- Email: [email protected]
A novel 3-D fabrication of platinum nanoparticles decorated micro carbon pillars electrode for high sensitivity detection of hydrogen peroxide
Sensors and Actuators B 181 (2013) 57–64
Contents lists available at SciVerse ScienceDirect
Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb
A novel 3-D fabrication of platinum nanoparticles decorated micro carbon pillars electrode for high sensitivity detection of hydrogen peroxide Junwei Su a , Fan Gao b , Zhiyong Gu b , Michael Pien c , Hongwei Sun a,∗ a b c
Department of Mechanical Engineering, University of Massachusetts Lowell, One University Avenue, Lowell, MA 01854, USA Department of Chemical Engineering, University of Massachusetts Lowell, One University Avenue, Lowell, MA 01854, USA ElectroChem, Inc, 400 West Cummings Park, Woburn, MA 01801, USA
a r t i c l e
i n f o
Article history: Received 18 December 2012 Received in revised form 30 January 2013 Accepted 6 February 2013 Available online xxx Keywords: Microscale carbon pillars Hydrogen dioxide detection Molding Electrochemical sensor Platinum/carbon nanoparticles
a b s t r a c t We report a novel fabrication technology for microscale carbon pillars (MCP) decorated with platinum/carbon nanoparticles (Pt/C NPs) for hydrogen peroxide (H2 O2 ) detection. The fabrication process involves three sequential steps: spray-coating of Pt/C nanoparticles, polydimethylsiloxane (PDMS) soft molding, and high temperature carbonization of polyacrylonitrile (PAN), which has great potential for large scale and high throughput manufacturing. The amperometric response of MCP based H2 O2 sensor exhibits superior electrochemical activity and charge transfer ability with a sensitivity of 1280–1750 AmM−1 cm−2 and a linear dynamic range (LDR) to 7000 M. Most importantly, the developed fabrication method can be used for a wide variety of functional nanomaterials for different applications such as electrochemical analysis, sustainable energy and biosensing. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Today’s analytical chemistry calls for complex instrumentation and considerable support, including special laboratory facilities and highly skilled personnel. However, chemical and biological sensing devices that are robust, portable and easy to use will lead to more innovative strategies for analytical instrumentation. In addition, they have other advantages such as requiring fewer reagents to operate and yielding reliable information continuously [1]. For instance, amperometric biosensors, usually formed by biologically surface-modified voltammetric electrodes, are gaining increasing importance owing to their high reliability, robustness and high sensitivity [2]. The early actual electrochemical sensors were based on the chemical modification of carbon paste. Complex functional groups such as silver ion were introduced to the carbon surface for the measurement of concentration of silver [3]. Another modification method was developed by Yao and Musha who dissolved anthraquinone (AQ) in the pasting liquid for immobilization of electroactive species [4]. These days most of electrochemical sensors were fabricated on glassy carbon electrode. Different materials, such as, anthraquinone [5], ordered mesoporous carbon modification [6], and nickel hexacyanoferrate/chitosan/carbon nanotubes (NiHCF/CS/CNTs) nanocomposite films [7], were immobilized on
∗ Corresponding author. Tel.: +1 978 934 4391; fax: +1 978 934 3084. E-mail address: Hongwei [email protected] (H. Sun). 0925-4005/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.snb.2013.02.023
the surface of glassy carbon electrodes to improve their electrochemical response. Nanostructured materials are becoming the focus of electrochemical sensor research. For instance, thionin nanowires [8] and CuO-MWCNTs [9] have been reported to improve hydrogen peroxide sensitivity of sensors. The performance of electrochemical sensor can be future improved by the use of so called hybrid nanostructures. He et al. built Ag nanoparticle 3D catalyst on a graphite substrate using Na2 Ti3 O7 nanowire as 3D frames for loading Ag nanoparticles [10]. Fe3 O4 –Ag hybrid submicrosphere was synthesized and developed as hydrogen peroxide sensor by Liu et al. [11]. Gold nanoparticles were attached on ordered mesoporous carbon (OMC) as hydrogen peroxide sensors [12]. A synthetic method to incorporate copper sulfide nanoparticles inside the mesopores of OMC is also reported for hydrogen peroxide sensor development [13]. However, random hybrid nanostructures hardly provide repeatable performance. There is no report for organized hybrid micro and nanoscale structures on the working electrode surface to further enhance sensor performance and repeatability. It is easy to understand that a 3-D organized microscale structure embedded with nanostructures such as catalyst metal nanoparticles can take advantage of both unique catalytic properties of metal NPs and high surface area provided by the microstructures, to achieve high performance and repeatable electrochemical sensing. In this paper, a novel method to fabricate platinum/carbonnanoparticles-decorated micro carbon pillars (Pt/C NPs MCP)
58
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
Fig. 1. Schematic of the fabrication process of Pt/C nanoparticles decorated micro carbon pillars.
electrodes is presented. The Pt/C NPs MCP sensor demonstrates superior properties well suitable for electrochemical sensing including low density, low electrical resistance, high resistance to chemical attack and impermeability to gases and liquid [14]. In addition, the surface to volume ratio of Pt/C NPs MCP is much larger than the traditional glassy carbon electrode thanks to high density micro pillar structure. Another critical feature of Pt/C NPs MCP electrode is its superhydrophilic property after thermal treatment, which can effectively prevent the gas bubbles from generating on electrode during hydrogen dioxide detection. In the first section, the process, materials and key parameters for Pt/C NPs MCP fabrication are introduced and discussed. Then the stabilization of polyacrylonitrile (PAN) material is described and analyzed using Fourier Transform Infrared Spectroscopy (FTIR) technique. In the meantime, the cyclic voltammetry (CV) measurement was conducted to characterize the charge transfer and electrochemical activity of developed Pt/C NPs MCP structure. At last, serious of experiments were performed to evaluate the performance of Pt/C NP MCP based hydrogen peroxide sensor. 2. Experimental 2.1. Materials and instruments Polyacrylonitrile (PAN) with molecule weight of 150,000 and N,N-dimethylformamide (DMF) were ordered from Sigma–Aldrich. The assay of DMF is greater than 99.8%. The polydimethylsiloxane (PDMS) for mold fabrication of soft molding was purchased from Dow Corning Corp. The Pt nanoparticles (2.9 nm) and carbon black nanoparticle (50 nm) were ordered from Alfa Aesar. Bruker Tensor 27 was used for the FTIR measurement for PAN samples after stabilization. The morphology of Pt/C NPs MCP was studied using a field emission scanning electron microscope (JEOLSEM, JSM-7401F). Cyclic voltammetry (CV), and amperometric measurements were conducted using a VersaSTAT 3 potentiostat (Princeton Applied Research, TN). A Pt wire (99.9%, Alfa Aesar) and Ag/AgCl electrode (BASi) were used as counter and reference electrodes in the measurement, respectively. 2.2. Pt/C NPs MCP fabrication The complete fabrication process is illustrated in Fig. 1 and consists of four steps: (1) PDMS mold fabrication; (2) preparation of
molding material; (3) spray-coating of Pt/C ink and molding process; and (4) stabilization and carbonization of PAN pillars. These steps are described in detail below. 2.2.1. PDMS mold fabrication PDMS was chosen as the mold material for molding of micro pillar arrays for two reasons: (1) PDMS has a lower surface energy than other substrates such as silicon and quartz, which can ease the mold removal after molding; (2) PDMS based molding method is well suited for low cost, large scale and high throughput nanomanufacturing. The PDMS was prepared by mixing the PDMS base and curing agent at a 10:1 ratio for 5 min in a glass container. A vacuum oven was used to remove the bubbles generated from mixing. A silicon substrate fabricated by deep reactive ion etching (DRIE) process to form micro pillar arrays was used as mothermold. Then the PDMS was evenly poured onto the surface of silicon substrate prior to being cured in a pressurized, heat-treated vacuum chamber for 60 min at 75 ◦ C. After the PDMS film was peeled off from the silicon substrate, the micro holes arrays were formed on the PDMS mold and the PDMS mold is ready for spray coating process. 2.2.2. Preparation of molding material The PAN (Sigma–Aldrich, molecular weight: 150,000) was dissolved in DMF solvent with a weight ratio of 2:8 and stirred at 80 ◦ C for 4 h. The dissolving time depends on the PAN concentration and amount. The higher concentration, the longer dissolving time is required. After a successful mixing, the PAN solution is clear and transparent. Commercially available Pt/C nanopowders were dispersed in DI water with a concentration of 5% to produce “Pt/C ink”. An ultra-sonication of 5–10 min for “Pt/C ink” is required before the spraying process. 2.2.3. Spray-coating of Pt/C ink and molding process The ultrasonic “Pt/C ink” was uniformly sprayed on the surface of PDMS mold using an air brush system (average droplet size: 6 m) at room temperature (20 ◦ C). During the spray process, the tiny Pt/C droplets were formed at the nozzle exit of the air brush system under pressured air. The distance between the mold surface and the nozzle exit was optimized to obtain the smallest droplets for a rapid solvent evaporation. After each coating, the cotton Q-tip with ink was used to repaint the mold surface to limit coffee stain effects [15].
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
59
After spray coating, the mold coated with Pt/C ink layer was pre-baked on a hot plate at 80–90 ◦ C for 1–2 min to evaporate the DI water. As a result, a layer of Pt/C NPs was left on the surface of these PDMS micro holes. Then the PAN/DMF solution was spincoated on the PDMS mold at the spinning speed of 3000 rpm in order to obtain thickness of 20 m. It is anticipated that the solution was uniformly filled into micro holes in PDMS mold. After this step, the PDMS mold with PAN/DMF film was heated again on a hotplate for 5 min at 100 ◦ C to evaporate the DMF in the film. Once the PAN solution layer was completely dry, the PAN layer with Pt/C NPs was removed from the PDMS mold. As a result, micro pillars decorated with Pt/C nanoparticles were generated on a PAN film surface. 2.2.4. Stabilization and carbonization of PAN micro pillars Oxidative stabilization is very critical to obtain production quality of carbon from PAN material. The stabilization was carried out at 250 ◦ C for 4 h to render them infusible and flameproof. Temperature below 200 ◦ C is impractical considering a very long period is required while temperatures above 300 ◦ C cause violent exothermic reactions [16,17], resulting in significant weight loss [18] and formation of tarry substances. On the other hand, improper stabilization could result in blowout of a core portion during carbonization due to its incomplete oxidation [19]. Proper conditions of the heating rate, time, and temperature were established for an optimal stabilization. The FTIR was used to monitor the stabilization procedure [20] and the results for transmittance of the PAN material after stabilization are presented in Fig. 2. In this optimization experiment, the stabilization temperature was varied from 220 ◦ C to 280 ◦ C while heating time is increased from 1 to 12 h. In Fig. 2, the band at 2243 cm−1 represents the C N stretching of acrylonitrile unit in the polymer chain and 2939 cm−1 stands for C H in CH2 . As the stabilization proceeds, the 2243 cm−1 band of the nitrile groups and 2939 cm−1 band of methylene continue to decrease in intensity. Besides, there were three new bands (at 1718, 1660 and 1595 cm−1 ) in carbonyl-stretch frequency region. The band 1595 cm−1 appears due to the combination vibration of C C and C N stretching, and NH in-plane bending of the ladder-frame structure of stabilized PAN [21]. The 1595 cm−1 band continues to increase, showing an enhancement of the extent of the stabilization reactions. As suggested by Shimada and Sen [21,22], the band at 1718 cm−1 is due to the free ketones in hydronaphthyridine rings and the 1660 cm−1 band is assigned to the conjugated ketones in acridone rings. These groups were generated by the oxygen uptake reactions. The original PAN sample contains a 1660 cm−1 peak due to the DMF solvent. It is apparent that the spectra does not change any more after heating for 4 h at 250 ◦ C, which implies a compete stabilization. It was observed that after the stabilization process, the color of PAN turned to light brown [23]. The carbonization was conducted in a temperature controllable tube furnace (Thermo Scientific, Lindberg/Blue M Mini-Mite Tube Furnace) with inert (nitrogen) gas purging through the quartz tube. The reaction formula during the stabilization and carbonization is shown in Fig. 3. The flow rate of nitrogen gas was controlled at 100 ml/min in the initial purging step. Then a high temperature heating was performed at 900 ◦ C for 20 min to carbonize the sample with the 60 ml/min N2 purging.
Fig. 2. (A) FTIR spectra of PAN material heated at 250 ◦ C for different durations: (a) original PAN, (b) 1 h, (c) 4 h, and (d) 12 h; (B) FTIR spectra of PAN heated for 4 h at different temperatures: (a) original PAN, (b) 220 ◦ C, (c) 250 ◦ C, and (d) 280 ◦ C.
investigating bubbling effect on sensor performance (see Sections 3.4 and 3.5). Fig. 4A and B shows the SEM images of micro carbon pillars and Pt/C NPs distribution on the pillar surfaces using COMPO mode of JEOL-SEM. The elements in these pillars material are shown with different contrasts based on the atomic number of
3. Result and discussion 3.1. Characteristics of Pt/C NPs MCP The micro carbon pillars in this research are 4 m in diameter, and 15 m in height with the spacing of 3 m, 5 m and 9 m, respectively. The three kinds of pillars were designed mainly for
Fig. 3. Reaction mechanisms of PAN stabilization and carbonization.
60
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
Fig. 4. SEM images of PAN based micro carbon pillar arrays including (A) bare micro carbon pillar arrays with spacing of 3 m and (B) Pt/C nanoparticles decorated micro carbon pillar arrays with spacing of 3 m (Pt nanoparticles are seen as very small bright spots on the background of the pillar arrays).
each element. The Pt/C NPs can be clearly seen on the pillar surface and inside the pillars (small white spots in Fig. 4B) compared to the bare micro carbon pillars (Fig. 4A). The composition of Pt/C NPs MCP was further characterized by energy dispersive X-ray (EDX) spectroscopy. For the bare micro carbon pillars, the EDX spectrum (Fig. 5A) shows the peaks corresponding to carbon and a small portion of oxygen, which is thought to be generated during carbonization. The EDX spectrum (Fig. 5B) of Pt/C NPs MCP shows the peaks corresponding to C, O and Pt elements, confirming the existence of Pt NPs on the surface of micro carbon pillar. The weight percentages of carbon, oxygen, silicon and platinum are 40.43% 4.55%, 5.99%, and 49.03%, respectively. The existence of a small portion of silicon was believed to be due to the impurity in Pt/C nanoparticles. Fig. 6 shows the high density micro carbon pillar array with embedded Pt/C NPs. It is worth noting that conventional coating techniques such as spray, vacuum deposition and spin-coating cannot form permanent bonding of particles with pillar surface. On the other side, other coating techniques such as E-beam deposition and ion beam assistant deposition require complex and expensive equipment. The developed approach in this research does not require any expensive equipment and can produce the hybrid
Fig. 5. EDX images of (A) bare micro carbon pillar and (B) Pt/C NPs micro carbon pillar.
Fig. 6. SEM micrographs of high aspect ratio and high density Pt/C nanoparticles decorated micro carbon pillars (spacing is 3 m).
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
Fig. 7. CV curves of bare and Pt/C catalyst decorated micro carbon pillars electrodes.
pillar-nanoparticle structure in a simple and reliable way. Furthermore, the Pt/C nanoparticles were permanently embedded in micro carbon pillars, which make this method reliable and robust for sensor applications. 3.2. Hydrogen peroxide (H2 O2 ) detection Rapid and accurate detection of H2 O2 is of great importance in clinical analyses and biosensors development in particular [24]. To evaluate the performance of Pt/C NPs MCP electrode for electrochemical sensing, a hydrogen peroxide (H2 O2 ) detection system based on conventional three-electrode cell was constructed with an Ag/AgCl (under saturated NaCl) electrode as the reference electrode, a platinum wire as the counter electrode, and Pt/C NPs MCP as the working electrode. Due to these micro pillar structures, the carbon based surface area has been increased from 0.2 cm2 to the 0.97 cm2 (3 m pillars spacing) and 0.67 cm2 (5 m pillars spacing) and 0.42 cm2 (9 m pillars spacing). The electrocatalytic activity of Pt/C NPs MCP electrode toward H2 O2 detection was evaluated by comparing its CV response with that of the bare micro carbon pillar electrode. Fig. 7 presents the steady-state CVs for the bare MCP and Pt/C NPs MCP electrodes in phosphate buffered solution (PBS) (pH = 7.2) between −1 and 1 V potential range. As expected, the Pt/C NPs MCP electrode displays a much better catalytic activity and facilitates a faster electron transfer at −0.4 V vs. Ag/AgCl electrode. In terms of the structural effect, the Pt/C nanoparticles were distributed along the surface of the micro carbon pillars with a thin layer which is believed to produce a catalytic activity. Comparing the areas under the two CV curves for bare MCP and Pt/C NPs MCP (Fig. 7), it is apparent that the total charge transfer of the Pt/C NPs MCP electrode is considerably greater than that of bare MCP. In Fig. 7, the modification efficiency of Pt/C is 225.8% at working potential (−0.4 V) compared to the bare microscale carbon pillars (MCP) electrode. Many metals such as Au, Ag, Pt, Pt, and Cu can act as catalyst for the decomposition of H2 O2 [12,13,27,29]. The overall reaction of the H2 O2 decomposition is: Catalyst
2H2 O2 −→ 2H2 O + O2 In the presence of Pt nanoparticles, H2 O2 experiences electrochemical redox reactions on the electrode. The reaction mechanism is proposed as follows. H2 O2
Pt nanoparticles
−→
2H+ + O2 + 2e−
(anodic)
61
Fig. 8. CV responses of Pt/C NPs MCP electrode to various concentrations (0.5, 1.0, 1.5, 2.0, 2.5 and 3 mM) of H2 O2 in a stirred PBS solution (pH = 7.2).
