P25粉末制备高效染料敏化太阳能电池

Electrochimica Acta 94 (2013) 277 – 284Contents lists available at SciVerse ScienceDirectElectrochimica Actaj o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e /e l e ct a c t aFabrication of mesoporous TiO 2 electrodes by chemical technique fordye-sensitized solar cellsYuan Yan, Jinzhong Wang ? , Quanhong Chang, Musbah Babikier, Huixin Wang, Hongtao Li,Qingjiang Yu, Shiyong Gao, Shujie JiaoDepartment of Opto-electric Information Materials and Quantum Devices, School of Materials Science and Engineering, Harbin Institute of Technology, 150001 Harbin, Chinaa r t i c l e i n f oArticle history:Received 15 November 2012Received in revised form26 December 2012Accepted 4 February 2013Available online xxxKeywords:Dye-sensitized solar cellsChemical dispersionP25 powderTransparent electrodesPACS:75.40.-s71.20.LPa b s t r a c tA new chemical technique for preparing screen-printing paste from commercially-available P-25 powderwas proposed to fabricate TiO 2 electrodes for dye-sensitized solar cells (DSSCs). The TiO 2 powder andacetic acid were dispersed in ethanol, following that the reactants solutions kept at 80 ?C for 12 h, thendried in oven for 6 h. This process made the TiO 2 particle positively charged and highly dispersed in water.Additionally ?occulating reaction using hydrochloric acid was done to change the dispersed TiO 2 colloidinto viscous paste, which omitted the process of adding polymers as thickener, and kept the purity andporosity of electrodes. The mesoporous transparent TiO 2 ?lms about 14 ? m thickness without crackingand peeling-off were fabricated on conducting glass substrates. A high conversion ef?ciency of 8.07% wasobtained under 100 mW cm - 2 AM 1.5 illumination.? 2013 Elsevier Ltd. All rights reserved.1. IntroductionSince the ?rstly typical nanocrystalline TiO 2 dye-sensitizedsolar cells (DSSCs) reported by Gr?tzel, this type of DSSCs hasbeen considered as a promising solution for many impendingenergy and environmental problems due to its low cost, eco-friendly characteristics, and reasonable ef?ciency (>11%) [1 – 5].To date, much research efforts to improve the overall conver-sion ef?ciency of TiO 2 nanocrystalline-based DSSCs has beenintensively made. Particularly, more and more researchers havefocused on TiO 2 nanocrystalline porous electrodes with largesurface area, where more dyes can be suf?ciently adsorbed andresult in a higher light-to-electricity conversion ef?ciency. Inorder to fabricate excellent TiO 2 porous electrodes, the challengeis to ?nd an optimal method to fabricate the correspondingTiO 2 paste with unique quality and characteristics. For the best-performing TiO 2 electrodes, the synthesis of TiO 2 paste involveshydrolysis of Ti(OCH(CH 3 ) 2 )4 in water at 250 ? C (70 atm) for 12 h,followed by conversion of the water to ethanol by three-times? Corresponding author. Tel.: +86 451 86418745; fax: +86 451 86418745.E-mail address: jinzhong wang@hit.edu.cn (J. Wang).centrifugation. Finally, the ethanol is exchanged with ? -terpineolby sonication and evaporation. In words, this method needs thespecialized and complicated equipments and techniques, which areeconomically unsuitable for industrial production. In order to sim-plify the technology, several reports on fabricating screen-printingpastes from a commercially-available TiO 2 powder (P25, Degussa)were published [6 – 12]. Most of the pastes were based on waterand ethanol, which induce TiO 2 aggregation and yielding poorlyreproducible results in long-term experiments. As known, ? -terpineol-based pastes are very stable and show very reproducibleresults. However, a reported paste made from P25 and ? -terpineolhad not given better DSSCs than that of a water-based paste[13 – 15] .One simply chemical technique that would replace the conven-tional complicated methods to prepare water-based paste froma commercially-available TiO 2 powder (P25, Degussa) has beenproposed in this paper. Firstly, the TiO 2 powder was modi?ed toCH 3 COO - /TiO 2 with acetic acid (HAc) at a temperature of 80 ? Cin ethanol, and then evaporated to dryness at 60 ? C. Secondly,water-based viscous paste for screen-printing was obtained byadding hydrochloric acid to neutralize the charges of TiO 2 colloid.Finally, the dispersion quality of CH 3 COO - /TiO 2 powder in waterwas studied by DLS and TEM. The porosity and photovoltaic char-acteristics of TiO 2 electrodes were investigated by BET and solarsimulator.0013-4686/$ – see front matter ? 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.electacta.2013.02.019278 Y. Yan et al. / Electrochimica Acta 94 (2013) 277 – 284Fig. 1. Fabrication scheme of screen-printing paste from P25 powder.2. Experimental2.1. MaterialsP-25 TiO 2 powders (av. 25 nm, 80% anatase and 20% rutile,Degussa); FTO glasses (15 / , 2.2 mm thickness, LOF Industries);N-719, BMII, GNCS, I 2 , were purchased from Dyesol company; 4-tert-butylpyridine (Aldrich), acetonitrile (Fluka) and valeronitrile(Fluka) were distilled before use; TiCl 4 (Fluka) was diluted withwater to 2 M at 0 ? C to make a stock solution, which was kept ina freezer and freshly diluted to 40 mM with water for each TiCl 4treatment.2.2. Preparation of porous electrodesThe fabrication scheme for screen-printing paste was given inFig. 