Health & Fitness

A General Method for the Anodic Formation of Crystalline Metal Oxide Nanotube Arrays without the Use of Thermal Annealing

Description
A General Method for the Anodic Formation of Crystalline Metal Oxide Nanotube Arrays without the Use of Thermal Annealing
Published
of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  DOI: 10.1002/adma.200800815 A General Method for the Anodic Formation of Crystalline Metal Oxide Nanotube Arrayswithout the Use of Thermal Annealing** By  Nageh K. Allam, Karthik Shankar,  and  Craig A. Grimes* Valve metal oxides are versatile in their range of applications,which include high- K   dielectrics, [1] gas sensing, [2] biomedicalimplants, [3,4] field emitters, [5] and photovoltaic cells. [6] It is nowgenerally recognized that nanoscale control of metal oxidearchitectures permits significant enhancement of the proper-ties utilized in the above applications. In particular, TiO 2 nanotube arrays formed by anodization [7] have demonstratedoutstanding performance in gas sensing, [8] photocatalytic, [9] and photovoltaic applications. [10–12] Review papers on thesubject are available. [13,14] To date, amorphous nanotubearrays have been synthesized by Ti anodization with anelevated-temperature heat treatment, with temperaturestypically greater than 350 8 C being required to induce crystal-linity. [15] With regard to photoelectrochemical water splittingusing thick-film Ti foil samples, [16–18] annealing at tempera-tures sufficient to induce crystallinity usually leads to theformation of a thick barrier layer, separating the nano-tube-array film from the underlying metal substrate, whererecombination losses can occur. This barrier layer acts tohinderelectrontransfertothemetalelectrode(cathode)wherewater reduction takes place, in turn reducing the overallwater-splitting efficiency. The need for high-temperaturecrystallization limits nanotube array use with temperature-sensitive materials, such as polymers, for applications such asphotocatalytic membranes. Therefore, low-temperature syn-thetic routes, where a high-temperature annealing step forcrystallization is not required, are needed to obtain the fullbenefit of this unique material architecture. Control of thenanotube-array crystallinity is also important for theirapplication in dye-sensitized solar cells and photocatalysis,where charge-carrier transport improves with fewer grainboundaries. Highly crystalline structures offer unique advan-tages over amorphous architectures by providing a direct andrapid pathway for charge transport, thus decreasing thecarrier-path length which in turn reduces recombinationlosses. [19,20] Various methods for the synthesis of crystalline TiO 2 architectures have been reported in the literature. Among thewidely used processing routes to fabricate crystalline TiO 2  arehydrothermal, sol–gel, and calcination processes. However,crystallization by hydrothermal treatment leads to a strongreduction of the textural properties due to excessivecoalescence of the inorganic framework, and structuraldamage results when hydrothermal treatment is performedon mesostructured TiO 2 . [21] With sol–gel synthesis, TiO 2 nanoparticles usually exhibit a high tendency to aggregate. [22] Similarly, with calcination the thus-generated TiO 2  crystals areusually too large to be accommodated within mesopore walls,resulting in structural collapse. [21] Herein we report a facile and novel method to fabricatecrystalline TiO 2  nanotube arrays up to 1.4  m m in length at80–120 8 C, and their use in water photoelectrolysis andliquid–junction