H2 O2 + 2e− → 2OH−
(cathodic)
In terms of the electronic effect, since PAN based carbon is a material with good electrical conductivity, the electron could transfer in two ways: directly to the PAN based carbon and through the Pt nanoparticles. The Pt nanoparticles increase effective surface area of electrode and facilitate indirect electron transfer via PAN based carbon as the mediator and showed the positive effect on the amperometric signals. It was suggested that a significantly increased charge transfer rate from H2 O2 to the PAN based carbon where Pt nanoparticles covalently linked. Hence, the electrons are easily transferred between the decorated Pt/C layer and carbon base which may enhance the sensing signals as well. It is anticipated that Pt/C NPs MCP electrode will generate enhanced signal to H2 O2 , which should be attributed by significantly increased catalytic surface, high conductivity and better charge transfer of platinum NPs. Fig. 8 shows the responses of Pt/C NPs MCP electrodes to H2 O2 at different concentrations. As can be seen, the well-defined and resolved cathodic and anodic peaks can be clearly distinguished on the forward and reverse scans. The cathodic peak is smaller than the anodic peak and the anodic peak current density increased linearly as the H2 O2 concentration increased. These features demonstrate that the Pt/C NPs MCP electrode is well suited for the detection of H2 O2 . In addition, the working potential of −0.4 V can be used as a static applied potential for the following amperometric detection measurement. 3.3. Amperometric detection of hydrogen peroxide The amperometric response of the Pt/C NPs MCP sensor to the successive additions of 500 M H2 O2 into PBS solution (pH = 7.2) are shown in Fig. 9 (with a constant potential of −0.4 V applied to the working electrode). The background response of the sensor was allowed decayed to a steady state after 50 s with magnetic stirring. The distinct potential signals responding to the increase of H2 O2 concentration was obtained in 5–10 s. The curve clearly shows the current increasing with the increasing of H2 O2 concentrations within a linear dynamic range (LDR) to 7000 M. To study the effect of the pillar spacing on sensing performance, the amperometric responses vs. step-wise increase of H2 O2 concentration for pillars with spacing of 3 m, 5 m, and 9 m were obtained and shown in Fig. 9 and Table 1. The MCP electrode with 3 m spacing has the largest surface area and most Pt NPs for electrochemical reaction, therefore lower limit of detection (LOD) and
62
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
Table 1 Amperometric response of Pt/C NPs MCP electrode with different spacing to step-wise change of H2 O2 concentration. Samples
I (A)
LOD (M)
Sensitivity (AmM−1 cm−2 )
R2
Associated error (%)
Pillar space = 3 m Pillar space = 5 m Pillar space = 9 m
174.61 143.66 128.07
9.60 15.04 17.71
1750 1440 1280
0.998 0.995 0.9997
2.34 8.46 5.08
higher sensitivity were obtained, as shown in Table 1. The LOD was determined when the limiting current for hydrogen peroxide from Pt/C NPs MCP electrode becomes three times greater than the standard deviation of a blank PBS (S/N = 3). The comparison of electroanalytical characteristics of Pt/C NPs MCP electrode for H2 O2 detection with that of other electrodes in literature was summarized in Table 2. As can be seen, the Pt/C NPs MCP electrode exhibits a rapid and sensitive response to the change of H2 O2 concentration. The response of the Pt/C NPs MCP electrode to H2 O2 behaves linearly ranging to 7000 M with sensitivities of 1280–1750 AmM−1 cm−2 . The sensitivity of Pt/C NPs MCP is higher than the most of Pt modified electrodes. Only the GC/CNT + Ptnano modified electrode [30] shows a higher sensitivity than that of Pt/C NP MCP. But Pt/C NPs MCP electrode demonstrates
wider range of LDR compared to GC/CNT + Ptnano modified electrode and is more suitable for large scale manufacturing. In addition, the anti-interference ability of the Pt/C NPs MCP sensor was investigated by adding 500 M of UA (uric acid), CA (citric acid), AA (ascorbic acid) and glucose sequentially into the solution and the sensor responses are shown in Fig. 10. The results demonstrate that the responses from UA, CA, AA and glucose are negligible compared with that observed for 500 M H2 O2 , indicating a high selectivity and anti-interference properties of Pt/C NPs MCP sensor. 3.4. Reproducibility and stability of Pt/C NPs MCP sensor To check the sensor reproducibility, five Pt/C NPs MCP sensors were prepared with the same fabrication method and tested for 0.5 mM hydrogen peroxide. A relative standard deviation of 5.15% was obtained. The operational stability of the modified electrode was investigated in continuous operation mode in which the data from the repeated experiments demonstrate a 2.6% deviation in sensing signal. Another experiment has been conducted to further check the stability of sensor performance by exposing the sensor to environment for 150 consecutive days. It is shown that the sensor still maintains 96.4% accuracy and demonstrates an excellent material stability and the bonding between Pt nanoparticles and carbon pillar. 3.5. Pt/C NPs MCP sensor surface fouling The closely spaced pillars provide more surface area for sensing and therefore higher sensitivity. However, it was found that when the H2 O2 concentration rise above 7000 M, an obvious bubbling process occurs, in which bubbles block the fresh chemicals to access to the sensing surface and results in significant oscillation in CV signal [32]. As a result, anodic peak will become undistinguished in the CV curve and sensor lost detection capability. Optimization is
Fig. 9. (A) Amperometric response curves of Pt/C NPs MCP electrode to step-wise change of 0.5 mM of H2 O2 ; (B) Calibration curve for step-wise change of 0.5 mM H2 O2 in a stirred PBS solution (pH 7.2) with operating potential of −0.4 V for sensors with pillar spacing of (a) 3 m, (b) 5 m, and (c) 9 m.
Fig. 10. Anti-interference property of the Pt/C NPs MCP electrode with addition of 500 M AA, UA, glucose, CA, and 500 M H2 O2.
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
63
Table 2 Comparison of electroanalytical characteristics of various modified electrodes toward H2 O2 detection. Electrode
Eapp (mV)
Graphene nanosheet encrusted MCP Fe3 O4 –Ag hybrid submicrosphere Pt NPs embedded carbon film GCE/CNTs/CB/sol–gel NiHCF/CS/CNTs AgPs-SWCNT
+400 −500 +600 −200 −200 −300
GC/CNT + Ptnano
+700
Macroporous Au-/nPts GC/RGO/PB/PTBO Ag-Na2 Ti3 O7 NWs Gold NPs on ordered mesoporous carbon Pt NPs MCP
+100 +200 +600 −150 −400
H2 O2
Reference
LDRa (M)
LODa (M)
Sensitivity (AmM−1 cm−2 )
250–5500 1.2–3500 nr 100–1000 40–5600 16–18,085 0.025–10 100–2000 ∼10,000 5–600 50–2500 2–3920 ∼7000
nr 1.2 0.075 10 0.28 2.76
0.07 12 3.93 ± 0.15b 1.055b 654 1092 3570 1850 264 420 1 22 1280–1750
0.025 50 1.5 1.0 0.49 9.6–17.7
[25] [11] [27] [26] [7] [28] [30] [29] [31] [10] [12] This work
a LDR: linear dynamic range; LOD: limit of detection; CNT: carbon nanotube; CS: chitosan; NiHCF: nickel hexacyanoferrate; AgPs: silver particles; PTBO: poly(toluidine blue O); PB: Prussian blue; RGO: reduced graphene oxide; MCP: micro carbon pillars; Ptnano : platinum nanoparticles; GCE, CG: glassy carbon electrode; CB: Celestine blue; NP: nanoparticle; nr: not reported. b AmM−1 .
Table 3 Detection of H2 O2 in Merrimack River water. Real sample
Merrimack River pH = 5.98
H2 O2 added (M)
H2 O2 found (M)
Recovery (%)
40 80 120
38.6 81.3 122.4
96.5% 101.6% 101.9%
ongoing for pillar diameter, spacing and height to diminish surface fouling of Pt/C NP MCP sensors for a wider range of LDR. 3.6. Detection of H2 O2 in Merrimack River with Pt/C NP MCP sensor A real water sample (pH = 5.98) was taken from the Merrimack River in Lowell, MA and was put into a glass beaker for two days for settlement. After that, the response of Pt/C NP MCP sensor was measured by adding hydrogen peroxide with a known concentration into the sample. Table 3 presents the comparison of prediction results from the sensor with the actual concentration in the water sample. The sensor has demonstrated an excellent sensing capability in detecting H2 O2 in a real sample although the limit of detection is decreased to 1.82 M (S/N = 3). 4. Conclusions In this paper, we demonstrated a novel, low cost and fast fabrication method to generate nanoparticles decorated micro carbon pillar structure. The FTIR was used to analyze and optimize the stabilization process. The structure and morphology of these pillars were identified by SEM and EDX. Besides of platinum nanoparticles, micro carbon pillars could be decorated by other particles for different applications. The CV studies confirmed the Pt/C NPs MCP electrode have faster charge transfer and higher electrochemical activity and selectivity toward H2 O2 than existing sensors. The results of this study indicate that the fabrication method of Pt/C NPs MCP is a promising manufacturing technology for highly sensitive electrochemical sensor development. Acknowledgments The authors are grateful for the materials support from ElectroChem Inc. This work was funded by National Science Foundation
(ECCS 0731125 and CMMI 0923403). We are grateful for the assistance in FTIR measurements from Professor David Ryan and his student Yipei Zhang at University of Massachusetts Lowell.
References [1] S. Alegret, Rigid carbon-polymer biocomposites for electrochemical sensing, Analyst 121 (1996) 1751–1758. [2] F. Scheller, F. Schubert, Biosensors, 1st ed., Elsevier, Amsterdam, 1992. [3] G.T. Cheek, R.F. Nelson, Applications of chemically modified electrodes to analysis of metal ions, Analytical Letters 11 (1978) 393–402. [4] T. Yao, S. Musha, Electrochemical enzymatic determinations of ethanol and l-lactic acid with a carbon paste electrode modified chemically with nicotinamide adenine dinucleotide, Analytica Chimica Acta 110 (1979) 203–209. [5] Y. Dou, S. Haswell, J. Greenman, J. Wadhawan, Immobilized anthraquinone for redox mediation of horseradish peroxidase for hydrogen peroxide sensing, Electrochemistry Communications 11 (2009) 1976–1981. [6] L. Zhu, C. Tian, D. Yang, X. Jiang, R. Yang, Bioanalytical application of the ordered mesoporous carbon modified electrode, Electroanalysis 20 (2008) 2518–2525. [7] Z. Wang, X. Hao, Z. Zhang, S. Liu, Z. Liang, G. Guan, One-step unipolar pulse electrodeposition of nickel hexacyanoferrate/chitosan/carbon nanotubes film and its application in hydrogen peroxide sensor, Sensors and Actuators B 162 (2012) 353–360. [8] H. Lin, H. Cheng, X. Miao, P. Papakonstantinou, D. Mihailovic, M. Li, A novel hydrogen peroxide amperometric sensor based on thionin incorporated onto a Mo6 S9 -xIx nanowire modified glassy carbon electrode, Electroanalysis 21 (2009) 2602–2606. [9] K.Y. Zhang, N. Zhang, H. Cai, C. Wang, A novel non-enzyme hydrogen peroxide sensor based on an electrode modified with carbon nanotube-wired CuO nanoflowers, Microchimica Acta 176 (2012) 137–142. [10] X. He, C. Hu, H. Liu, G. Du, Y. Xi, Y. Jiang, Building Ag nanoparticle 3D catalyst via Na2 Ti3 O7 nanowires for the detection of hydrogen peroxide, Sensors and Actuators B 144 (2010) 289–294. [11] Z. Liu, B. Zhao, Y. Shi, C. Guo, H. Yang, Z. Li, Novel nonenzymatic hydrogen peroxide sensor based on iron oxide–silver hybrid submicrospheres, Talanta 81 (2010) 1650–1654. [12] L. Wang, X. Bo, J. Bai, L. Zhu, L. Guo, Gold nanoparticles electrodeposited on ordered mesoporous carbon as an enhanced material for nonenzymatic hydrogen peroxide sensor, Electroanalysis 22 (2010) 2536–2542. [13] X. Bo, J. Bai, L. Wang, L. Guo, In situ growth of copper sulfide nanoparticles on ordered mesoporous carbon and their application as nonenzymatic amperometric sensor of hydrogen peroxide, Talanta 81 (2010) 339–345. [14] D. Robson, F.Y.I. Assabghy, D.J.E. Ingram, Some electronic properties of polyacrylonitrile-based carbon fibres, Journal of Physics D: Applied Physics 5 (1972) 169–179. [15] M. Majumder, C. Rendall, M. Li, N. Behabtu, J.A. Eukel, R.H. Hauge, H. Schmidt, M. Pasquali, Insights into the physics of spray coating of SWNT films, Chemical Engineering Science 65 (2010) 2000–2008. [16] L.M. Manocha, O.P. Bhal, DTA studies on polyacrylonitrile fiber, Die Angewandte Makromolekulare Chemie. 64 (1977) 115–122. [17] E.V. Thompson, The thermal behavior of acrylonitrile polymers. I. On the decomposition of polyacrylonitrile between 250 and 350 ◦ C, Journal of Polymer Science Part B: Polymer Letters 4 (1966) 361–366.
64
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
[18] W.N. Turner, F.C. Johnson, The pyrolysis of acrylic fiber in inert atmosphere. I. Reactions up to 400 ◦ C, Journal of Applied Polymer Science 13 (1969) 2073–2084. [19] J.W. Johnson, P.G. Rose, G. Scott, Stabilization of polyacrylonitrile by oxidative transformation, British Polymer Journal 4 (1972) 527–540. [20] Q. Ouyang, L. Cheng, H. Wang, K. Li, Mechanism and kinetics of the stabilization reactions of itaconic acid-modified polyacrylonitrile, Polymer Degradation and Stability 93 (2008) 1415–1421. [21] I. Shimada, T. Takahagi, M. Fukuhara, K. Morita, A. Ishitani, FT-IR study of the stabilization reactions of polyacrylonitrile in the production of carbon fibers, Journal of Polymer Science Part A: Polymer Chemistry 24 (1986) 1989–1995. [22] K. Sen, P. Bajaj, T.V. Sreekumar, Thermal behavior of drawn acrylic fibers, Journal of Polymer Science Part B: Polymer Physics 41 (2003) 2949–2958. [23] M.K. Jain, A.S. Abhiraman, Conversion of acrylonitrile-based precursor fibers to carbon fibers, Journal of Materials Science 22 (1987) 278–300. [24] D. Lee, V.R. Erigala, M. Dasari, J. Yu, R.M. Dickson, N. Murthy, Detection of hydrogen peroxide with chemiluminescent micelles, International Journal of Nanomedicine 3 (2008) 471–476. [25] V. Penmatsa, T. Kim, M. Beidaghi, H. Kawarada, L. Gu, Z. Wang, C. Wang, Three-dimensional graphene nanosheet encrusted carbon micropillar arrays for electrochemical sensing, Nanoscale 4 (2012) 3673–3678. [26] A. Noorbakhsh, A. Salimi, E. Sharifi, Fabrication of glucose biosensor based on encapsulation of glucose-oxidase on sol–gel composite at the surface of glassy carbon electrode modified with carbon nanotubes and Celestine blue, Electroanalysis 20 (2008) 1788–1797. [27] T. You, O. Niwa, M. Tomita, S. Hirono, Characterization of platinum nanoparticle-embedded carbon film electrode and its detection of hydrogen peroxide, Analytical Chemistry 75 (2003) 2080–2085. [28] M.P.N. Bui, X.H. Pham, K.N. Han, C.A. Li, Electrocatalytic reduction of hydrogen peroxide by silver particles patterned on single-walled carbon nanotubes, Sensors and Actuators B 150 (2010) 436–441. [29] Y.J. Lee, J.Y. Park, Y. Kim, J.W. Ko, Amperometric sensing of hydrogen peroxide via highly roughened macroporous gold/platinum nanoparticles electrode, Current Applied Physics 11 (2011) 211–216. [30] S. Hrapovic, Y. Liu, B. Keith, J.H.T. Luong, Eletrochemical biosensing platforms using platinum nanoparticles and carbon nanotubes, Analytical Chemistry 76 (2004) 1083–1088. [31] X. Bai, G. Chen, K.K. Shiu, Electrochemical biosensor based on reduced graphene oxide modified electrode with Prussian blue and poly(toluidine blue O) coating, Electrochimica Acta 89 (2013) 454–460. [32] D.J. Brodrecht, J.J. Rusek, Aluminum hydrogen peroxide fuel-cell studies, Applied Energy 74 (2003) 113–124.
Biographies Junwei Su is a Ph.D. candidate in the Department of Mechanical Engineering at University of Massachusetts Lowell (UML). He received his Bachelor Degree at Harbin Institute of Technology. His research interest is in micro/nanofabrication technology, functional micro/nanostructure and their applications in biological analysis and energy areas. Fan Gao is currently a postdoctoral researcher at the University of Massachusetts Lowell. She received her B.S. from the Beijing University of Chemical Technology, China in 2007, her M.S. and Ph.D. in Chemical Engineering from the University of Massachusetts Lowell, USA in 2009 and 2012, respectively. Her research interests include the nanowire and nanoparticle-based nanosolders, nanojoining, self-assembly for electronics and electrochemical sensors. Zhiyong Gu is an associate professor in the Department of Chemical Engineering at the University of Massachusetts Lowell (UML). He is also affiliated with the CHN/NCOE Nanomanufacturing Center. He received his B.E. from Qingdao Institute of Chemical Technology (now Qingdao University of Science and Technology), China, in 1996, his M.S. from the University of Notre Dame in 2001, and his Ph.D. from the State University of New York at Buffalo in 2004, respectively. He was a Postdoctoral Fellow at the Johns Hopkins University from 2004 to 2006. He joined UML as an Assistant Professor in 2006 and was promoted to Associate Professor with tenure in 2012. He has published 4 book chapters and over 50 refereed papers. His current research interests include synthesis of nanoparticles and nanowires, self-assembly, block copolymers, nanocomposites, and nanoscale-integration for electronics and sensors. Michael Pien is the Vice President of R&D at ElectroChem Inc. He has been involved in extensive developments of PEM fuel cells, MEAs, separator plates and stack design, and automated test station development. Hongwei Sun is an associate professor in the Department of Mechanical Engineering at University of Massachusetts Lowell (UML). He graduated with a Ph.D. from Institute of Engineering Thermophysics at Chinese Academy of Science in 1998. Prior to joining UML in 2005, he worked as a postdoctoral researcher at University of Rhode Island (URI) and later a research scientist at Massachusetts Institute of Technology (MIT). His research interests are in the areas of Power Microelectromechanical systems (Power MEMS), MEMS acoustic sensors, and microscale cooling systems. His other interests are in micro/nanofabrication technology, fundamental understanding of micro/nanoscale fluidics and their applications in biological analysis and energy areas.
Contents lists available at SciVerse ScienceDirect
Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb
A novel 3-D fabrication of platinum nanoparticles decorated micro carbon pillars electrode for high sensitivity detection of hydrogen peroxide Junwei Su a , Fan Gao b , Zhiyong Gu b , Michael Pien c , Hongwei Sun a,∗ a b c
Department of Mechanical Engineering, University of Massachusetts Lowell, One University Avenue, Lowell, MA 01854, USA Department of Chemical Engineering, University of Massachusetts Lowell, One University Avenue, Lowell, MA 01854, USA ElectroChem, Inc, 400 West Cummings Park, Woburn, MA 01801, USA
a r t i c l e
i n f o
Article history: Received 18 December 2012 Received in revised form 30 January 2013 Accepted 6 February 2013 Available online xxx Keywords: Microscale carbon pillars Hydrogen dioxide detection Molding Electrochemical sensor Platinum/carbon nanoparticles
a b s t r a c t We report a novel fabrication technology for microscale carbon pillars (MCP) decorated with platinum/carbon nanoparticles (Pt/C NPs) for hydrogen peroxide (H2 O2 ) detection. The fabrication process involves three sequential steps: spray-coating of Pt/C nanoparticles, polydimethylsiloxane (PDMS) soft molding, and high temperature carbonization of polyacrylonitrile (PAN), which has great potential for large scale and high throughput manufacturing. The amperometric response of MCP based H2 O2 sensor exhibits superior electrochemical activity and charge transfer ability with a sensitivity of 1280–1750 AmM−1 cm−2 and a linear dynamic range (LDR) to 7000 M. Most importantly, the developed fabrication method can be used for a wide variety of functional nanomaterials for different applications such as electrochemical analysis, sustainable energy and biosensing. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Today’s analytical chemistry calls for complex instrumentation and considerable support, including special laboratory facilities and highly skilled personnel. However, chemical and biological sensing devices that are robust, portable and easy to use will lead to more innovative strategies for analytical instrumentation. In addition, they have other advantages such as requiring fewer reagents to operate and yielding reliable information continuously [1]. For instance, amperometric biosensors, usually formed by biologically surface-modified voltammetric electrodes, are gaining increasing importance owing to their high reliability, robustness and high sensitivity [2]. The early actual electrochemical sensors were based on the chemical modification of carbon paste. Complex functional groups such as silver ion were introduced to the carbon surface for the measurement of concentration of silver [3]. Another modification method was developed by Yao and Musha who dissolved anthraquinone (AQ) in the pasting liquid for immobilization of electroactive species [4]. These days most of electrochemical sensors were fabricated on glassy carbon electrode. Different materials, such as, anthraquinone [5], ordered mesoporous carbon modification [6], and nickel hexacyanoferrate/chitosan/carbon nanotubes (NiHCF/CS/CNTs) nanocomposite films [7], were immobilized on
∗ Corresponding author. Tel.: +1 978 934 4391; fax: +1 978 934 3084. E-mail address: Hongwei [email protected] (H. Sun). 0925-4005/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.snb.2013.02.023
the surface of glassy carbon electrodes to improve their electrochemical response. Nanostructured materials are becoming the focus of electrochemical sensor research. For instance, thionin nanowires [8] and CuO-MWCNTs [9] have been reported to improve hydrogen peroxide sensitivity of sensors. The performance of electrochemical sensor can be future improved by the use of so called hybrid nanostructures. He et al. built Ag nanoparticle 3D catalyst on a graphite substrate using Na2 Ti3 O7 nanowire as 3D frames for loading Ag nanoparticles [10]. Fe3 O4 –Ag hybrid submicrosphere was synthesized and developed as hydrogen peroxide sensor by Liu et al. [11]. Gold nanoparticles were attached on ordered mesoporous carbon (OMC) as hydrogen peroxide sensors [12]. A synthetic method to incorporate copper sulfide nanoparticles inside the mesopores of OMC is also reported for hydrogen peroxide sensor development [13]. However, random hybrid nanostructures hardly provide repeatable performance. There is no report for organized hybrid micro and nanoscale structures on the working electrode surface to further enhance sensor performance and repeatability. It is easy to understand that a 3-D organized microscale structure embedded with nanostructures such as catalyst metal nanoparticles can take advantage of both unique catalytic properties of metal NPs and high surface area provided by the microstructures, to achieve high performance and repeatable electrochemical sensing. In this paper, a novel method to fabricate platinum/carbonnanoparticles-decorated micro carbon pillars (Pt/C NPs MCP)
58
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
Fig. 1. Schematic of the fabrication process of Pt/C nanoparticles decorated micro carbon pillars.
electrodes is presented. The Pt/C NPs MCP sensor demonstrates superior properties well suitable for electrochemical sensing including low density, low electrical resistance, high resistance to chemical attack and impermeability to gases and liquid [14]. In addition, the surface to volume ratio of Pt/C NPs MCP is much larger than the traditional glassy carbon electrode thanks to high density micro pillar structure. Another critical feature of Pt/C NPs MCP electrode is its superhydrophilic property after thermal treatment, which can effectively prevent the gas bubbles from generating on electrode during hydrogen dioxide detection. In the first section, the process, materials and key parameters for Pt/C NPs MCP fabrication are introduced and discussed. Then the stabilization of polyacrylonitrile (PAN) material is described and analyzed using Fourier Transform Infrared Spectroscopy (FTIR) technique. In the meantime, the cyclic voltammetry (CV) measurement was conducted to characterize the charge transfer and electrochemical activity of developed Pt/C NPs MCP structure. At last, serious of experiments were performed to evaluate the performance of Pt/C NP MCP based hydrogen peroxide sensor. 2. Experimental 2.1. Materials and instruments Polyacrylonitrile (PAN) with molecule weight of 150,000 and N,N-dimethylformamide (DMF) were ordered from Sigma–Aldrich. The assay of DMF is greater than 99.8%. The polydimethylsiloxane (PDMS) for mold fabrication of soft molding was purchased from Dow Corning Corp. The Pt nanoparticles (2.9 nm) and carbon black nanoparticle (50 nm) were ordered from Alfa Aesar. Bruker Tensor 27 was used for the FTIR measurement for PAN samples after stabilization. The morphology of Pt/C NPs MCP was studied using a field emission scanning electron microscope (JEOLSEM, JSM-7401F). Cyclic voltammetry (CV), and amperometric measurements were conducted using a VersaSTAT 3 potentiostat (Princeton Applied Research, TN). A Pt wire (99.9%, Alfa Aesar) and Ag/AgCl electrode (BASi) were used as counter and reference electrodes in the measurement, respectively. 2.2. Pt/C NPs MCP fabrication The complete fabrication process is illustrated in Fig. 1 and consists of four steps: (1) PDMS mold fabrication; (2) preparation of
molding material; (3) spray-coating of Pt/C ink and molding process; and (4) stabilization and carbonization of PAN pillars. These steps are described in detail below. 2.2.1. PDMS mold fabrication PDMS was chosen as the mold material for molding of micro pillar arrays for two reasons: (1) PDMS has a lower surface energy than other substrates such as silicon and quartz, which can ease the mold removal after molding; (2) PDMS based molding method is well suited for low cost, large scale and high throughput nanomanufacturing. The PDMS was prepared by mixing the PDMS base and curing agent at a 10:1 ratio for 5 min in a glass container. A vacuum oven was used to remove the bubbles generated from mixing. A silicon substrate fabricated by deep reactive ion etching (DRIE) process to form micro pillar arrays was used as mothermold. Then the PDMS was evenly poured onto the surface of silicon substrate prior to being cured in a pressurized, heat-treated vacuum chamber for 60 min at 75 ◦ C. After the PDMS film was peeled off from the silicon substrate, the micro holes arrays were formed on the PDMS mold and the PDMS mold is ready for spray coating process. 2.2.2. Preparation of molding material The PAN (Sigma–Aldrich, molecular weight: 150,000) was dissolved in DMF solvent with a weight ratio of 2:8 and stirred at 80 ◦ C for 4 h. The dissolving time depends on the PAN concentration and amount. The higher concentration, the longer dissolving time is required. After a successful mixing, the PAN solution is clear and transparent. Commercially available Pt/C nanopowders were dispersed in DI water with a concentration of 5% to produce “Pt/C ink”. An ultra-sonication of 5–10 min for “Pt/C ink” is required before the spraying process. 2.2.3. Spray-coating of Pt/C ink and molding process The ultrasonic “Pt/C ink” was uniformly sprayed on the surface of PDMS mold using an air brush system (average droplet size: 6 m) at room temperature (20 ◦ C). During the spray process, the tiny Pt/C droplets were formed at the nozzle exit of the air brush system under pressured air. The distance between the mold surface and the nozzle exit was optimized to obtain the smallest droplets for a rapid solvent evaporation. After each coating, the cotton Q-tip with ink was used to repaint the mold surface to limit coffee stain effects [15].
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
59
After spray coating, the mold coated with Pt/C ink layer was pre-baked on a hot plate at 80–90 ◦ C for 1–2 min to evaporate the DI water. As a result, a layer of Pt/C NPs was left on the surface of these PDMS micro holes. Then the PAN/DMF solution was spincoated on the PDMS mold at the spinning speed of 3000 rpm in order to obtain thickness of 20 m. It is anticipated that the solution was uniformly filled into micro holes in PDMS mold. After this step, the PDMS mold with PAN/DMF film was heated again on a hotplate for 5 min at 100 ◦ C to evaporate the DMF in the film. Once the PAN solution layer was completely dry, the PAN layer with Pt/C NPs was removed from the PDMS mold. As a result, micro pillars decorated with Pt/C nanoparticles were generated on a PAN film surface. 2.2.4. Stabilization and carbonization of PAN micro pillars Oxidative stabilization is very critical to obtain production quality of carbon from PAN material. The stabilization was carried out at 250 ◦ C for 4 h to render them infusible and flameproof. Temperature below 200 ◦ C is impractical considering a very long period is required while temperatures above 300 ◦ C cause violent exothermic reactions [16,17], resulting in significant weight loss [18] and formation of tarry substances. On the other hand, improper stabilization could result in blowout of a core portion during carbonization due to its incomplete oxidation [19]. Proper conditions of the heating rate, time, and temperature were established for an optimal stabilization. The FTIR was used to monitor the stabilization procedure [20] and the results for transmittance of the PAN material after stabilization are presented in Fig. 2. In this optimization experiment, the stabilization temperature was varied from 220 ◦ C to 280 ◦ C while heating time is increased from 1 to 12 h. In Fig. 2, the band at 2243 cm−1 represents the C N stretching of acrylonitrile unit in the polymer chain and 2939 cm−1 stands for C H in CH2 . As the stabilization proceeds, the 2243 cm−1 band of the nitrile groups and 2939 cm−1 band of methylene continue to decrease in intensity. Besides, there were three new bands (at 1718, 1660 and 1595 cm−1 ) in carbonyl-stretch frequency region. The band 1595 cm−1 appears due to the combination vibration of C C and C N stretching, and NH in-plane bending of the ladder-frame structure of stabilized PAN [21]. The 1595 cm−1 band continues to increase, showing an enhancement of the extent of the stabilization reactions. As suggested by Shimada and Sen [21,22], the band at 1718 cm−1 is due to the free ketones in hydronaphthyridine rings and the 1660 cm−1 band is assigned to the conjugated ketones in acridone rings. These groups were generated by the oxygen uptake reactions. The original PAN sample contains a 1660 cm−1 peak due to the DMF solvent. It is apparent that the spectra does not change any more after heating for 4 h at 250 ◦ C, which implies a compete stabilization. It was observed that after the stabilization process, the color of PAN turned to light brown [23]. The carbonization was conducted in a temperature controllable tube furnace (Thermo Scientific, Lindberg/Blue M Mini-Mite Tube Furnace) with inert (nitrogen) gas purging through the quartz tube. The reaction formula during the stabilization and carbonization is shown in Fig. 3. The flow rate of nitrogen gas was controlled at 100 ml/min in the initial purging step. Then a high temperature heating was performed at 900 ◦ C for 20 min to carbonize the sample with the 60 ml/min N2 purging.
Fig. 2. (A) FTIR spectra of PAN material heated at 250 ◦ C for different durations: (a) original PAN, (b) 1 h, (c) 4 h, and (d) 12 h; (B) FTIR spectra of PAN heated for 4 h at different temperatures: (a) original PAN, (b) 220 ◦ C, (c) 250 ◦ C, and (d) 280 ◦ C.
investigating bubbling effect on sensor performance (see Sections 3.4 and 3.5). Fig. 4A and B shows the SEM images of micro carbon pillars and Pt/C NPs distribution on the pillar surfaces using COMPO mode of JEOL-SEM. The elements in these pillars material are shown with different contrasts based on the atomic number of
3. Result and discussion 3.1. Characteristics of Pt/C NPs MCP The micro carbon pillars in this research are 4 m in diameter, and 15 m in height with the spacing of 3 m, 5 m and 9 m, respectively. The three kinds of pillars were designed mainly for
Fig. 3. Reaction mechanisms of PAN stabilization and carbonization.
60
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
Fig. 4. SEM images of PAN based micro carbon pillar arrays including (A) bare micro carbon pillar arrays with spacing of 3 m and (B) Pt/C nanoparticles decorated micro carbon pillar arrays with spacing of 3 m (Pt nanoparticles are seen as very small bright spots on the background of the pillar arrays).
each element. The Pt/C NPs can be clearly seen on the pillar surface and inside the pillars (small white spots in Fig. 4B) compared to the bare micro carbon pillars (Fig. 4A). The composition of Pt/C NPs MCP was further characterized by energy dispersive X-ray (EDX) spectroscopy. For the bare micro carbon pillars, the EDX spectrum (Fig. 5A) shows the peaks corresponding to carbon and a small portion of oxygen, which is thought to be generated during carbonization. The EDX spectrum (Fig. 5B) of Pt/C NPs MCP shows the peaks corresponding to C, O and Pt elements, confirming the existence of Pt NPs on the surface of micro carbon pillar. The weight percentages of carbon, oxygen, silicon and platinum are 40.43% 4.55%, 5.99%, and 49.03%, respectively. The existence of a small portion of silicon was believed to be due to the impurity in Pt/C nanoparticles. Fig. 6 shows the high density micro carbon pillar array with embedded Pt/C NPs. It is worth noting that conventional coating techniques such as spray, vacuum deposition and spin-coating cannot form permanent bonding of particles with pillar surface. On the other side, other coating techniques such as E-beam deposition and ion beam assistant deposition require complex and expensive equipment. The developed approach in this research does not require any expensive equipment and can produce the hybrid
Fig. 5. EDX images of (A) bare micro carbon pillar and (B) Pt/C NPs micro carbon pillar.
Fig. 6. SEM micrographs of high aspect ratio and high density Pt/C nanoparticles decorated micro carbon pillars (spacing is 3 m).
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
Fig. 7. CV curves of bare and Pt/C catalyst decorated micro carbon pillars electrodes.
pillar-nanoparticle structure in a simple and reliable way. Furthermore, the Pt/C nanoparticles were permanently embedded in micro carbon pillars, which make this method reliable and robust for sensor applications. 3.2. Hydrogen peroxide (H2 O2 ) detection Rapid and accurate detection of H2 O2 is of great importance in clinical analyses and biosensors development in particular [24]. To evaluate the performance of Pt/C NPs MCP electrode for electrochemical sensing, a hydrogen peroxide (H2 O2 ) detection system based on conventional three-electrode cell was constructed with an Ag/AgCl (under saturated NaCl) electrode as the reference electrode, a platinum wire as the counter electrode, and Pt/C NPs MCP as the working electrode. Due to these micro pillar structures, the carbon based surface area has been increased from 0.2 cm2 to the 0.97 cm2 (3 m pillars spacing) and 0.67 cm2 (5 m pillars spacing) and 0.42 cm2 (9 m pillars spacing). The electrocatalytic activity of Pt/C NPs MCP electrode toward H2 O2 detection was evaluated by comparing its CV response with that of the bare micro carbon pillar electrode. Fig. 7 presents the steady-state CVs for the bare MCP and Pt/C NPs MCP electrodes in phosphate buffered solution (PBS) (pH = 7.2) between −1 and 1 V potential range. As expected, the Pt/C NPs MCP electrode displays a much better catalytic activity and facilitates a faster electron transfer at −0.4 V vs. Ag/AgCl electrode. In terms of the structural effect, the Pt/C nanoparticles were distributed along the surface of the micro carbon pillars with a thin layer which is believed to produce a catalytic activity. Comparing the areas under the two CV curves for bare MCP and Pt/C NPs MCP (Fig. 7), it is apparent that the total charge transfer of the Pt/C NPs MCP electrode is considerably greater than that of bare MCP. In Fig. 7, the modification efficiency of Pt/C is 225.8% at working potential (−0.4 V) compared to the bare microscale carbon pillars (MCP) electrode. Many metals such as Au, Ag, Pt, Pt, and Cu can act as catalyst for the decomposition of H2 O2 [12,13,27,29]. The overall reaction of the H2 O2 decomposition is: Catalyst
2H2 O2 −→ 2H2 O + O2 In the presence of Pt nanoparticles, H2 O2 experiences electrochemical redox reactions on the electrode. The reaction mechanism is proposed as follows. H2 O2
Pt nanoparticles
−→
2H+ + O2 + 2e−
(anodic)
61
Fig. 8. CV responses of Pt/C NPs MCP electrode to various concentrations (0.5, 1.0, 1.5, 2.0, 2.5 and 3 mM) of H2 O2 in a stirred PBS solution (pH = 7.2).
