Effect of TiO2 nanotube parameters on field emission properties

Home Search Collections Journals About Contact us My IOPscienceEffect of TiO2 nanotube parameters on field emission propertiesThis article has been downloaded from IOPscience. Please scroll down to see the full text article.2010 Nanotechnology 21 505706(/0957-4484/21/50/505706)View the table of contents for this issue, or go to the journal homepage for moreDownload details:IP Address: 202.115.35.14The article was downloaded on 22/02/2011 at 07:01Please note that terms and conditions apply.IOP P UBLISHING N ANOTECHNOLOGY Nanotechnology21(2010)505706(5pp)doi:10.1088/0957-4484/21/50/505706Effect of TiO2nanotube parameters onfield emission propertiesYahya Alivov1,Michael Klopfer and Sabee MolloiDepartment of Radiological Sciences,University of California,Irvine,CA92697,USAE-mail:yalivov@ and symolloi@Received19July2010,infinal form28October2010Published23November2010Online at /Nano/21/505706AbstractThe dependence offield emission properties of titanium dioxide(TiO2)nanotubes(NTs)hasbeen studied as a function of NT diameter(D)and height(h),which varied in the ranges18–500nm and500–12000nm,respectively.The studies showed a strong dependence of thefield emission on these parameters.With an increase of NT diameter,thefield enhancementfactor increased monotonically from120to3800;the current density also increased untilD=320(current density∼3.8mA cm−2),with subsequent decrease for larger diameters.Thefield emission properties initially improved with NT height until h=5μm,and later remainedunchanged with further increases in h.(Somefigures in this article are in colour only in the electronic version)Nanoscalefield emission electron sources,such as nanotubes, nanorods and nanofibers,have recently attracted a considerable degree of interest due to their applications inflat panel displays and x-ray tubes.Carbon nanotubes(CNTs)and carbon nanofibers(CNFs)have been widely explored due to their small tip radius of curvature,high aspect ratio and mechanical toughness[1–5].Despite the progress made in this area, these kinds of cold cathodes still suffer from fast degradation, which leads to lower longevity.This degradation occurs primarily for two reasons:(1)CNTs are oxidized because residual oxygen is always present in vacuum chambers even at10−7–10−8Pa(an amount sufficient to oxidize CNTs[6]);(2)poor adhesion of CNTs to conductive substrates,which results in enhanced electrical resistivity of the interface layer, thus leading to heating effects[7].Electrochemically grown TiO2NTs seem to be an excellent solution to these problems. First,since TiO2is a natural oxide,its exposure to oxygen does not affect its properties.Second,since TiO2nanotubes are grown on titanium(Ti)sheet,as the latter oxidizes during anodization,a good electrical contact between the TiO2nanotubefilm and conductive Ti sheet is intrinsically guaranteed[8].Furthermore,the electrochemical growth process is very simple and it does not require expensive tools for nanotube growth.The employment of this type of nanotubes should further decrease the cost of cold cathodes. By contrast,TiO2also has a lower work function range 1Author to whom any correspondence should be addressed.(3.9–4.5eV)[9–11]compared to CNTs(5.0eV)[12,13] and a higher degree of NT array uniformity.This ensures a narrower electron kinetic energy distribution and,therefore,a better spatial resolution due to more uniformfield emission conditions.The latter significantly contrasts with CNTs,which grow in different diameters,helicities and orientations on the same growth run[7].Additionally,oxide materials have been proven to be very radiation tolerant[14].There have been few previous reports on thefield emission studies of TiO2nanotubes and nanorods[15,16,11].We have recently demonstrated an x-ray tube employing the TiO2NTfield emitter[17].It is well known that the field emission density and thefield enhancement are critical factors for evaluating the performance offield emitters.They are greatly influenced by the nanotube parameters such as diameter,height,wall thickness and density.It is important to understand the behavior of thefield emission of TiO2 NTs with geometrical parameters to improve thefield emitter performance.There have been no reports yet on experimental studies of the effect of