dye-sensitized solar cells. A schematic of thistwo-step process is shown in Figure 1. The Ti foil sample is firsttreatedwithanelectrolytecontaininganoxidizingagent(H 2 O 2 at 80 8 C, or (NH 4 ) 2 S 2 O 8  at 120 8 C), then anodized at constantvoltage in a fluoride-containing electrolyte. This process isquite general, hence may be extended to other valve metaloxides. We have confirmed the crystallinity of the as-anodizedarchitectures using XRD and TEM measurements. Interestingly,the two steps are not symmetrical; treating as-synthesizedamorphous TiO 2  nanotube arrays with the oxidizing-agent-containing electrolyte destroys the nanotubes. Our ability toachieve crystalline nanotube arrays at low temperatures issignificant, enabling the TiO 2  nanotube architecture to be usedin combination with flexible polymeric substrates as well asother temperature-sensitive substrates intended for semicon-ductor devices.After the initial treatment of the Ti foil sample in an 80 8 Cperoxide electrolyte, the resulting surfaces were anodized inaqueous solutions containing NH 4 F. We note that both lowerand higher temperatures resulted in thin oxide films. Figure 2shows the glancing-angle x-ray diffraction (GAXRD) results     C    O    M    M    U    N    I    C    A    T    I    O    N [ * ] Prof. C. A. Grimes, Dr. K. ShankarThe Materials Research Institute, The Pennsylvania State UniversityUniversity Park, Pennsylvania 16802 (USA)E-mail: cgrimes@engr.psu.eduProf. C. A. GrimesThe Department of Electrical Engineering,The Pennsylvania State UniversityUniversity Park, Pennsylvania 16802 (USA)Dr. N. K. Allam [+] Physical Chemistry Department, National Research CenterDokki, Cairo 12622 (Egypt)[+] Current Address: The Department of Materials Science andEngineering, The Pennsylvania State University, University Park,Pennsylvania 16802, USA[ ** ] Support of this work by the Department of Energy under grantDE-FG02-06ER15772 is gratefully acknowledged. Nageh K. Allamgratefully acknowledges support under the IFP fellowship providedby the Ford Foundation. We thank Dr. Joe Kulik of the Penn StateMaterials Research Institute for help with the TEM analysis, and thereferees for their helpful comments. 3942    2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim  Adv. Mater.  2008 ,  20,  3942–3946   C  OMM UNI   C AT I   ON for an as-anodized sample, indicating that the nanotube arraysare purely anatase. The inset of Figure 2 shows a field-emissionscanning electron microscopy (FESEM) top-view image of thenanotube-array architecture. Figure 3a shows a transmissionelectron microscopy (TEM) image for a similarly fabricatednanotube-array sample, and Figure 3b is the correspondingdiffraction pattern that indicates a best fit to anatase. Theintensity of a rotationally averaged pattern with expectedBragg lines from anatase is shown in Figure 3c. Since theanodic formation of self-organized valve metal oxide nano-tubes has several common mechanistic aspects, [23] wehypothesize that the two-step process may be readily extendedto the formation of nanoporous/nanotubular metal oxides of other valve metals as well.The 80 8 C peroxide pretreatment of the Ti foil produces acrystalline oxide layer approximately 1.2 m m thick. Subsequentanodization in the fluoride-containing electrolyte initiallyproduces a crystalline nanotube array structure by structuringof the crystalline oxide layer. As the anodization continueswithtime,duetofield-assistedoxidationanddissolutionaswellas chemical dissolution, the Ti/oxide interface gradually movesdeeper into the metallic Ti, and the outer surface dissolves.Therefore, with increasing anodization time (above 3h) thecrystalline oxide layer that is initially present is replaced withan amorphous nanotube-array structure, hence the GAXRD-observed crystallinity of the nanotube arrays decreases forextended anodization durations. This anodization-time restric-tion limits the tube length to about 800nm. One strategy to Figure 1.  