H2 O2 + 2e− → 2OH−
(cathodic)
In terms of the electronic effect, since PAN based carbon is a material with good electrical conductivity, the electron could transfer in two ways: directly to the PAN based carbon and through the Pt nanoparticles. The Pt nanoparticles increase effective surface area of electrode and facilitate indirect electron transfer via PAN based carbon as the mediator and showed the positive effect on the amperometric signals. It was suggested that a significantly increased charge transfer rate from H2 O2 to the PAN based carbon where Pt nanoparticles covalently linked. Hence, the electrons are easily transferred between the decorated Pt/C layer and carbon base which may enhance the sensing signals as well. It is anticipated that Pt/C NPs MCP electrode will generate enhanced signal to H2 O2 , which should be attributed by significantly increased catalytic surface, high conductivity and better charge transfer of platinum NPs. Fig. 8 shows the responses of Pt/C NPs MCP electrodes to H2 O2 at different concentrations. As can be seen, the well-defined and resolved cathodic and anodic peaks can be clearly distinguished on the forward and reverse scans. The cathodic peak is smaller than the anodic peak and the anodic peak current density increased linearly as the H2 O2 concentration increased. These features demonstrate that the Pt/C NPs MCP electrode is well suited for the detection of H2 O2 . In addition, the working potential of −0.4 V can be used as a static applied potential for the following amperometric detection measurement. 3.3. Amperometric detection of hydrogen peroxide The amperometric response of the Pt/C NPs MCP sensor to the successive additions of 500 M H2 O2 into PBS solution (pH = 7.2) are shown in Fig. 9 (with a constant potential of −0.4 V applied to the working electrode). The background response of the sensor was allowed decayed to a steady state after 50 s with magnetic stirring. The distinct potential signals responding to the increase of H2 O2 concentration was obtained in 5–10 s. The curve clearly shows the current increasing with the increasing of H2 O2 concentrations within a linear dynamic range (LDR) to 7000 M. To study the effect of the pillar spacing on sensing performance, the amperometric responses vs. step-wise increase of H2 O2 concentration for pillars with spacing of 3 m, 5 m, and 9 m were obtained and shown in Fig. 9 and Table 1. The MCP electrode with 3 m spacing has the largest surface area and most Pt NPs for electrochemical reaction, therefore lower limit of detection (LOD) and
62
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
Table 1 Amperometric response of Pt/C NPs MCP electrode with different spacing to step-wise change of H2 O2 concentration. Samples
I (A)
LOD (M)
Sensitivity (AmM−1 cm−2 )
R2
Associated error (%)
Pillar space = 3 m Pillar space = 5 m Pillar space = 9 m
174.61 143.66 128.07
9.60 15.04 17.71
1750 1440 1280
0.998 0.995 0.9997
2.34 8.46 5.08
higher sensitivity were obtained, as shown in Table 1. The LOD was determined when the limiting current for hydrogen peroxide from Pt/C NPs MCP electrode becomes three times greater than the standard deviation of a blank PBS (S/N = 3). The comparison of electroanalytical characteristics of Pt/C NPs MCP electrode for H2 O2 detection with that of other electrodes in literature was summarized in Table 2. As can be seen, the Pt/C NPs MCP electrode exhibits a rapid and sensitive response to the change of H2 O2 concentration. The response of the Pt/C NPs MCP electrode to H2 O2 behaves linearly ranging to 7000 M with sensitivities of 1280–1750 AmM−1 cm−2 . The sensitivity of Pt/C NPs MCP is higher than the most of Pt modified electrodes. Only the GC/CNT + Ptnano modified electrode [30] shows a higher sensitivity than that of Pt/C NP MCP. But Pt/C NPs MCP electrode demonstrates
wider range of LDR compared to GC/CNT + Ptnano modified electrode and is more suitable for large scale manufacturing. In addition, the anti-interference ability of the Pt/C NPs MCP sensor was investigated by adding 500 M of UA (uric acid), CA (citric acid), AA (ascorbic acid) and glucose sequentially into the solution and the sensor responses are shown in Fig. 10. The results demonstrate that the responses from UA, CA, AA and glucose are negligible compared with that observed for 500 M H2 O2 , indicating a high selectivity and anti-interference properties of Pt/C NPs MCP sensor. 3.4. Reproducibility and stability of Pt/C NPs MCP sensor To check the sensor reproducibility, five Pt/C NPs MCP sensors were prepared with the same fabrication method and tested for 0.5 mM hydrogen peroxide. A relative standard deviation of 5.15% was obtained. The operational stability of the modified electrode was investigated in continuous operation mode in which the data from the repeated experiments demonstrate a 2.6% deviation in sensing signal. Another experiment has been conducted to further check the stability of sensor performance by exposing the sensor to environment for 150 consecutive days. It is shown that the sensor still maintains 96.4% accuracy and demonstrates an excellent material stability and the bonding between Pt nanoparticles and carbon pillar. 3.5. Pt/C NPs MCP sensor surface fouling The closely spaced pillars provide more surface area for sensing and therefore higher sensitivity. However, it was found that when the H2 O2 concentration rise above 7000 M, an obvious bubbling process occurs, in which bubbles block the fresh chemicals to access to the sensing surface and results in significant oscillation in CV signal [32]. As a result, anodic peak will become undistinguished in the CV curve and sensor lost detection capability. Optimization is
Fig. 9. (A) Amperometric response curves of Pt/C NPs MCP electrode to step-wise change of 0.5 mM of H2 O2 ; (B) Calibration curve for step-wise change of 0.5 mM H2 O2 in a stirred PBS solution (pH 7.2) with operating potential of −0.4 V for sensors with pillar spacing of (a) 3 m, (b) 5 m, and (c) 9 m.
Fig. 10. Anti-interference property of the Pt/C NPs MCP electrode with addition of 500 M AA, UA, glucose, CA, and 500 M H2 O2.
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
63
Table 2 Comparison of electroanalytical characteristics of various modified electrodes toward H2 O2 detection. Electrode
Eapp (mV)
Graphene nanosheet encrusted MCP Fe3 O4 –Ag hybrid submicrosphere Pt NPs embedded carbon film GCE/CNTs/CB/sol–gel NiHCF/CS/CNTs AgPs-SWCNT
+400 −500 +600 −200 −200 −300
GC/CNT + Ptnano
+700
Macroporous Au-/nPts GC/RGO/PB/PTBO Ag-Na2 Ti3 O7 NWs Gold NPs on ordered mesoporous carbon Pt NPs MCP
+100 +200 +600 −150 −400
H2 O2
Reference
LDRa (M)
LODa (M)
Sensitivity (AmM−1 cm−2 )
250–5500 1.2–3500 nr 100–1000 40–5600 16–18,085 0.025–10 100–2000 ∼10,000 5–600 50–2500 2–3920 ∼7000
nr 1.2 0.075 10 0.28 2.76
0.07 12 3.93 ± 0.15b 1.055b 654 1092 3570 1850 264 420 1 22 1280–1750
0.025 50 1.5 1.0 0.49 9.6–17.7
[25] [11] [27] [26] [7] [28] [30] [29] [31] [10] [12] This work
a LDR: linear dynamic range; LOD: limit of detection; CNT: carbon nanotube; CS: chitosan; NiHCF: nickel hexacyanoferrate; AgPs: silver particles; PTBO: poly(toluidine blue O); PB: Prussian blue; RGO: reduced graphene oxide; MCP: micro carbon pillars; Ptnano : platinum nanoparticles; GCE, CG: glassy carbon electrode; CB: Celestine blue; NP: nanoparticle; nr: not reported. b AmM−1 .
Table 3 Detection of H2 O2 in Merrimack River water. Real sample
Merrimack River pH = 5.98
H2 O2 added (M)
H2 O2 found (M)
Recovery (%)
40 80 120
38.6 81.3 122.4
96.5% 101.6% 101.9%
ongoing for pillar diameter, spacing and height to diminish surface fouling of Pt/C NP MCP sensors for a wider range of LDR. 3.6. Detection of H2 O2 in Merrimack River with Pt/C NP MCP sensor A real water sample (pH = 5.98) was taken from the Merrimack River in Lowell, MA and was put into a glass beaker for two days for settlement. After that, the response of Pt/C NP MCP sensor was measured by adding hydrogen peroxide with a known concentration into the sample. Table 3 presents the comparison of prediction results from the sensor with the actual concentration in the water sample. The sensor has demonstrated an excellent sensing capability in detecting H2 O2 in a real sample although the limit of detection is decreased to 1.82 M (S/N = 3). 4. Conclusions In this paper, we demonstrated a novel, low cost and fast fabrication method to generate nanoparticles decorated micro carbon pillar structure. The FTIR was used to analyze and optimize the stabilization process. The structure and morphology of these pillars were identified by SEM and EDX. Besides of platinum nanoparticles, micro carbon pillars could be decorated by other particles for different applications. The CV studies confirmed the Pt/C NPs MCP electrode have faster charge transfer and higher electrochemical activity and selectivity toward H2 O2 than existing sensors. The results of this study indicate that the fabrication method of Pt/C NPs MCP is a promising manufacturing technology for highly sensitive electrochemical sensor development. Acknowledgments The authors are grateful for the materials support from ElectroChem Inc. This work was funded by National Science Foundation
(ECCS 0731125 and CMMI 0923403). We are grateful for the assistance in FTIR measurements from Professor David Ryan and his student Yipei Zhang at University of Massachusetts Lowell.