TiO2NT parameters on theirfield emission properties.In this paper,we present thefirst experimental results on the effect of NT diameter and height on thefield emission properties of TiO2NTs grown by electrochemical oxidation.It should be noted that the NT parameters such as diameter,height,and wall thickness can be tuned with high precision by varying the growth conditions[18,19].Figure1.SEM images of the TiO2NTs with different average diameters:(a)20nm,(b)40nm,(c)80nm,(d)170nm,(e)320nm,and (f)550nm.The diameter of TiO2NTs during electrochemical growthlinearly depends on the applied voltage[19].Samples withaverage diameters of18,40,85,170,320and550nm weregrown via electrochemical oxidation of Ti sheets in a glycerol +0.5%NHF4electrolyte by varying the anodization voltage. These NT diameters were estimated by averaging about20neighbouring nanotubes in a sample.The correspondinganodization voltages used to grow these diameter nanotubeswere15,30,60,120and240V[20].Representativescanning electron microscope(SEM)images of the grown NT arrays with different diameters and spacings are shown infigures1(a)–(f).A well-defined and highly aligned tubular structure of arrays can be seen in thisfigure.As-grown amorphous TiO2NT arrays were then annealed at500◦C in air for1h to convert to crystal phase.Annealing in such conditions leads to anatase modification,as shown in the previous reports of x-ray diffraction studies[20].It should be noted that normally there are difficulties in applying high voltages during electrochemical growth because high electric currents through the electrolyte lead to heating effects.The critical value of the electrical current depends on the type of solvent and acid in the electrolyte.For this reason,a ‘multi-stage’growth method was developed to grow samples with550nm NT diameter.In this method,high voltage was employed ramped to240V in50V increments.The time interval between two neighboring voltage values was10min. In this case the forming semiconducting TiO2layer at lower voltages prevents the electrolyte from rapid current increase. It should be noted that in this‘multi-step’growth method there is still a limitation for the maximum applied voltage. Breakdown of the electrolyte current occurs at very high voltages and no NT growth can be initiated.In our experiments the highest applicable voltage changed from sample to sample and ranged from230–270V.We want to note,however,that the reproducibility at such high anodization voltages is not high enough and the samples frequently fail to grow with well-defined tubular structure.The possible reason for this effect may be too strong driving forces leading to deviation from the normal nanotube growth mechanism.To study the dependence of thefield emission on the NT height,the length of the NTsFigure 2.Schematic diagram of the experimental setup for field emission measurements.was varied within the 0.5–12μm range by varying the growth time.The 5×5mm 2size TiO 2NT array samples were bonded to an aluminum backplane with silver paste.The field emission measurements were performed in a vacuum chamber,which was pumped down to the base pressure of 6.6×10−5Pa by an ion pump.The schematic of the experimental configuration for field emission current measurements is shown in figure 2.The spacing d between cathode and anode was achieved by using a 150μm thick glass plate,and this parameter refers to the distance between the top of the nanotubes and the anode.The applied voltage was in the range 0–1000V that corresponds to an electric field range of 0–6.6V μm −1.The maximum voltage applied was limited by measurement setup.As an anode,a copper grid with a 30μm diameter wire and 70%open area was used.The use of a copper grid was motivated by the need to improve gas evacuation from the anode–cathode space,and to reduce the arcing effects.The current measurements were performed by using a Fluke 187multimeter.Figure 3presents current–voltage (I –V )characteristics for all the TiO 2NTs studied,which exhibit exponential bels 1,2,3,4,5,and 6in this figure refer to 320,550,170,85,40,and 18nm diameter nanotube arrays,respectively.The analysis of I –V characteristics of the field emission was done using the simplified Fowler–Nordheim equation [21]:J =A β2E 2φ exp −B φ3/2βE(1)lnJ E 2 =A ln β2φ −−B φ3/2β 1E (2)where A and B are constants with values 1.56×10−6A V −2and 6.83×103V eV −3/2μm −1,respectively;E ,β,and φrefer to the electric field,field enhancement factor,and work function of the TiO 2.The corresponding FN plot is shown in the inset to figure 3,where a nearly linear relationship betweenln (J E 2)and 1E indicates the field emission nature of the TiO 2NT cathode.The FN plot in this figure is shown for the high electric field region of the I –V characteristic,>3V μm −1,Figure 3.