Schematic presentation of the two-step fabrication process usedto directly synthesize crystalline TiO 2 -nanotube arrays. Figure 2.  2 8  GAXRD pattern of well-developed nanotube arrays formed byanodizing a peroxide-treated Ti foil sample in an aqueous electrolytecontaining 0.25 M  NH 4 F, 0.1 M  H 3 PO 3 , and 0.05 M  H 2 O 2 . The inset showsan FESEM top-view image of this same sample. Figure 3.  a) TEM image of nanotubes formed by the described technique;b) corresponding selected-area diffraction pattern; and c) intensity fromrotationally-averaged pattern compared with expected Bragg lines foranatase.  Adv. Mater.  2008 ,  20,  3942–3946    2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim  www.advmat.de 3943      C    O    M    M    U    N    I    C    A    T    I    O    N overcome this limitation is the use of strong oxidizing agents toproduce a thicker crystalline oxide layer, which is thenconverted into a nanotube architecture by subsequentanodization. To this end, we immersed Ti foil samples inammonium persulfate ((NH 4 ) 2 S 2 O 8 ), a stronger oxidizingagent, at 120 8 C for 3h. This pretreatment of the Ti foilproduces a crystalline oxide layer approximately 1.6 m m thick;temperatures lower than 120 8 C resulted in thinner films, whilehigher temperatures resulted in an insulator-like film that wewere unable to anodize. Subsequent anodization of the resultingsurfaces in aqueous solutions containing NH 4 F enabledfabrication of nanotube arrays 1.4 m m in length. The GAXRDresultsforanas-oxidizedpersulfatesampleareshowninFigure4, and indicate that the nanotube arrays are purely anatase.The inset of Figure 4 shows an FESEM top-view image of thenanotube-array architecture achieved with the persulfatesample.A preliminary, proof-of-concept photoelectrochemical activ-ity testforwaterphotoelectrolysisusing the as-synthesized TiO 2 nanotube arrays was carried out. Figure 5a shows photo-current–photovoltage characteristics and photoconversionefficiencies for the as-anodized crystalline nanotubular TiO 2 electrodes under 95mW cm  2 UV illumination (320–400nm)in 1  M  KOH. The current–voltage characteristics of anilluminated TiO 2  electrode in contact with a redox electrolytecan be described using the following equation: [23] i  ¼  i  ph    i 0  exp e 0 V kT       1    (1) where  i  is the net current obtained by adding the majority andminority current components,  i 0  is the reverse-bias saturationcurrent,  i ph  is the illumination current, which is proportional tothe photon flux,  k  is Boltzmann’s constant,  T   is the absolutetemperature, and V is the potential. The tested nanotube-arrayelectrodes show n-type behavior, i.e., positive photocurrents atanodic potentials. For this type of semiconductor, the surfaceelectron density ( N  s ) decreases with the applied anodicpotentials ( E  a ) as [24]  N  s  ¼  N  b  exp   e E  a    V   fb kT      (2) where  N  b  is the bulk electron density in the semiconductor,  V  fb is its flat-band potential, and  e  is the elementary charge. Notethat  N  s < N  b  for n-type semiconductors at all potentialspositive of   V  fb .The corresponding light-energy to chemical-energy conver-sion (photoconversion) efficiency  h  was calculated as fol-lows: [25,26] h % ¼  j  p ½ð E  0 rev    E  appl  Þ =  I  0    100 (3) Figure 4.  