d , Defense ARTICLE INFO Article history: Received 25 November 2010 Accepted 18 April 2011 Available online 27 April 2011 silicon-based solar cells, the fabrication costs of DSSCs are must overcome grain boundaries to be collected by a conduc- tive glass. The low transfer efficiency of photo-induced elec- trons across a TiO 2 matrix represents a major limitation of such nanostructured photoanodes. The inefficient charge transfer paths cause photo-induced electrons to recombine acid-treated multi-walled carbon nanotubes (acid-MWCNTs) carbon nanotubes (CNTs). Electron transfer efficiency was also limited by the resistance of the contacts between interacting CNTs [4]. Recently, two-dimensional (2-D) structured graphene has emerged as a conductive nanomaterial showing considerate CARBON 49 (2011) 3597–3606 available at www.sciencedirect.com * Corresponding author: Fax: +886 35715408. lower and the performance is higher. Despite the 11% solar conversion efficiency already achieved with these cells, any further improvement in performance would provide consider- ate benefits [2]. Within the photoanode of a DSSC, photo-induced electrons into the TiO 2 matrix of the photoanode of DSSCs can improve the performance of the cell by improving the electron conduc- tion paths and distribution of the pores. The improvement in performance thus obtained was limited by the restricted area of contact between the TiO 2 nanoparticles and the cylindrical 1. Introduction Since their initial introduction by Gra¨tzel et al., dye-sensitized solar cells (DSSCs) have been attracting interest from aca- demics and practitioners alike [1]. Compared with traditional with the oxidizing species or tri-iodide ions present in the electrolyte, resulting in a decrease in photocurrent and photo- conversion efficiency. Preventing charge recombination can therefore improve the photo-induced transfer of electrons. In a previous study [3], we reported that the introduction of 0008-6223/$ - see front matter Crown Copyright doi:10.1016/j.carbon.2011.04.062 E-mail address:
[email protected] (C.M. ABSTRACT This study employed a solution-based method to prepare a 3-D hybrid material comprising graphene and acid-treated multi-walled carbon nanotubes (MWCNTs). The adsorption of MWCNTs on graphene reduces the p–p interaction between graphene sheets resulting from steric hindrance, providing a subsequent reduction in aggregation. Optimal proportions of MWCNTs to graphene (2:1) enabled the even distribution of individual MWCNTs deposited on the surface of the graphene. The hybrid 3-D material was incorporated within a TiO 2 matrix and used as a working electrode in dye-sensitized solar cells (DSSCs). The hybrid material provides a number of advantages over electrodes formed of either MWCNTs or graphene alone, including a greater degree of dye adsorption and lower levels of charge recombination. In this study, DSSCs incorporating 3-D structured hybrid materials demon- strated a conversion efficiency of 6.11%, which is 31% higher than that of conventional TiO 2 -based devices. Crown Copyright C211 2011 Published by Elsevier Ltd. All rights reserved. 33448, Taiwan Preparation of graphene/multi-w and its use as photoanodes of Ming-Yu Yen a , Min-Chien Hsiao a , Shu-Hang Chen-Chi M. Ma a, * , Nen-Wen Pu b , Ming-Der a Department of Chemical Engineering, National Tsing Hua University b Department of Applied Chemistry and Materials Science, National journal homepage: www.elsevier.com/locat C211 2011 Published by Ma). alled carbon nanotube hybrid ye-sensitized solar cells Liao a , Po-I Liu a , Han-Min Tsai a , Ger b No. 101, Section 2, Kuang-Fu Road, Hsinchu 30013, Taiwan University, No. 1000, Shingfeng Road, Bade, Taoyuan County e/carbon Elsevier Ltd. All rights reserved. promise for use in organic solar cells [5–8] and DSSCs [9–12]. Graphene has been progressively developing in the applica- tion of photoanodes [10,11,13]. Yang et al. introduced graph- ene into a TiO 2 matrix as a 2-D bridge for the photoanode of DSSCs. They demonstrated lower recombination rates, increased efficiency in electron transport, and enhanced light scattering through the incorporation of graphene into the photoanode, contributing to improved cell performance [10]. Sun et al. used the coagulation of commercial TiO 2 (P25) on to Nafion-coated graphene to prepare a photoanode with good interfacial binding. They demonstrated that the intro- duction of graphene to TiO 2 -based working electrodes improves in both dye adsorption and electron lifetime [13]. Tan et al. reported on the preparation of a photoanode in which chemically exfoliated graphene was incorporated into aTiO 2 matrix, using a method of molecular grafting. They determined that the implanted graphene provided additional transport pathways for photo-induced carriers, leading to an increase in photocurrent [11]. Due to high specific surface area and p–p interactions, dispersing graphene is very diffi- cult; however, improving the dispersion of graphene in the matrix is critical to its successful application in photoanodes. Two strategies to improve the dispersion of graphene have previously been reported, namely covalent functionalization [14–16] and non-covalent functionalization using aromatic molecules [17,18]. The covalent functionalization of graphene adversely influences the electrical properties of the matrix, due to the strength of the chemical bonding and the defects introduced to the surface of the graphene. The non-covalent functionalization of graphene with aromatic molecules leads to poor electron transport at the photoanode, due to the high resistivity of the contacts [19]. Recently, a 3-D hybrid material was reported comprising pristine MWCNTs (1-D) and chemi- cally converted graphene (2-D). In this system, the combined use of these two materials in the matrix resulted in a twin ef- fect influencing both the electrical conductivity and electro- chemical performance of the photoanode. In these reports [20–23], 3-D hybrid material exhibited stable dispersion in the polar solvents H 2 O, DMF and hydrazine; however, the fab- rication procedure was somewhat complex, thereby hinder- ing reproducibility, and the presence of a residual reducing agent between the graphene sheets following chemical reduc- tion was another drawback of this method. 3-D structured nano-materials have also been prepared by aligning CNTs onto a graphene surface using chemical vapor deposition (CVD) at high temperatures (exceeding 750 C176C) [24,25]. In the study described herein, we report a simple approach to preparing a 3-D hybrid graphene material using a two-step solution-based method at room temperature. First, graphite oxide (GO) was fabricated using Staudenmaier’s method [26], followed by heat treatment (1050 C176C) for 30 s in an inert atmosphere to form graphene. The graphene was then dis- persed in an ethanol solution containing acid-MWCNTs fol- lowed by sonication for 1 h at room temperature to yield a novel hybrid material, as shown in Fig. 1. The prepared 3-D hybrid was then incorporated into TiO 2 -based photoanodes 3598 CARBON 49 (2011) 3597–3606 Fig. 1 – The mechanism of acid-MWCNTs in improving the dispersibility of graphene. (87.5 ml) and nitric (45 ml) acid, and the mixture was stirred. Following the uniform dispersal of graphite, potassium chlo- a surfactant, and anhydrous ethanol (99.5%, Acros Co., USA) were mixed. Dispersed solutions were obtained using ultra- electrodes were also measured using a 2400 digital source me- sonic horn. The solution was concentrated in an rotary evap- orator to a final concentration of 20 wt.%, thus the paste was obtained. To prepare the DSSC photoanodes, fluorine-doped tin oxide glass (TEC-7, 7 X square C01 , Hartford Glass Co., USA) plates were coated with a layer of paste. The coated electrodes were heated in an air atmosphere at 500 C176C for 30 min. The thickness and active cell area of the TiO 2 film were approximately 10 lm and 0.16 cm 2 , respectively. Four composite photoanodes were prepared, each containing rate (55 g) was slowly added to the mixture and the stirring was continued for at least 96 h to sure complete oxidation. The mixture was then added to deionized water, and filtered. The GO was rinsed repeatedly using deionized water and re- dispersed three times in a 5% HCl (aq) solution. It was then washed continuously using deionized water until the pH of the filtrate was neutral. The GO slurry was dried and pulver- ized twice. The GO was heated to 1050 C176C in an inert atmo- sphere and held at this temperature for 30 s to form graphene. The acid treatment of MWCNTs was performed according to a nitric acid washing procedure; one gram of raw-MWCNTs was boiled in 50 ml of concentrated nitric acid for 4 h. The MWCNTs were then filtered, washed with deion- ized water several times to remove residual acid, and dried in an oven at 105 C176C. This was the method used to produce the hybrid material. MWCNTs and graphene were combined in various proportions in 10 ml ethanol followed by ultrasonica- tion for 1 h. 2.2. Fabrication of the DSSCs Various composite pastes were prepared for doctoral blading as described below. TiO 2 colloid (prepared according to a re- lated method [27]), nanostructural material (loading amount: 0.1 wt.%), ethyl cellulose (10 wt.%, TCI Chemical Co., USA) as a binder, a-terpineol (Showa Chemical Industry Co., Japan) as for application in DSSCs. To the best of our knowledge, this is the first report of a 3-D hybrid photoanode. The dispersibility of the hybrid material was investigated using UV–Vis spec- troscopy and transmission electron microscopy (TEM). We also investigated the photoelectric properties of the 3-D hybrid photoanode, revealing that the performance of this hybrid photoanode was better than that of previously re- ported acid-MWCNT- and graphene-containing photoanodes [3,10,11,13]. 2. Experimental section 2.1. Synthesis of graphene Natural graphite powder (particle size C2470 lm, Alfa Aesar) was oxidized using Staudenmaier’s method [26] to produce GO. In this method, graphite (5 g) was first mixed with sulfuric CARBON 49 (2011) MWCNTs (denoted electrode 1, C tube 100, CNT Co. Ltd., Korea), hybrid material (electrode 2, 0.07 wt.% MWCNTs ter (Keithley, USA) under illumination with a Class A sunlight simulator at 100 mW cm C02 (91160A, AM 1.5, Oriel, Newport Corporation, USA), equipped with an AM 1.5G filter (81088A, Oriel, Newport Corporation, USA) and a 300 W xenon lamp (6258, Oriel, Newport Corporation, USA). The intensity of the simulated incident light was calibrated to 100 mW cm C02 using a reference Si solar cell, calibrated at the National Renewable Energy Laboratory (NREL) (USA) institutes. Electrochemical impedance spectra (EIS) were obtained using a potentiostat/ galvanostat equipped with a frequency response analysis (FRA) module (PGSTAT 302N, Autolab, EcoChemie, Nether- lands), under illumination of 100 mW cm C02 . The scanned fre- quency range was from 10 5 to 10 C02 Hz, with an applied voltage of 10 mV. Impedance spectra were analyzed using an equiva- lent circuit model with Autolab FRA software (v4.9, Eco Chemie B.V.). The incident photon-to-current conversion effi- and 0.03 wt.% graphene), graphene (electrode 3) or prinstine TiO 2 (electrode 4), respectively. Prior to the fabrication of the DSSCs, the electrodes were sensitized by soaking for 24 h in a3· 10 C04 M solution of ruthenium dye (cis-dithiocyanato- N,N 0 -bis(2,2 0 -bipyridyl-4-carboxylicacid-4 0 -tetra-butylammo- nium carboxylate) ruthenium(II); N719, Solaronix SA, Swit- zerland) in acetonitrile/t-butyl alcohol (v:v = 1:1). The sensitized electrodes were then immersed in acetonitrile for 12 h. The DSSC comprised a sensitized working electrode, a platinized counter (Pt) electrode and an electrolyte with a 25 lm hot-melt sealing foil (SX1170-60, Solaronix SA, Swit- zerland) between each layer. The electrolyte used in this study consisted of 0.6 M BMII, 0.1 M guanidinium thiocya- nate, 0.03 M iodine, and 0.5 M TBP in acetonitrile/valeronitri- le (volume ratio: 85:15) [27]. 2.3. Characterization of the dispersion of carbon materials and composite working electrodes We investigated the dispersibility of the carbon materials in ethanol. We first obtained a calibration curve by measuring (at 270 nm) the adsorption (Varian, Cary 50, Varian, USA) of the solution of graphene and MWCNTs at a known concentra- tion [28]. The tested carbon materials were then suspended in ethanol with various ratios of MWCNTs to graphene (10 ml) and sonicated for 1 h. The solutions were centrifuged at 3000 rpm for 15 min, whereupon the absorption of the super- natants was measured at 270 nm [19,20]. The morphology of the composites working electrode was then studied using scanning electron microscopy (SEM) (Hitachi S-4700I, Japan), and the microstructures of the carbon materials was also investigated using transmission electron microscopy (TEM) (JEOL 2100F, Japan) at an accelerating voltage of 200 kV. The quantity of dye adsorbed into the electrode was determined by desorbing the dye from the composites into a 0.1 M KOH aqueous solution, and measuring the absorbance of the resulting solution using UV–vis spectroscopy (Varian, Cary 50, Varian, USA). Photoluminescence (PL) spectra were ob- tained by using an F-4500 Luminescence Spectrometer (Hit- achi, Japan). The photocurrent–voltage characteristics of the 3597–3606 3599 ciency (IPCE) was measured using an IPCE instrument (Model QEX7, PV measurement Co., USA). 3. Results and discussion 3.1. Dispersion of the hybrid materials In our previous work [3], we showed that the aggregation of acid-MWCNTs seriously restricts the charge transport and adsorption of dye by the photoanode. In the present study, acid-MWCNT-graphene hybrids, composed of various ratios of the two components, were prepared to investigate their microstructure. As shown in Fig. 2, both the pristine MWCNTs and the graphene were found to form unstable suspensions following sonication, leading to aggregation and subsequent precipita- tion. The aggregation of pristine MWCNTs and graphene are known to be due to a lack of hydrophilic groups in their struc- ture, prohibiting interaction with the polar solvent. Although the presence of hydrophilic groups (e.g. hydroxyl, epoxy, or carboxyl) on the surface of graphene and edges of graphene can improve dispersion, they are generally insufficient to al- low complete dispersion due to the large hydrophobic surface area of graphene [29]. In contrast, Fig. 1 shows the nearly 3600 CARBON 49 (2011 complete dispersion of both acid-MWCNTs and hybrid mate- rials in solution following sonication. After one week, much of the suspended acid-MWCNT material had aggregated and precipitated. This was due to the attraction produced by the large hydrophobic specific surface of the CNTs, even after acid-treatment [20]. On the other hand, the hybrid materials exhibited stable dis
中国文库网