(a)Field emission I –V characteristics of TiO 2nanotubes as a function of NT diameter.(b)Corresponding FN plots with linear fits for each NT diameter sample.The NT height in thesemeasurements was kept constant and was equal to ∼2μbels 1,2,3,4,5,and 6in both figures refer to 550nm,320nm,170nm,85nm,40nm,and 18nm diameter nanotube arrays,respectively.Because no reasonable current was observed for the sample with 18nm NT diameter within the studied electric field,the corresponding FN plot is not shown in plot (b).where effective electron field emission starts.Field emission is the extraction of electrons from a solid by tunneling through the triangular shape surface potential barrier when the width of the barrier is comparable to the electron wavelength.This tunneling is possible in strong electric fields,which can be achieved in the top of the nanotubes.The local electric field E is greater than the macroscopic field V /d ,where d is the distance between anode and cathode,by the field amplification factor β.The threshold voltage was estimated as the J =0intercept value of the extrapolation of the high current I –V characteristics performed on the linear scale (figure 3).This threshold voltage changed from sample to sample within the range 2.0–5.0V μm −1.No field emission was observed for the sample with 18nm NT diameter within the studied electric field range (0–6.6V μm −1),which can be explained by too low electric field enhancement due to the large electric field screening effect.Although no strong correlation was observed between the threshold voltage and the NT diameter,there was aFigure4.Field emission current density andfield enhancement factor as a function of TiO2NT diameter.The NT height in these measurements was kept constant and was equal to∼2μm. tendency for the threshold voltage to decrease with increases in NT diameter.Thefield amplification factor can be determined from the slope of the FN plot assuming the work functionφis known.The work function of anatase TiO2was taken to be 4.2eV[10],an intermediate value of the reported ones in the range3.9–4.5eV[9–11].Thefield enhancement factor and the current density for all samples corresponding to an electricfield of6Vμm−1 are summarized infigure4.As can be seen in thisfigure,the field enhancement factor linearly increases with NT diameter from144to3495as the latter increases from40to550nm. Because no reasonable current was observed for the18nm diameter sample within the studied electricfield range,the corresponding FN plot is not shown in the plot.On the other hand,the current densityfirst grows from0to∼3.8mA cm−2, when D increases from18to320nm,but then starts to decrease with further increase in D(figure4).Small diameter nanotubes are relatively dense,which increases the screening effects.The latter,in turn,reduce thefield enhancement factor, which causes a reduction in current density.The opposite is true for larger diameter NT arrays,where a larger open area of NTs leads to reduced screening effects.The induced charges on top of NTs are increased with D,resulting in a largerfield enhancement factor.The tradeoff between these two factors results in a peak position near D=320nm in the current density–diameter dependence.Thefield emission current densities for320nm and550nm are3.2mA cm−2 and4.0mA cm−2,respectively.These results indicate that the larger diameter TiO2nanotubes have more favorablefield emission properties.Similar results were reported for carbon nanotubefield emitters.In a previous report[22],it was shown that‘short and stubby’nanotubes with diameters of200nm showed the best emission characteristics.The dependence of thefield emission density J and the field enhancement factorβon the NT height h was studied for NTs with a diameter of320nm,which corresponds to the highest emission current density(figure4).The results are shown infigure5.As seen from