2 8  GAXRD pattern of well-developed nanotube array sampleformed by anodizing a persulfate-treated Ti foil sample in an aqueouselectrolyte containing 0.25 M  NH 4 F, 0.1 M  H 3 PO 3 , and 0.05 M  H 2 O 2 . Theinset shows an FESEM top-view image of a similar sample. Figure 5.  Data measured from application of anatase nanotubes synthe-sized by described technique, either 800nm long via peroxide or 1.4 m mlong via persulfate: a) current-voltage characteristics and photoconversionefficiency for a water-splitting photoelectrochemical cell, and b) electricalcharacteristics of a N-719-dye-sensitized solar cell under AM 1.5 illumina-tion (through cathode). 3944 www.advmat.de    2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim  Adv. Mater.  2008 ,  20,  3942–3946   C  OMM UNI   C AT I   ON where  j p  is the photocurrent density (mA cm  2 ),  j p E 0 rev  is thetotal power output,  j p j E  appl j  is the electrical power input, and  I  0 is the power density of incident light (mW cm  2 ).  E  0 rev  is thestandard reversible potential, 1.23  V  NHE  (where NHE is thenormal hydrogen electrode), and the applied potential  E  appl  ¼ E  meas   E  aoc ,where E  meas  istheelectrodepotential(versusAg/AgCl) of the working electrode at which photocurrent wasmeasured,and E  aoc istheelectrodepotential(versusAg/AgCl)of the same working electrode at open-circuit conditions undersame illumination and in the same electrolyte. The photo-conversion efficiencies for the as-synthesized nanotube arrays,under 320–400nm illumination, are   3% for the H 2 O 2 -fabricated nanotubes and 4.2% for the longer (NH 4 ) 2 S 2 O 8 -fabricated nanotubes.The  I  – V   characteristics of typical N-719 ( cis -bis(isothiocyanato)bis(2,2 0 -bipyridyl-4,4 0 -dicarboxylato)-ruthenium(II)bis-tetrabutylammonium)-sensitized TiO 2  nanotube-array-electrode solarcells, comprised of either a H 2 O 2  800nm long nanotube-arraysample or a (NH 4 ) 2 S 2 O 8  1.4 m m nanotube-array sample, areshown in Figure 5b. The active area of the devices, typically0.4cm 2 to 0.5cm 2 , was measured using digital Vernier calipersand verified by examination under a calibrated opticalmicroscope. The typical peroxide solar cell showed a short-circuit photocurrent density (  J  sc ) of 1.75mA cm  2 and anopen-circuit potential ( V  oc ) of 730mV for an overall photo-conversion efficiency of 0.46%.The typical persulfate solarcellshowed a  J  sc  of 3.72mA cm  2 and a  V  oc  of 0.752mV for anoverall photoconversion efficiency of 1.31%. Using dye-desorption measurements, the coverage of N-719 dye on thesurface of the as-anodized anatase 800nm long nanotubearrays was determined to be 12.6 nmol cm  2 . The surfacedye coverage of a thermally annealed 6.6 m m long TiO 2 nanotube array formed in aqueous electrolyte was previouslydetermined to be 50 nmol cm  2 , while that of a 10 m m thicknanoparticulate film was reported to be 130 nmol cm  2 . [24] With the perspective provided by these numbers, we note thatthe surface dye coverage of the as-anodized anatase samples isbetter than that expected for a 800nm long nanotube array,indicating that the surface of these nanotubes are nanoscopi-cally rougher and therefore provide more surface sites for dyeadsorption; similar results were found for the persulfatenanotubes.In conclusion, the low-temperature synthesis of crystallineTiO 2 -nanotube arrays of up to 1.4 m m length using a two-stepprocess has been demonstrated. The two-step process consistsof initial treatment