References [1] S. Alegret, Rigid carbon-polymer biocomposites for electrochemical sensing, Analyst 121 (1996) 1751–1758. [2] F. Scheller, F. Schubert, Biosensors, 1st ed., Elsevier, Amsterdam, 1992. [3] G.T. Cheek, R.F. Nelson, Applications of chemically modified electrodes to analysis of metal ions, Analytical Letters 11 (1978) 393–402. [4] T. Yao, S. Musha, Electrochemical enzymatic determinations of ethanol and l-lactic acid with a carbon paste electrode modified chemically with nicotinamide adenine dinucleotide, Analytica Chimica Acta 110 (1979) 203–209. [5] Y. Dou, S. Haswell, J. Greenman, J. Wadhawan, Immobilized anthraquinone for redox mediation of horseradish peroxidase for hydrogen peroxide sensing, Electrochemistry Communications 11 (2009) 1976–1981. [6] L. Zhu, C. Tian, D. Yang, X. Jiang, R. Yang, Bioanalytical application of the ordered mesoporous carbon modified electrode, Electroanalysis 20 (2008) 2518–2525. [7] Z. Wang, X. Hao, Z. Zhang, S. Liu, Z. Liang, G. Guan, One-step unipolar pulse electrodeposition of nickel hexacyanoferrate/chitosan/carbon nanotubes film and its application in hydrogen peroxide sensor, Sensors and Actuators B 162 (2012) 353–360. [8] H. Lin, H. Cheng, X. Miao, P. Papakonstantinou, D. Mihailovic, M. Li, A novel hydrogen peroxide amperometric sensor based on thionin incorporated onto a Mo6 S9 -xIx nanowire modified glassy carbon electrode, Electroanalysis 21 (2009) 2602–2606. [9] K.Y. Zhang, N. Zhang, H. Cai, C. Wang, A novel non-enzyme hydrogen peroxide sensor based on an electrode modified with carbon nanotube-wired CuO nanoflowers, Microchimica Acta 176 (2012) 137–142. [10] X. He, C. Hu, H. Liu, G. Du, Y. Xi, Y. Jiang, Building Ag nanoparticle 3D catalyst via Na2 Ti3 O7 nanowires for the detection of hydrogen peroxide, Sensors and Actuators B 144 (2010) 289–294. [11] Z. Liu, B. Zhao, Y. Shi, C. Guo, H. Yang, Z. Li, Novel nonenzymatic hydrogen peroxide sensor based on iron oxide–silver hybrid submicrospheres, Talanta 81 (2010) 1650–1654. [12] L. Wang, X. Bo, J. Bai, L. Zhu, L. Guo, Gold nanoparticles electrodeposited on ordered mesoporous carbon as an enhanced material for nonenzymatic hydrogen peroxide sensor, Electroanalysis 22 (2010) 2536–2542. [13] X. Bo, J. Bai, L. Wang, L. Guo, In situ growth of copper sulfide nanoparticles on ordered mesoporous carbon and their application as nonenzymatic amperometric sensor of hydrogen peroxide, Talanta 81 (2010) 339–345. [14] D. Robson, F.Y.I. Assabghy, D.J.E. Ingram, Some electronic properties of polyacrylonitrile-based carbon fibres, Journal of Physics D: Applied Physics 5 (1972) 169–179. [15] M. Majumder, C. Rendall, M. Li, N. Behabtu, J.A. Eukel, R.H. Hauge, H. Schmidt, M. Pasquali, Insights into the physics of spray coating of SWNT films, Chemical Engineering Science 65 (2010) 2000–2008. [16] L.M. Manocha, O.P. Bhal, DTA studies on polyacrylonitrile fiber, Die Angewandte Makromolekulare Chemie. 64 (1977) 115–122. [17] E.V. Thompson, The thermal behavior of acrylonitrile polymers. I. On the decomposition of polyacrylonitrile between 250 and 350 ◦ C, Journal of Polymer Science Part B: Polymer Letters 4 (1966) 361–366.
64
J. Su et al. / Sensors and Actuators B 181 (2013) 57–64
[18] W.N. Turner, F.C. Johnson, The pyrolysis of acrylic fiber in inert atmosphere. I. Reactions up to 400 ◦ C, Journal of Applied Polymer Science 13 (1969) 2073–2084. [19] J.W. Johnson, P.G. Rose, G. Scott, Stabilization of polyacrylonitrile by oxidative transformation, British Polymer Journal 4 (1972) 527–540. [20] Q. Ouyang, L. Cheng, H. Wang, K. Li, Mechanism and kinetics of the stabilization reactions of itaconic acid-modified polyacrylonitrile, Polymer Degradation and Stability 93 (2008) 1415–1421. [21] I. Shimada, T. Takahagi, M. Fukuhara, K. Morita, A. Ishitani, FT-IR study of the stabilization reactions of polyacrylonitrile in the production of carbon fibers, Journal of Polymer Science Part A: Polymer Chemistry 24 (1986) 1989–1995. [22] K. Sen, P. Bajaj, T.V. Sreekumar, Thermal behavior of drawn acrylic fibers, Journal of Polymer Science Part B: Polymer Physics 41 (2003) 2949–2958. [23] M.K. Jain, A.S. Abhiraman, Conversion of acrylonitrile-based precursor fibers to carbon fibers, Journal of Materials Science 22 (1987) 278–300. [24] D. Lee, V.R. Erigala, M. Dasari, J. Yu, R.M. Dickson, N. Murthy, Detection of hydrogen peroxide with chemiluminescent micelles, International Journal of Nanomedicine 3 (2008) 471–476. [25] V. Penmatsa, T. Kim, M. Beidaghi, H. Kawarada, L. Gu, Z. Wang, C. Wang, Three-dimensional graphene nanosheet encrusted carbon micropillar arrays for electrochemical sensing, Nanoscale 4 (2012) 3673–3678. [26] A. Noorbakhsh, A. Salimi, E. Sharifi, Fabrication of glucose biosensor based on encapsulation of glucose-oxidase on sol–gel composite at the surface of glassy carbon electrode modified with carbon nanotubes and Celestine blue, Electroanalysis 20 (2008) 1788–1797. [27] T. You, O. Niwa, M. Tomita, S. Hirono, Characterization of platinum nanoparticle-embedded carbon film electrode and its detection of hydrogen peroxide, Analytical Chemistry 75 (2003) 2080–2085. [28] M.P.N. Bui, X.H. Pham, K.N. Han, C.A. Li, Electrocatalytic reduction of hydrogen peroxide by silver particles patterned on single-walled carbon nanotubes, Sensors and Actuators B 150 (2010) 436–441. [29] Y.J. Lee, J.Y. Park, Y. Kim, J.W. Ko, Amperometric sensing of hydrogen peroxide via highly roughened macroporous gold/platinum nanoparticles electrode, Current Applied Physics 11 (2011) 211–216. [30] S. Hrapovic, Y. Liu, B. Keith, J.H.T. Luong, Eletrochemical biosensing platforms using platinum nanoparticles and carbon nanotubes, Analytical Chemistry 76 (2004) 1083–1088. [31] X. Bai, G. Chen, K.K. Shiu, Electrochemical biosensor based on reduced graphene oxide modified electrode with Prussian blue and poly(toluidine blue O) coating, Electrochimica Acta 89 (2013) 454–460. [32] D.J. Brodrecht, J.J. Rusek, Aluminum hydrogen peroxide fuel-cell studies, Applied Energy 74 (2003) 113–124.
Biographies Junwei Su is a Ph.D. candidate in the Department of Mechanical Engineering at University of Massachusetts Lowell (UML). He received his Bachelor Degree at Harbin Institute of Technology. His research interest is in micro/nanofabrication technology, functional micro/nanostructure and their applications in biological analysis and energy areas. Fan Gao is currently a postdoctoral researcher at the University of Massachusetts Lowell. She received her B.S. from the Beijing University of Chemical Technology, China in 2007, her M.S. and Ph.D. in Chemical Engineering from the University of Massachusetts Lowell, USA in 2009 and 2012, respectively. Her research interests include the nanowire and nanoparticle-based nanosolders, nanojoining, self-assembly for electronics and electrochemical sensors. Zhiyong Gu is an associate professor in the Department of Chemical Engineering at the University of Massachusetts Lowell (UML). He is also affiliated with the CHN/NCOE Nanomanufacturing Center. He received his B.E. from Qingdao Institute of Chemical Technology (now Qingdao University of Science and Technology), China, in 1996, his M.S. from the University of Notre Dame in 2001, and his Ph.D. from the State University of New York at Buffalo in 2004, respectively. He was a Postdoctoral Fellow at the Johns Hopkins University from 2004 to 2006. He joined UML as an Assistant Professor in 2006 and was promoted to Associate Professor with tenure in 2012. He has published 4 book chapters and over 50 refereed papers. His current research interests include synthesis of nanoparticles and nanowires, self-assembly, block copolymers, nanocomposites, and nanoscale-integration for electronics and sensors. Michael Pien is the Vice President of R&D at ElectroChem Inc. He has been involved in extensive developments of PEM fuel cells, MEAs, separator plates and stack design, and automated test station development. Hongwei Sun is an associate professor in the Department of Mechanical Engineering at University of Massachusetts Lowell (UML). He graduated with a Ph.D. from Institute of Engineering Thermophysics at Chinese Academy of Science in 1998. Prior to joining UML in 2005, he worked as a postdoctoral researcher at University of Rhode Island (URI) and later a research scientist at Massachusetts Institute of Technology (MIT). His research interests are in the areas of Power Microelectromechanical systems (Power MEMS), MEMS acoustic sensors, and microscale cooling systems. His other interests are in micro/nanofabrication technology, fundamental understanding of micro/nanoscale fluidics and their applications in biological analysis and energy areas.