1. In order to make acetic acid adsorb on TiO 2 surface suf?-ciently by chemisorptions, the TiO 2 powder and HAc were mixedwith ethanol, and then kept at 80 ? C for 12 h. After that, the mix-ture was dried in a drying oven at 60 ? C for 6 h, and CH3 COO - /TiO 2powder was obtained. The coordination of CH 3 COOH with hydrox-ides ( OH) on the TiO 2 surface would take place in the heatingprocess. The ethanol can facilitate the coordination in the heatingprocess and avoid the aggregation during the dryness. Due to thehydrophilic group, the CH 3 COO - /TiO 2 powder was easy to formwater-based TiO 2 colloid with the sonication assistance. At last, lit-tle hydrochloric acid was necessary to change the TiO 2 colloid intoviscous TiO 2 paste by ?occulating reaction.The FTO glasses were ?rstly cleaned in a detergent solution usingan ultrasonic bath for 15 min, and then respectively rinsed withethanol and deionized water. The TiO 2 electrodes were fabricatedby a screen-printing method and gradually sintered at 450 ?C for10 min after each deposition. The thicknesses of one screen-printcoating was controlled as 3 ? m by Alphastep. Five times of screen-printing procedure were required to get an appropriate thickness of14– 15 ? m for the electrodes. Finally, the electrodes were sinteredat 500 ?C for 30 min in air.2.3. Fabrication of dye-sensitized solar cellsAfter sintered at 500 ? C for 30 min, the TiO 2 electrodes wereimmersed into a 40 mM aqueous TiCl 4 solution at 70 ?C for 30 min,then rinsed with ethanol and deionized water. Subsequently, theelectrodes were sintered at 500 ?C for 30 min again. When thetemperature reached 80 ? C in the cooling process, the electrodeswere immersed into 0.5 mM N-719 dye solution in a mixture ofacetonitrile and tert-butyl alcohol (volume ratio: 1:1), then keptat room temperature for 24 h to complete the sensitizer uptake.The dye-sensitized TiO 2 electrodes were subsequently rinsed withacetonitrile, and then further sandwiched with the thermally pla-tinized FTO positive electrodes, separated by a 30 ? m thick Bynel(DuPont) hot-melt spacer to assemble the solar cells. The inter-vening space was ?lled with a drop of the electrolyte (0.60 MBMII, 0.03 M I2 , 0.10 M guanidinium thiocyanate and 0.50 M 4-tert-butylpyridine in the mixture of acetonitrile and valeronitrile(volume ratio: 85:15)).2.4. MeasurementsP-25 dispersion characteristics in water were analyzed bydynamic laser-light scattering (Otsuka Electronics, DLS-70SAr).The morphology of the TiO 2 aggregations was characterized bytransmission electron microscopy (TEM, JEOL CM200). The trans-mittance spectra were recorded by UV– vis spectrophotometer(Phenix, UV1700PC). The morphology of the porous TiO 2 ?lm wascharacterized by scanning electron microscope (SEM, Hitachi S-5200). The porosity of porous TiO 2 ?lms was measured by BET(BELSORP-max). Photovoltaic measurements were completed withan AM 1.5 solar simulator (Solar Light Company) in combinationwith a metal mesh employed to give an irradiance of 100 mW cm - 2 .The light intensity was tested using a PMA2144 pyranometer and acalibrated PMA 2100 dose control system. During measurements,the total active area of the cells was 0.25 cm - 2 , the illuminatedpart corresponding to the aperture area of the metal mask was0.16 cm - 2 . I– V curves were recorded using a Keithley model 2400digital source meter.3. Results and discussion3.1. Effects of adsorption treatments on TiO 2 particles dispersionTiO 2 particles with a good dispersion are one key factor forfabricating high quality paste. Therefore, the dispersion charac-teristics of TiO 2 particles were investigated in detail. Fig. 2 (a)shows the size-distribution of P25 powder without adsorptiontreatment. As seen, there are two size-distribution regions around100 nm and 400 nm, and the particles around 400 nm dominatedthe region. These phenomena revealed the presence of large sizeTiO 2 aggregations. After the adsorption treatments with HAc, onlyone size-distribution region was observed. Further, as shown inFig. 