of the Ti foil in an oxidizing agent, anelectrolyte containing either H 2 O 2  or (NH 4 ) 2 S 2 O 8 , followed bypotentiostatic anodization of the resulting foil in anNH 4 F-containing electrolyte. The crystallinity of the nano-tube-array films is confirmed using GAXRD and TEMmeasurements. The as-synthesized crystalline nanotube arrayswere successfully tested as anode electrodes for waterphotoelectrolysis, with performances comparable to samplesannealed at high temperatures, and for liquid-junction dye(N 719 dye)-sensitized solar cells. Experimental  Pure titanium foil (0.25mm thick) was purchased from Sigma–Aldrich. Prior to anodization, rectangular samples (2.0cm   2.0cm)were polished using 500 grade silicon carbide paper (to create roughsurfaces) and cleaned with acetone followed rinsing with deionized(D.I.) water. The samples were immersed in a solution similar to thatreported in literature to obtain crystalline TiO 2  films [27–29], with onlya slight modification in the electrolyte composition: either a) 30%H 2 O 2  þ 5m M  Na 2 SO 4  þ 0.5 M  H 3 PO 4  at 80 8 C for 50–72h, or b) 0.05 M (NH 4 ) 2 S 2 O 8  at 120 8 C for 3–5h, resulting in crystalline oxide filmsapproximately 1.2 m m and 1.6 m m thick, respectively. The resultingsamples were then washed with 1 M  HCl for at least two hours at roomtemperature. The anodization was performed using a two-electrodecell with the titanium foil as the working electrode and a platinum foilas the counter electrode, under constant applied voltage at roomtemperature (approximately 22 8 C). Electrolyte NH 4 F concentrationsranging from 0.2–0.4 M  with 0.1 M  H 3 PO 4  and 0.05 M  H 2 O 2  were used.The as-anodized samples were washed with D.I. water and dried usinga nitrogen stream. Sample morphology was examined using FESEM(JEOL JSM-6300) and high-resolution TEM (JEOL 2010F). Sampleswere prepared for TEM by scraping the substrate with a needle.Materialwasallowedtofallontoacoppergridwith alaceycarbonfilm.The crystalline phases were detected and identified by GAXRD on aPhilips X’pert MRD PRO x-ray diffractometer (Almelo, TheNetherlands) as well as by TEM.The photoelectrochemical properties of the nanotube arrays wereinvestigated using a three-electrode configuration with a TiO 2 nanotube sample as a photoanode, saturated Ag/AgCl as a referenceelectrode, and platinum foil as a counter electrode. A 1.0 M  KOHsolution was used as the electrolyte. A scanning potentiostat (CHInstruments, model CHI 600B) was used to measure dark andilluminated currents at a scan rate of 10mV s  1 . A 50W metal-halidelamp(ExfoLite)wasusedasthelightsource,withopticalfiltersusedtorestrict the incident light to UV wavelengths between 320 and 400nm.The incident power was determined as 100 mW cm  2 using athermopile detector (Spectra Physics, CA, USA) after eliminating thelight reflection and absorption effects at the Pyrex glass window.For solar-cell measurements, the as-anodized crystalline nanotu-be-array samples (800nm long via peroxide treatment, 1.4 m m long viapersulfate treatment) were coated with dye by leaving them overnightin a 0.5m M  solution of the N-719 dye (Solaronix, Switzerland). Forsolar-cellfabrication,electrodespacingwasensuredbytheuseof25 m mthick SX-1170 spacer (Solaronix Inc., Switzerland). A liquid-junctionsolar cell was prepared by infiltrating the dye-coated TiO 2  electrodewith commercially available I 3  /I 2  redox electrolyte MPN-100(Solaronix, Switzerland). A conductive glass slide sputter-coated with0.5nmofPtwasusedasthecounterelectrodethroughwhichthedevicewas illuminated. The electrolyte was