thefigure,the enhancement factorfirst increases with the height h,reaches thelargest Figure5.Field emission current density andfield enhancement factor as a function of TiO2NT height.The NT diameter in these measurements was kept constant and was equal to∼320nm.value of3112at h=5μm,and then remains almost unchanged with further increase of h,resembling‘saturation’. The corresponding current density also changes in the same way growingfirst from2.15to4.53mA cm−2when NT height increases from0.5to4μm,and then becoming independent with further increase of NT height.The initial increase of field enhancement factor and current density with h can be explained by screening of the electricfield.At the beginning as the nanotube height grows,thefield enhancement increases as it is proportional to the nanotube height;however,at some point it becomes insensitive to the NT height because of electricfield screening.This‘saturation’effect offield emission properties possibly occurs because at the bottom of the NT arrays the nanotubes are normally interconnected.Another important factor influencing the efficiency of electronfield emission is the spacing between nanotubes, which determines thefield enhancement factor.Unfortunately, the control of this parameter in electrochemically grown TiO2 NT arrays,similar to many other growth methods,has been a challenge.Recently,some efforts have been made to increase the spacing of TiO2NTs by selection of electrolyte.Grimes et al[23,24]reported that with the use of diethylene glycol as a solvent in the electrolyte,the NT spacing could be increased to some extent[23,24].Although our attempts to reproduce these results have not been successful to date,we believe that by better adjusting the chemistry of the electrolyte, significant progress could be made in this area.In addition, this challenge could be further addressed by technological procedures,such as photo/electron-beam lithography,wet/dry etching,and postgrowth and plasma/thermal treatments.In conclusion,thefield emission properties of TiO2 NTs were studied as a function of NT diameter and height, which were varied in the ranges of20–500nm and500–12000nm,respectively.The studies showed a strong effect of these parameters on thefield emission properties.Thefield enhancement factor increased with NT diameter monotonically from120to3800,while the current density peaked at320nm with the value of∼3.8mA cm−2.Thefield emission properties initially improved with an increase of nanotube diameter andheight,but then‘saturated’.The results are reasonable in light of the screening effects.References[1]Sugie H,Tanemura M,Filip V,Iwata K,Takahashi K andOkuyama F2001Appl.Phys.Lett.782578[2]Yue G Y,Qiu Q,Gao B,Chang Y,Zhang J,Shimoda H,Chang S,Lu J P and Zhou O2002Appl.Phys.Lett.81355[3]Matsumoto T and Mimura H2003Appl.Phys.Lett.821637[4]Senda S,Sakai Y,Mizuta Y,Kita S and Okuyama F2004Appl.Phys.Lett.855679[5]Sakai Y,Haga A,Sugita S,Kita S,Tanaka S-I,Okuyama F andKobayashi N2007Rev.Sci.Instrum.78013305[6]Collins P G,Bradley K,Ishigami M and Zettl A2000Science2871801[7]Bonard J M,Kind H,Stokli T and Nilsson L-O2001SolidState Electron.45893[8]Macak J M,Tsuchiya H and Ghicov A2007Curr.Opin.SolidState Mater.Sci.113[9]Rothenberger G,Fitzmaurice D and Graetzel M1992J.Phys.Chem.965983[10]Tan T T Y,Yip C K,Beydoun D and Amal R2003Chem.Eng.J.95179[11]Liu G,Li F,Wang D-W,Tang D-M,Liu C,Ma X,Lu G andCheng H-M2008Nanotechnology19025606[12]Shiraishi M and Ata M2001Carbon391913[13]Bonard J-M,Maier F,Stockli T,Chatelain A,de Heer W A,Salvetat J-P and Forro L1998Ultramicroscopy737–15 [14]Sickafus K E,Minervini L and Grimes R W2000Science289748[15]Yang Y,Wang X and Sun C2008J.Am.Ceram.Soc.914109[16]Miyauchi M et al2006Appl.Phys.Lett.89043114[17]Alivov Y,Klopfer M and Molloi S2010Appl.Phys.Lett.96243502[18]Mor G K,Varghese O K,Paulose M,Shankar K andGrimes C A2006Sol.Energy Mater.Sol.Cells902011 [19]Grimes C A,Varghese O K and Ranjan S2008Light,Water,Hydrogen:The Solar Generation of Hydrogen by WaterPhotoelectrolysis(New York:Springer)[20]Alivov Y,Fan Z Y and Johnstone D2009J.Appl.Phys.106034314[21]Stratton R1964Phys.Rev.135A794[22]Chhowalla M,Ducati C,Rupesinghe N L,Teo K B K andAmaratunga G A J2001Appl.Phys.Lett.792079 [23]Yoriya S,Mor G K,Sharma S and Grimes C A2008J.Mater.Chem.183332[24]Yoriya S and Grimes C A2010Langmuir26417。