2 , the size-distribution region moves to small diameter sidewith increase in the amount of HAc. When the amount of HAc was2 ml, the best size-distribution around 30 nm was obtained, whichis also coincided well with that of primary particles of TiO 2 powder(80% anatase (d = 21 nm) and 20% rutile (d = 50 nm)). All above indi-cated that 2 ml HAc is enough to achieve monodispersion of 10 gP-25 powder and the surplus HAc will be evaporated in the dryingprocess.In order to study the dispersion mechanism of HAc, 100 mg TiO 2particles samples were dispersed in 20 ml water and the relatedzeta potential was measured. As shown in Fig. 2 (e), with an increasein HAc, the zeta potential increases abruptly. Further, the incrementgradually reduces and the zeta potential approximately reaches amaximum when the amount of HAc is 2 ml. Compared with theuntreated TiO 2 particles, the zeta potential increases from +5.35 mVto +46.1 mV for the particles treated with 2 ml HAc. All these indi-cate that the CH 3 COO - /TiO 2 particles can be stabilized againstagglomeration due to the electrostatic repulsion causing by thecharges absorbed on the particle surface.Fig. 3 shows the TEM images of TiO 2 samples treated with dif-ferent amount of HAc. As seen in Fig. 3 (a), large aggregationsexist, whereas more and more small aggregations is respectivelyY. Yan et al. / Electrochimica Acta 94 (2013) 277 – 284 279Fig. 2. The size distribution (a – d)and zeta potential (e) of TiO 2 power treated with (a) 0 ml, (b) 0.5 ml, (c) 1 ml, (d) 2 ml HAc.observed in Fig. 3(b – d). Further, the smallest aggregation with thesize of 30 nm is observed when the amount of HAc is 2 ml. Thevariation indicates that the size of aggregations decreases againstthe amount of HAc. Additionally, the TEM results are in agree-ment with that measured by DLS. All these prove that this kindof surface modi?cation could achieve commercially available P25powder with excellent dispersion quality. It is known that the cur-rent produced by the DSSCs is directly linked to the number of dyemolecules adsorbed on the TiO 2 electrode. If large aggregations(Fig. 3 a) existed in TiO 2 ?lm, dye molecules would be preventedfrom adsorbing on the pore walls, which is the main reason whyneeds TiO 2 paste with good dispersion quality.The model of the chemical technique for fabricating TiO 2 viscouspaste is shown in Fig. 4 . The CH 3 COOH is adsorbed on TiO 2 sur-face by the ester bond. According to the theory of DLVO [16] , whenthe powder (CH 3 COO - /TiO 2 ) was dispersed in water, the groups ofCH3 COO - ionize in water and form a negative anion layer on parti-cle, which was called “ Stern layer ”.Additionally, the other positiveions such as protons (H + ) are attracted to counterbalance the nega-tive layer and formed one “ Diffusion layer ”.The positive “ Diffusionlayer ” shifted the Zeta potential to positive, resulting in repellingthe particles from each other. In order to form a negative anion layeron the surface, the powder (CH 3 COO - /TiO 2 ) must be mixed withwater ?rstly. If not, the ionization process of CH3 COO - won’t hap-pen, and the repulsive force won’t exist. Then it is hard to preventaggregation.When the stable TiO 2 colloid had formed, a little hydrochloricacid, which would release numerous ions by ionization, was addedto destroy the positive “ Diffusion layer ” ( Fig. 4 ). Without the elec-trostatic force, the TiO 2 particles would gather together by the Vander Waals force and turned into viscous paste. Due to the abun-dance ethanol existing in the paste, numerous pores were formedin the quick evaporating process. Acetic acid and hydrochloric acidwere volatilized in the sinering process at 500 ?C and pure porouselectrodes were obtained.3.2. Morphology of nanocrystalline porous TiO 2 ?lmsFig. 5 (a) shows the photographs of transparent and opaque elec-trodes. The electrode fabricated from the CH3 COO - /TiO 2 powderwith 2 ml HAc treatment was transparent and the picture wasclearly observed through the ?lm. The opaque one was fabricated280 Y. Yan et al. / Electrochimica Acta 94 (2013) 277 – 284Fig. 3. TEM images of TiO 2 aggregations treated with (a) 0 ml, (b) 0.5 ml, (c) 1 ml (d) 2 ml HAc.from TiO 2 powder without the treatment, which is obviously dif-ferent from the treated one. Fig. 5 (b) shows the transmittancespectra of the two different electrodes. As seen, the substrate ofFTO transmitted more than 75% of incident light above 500 nm; thetransparent one had the transmittance more than 50% of the inci-dent light above 500 nm. The opaque one blocked the incident lightbelow 500 nm, and only let the light slightly pass through abovethat. It was reported that the particles over 100 nm can diffuse thevisible light effectively [17,18] . This is the reason why the electrodewas white and opaque.Fig. 5 (c – h) shows SEM images of the transparent and opaqueelectrodes. As seen, the transparent electrode was dispersed homo-geneously over the large area without aggregations. Additionally,the ?lm showed a compact microstructure with numerous poresabout 20 nm, which supplied passageways for diffusion of elec-trolyte and adsorption of dye molecules. The opaque electrodeshowed a loose microstructure with numerous pores about 40 nmand aggregations. The loose microstructure is bene?cial for thediffusion of electrolyte, but not for the transmission of electronsbetween TiO 2 particles. The sizes of I3 - ion and dye molecule wereonly 1 nm and 1.5 nm, so the pore diameter of 20 nm is large enoughto satisfy the diffusion kinetics of electrolyte, larger pore size wouldonly increase the dark current of DSSCs. Fig. 5 (e and h) shows t