introduced into the clampedelectrodes by capillary action, and the solar cell was subjected to thebackside illumination geometry. Photocurrent (  I  ) and photovoltage( V  ) of the liquid-junction cell were measured using simulated sunlightat AM-1.5 produced by a 500W Oriel Solar Simulator whose outputwas measured using a National Renewable Energy Laboratorycalibrated standard Silicon solar cell.Received: March 24, 2008Revised: June 17, 2008 [1] J. Robertson,  Eur. Phys. J. Appl. Phys.  2004  , 28 , 265. [2] G.K.Mor,O.K.Varghese,M.Paulose,C. A.Grimes,  Sens.Lett.  2003  ,1 , 42. [3] M. Long, H. J. Rack,  Biomaterials  1998  , 19 , 1621.  Adv. Mater.  2008 ,  20,  3942–3946    2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim  www.advmat.de 3945      C    O    M    M    U    N    I    C    A    T    I    O    N [4] N. P. Huang, R. Michel, J. Voros, M. Textor, R. Hofer, A. Rossi, D. L.Elbert, J. A. Hubbell, N. D. Spencer,  Langmuir   2001  , 17  , 489. [5] P. L. Chen, W. J. Huang, J. K. Chang, C. T. Kuo, F. M. Pan, Electrochem. Solid-State Lett.  2005  , 8 , H83. [6] M. Lira-Cantu, F. C. Krebs,  Sol. Energy Mater. Sol. Cells  2006  , 90 ,2076. [7] D. Gong, C. A. Grimes, O. K. Varghese, W. C. Hu, R. S. Singh, Z.Chen, E. C. Dickey,  J. Mater. Res.  2001  , 16 , 3331. [8] M. Paulose, O. K. Varghese, G. K. Mor, C. A. Grimes, K. G. Ong, Nanotechnology  2006  , 17  , 398. [9] X. Quan, S. G. Yang, X. L. Ruan, H. M. Zhao,  Environ. Sci. Technol. 2005  , 39 , 3770. [10] K. Zhu, N. R. Neale, A. Miedaner, A. J. Frank,  Nano Lett.  2007  , 7  ,69. [11] G. K. Mor, K. Shankar, M. Paulose, O. K. Varghese, C. A. Grimes, Nano Lett.  2006  , 6 , 215. [12] G. K. Mor, K. Shankar, M. Paulose, O. K. Varghese, C. A. Grimes,  Appl. Phys. Lett.  2007  , 91,  152 111.[13] C. A. Grimes,  J. Mater. Chem.  2007  , 17  , 1451. [14] G.K.Mor,O.K.Varghese,M.Paulose,K.Shankar,C.A.Grimes, Sol.Energy Mater. Sol. Cells  2006  , 90 , 2011. [15] O. K. Varghese, D. W. Gong, M. Paulose, C. A. Grimes, E. C. Dickey,  J. Mater. Res.  2003  , 18 , 156. [16] G. K. Mor, K. Shankar, M. Paulose, O. K. Varghese, C. A. Grimes, Nano Lett.  2005  , 5 , 191. [17] O. K. Varghese, M. Paulose, K. Shankar, G. K. Mor, C. A. Grimes,  J.Nanosci. Nanotechnol.  2005  , 5 , 1158. [18] G. K. Mor, H. E. Prakasam, O. K. Varghese, K. Shankar, C. A.Grimes,  Nano Lett.  2007  , 7  , 2356. [19] H. Nishikiori, W. Qian, M. A. El-Sayed, N. Tanaka, T. Fujii,  J. Phys.Chem. C   2007  , 111 , 9008. [20] J. N. Kondo, K. Domen,  Chem. Mater.  2008  , 20 , 835. [21] D.P.Serrano, G. Calleja, R.Sanz, P.Pizarro,  J.Mater.Chem. 2007  ,17  ,1178. [22] Y. Li, T. J. White, S. H. Lim,  J. Solid State Chem.  2004  , 177  , 1372. [23] R. Memming,  Semiconductor Electrochemistry , Wiley-VCH, Wein-heim, Germany  2001 . [24] M. K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphrybaker, E. Muller,P. Liska, N. Vlachopoulos, M. Gratzel,  J. Am. Chem. Soc.  1993  , 115 ,6382. [25] O. K. Varghese, C. A. Grimes,  Sol. Energy Mater. Sol. Cells  2008  , 92 ,374. [26] C. A.Grimes, O.K.Varghese,S.Ranjan,  Light, Water, Hydrogen: TheSolar Production of Hydrogen by Water Photoelectrolysis , Springer,Norwell, MA  2007 . [27] Z. Miao, D. S. Xu, J. H. Ouyang, G. L. Guo, X. S. Zhao, Y. Q. Tang, Nano Lett.  2002  , 2 , 717. [28] X. X. Wang, S. Hayakawa, K. Tsuru, A. Osaka,  J. Biomed. Mater. Res. 2000  , 52 , 171. [29] J. M. Wu,  J. Cryst. Growth  2004  , 269 , 347. 3946 www.advmat.de    2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim  Adv. Mater.  2008 ,  20,  3942–3946
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks