation of carbohydrates. Because graphene has a large specific surface area and high electrical conductivity, it is interesting to deposit copper nanoparticles on it for electrochemical sensing of carbohydrates. To the best of our knowledge, the existing graphene-copper nanoparticle hybrids were prepared by encapsulating copper nanoparticles with graphene. In 2009, Bin et al. prepared graphene-encapsulated copper nanoparticle composite by reducing CuCl2-graphite intercalation compounds.36 Another approach was developed by Luechinger et al. to deposit ∼3 nm thick graphene layers on copper nanoparticles on a large scale based on a reducing flame technique.37 Because the graphene shells were prepared to protect the copper cores from being oxidized, these grapheneencapsulated copper nanoparticles are not suitable for electrochemical sensing of carbohydrates. In this work, the graphene-copper nanoparticle composite was prepared by reducing a mixture containing GO nanosheets and copper(II) ions. The copper particles were directly deposited on the surface of graphene. The prepared composite was then mixed with paraffin oil and packed into fused silica capillaries to form microdisc electrodes. The preparation details and characterization of the novel composite as well as the feasibility and performance of the composite electrodes are reported in the following sections in combination with the measurement of carbohydrates by cyclic voltammetry (CV) and CE.

nitrate and 6 g potassium permanganate were successively added into the mixture in an ice bath. Note that both compounds should be added slowly to prevent the temperature from exceeding 20 °C. After the reaction was allowed to proceed in a 35 °C water bath for 30 min, 92 mL of doubly distilled water was gradually added. Then, the temperature of the water bath was increased to 98 °C. The reaction mixture was maintained at this temperature for 40 min to increase the oxidation degree of the product. After the volume of the resultant suspension was adjusted to 280 mL with doubly distilled water, 6 mL of hydrogen peroxide solution (30%, w/w) was added while the color of the suspension changed from brown to bright-yellow. The prepared oxidized graphite could be easily isolated from the solution by vacuum filtration. The sulfuric acid and salt impurities in the crude product were removed by washing with 5% (w/w) hydrochloric acid, 1% (w/w) hydrochloric acid, and doubly distilled water successively with the aid of vacuum filtration. The wet product was dried in a vacuum at 50 °C. Preparation of Graphene-Copper Nanoparticle Composite. To prepare the graphene-copper nanoparticle composite, 0.4 g of oxidized graphite powder was dispersed in 100 mL of doubly distilled water and sonicated for 1 h using an ultrasonic cleaner (SKQ-2200, frequency 56 kHz, 100 W) to exfoliate oxidized graphite particles to GO sheets. After a volume of 20 mL of mixture solution containing 2.5 g of CuSO4 3 5H2O (0.01 mol) and 0.8 g of EDTA 3 2Na 3 2H2O was added, the mixture solution was sonicated for 10 min. Then, a volume of 20 mL aqueous solution containing 2.24 g of KOH (0.04 mol) and 0.27 g of KBH4 (0.005 mol) was gradually added to the mixture solution of GO, CuSO4, and EDTA 3 2Na. After the mixture solution was stirred for 30 min at 30 °C, its color changed from gray to black brown. Subsequently, 1.0 g of KBH4 was dissolved in the mixture. The prepared mixture was transferred into a 200 mL roundbottom flask that was assembled in a far-infrared (IR)-assisted reduction system (Figure S1 in the Supporting Information). It was then exposed to far IR radiation and refluxed for 60 min to reduce GO to graphene. The obtained black graphene-copper nanoparticle composite could be easily isolated from the solution by vacuum filtration and was purified by washing with copious amount of doubly distilled water. Finally, it was washed with absolute ethanol and dried in a vacuum. The content of graphene in the composite was determined to be 22.7% (w/w) by dissolving the copper constituent in it with 2 M HNO3 aqueous solution. Figure S1 in the Supporting Information illustrates the schematic and photograph of the far IR-assisted reduction system used in this work. It consists of an ac voltage regulator (1000 W, Shanghai Renmin Electrical Apparatus Works, Shanghai, China), a far IR generator (900 W at 220 V; wavelength range, 2.5 15 μm; Shanghai Shuocun Hardware & Electrical Appliance Co. Ltd., Shanghai, China), and a 200 mL single neck flask that was connected to an Allihn condenser. The far IR generator was assembled under the round-bottom flask. The ac voltage regulator was employed to adjust the ac voltage (0 220 V) applied to the far IR generator to control its output power. The distance between it surface and the bottom of the flask was ∼2 cm. In this work, the voltage applied to the infrared generator was 150 V and the calculated power was approximately 418.4 W. Preparation of Copper Nanoparticles and GraphiteCopper Nanoparticle Composite. Copper nanoparticles could be facilely prepared by dropping a mixture solution of 2.24 g of KOH (0.04 mol) and 0.27 g of KBH4 (0.005 mol) (20 mL,

’ EXPERIMENTAL SECTION Reagent and Solutions. Cupric sulfate pentahydrate (CuSO4 3 5H2O), potassium borohydride (KBH4), potassium hydroxide, ethylenediaminetetraacetic acid disodium salt dehydrate (EDTA 3 2Na 3 2H2O), paraffin oil, graphite powder, sodium nitrate, potassium permanganate, hydrogen peroxide solution (30% w/w), hydrazine hydrate (85% w/w), and sulfuric acid (98% w/w) were all supplied by SinoPharm (Shanghai, China). Mannitol, sucrose, lactose, glucose, and fructose were supplied by Sigma (St. Louis, MO). Other chemicals were analytical grade. The stock solutions (100 mM) of the five carbohydrates were prepared in doubly distilled water and were diluted to the desired concentration with 75 mM NaOH aqueous solution. Preparation of Oxidized Graphite. Oxidized graphite was synthesized from graphite by a modified Hummers method.38 Briefly, 2 g of graphite powder was dispersed in 46 mL of sulfuric acid (98% w/w) under agitation. Subsequently, 1.2 g sodium 172

dx.doi.org/10.1021/ac2022772 |Anal. Chem. 2012, 84, 171–178

Analytical Chemistry

ARTICLE

nanoparticle composite were also fabricated following the same procedures. The content of the paraffin oil in all pastes was 25% (w/w). Apparatus. The CE AD system used in this work has been described in our previous report.39 A (30 kV high-voltage dc power supply (Shanghai Institute of Nuclear Research, China) provided a separation voltage between the two ends of the capillary. The inlet of the capillary was held at a positive potential while the outlet of the capillary was maintained at ground. The separations were carried out in a 40 cm length of 25 μm i.d. and 360 μm o.d. fused silica capillary (Polymicro Technologies, Phoenix, AZ). A three-electrode electrochemical cell consisting of a 320 μm diameter paste microdisc detection electrode prepared in this work, a platinum auxiliary electrode, and an Ag/AgCl wire as the reference electrode was used in combination with a BAS LC-4C amperometric detector (Bioanalytical Systems Inc., West Lafayette, IN) for CE detection. The composite detection electrode was positioned opposite the outlet of the capillary with the aid of a 3-D adjustable amperometric setup and arranged in a wall-jet configuration (Figure S2 in the Supporting Information). The schematics and operation procedures of the 3-D adjustable alignment device can be found in the Supporting Information. The surface morphologies of graphene, copper nanoparticles, graphene-copper nanoparticle composite, and graphite-copper nanoparticle composite were observed by a scanning electron microscope (PHILIPS XL 30). X-ray diffraction (XRD) measurements were carried out using a Rigaku D/max-rB diffractometer (Rigaku, Tokyo, Japan) with CuK-α1 radiation (40 kV, 60 mA). The Fourier transform-infrared spectroscopy (FT-IR) spectra of GO, graphene, copper nanoparticles, and graphenecopper nanoparticle composite were measured using a FT-IR spectrometer (NEXUS470, Nicolet). Sample Preparation. An accurate volume of 10 mL of bovine milk (Bright Diary, Shanghai, China) was diluted with 75 mM NaOH aqueous solution to 100 mL. After vacuum filtration, the filtrate was diluted with 75 mM NaOH aqueous solution at a ratio of 25:1 for CE analysis. To prepare the sample solution of the honey peach or banana, 10 g of the fruit tissue was homogenized with 50 mL of doubly distilled water. The homogenate was collected and diluted to 100 mL. After it was centrifuged at 4000 rpm for 10 min, the supernatant was diluted with 75 mM NaOH aqueous solution at a ratio of 5:1 for the following CE analysis. CV Measurements. CV was performed using a CHI 830B electrochemical analyzer (Shanghai Chen-Hua Instruments Co., Shanghai, China) in combination with the three-electrode electrochemical cell mentioned above. The working electrode was a graphene-copper nanoparticle composite paste electrode, a graphite-copper nanoparticle composite paste electrode, a copper nanoparticle paste electrode, or a graphene paste electrode. Prior to electrochemical sensing, the electrodes were pretreated by cyclic voltammetric scanning over the potential range of 0 0.80 V at a scan rate of 50 mV/s. CE Procedures. Prior to CE analysis, the surface of the detection electrode was positioned carefully opposite the outlet of the separation capillary through a guiding metal tube (Figure 2(i),(j) in the Supporting Information). The gap distance between the disk electrode and the channel outlet was adjusted to be ∼50 μm. Both the separation and injection voltages were 12 kV for convenience. The potential applied to the detection electrode was +0.65 V (vs Ag/AgCl). Before use, the fused-silica capillary

Figure 1. (A C) Schematic illustrating the fabrication process of a graphene-copper nanoparticle composite paste electrode, (D) photograph of a graphene-copper nanoparticle composite paste electrode, and (E) microscopic photograph of the composite-filled end of the prepared electrode. (A) Inserting a piece of copper wire (b, 10 cm long, 150 μm diameter) into a 3.5 cm long fused-silica capillary (a, 320 μm i.d.  450 μm o.d.); (B) applying hot melt adhesive (c) to glue (b) in place, (C) filling the empty end of part (a) with a mixture of graphene-copper nanoparticle composite and paraffin oil (3:1, w/w) (d).

solution A) into 20 mL of solution containing 2.5 g of CuSO4 3 5H2O (0.01 mol) and 0.8 g of EDTA 3 2Na 3 2H2O (solution B) at 30 °C while the solution was violently stirred. The prepared copper nanoparticles could be easily isolated from the solution by vacuum filtration and were purified by washing with copious amount of doubly distilled water and absolute ethanol successively. Finally, they were dried in a vacuum. Graphite-copper nanoparticle composite could be prepared by dropping solution A into solution B with 0.2 g of graphite powder dispersed inside. Other procedures were the same as those of copper nanoparticles. Electrode Fabrication. The fabrication process of the composite electrodes was illustrated in Figure 1A C. Prior to fabrication, graphene-copper nanoparticle composite powder was mixed with paraffin oil at a ratio of 3:1 (w/w) on a glass plate. Subsequently, a piece of copper wire (Figure 1B, 10 cm long and 150 μm diameter) was inserted into a 3.5 cm long fused silica capillary (Figure 1A, 320 μm i.d.  450 μm o.d., Hebei Yongnian Ruipu Chromatogram Equipment Co., Ltd., Hebei, China) and a ∼2.5 mm long opening was left for the subsequent filling of the paste. Hot melt adhesive (Figure 1C) was applied to another end of the capillary to glue the copper wire in place. The mixture was then pressed into the empty end of the capillary to a depth of ∼4.5 mm. The paste should touch the end of the copper wire inside the capillary tightly for electrical contact. Finally, the paste electrode was smoothed on a piece of weighing paper to get a flat surface. For comparison, paste electrodes of graphene, copper nanoparticles, and graphite-copper 173

dx.doi.org/10.1021/ac2022772 |Anal. Chem. 2012, 84, 171–178

Analytical Chemistry

ARTICLE

Figure 2. SEM images of (a and b) graphene-copper nanoparticle composite, (c) graphene, and (d) copper nanoparticles. Conditions: accelerating voltage, 20 kV; magnification, (a)  10 000, (b d)  20 000.

for CE separation was rinsed with 75 mM NaOH aqueous solution for at least 10 min. Samples were injected electrokinetically at 12 kV for 6 s. The amperometric detector was on during the injection procedures. Moreover, sample solutions, standard solutions, and the separation medium were all filtered through a syringe polypropylene filter (0.22 μm) prior to their use. Peak identification was performed by the standard addition method.

’ RESULTS AND DISCUSSION In this work, graphene-copper nanoparticle composite was prepared by in situ reduction for sensing carbohydrates electrochemically. Figure 2 illustrates the scanning electron microscopy (SEM) images of graphene-copper nanoparticle composite, graphene, and copper nanoparticles. Obviously, the surface morphology of the composite is much different from those of graphene and copper nanoparticles. As shown in Figure 2a,b, a great amount of copper nanoparticles are observed on graphene. It can be seen clearly in the SEM images of the composite that copper nanoparticles are well dispersed and embedded throughout the graphene matrix and an interconnected hybrid network has formed. This conductive copper-graphene network may establish electrical conduction pathways throughout the whole composite, which is responsible for the electrical conductivity and electrochemical sensing. The FT-IR spectra of graphene, copper nanoparticles, graphenecopper nanoparticle composite, and GO are shown in Figure 3. Obviously, there is no observable absorption peak when the wavenumber is above 1000 cm l for copper nanoparticles (Figure 3b). As illustrated in Figure 3d, absorption bands of GO are observed at 3412, 1730, and 1053 cm l, which are attributed to the stretching vibrations of O H, CdO, and C O C, respectively. The peaks at 1624, 1407, and 1224 cm l correspond to the vibration of carboxyl groups.40,41 Figure 3a,d

Figure 3. FT-IR spectra of (a) graphene, (b) copper nanoparticles, (c) graphene-copper nanoparticle composite, and (d) GO.

indicates that these peaks become weak or disappear when GO is reduced to graphene by potassium borohydride. The FT-IR spectrum of graphene-copper nanoparticle composite is similar to that of graphene although the peaks of copper nanoparticle below 1000 cm l can be observed in it. The results imply that the number of epoxy, hydroxyl, carbonyl, and carboxyl groups in graphene and the graphene-copper nanoparticle composite substantially decreased after chemical reduction. Figure 4 illustrates the X-ray diffraction (XRD) patterns of graphene, copper nanoparticles, and graphene-copper nanoparticle composite. Diffraction peaks assigned to graphene at 24.2° and 42.9° (corresponding to the indices of (002) and (100))22 can be observed in the XRD patterns of graphene and graphenecopper nanoparticle composite, indicating that the graphene structure was not destroyed after the in situ chemical reduction. The three characteristic peaks of copper are also observed in the XRD pattern of the composite, while the second 174

dx.doi.org/10.1021/ac2022772 |Anal. Chem. 2012, 84, 171–178

Analytical Chemistry

ARTICLE

Figure 5. Cyclic voltammograms at (a) a graphene-copper nanoparticle composite paste electrode, (b) a copper nanoparticle paste electrode, (c) a graphite-copper nanoparticle composite paste electrode, and (d) a graphene paste electrode in 75 mM NaOH aqueous solution containing 2 mM glucose. Scan rate, 50 mV/s. The content of paraffin oil in the four pastes was 25% (w/w). Figure 4. XRD patterns of (a) graphene-copper nanoparticle composite, (b) copper nanoparticles, and (c) graphene.

Cyclic voltammograms of 2.0 mM glucose were measured on the paste electrodes of graphene-copper nanoparticle composite, copper nanoparticles, graphite-copper nanoparticle composite, and graphene. As illustrated in Figure 5a,b, the peak current for the oxidation of glucose on the graphene-copper nanoparticle composite paste electrode is much higher than that on the copper nanoparticle paste electrode at +0.65 V (vs Ag/AgCl). The oxidation of glucose occurs in the potential range of 0.40 0.8 V where the oxidation wave for Cu(II)/Cu(III) redox couple was reported,43 45 implying that the Cu(III) species participate in the catalytic oxidation process of glucose.46 The results indicate that copper nanoparticles deposited on the high specific surface area graphene sheets show higher electrocatalytic activity toward the oxidation of carbohydrates while graphenes improve the electron transduction. The synergism between the copper nanoparticles and the graphene nanosheets in the graphene-copper nanoparticle composite paste electrode can significantly enhance the current response of carbohydrates. However, the current response of graphite-copper nanoparticle composite paste electrode toward glucose (Figure 5c) is even lower than that of copper nanoparticle paste electrode (Figure 5b). Figure S3 in the Supporting Information illustrates the SEM images of graphite particles and graphitecopper nanoparticle composite. A great amount of copper nanoparticles are observed on the surface of graphite particles. The lower current response of glucose on the graphite-copper nanoparticle composite paste electrode might be attributed to the lower electrical conductivity of the graphite cores. As expected, no response of glucose is observed at the graphene paste electrode in the absence of copper constituent that is a crucial component for the electrochemical sensing of carbohydrates in a strong alkaline solution (Figure 5d). To demonstrate the analytical performance of the prepared graphene-copper nanoparticle composite paste electrode, it was coupled with a CE system as an end-column amperometric detector for the separation and detection of mannitol, sucrose, lactose, glucose, and fructose. It is necessary to investigate the electrochemical oxidation performances of the five carbohydrates at the fabricated electrode to obtain the optimum potential because the detection potential applied to the detection electrode directly affects the sensitivity and the detection limits of this method. Figure S4(b) (f) in the

peak of graphene partially overlap with the second peak of copper. On the basis of the peak of copper at 43.2° in the XRD spectra, the average crystallite sizes of the pure copper particles and the copper particles in the composite are estimated to be 15.4 and 20.8 nm, respectively, using the Scherrer formula.42 The XRD results indicate that the product prepared by the in situ chemical reduction of a mixture solution containing GO and cupric sulfate is a hybrid of graphene and copper nanoparticles. In this work, the prepared graphene-copper nanoparticle composite was employed to fabricate the electrode for sensing carbohydrates. It was mixed with an appropriate amount of paraffin oil to form a paste that was further packed into a fused silica capillary to fabricate a microdisc electrode. The fabrication process of the electrode was illustrated in Figure 1A C. It is quite simple and easy to operate, allowing for the rapid fabrication of the novel composite electrodes at low cost. Figure 1D shows the photography of a prepared graphenecopper nanoparticle composite paste electrode. The quality of the electrode was checked by a microscope. The microscopic photograph of the composite-filled end of the capillary-based electrode indicates that the end of the copper wire was well buried in the packed graphene-copper nanoparticle composite paste to realize electrical contact (Figure 1E). Carbohydrates can be detected by copper electrodes in strongly alkaline media based on electrocatalytic oxidation. The most accepted mechanism for the catalytic oxidation of carbohydrates at copper-based electrodes has been intensively investigated by Kuwana and Baldwin.43,44 According to their model, deprotonated carbohydrates adsorbed on the copper electrode surface in their enediol forms are oxidized by the Cu(III) active sites that can be regenerated by electrochemical oxidation. It has been revealed that the minimum structural requirement for facile oxidation is the presence of at least two nearby hydroxyl groups in the compound to be oxidized. The electrochemical oxidation of carbohydrates at copper electrode consists of many-electron processes with C C bond cleavage involved.44 In this work, the graphene-copper nanoparticle composite paste electrode was employed to measure carbohydrates by CV and CE AD. 175

dx.doi.org/10.1021/ac2022772 |Anal. Chem. 2012, 84, 171–178

Analytical Chemistry

ARTICLE

Figure 7. Electropherogram of a mixture containing 0.25 mM mannitol (a), 0.5 mM sucrose (b), 0.5 mM lactose (c), 0.5 mM glucose (d), and 0.5 mM fructose (e) at a 320 μm diameter copper nanoparticle composite paste electrode. Conditions as in Figure 6.

nanoparticle composite paste electrode showed satisfactory stability and high reproducibility at the selected potential. Figure 6A illustrates the electropherogram of a mixture containing 0.25 mM mannitol, 0.5 mM sucrose, 0.5 mM lactose, 0.5 mM glucose, and 0.5 mM fructose at a graphene-copper nanoparticle composite paste electrode. Under the selected conditions, the five analytes could be separated resulting in well-defined and resolved peaks within 9 min. A series of the standard mixture solutions of mannitol, sucrose, lactose, glucose, and fructose with concentration ranging from 0.001 to 2.0 mM were tested to determine the linearity at the graphene-copper nanoparticle composite paste electrode. The composite detection electrode showed a welldefined concentration dependence in the investigated concentration range. The linear equations were y = 0.0975 + 254.39x (R = 0.9996, mannitol), y = 0.1247 + 177.19x (R = 0.9995, sucrose), y = 0.1320 + 172.81x (R = 0.9990, lactose), y = 0.1033 + 203.51x (R = 0.9994, glucose), and y = 0.1158 + 148.25x (R = 0.9991, fructose), where y, x, and R were the concentration of the analytes (mM), the peak current (nA), and the correlation coefficient, respectively. On the basis of a signal-to-noise ratio of 3:1, the determination limits were evaluated to be 0.29, 0.42, 0.43, 0.37, and 0.51 μM for mannitol, sucrose, lactose, glucose, and fructose, respectively. For comparison, a 320 μm diameter copper nanoparticle paste electrode was employed to determine the five carbohydrates mentioned above in combination with CE under the same conditions. Figure 7 illustrates the obtained electropherogram of a mixture solution of 0.25 mM mannitol, 0.5 mM sucrose, 0.5 mM lactose, 0.5 mM glucose, and 0.5 mM fructose. The weight ratio of copper nanoparticle and paraffin oil in the paste was 3:1. In comparison with Figure 6A, the peak currents of the five carbohydrates at the copper naonparticle paste electrode are much lower than those at the graphene-copper nanoparticle composite paste electrode. It is well in agreement with the CV results (Figure 5a,b). The sensitivity and detection limits were determined to be 85.96 nA/mM and 0.87 μM for mannitol, 52.63 nA/mM and 1.42 μM for sucrose, 50.88 nA/mM and 1.47 μM for lactose, 63.16 nA/mM and 1.19 μM for glucose, and 45.61 nA/mM and 1.64 μM for fructose, respectively. The detection limits were also evaluated based on a signal-to-noise ratio of 3:1. Obviously, the carbohydrates could be detected with higher sensitivity and lower detection limits when the graphene-copper nanoparticle composite paste electrode was used. In addition, broader peaks (and, hence, inferior resolution) are observed at the copper nanoparticle paste electrode (Figure 7). As illustrated

Figure 6. (A) Electropherogram of a mixture containing 0.25 mM mannitol (a), 0.5 mM sucrose (b), 0.5 mM lactose (c), 0.5 mM glucose (d), and 0.5 mM fructose (e) and the typical electropherograms of (B) the diluted bovine milk and the diluted extracts from (C) honey peach and (D) banana. Conditions: fused-silica capillary, 25 μm i.d.  40 cm long; detection electrode, 320 μm diameter graphene-copper nanoparticle composite paste electrode; running buffer, 75 mM NaOH; separation and injection voltage, 12 kV; injection time, 6 s; detection potential, +0.65 V (vs Ag/AgCl).

Supporting Information illustrate the cyclic voltammograms of 1 mM mannitol, 2 mM sucrose, 2 mM lactose, 2 mM glucose, and 2 mM fructose in 75 mM aqueous solution on a graphene-copper nanoparticle composite paste electrode. The profiles of these cyclic voltammograms are similar. The electrocatalytic oxidation of these carbohydrates occurs in the potential region corresponding to the formation of Cu(III).43 45 The anodic currents of the five analytes increase rapidly when the applied potential is above +0.40 V. However, the oxidation current starts to decrease or increase more slowly upon increasing the potential above +0.65 V. In addition, Figure S4a in the Supporting Information indicates that the background current of the composite electrode increases substantially when the potential is above 0.7 V. In the amperometric detection of CE, a high background current leads to an unstable baseline, which is a disadvantage for the sensitive and stable detection. Considering the sensitivity and background current, a detection potential of +0.65 V was employed for the subsequent amperometric detection. The graphene-copper 176

dx.doi.org/10.1021/ac2022772 |Anal. Chem. 2012, 84, 171–178

Analytical Chemistry in Figure 6A, the sharp and well-resolved responses at the graphenecopper nanoparticle composite paste electrode result in smaller values of the half-peak widths for mannitol, sucrose, lactose, glucose, and fructose relative to the copper nanoparticle paste electrode (8.4 vs 11.0, 7.8 vs 12.0, 8.6 vs 10.8, 7.2 vs 11.4, and 7.4 vs 11.2 s, respectively). As an important constituent in the composite, graphene could enhance not only the current response but also the resolution for the investigated carbohydrates. It has been demonstrated that fast heterogeneous electrontransfer reactions on the detection electrodes of CE are essential for not only high sensitivity but also high separation efficiency.47 The ability of the graphene in the composite to promote electron-transfer reactions could be attributed to their special nanosheet structure and high electrical conductivity.4,11 The stability and the precision of the graphene-copper nanoparticle composite paste electrode in the amperometric detection of CE was examined from a series of 9 repetitive injections of a sample mixture containing mannitol, sucrose, glucose, and fructose (0.1 mM each) under the selected conditions. The time of each run was 20 min. The relative standard deviations (RSDs) of peak current were 2.3%, 2.7%, 3.1%, 2.5%, and 3.3% for mannitol, sucrose, lactose, glucose, and fructose, respectively. Such good repeatability reflects the reduced surface fouling of the detection electrode, indicating the stability of the graphene-copper composite electrode is suitable for the analysis of real samples. The feasibility of the graphene-copper nanoparticle composite paste electrode for measuring real samples was demonstrated by detecting carbohydrates in milk and fruits after CE separation. Figure 6B,C illustrates the typical electropherograms of the diluted bovine milk and the diluted extracts from honey peach and banana. The content of lactose in the bovine milk was determined to be 4.735 g/L (RSD 3.2%, n = 3) that was well in agreement with the label amount (4.7 g/L). In addition, the determined amounts of mannitol, sucrose, glucose, and fructose were 0.69 (RSD 3.6%, n = 3), 67.78 (RSD 2.8%, n = 3), 27.76 (RSD 3.1%, n = 3), and 33.85 (RSD 3.5%, n = 3) mg/g in honey peach and 0.00, 62.53 (RSD 3.1%, n = 3), 33.70 (RSD 2.4%, n = 3), and 33.05 (RSD 3.5%, n = 3) mg/g in banana, respectively. In addition, some unidentified peaks were also found in the electropherograms of the real samples. Because the samples were separated by CE in 75 mM NaOH aqueous solution after being diluted at high ratios, the interference of some coexisting compounds could be minimized. The graphene-copper nanoparticle composite paste electrode was successfully employed to detect the carbohydrates in the samples with satisfactory sensitivity and selectivity in combination with CE separation.

ARTICLE

the pure copper nanoparticles. The performance, feasibility, and advantages of the novel detection electrodes were also demonstrated by the separation and detection of five carbohydrates in real samples in combination with CE. The novel graphene-based CE detectors offer favorable signal-to-background characteristics, sharp peaks for the carbohydrates, as well as simple design and fabrication, indicating great promise for a wide range of analytical applications.

’ ASSOCIATED CONTENT

bS

Supporting Information. 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: +86-21-51980061.

’ ACKNOWLEDGMENT This work was financially supported by the NSFC (Grants 20875015 and 21075020), the State Oceanic Administration (Grant 201105007), the Shanghai Science Committee (Grant 2009JC1401400), and the Education Ministry of China (Grant NCET-08-0134). ’ REFERENCES (1) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang, Y.; Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Science 2004, 306, 666–669. (2) Rao, C. N. R.; Sood, A. K.; Subrahmanyam, K. S.; Govindaraj, A. Angew. Chem., Int. Ed. 2009, 48, 7752–7777. (3) Geim, A. K.; Novoselov, K. S. Nat. Mater. 2007, 6, 183–191. (4) Pumera, M.; Ambrosi, A.; Bonanni, A.; Chng, E. L. K.; Poh, H. L. Trends Anal. Chem. 2010, 29, 954–965. (5) Shao, Y. Y.; Wang, J.; Wu, H.; Liu, J.; Aksayb, I. A.; Lin, Y. H. Electroanalysis 2010, 22, 1027–1036. (6) Compton, O. C.; Nguyen, S. T. Small 2010, 6, 711–723. (7) Kim, H.; Abdala, A. A.; Macosko, C. W. Macromolecules 2010, 43, 6515–6530. (8) Hu, W. B.; Peng, C.; Luo, W. J.; Lv, M.; Li, X. M.; Li, D.; Huang, Q. ACS Nano 2010, 4, 4317–4323. (9) Dreyer, D. R.; Park, S.; Bielawski, C. W.; Ruoff, R. S. Chem. Soc. Rev. 2010, 39, 228–240. (10) Shen, J. F.; Shi, M.; Yan, B.; Ma, H. W.; Li, N.; Hu, Y. Z.; Ye, M. X. Colloids Surf. B 2010, 81, 434–438. (11) Pumera, M. Chem. Rec. 2009, 9, 211–223. (12) Ambrosi, A.; Sasaki, T.; Pumera, M. Chem. Asian J. 2010, 5, 266–271. (13) Shan, C.; Yang, H.; Song, J.; Han, D.; Ivaska, A.; Niu, L. Anal. Chem. 2009, 81, 2378–2382. (14) Zhou, M.; Zhai, Y. M.; Dong, S. J. Anal. Chem. 2009, 81, 5603–5613. (15) Tang, L. H.; Wang, Y.; Li, Y. M.; Feng, H. B.; Lu, J.; Li, J. H. Adv. Funct. Mater. 2009, 19, 2782–2789. (16) Li, H. J.; Chen, J.; Han, S.; Niu, W. X.; Liu, X. Q.; Xu, G. B. Talanta 2009, 79, 165–170. (17) Han, D. X.; Han, T. T.; Shan, C. S.; Ivaska, A.; Niu, L. Electroanalysis 2010, 22, 2001–2008. (18) Feng, X. M.; Li, R. M.; Ma, Y. W.; Chen, R. F.; Shi, N. E.; Fan, Q. L.; Huang, W. Adv. Funct. Mater. 2011, 21, 2989–2996. (19) Chen, B.; Zhang, L. Y.; Chen, G. Electrophoresis 2011, 32, 870–876.

’ CONCLUSIONS In summary, a facile approach based on in situ chemical reduction has been successfully developed for the facile preparation of graphene-copper nanoparticle composite for electrochemical sensing. SEM, XRD, and FT-IR spectra offered insights into the structure of the material. The results indicated that copper nanoparticles in the composite were well dispersed and embedded throughout the graphene matrix to form an interconnected hybrid network. The novel composite was mixed with paraffin oil and packed into the bore of fused silica capillaries to fabricate paste microdisc electrodes. The results of CV and CE indicated that the graphene in the composite can significantly enhance the current response of the electrodes toward carbohydrates in comparison to 177

dx.doi.org/10.1021/ac2022772 |Anal. Chem. 2012, 84, 171–178

Analytical Chemistry

ARTICLE

(20) Lu, Y.; Wang, X.; Chen, D. F.; Chen, G. Electrophoresis 2011, 32, 1906–1912. (21) Wang, X.; Li, J. Y.; Qu, W. D.; Chen, G. J. Chromatogr., A 2011, 1218, 5542–5548. (22) Wang, Y.; Li, Y. M.; Tang, L. H.; Lu, J.; Li, J. H. Electrochem. Commun. 2009, 11, 889–892. (23) Wu, J. F.; Xu, M. Q.; Zhao, G. C. Electrochem. Commun. 2010, 12, 175–177. (24) Xu, H. F.; Dai, H.; Chen, G. N. Talanta 2010, 81, 334–338. (25) Reinhold, V.; Sheeley, D.; Kuei, J.; Her, G. Anal. Chem. 1988, 60, 2719–2722. (26) El Rassi, Z.; Mechref, Y. Electrophoresis 1996, 17, 275–301. (27) Suzuki, S.; Honda, S. Electrophoresis 2003, 24, 3577–3582. (28) Johnson, D. C.; LaCourse, W. R. Electroanalysis 1992, 4, 367–380. (29) O’Shea, T. J.; Lunte, S. M.; LaCourse, W. R. Anal. Chem. 1993, 65, 948–951. (30) Ye, J. N.; Baldwin, R. P. J. Chromatogr., A 1994, 687, 141–148. (31) Luo, P.; Prabhu, S. V.; Baldwin, R. P. Anal. Chem. 1990, 62, 752–755. (32) Garcia, M.; Escarpa, A. Biosens. Bioelectron. 2011, 26, 2527–2533. (33) Garcia, M.; Escarpa, A. Anal. Bioanal. Chem. 2011, DOI: 10.1007/s00216-011-5453-x. (34) Cheng, X.; Zhang, S.; Zhang, H. Y.; Wang, Q. J.; He, P. G.; Fang, Y. Z. Food Chem. 2008, 106, 830–835. (35) Batchelor-McAuley, C.; Du, Y.; Wildgoose, G, G.; Compton, R. G. Sens. Actuators, B 2008, 135, 230–235. (36) Bin, X. B.; Chen, J. Z.; Cao, H.; Chen, L. T.; Yuan, J. Z. J. Phys. Chem. Solids 2009, 70, 1–7. (37) Luechinger, N. A.; Athanassiou, E. K.; Stark, W. J. Nanotechnology 2008, 19, 445201. (38) Chen, C. M.; Yang, Q. H.; Yang, Y. G.; Lv, W.; Wen, Y. F.; Hou, P. X.; Wang, M. Z.; Cheng, H. M. Adv. Mater. 2009, 21, 3007–3011. (39) Chen, Q. W.; Gan, Z. B.; Wang, J.; Zhang, L. Y. Chem.—Eur. J. 2011, 17, 12458–12464. (40) Das, B.; Prasad, K. E.; Ramamurty, U.; Rao, C. N. R. Nanotechnology 2009, 20, 125705. (41) Hu, H. T.; Wang, X. B.; Liu, F. M.; Wang, J. C.; Xu, C. H. Synth. Met. 2011, 161, 404–410. (42) Kryszewski, M.; Jeszka, J. K. Synth. Met. 1998, 94, 99–104. (43) Marioli, J. M.; Kuwana, T. Electrochim. Acta 1992, 37, 1187–1197. (44) Luo, M. Z.; Baldwin, R. P. J. Electroanal. Chem. 1995, 387, 87–94. (45) Burke, L. D.; Ahern, M. J. G.; Ryan, T. G. J. Electrochem. Soc. 1990, 137, 553–561. (46) Colon, L. A.; Dadoo, R.; Zare, R. N. Anal. Chem. 1993, 65, 476–481. (47) Wang, J.; Tian, B. M.; Chatrathi, M. P.; Escarpa, A.; Pumera, M. Electrophoresis 2009, 30, 3334–3338.

178

dx.doi.org/10.1021/ac2022772 |Anal. Chem. 2012, 84, 171–178