zn ofilmlidoped
TRANSCRIPT
Published: April 29, 2011
r 2011 American Chemical Society 10252 dx.doi.org/10.1021/jp200815d | J. Phys. Chem. C 2011, 115, 10252–10255
ARTICLE
pubs.acs.org/JPCC
Local Electronic Structure of Lithium-Doped ZnO Films Investigatedby X-ray Absorption Near-Edge SpectroscopyShu-Yi Tsai,† Min-Hsiung Hon,† and Yang-Ming Lu*,‡
†Department of Materials Science and Engineering, National Cheng Kung University, Tainan, Taiwan‡Department of Electrical Engineering, National University of Tainan, Tainan, Taiwan
1. INTRODUCTION
Transparent electronics is an advanced technology concerningthe realization of invisible electronic devices. Recently, researchon ZnO thin films has been increasing due to their low cost,nontoxicity, and high stability in hydrogen plasma. ZnO is one ofthe most important semiconductor materials for optoelectronicapplications based on its wide band gap (3.37 eV) and largeexciton binding energy (60 meV). Its considerable applicationsin solar cells,1 sensors,2,3 photocatalytics,4 and optoelectronicdevices5,6 have also triggered wide research interest. However,the fabrication of p-type ZnO, which is an essential step for p�njunction-based devices, is still a bottleneck because of a self-compensation effect from native defects, such as oxygen vacan-cies and zinc interstitials on doping.7�9 p-Type ZnO is achievedby the doping of elements from group I (Li, Na, K) and fromgroup V (N, P, As) dopants. The theoretical studies demon-strated, the group I elements might be better p-type dopants thangroup V elements for introducing shallowness of acceptorlevels.10 Lu et al. proposed that Li can be expected to substituteZn in its site, thus shifting the (002) position to the higher 2θvalues and reducing the c-axis length,11 whereas Wardle et al.suggested that lithium doping may be limited by the formation ofcomplexes, such as LiZn�Lii, LiZn�H, and LiZn�AX.12 Never-theless, there remain a lot of open questions and controversialopinions. A determination of the dominating mechanism of thelocal electronic structure of lithium-doped ZnO and its valencestate is necessary, preferably from experimental results ratherthan a theoretical approach. Away to identify these issues is X-rayabsorption spectroscopy (XAS). XAS is a powerful tool toinvestigate the local arrangement of atoms inmaterials, providingelement-specific information about chemistry, site occupancy,and the neighboring environment.13,14 In this work, we describethe local environment around Zn and its chemical valence state in
ZnO:Li thin films in detail using X-ray absorption spectroscopy(XAS), which allows us to understand the primary mechanism ofthe p-type behavior of ZnO.
2. EXPERIMENTAL DETAILS
We have deposited the Li-doped ZnO thin films on a glasssubstrate (Corning 1737F) at room temperature by radiofrequency (rf) sputtering in a mixture of oxygen and argongases. The target material was zinc metal (99.99% purity). Ar(99.995%) andO2 (99.99%) with a ratio of 10:1 were introducedas the sputtering gases at a total pressure of 1.33 Pa. The contentof Li in the ZnO thin films was adjusted by placing Li2CO3 diskson the target surface. The thickness and diameter of the Li2CO3
disks were controlled to be 0.2 and 1 cm, respectively. TheLi2CO3 disks were made by sintering at high temperature; theywill be dissociated into Li2O and CO2 at decomposition.15,16 Arotating substrate holder was used to obtain uniform composi-tion distributions in the films. After being deposited, the filmswere annealed at 450 �C in Ar ambient for 3 h with heating andcooling rates of 3 and 2 �C/min, respectively. The film thicknesswas measured using a conventional stylus surface roughnessdetector (Alpha-step 200, Tencor, USA). All samples wereanalyzed in the same thickness of about 200 nm. The filmcomposition was determined by a high resolution hyper probe(JXA-8500F Fe-EPMA) equipped with a wavelength-dispersiveX-ray spectrometer (WDS) and by an inductively coupledplasma mass spectrometer (Hewlett-Packard 4500 ICP-MS).The crystalline structure of the films was confirmed by glanc-ing incident angle XRD (GIAXRD) using a Cu KR radiation
Received: January 26, 2011Revised: April 18, 2011
ABSTRACT: Lithium-doped ZnO films were deposited by radio frequency magnetronsputtering on Corning 1737 glass substrates. The Li content in the films varied from 0 to 10at. %, as determined by wavelength-dispersive X-ray analysis and inductively coupled plasmamass spectrometry. The effect of Li content on the microstructure and electrical properties wasstudied. The XRD results indicated that all the samples have a ZnO wurtzite structure, and nosecondary phase formed as the Li atoms were incorporated into ZnO thin films. The Hall andelectrical resistance measurements revealed that the resistivity is decreased by Li doping. TheEXAFS measurement showed that the bonding length of both Zn�O and Zn�Zn wasdecreased after converting to p-type conduction due to incorporation of lithium atoms. Allthe results confirmed that the Li ions were well incorporated into the ZnO lattices as a result ofsubstituting Zn sites without changing the wurtzite structure, and no secondary phase appeared in the Li-doped ZnO thin film.
10253 dx.doi.org/10.1021/jp200815d |J. Phys. Chem. C 2011, 115, 10252–10255
The Journal of Physical Chemistry C ARTICLE
(λ = 0.15406 nm). The Zn K-edge (9659 eV) XAS spectra wererecorded on a wiggler C (BL-17C) beamline at the NationalSynchrotron Radiation Research Center (NSRRC) of Taiwan.The XAS data analyses were performed using standard methodsand WinXAS software. The fittings of the EXAFS were per-formed using least-squared fittings from outputs from FEFF8.0software. General EXAFS data analysis has been described inthe literature.17�19 The parameters calculated from the fittingswere the interatomic distances, coordination numbers, and theDebye�Waller factors. The resistivity and carrier concentrationsof the ZnO:Li thin film at room temperature were measured by aHall-effect measurement system (Lake Shore, model 7662) usingthe van der Pauw method.
3. RESULTS AND DISCUSSION
Figure 1 shows the XRD pattern of ZnO thin films withdifferent Li-doping contents on glass substrates prepared by an rfmagnetron sputtering method. The compositions of the dopedZnO films were determined by both WDS and ICP-MS. TheO/Zn atomic ratios were obtained from WDS, and the Li/Znatomic ratios were measured by ICP-MS. The Li content in thefilms increased with an increasing number of Li2CO3 disksmounted on the Zn target surface. The maximum Li contentobtained in this study was approximately 10 at. %. A similarcontent has been reported by Wang.20�22The solubility of Li insingle-crystal ZnO is very high, with up to around 30% of the Znsites being occupied by Li has been reported by Onodera et al.23
Only one peak corresponding to a (002) plane was observed forall the samples, and no diffraction peaks reflected from other
crystalline phases was seen, suggesting good crystallinity with ahigh preferential c-axis orientation and formation of LiZn in thefilms.With the Li-doped content increasing, the full width at half-maximum (fwhm) became weak and broad, and the diffractionangle shifted toward the high angle direction, as shown inFigure 1b. It is known generally that dopants can be substitutedor inserted, depending on the doping ions' size. Yamamoto24 andOnodera22 reported that most doping ions substituted for Zn ionsites in the doping case due to a decrease in theMadelung energy.If Liþ ions interstitial to Zn2þ ions, the lattice parameter of theZnO crystal increases and the (0 0 2) peak should shift to lowangle. In addition, Li at a substitutional site creates an energylevel at 0.09 eV. However, Li at an interstitial site creates anenergy level at 1.58 eV, and it is more stable, according to Parket al. According to their result, the XRD peak shifts toward highangle, which implies that the highly incorporated Liþ ions exist inthe substitutional sites, not in the interstitial sites.
The electrical resistivity values of ZnO:Li films with differentLi dopant contents can be seen in Figure 2. The Hall coefficientand hot probe measurements method were employed to identifythe type of conduction in these films. The p-type conductivitybehavior could be achieved only in the Li content from 1 to 5 at. %.As is well-known, in Li-doped ZnO specimens, Li doping mainlyoccurs as follows25
Li2O sfZnO
Li0Zn þ Li•i þOo
where LiZn represents lithium on the zinc lattice site, Lii lithiumin an interstitial position, and Oo oxygen on the lattice site ofitself. Significantly, LiZn is theoretically predicted to have ashallow acceptor level.10 For the 1 at. % Li-doped ZnO films,weak p-type conduction was found to have high resistivityand low carrier concentration due to the fact that holes maybe compensated for by n-type native defects. For the 3 at. %Li-doped ZnO film, more Li atoms substituted for Zn, whichacted as an effective acceptor, thus achieving optimized p-typeconduction. By doping a I group impurity into the II�VIsemiconductor of ZnO, the impurity became the acceptor, andthe electrons decreased, thus transforming the film from ann-type to a p-type conductive behavior. In a certain amount ofdoping, the electronic holes increased with doping Li concentra-tions. The optimized doping amount obtained in this study is at3 at. % Li-doped ZnO thin films with 0.11 Ω 3 cm in electrical
Figure 1. (a) XRD diffraction patterns of undoped ZnO and ZnO:Lithin films with different Li contents. (b) Positions of the (002) peak andfull width at half-maxima (fwhm) of ZnO:Li thin films.
Figure 2. Resistivity (σ), Hall mobility (μ), and carrier concentration(n) as functions of Li content for ZnO:Li thin films deposited on a glasssubstrate.
10254 dx.doi.org/10.1021/jp200815d |J. Phys. Chem. C 2011, 115, 10252–10255
The Journal of Physical Chemistry C ARTICLE
resistivity, 0.22 cm2/V 3 s in Hall mobility, and 3.13� 1018 cm�3
in concentration. The conversion of the conducting type fromp-type to n-type at a higher doping level (5 at. %), which may beattributed to the formation of the defects (Lii or LiZn�Lii) actingas donors. Theymay act as a compensative and scattering centersthat reduce the hole concentration and result in further deteriora-ting of hole mobility and depress the p-type behavior of ZnO.Wardle et al. suggested that excess lithiummay occupy interstitialsites as well and lead to the formation of electrically inactiveLiZn�Lii pairs.
12
ZnO:Li thin films have previously been partly characterized byX-ray powder diffraction and transmission electron microscopy.26
Whereas X-ray diffraction yields information on long-range
structural aspects, X-ray spectroscopy (XAS) provides comple-mentary details on the electronic environments of the metals andon the short-range structure. The XAS, including X-ray absorp-tion near-edge structure (XANES) and extended X-ray absorp-tion fine structure (EXAFS), is a nonintrusive techniqueintended to investigate the molecular environment around atarget element in various matrices of different states. Forexample, XANES can be used to determine the oxidation stateof an absorbing element by measuring the energy shift of theabsorption edge. With higher oxidation states, the absorptionedge shifts to higher energy by a few electronvolts. Furthermore,the shape of the XANES profile often reflects the geometry of thefirst coordination sphere of many transition elements withunfilled d orbitals and can be used to qualitatively assess thecoordination environment of the absorbing atom. Figure 3illustrates the normalized Zn K-edge spectra of undoped and 3at. % Li-doped ZnO films. The result shows a sharp increase inabsorption edge energy of 9664 eV, caused by excitation of Zn 1selectrons.27 The XANES in Figure 3 for both samples arevirtually identical, indicating that the Zn is predominantlypresent in a formal 2þ oxidation state in tetrahedral coordina-tion. As the amount of doped Li increased, the edge energycorresponding to the Zn2þ oxidation state has a small structuraldistortion. The enlarged near-edge spectra are shown in the insetof Figure 3. Because the intensity is approximately proportionalto the density of the unoccupied Zn 3d-derived states, the resultsindicate that increases in the absorption intensity will decreasethe number of 3d electrons in Zn.
For the purpose of studying in more detail the local structureof the ZnO host lattice upon Li incorporation, we performedextended X-ray absorption fine structure (EXAFS) measure-ments at the Zn K edges. The Zn K-edge EXAFS spectrum wasquantitatively simulated using the FEFF 8.0 program.19 Both theexperimental results and the fitting curve are displayed in R-spaceand are provided in Figure 4. In the simulation, Liþ is assumed tosubstitute for the Zn2þ site in the ZnO lattice. The first shell ofthe radial distribution function indicates the position of theZn�O bonding distance, and the second shell peak denotes acombination of Zn�Zn bonding distances. From the results, thefitting curve was shown to be in good agreement with theexperimental results, which provided evidence that Li occupiedZn sites in the ZnO lattice without forming impurity phases. Inthe case of the Li-doped ZnO, the intensity of the second peakdecreased, revealing degradation in the crystal structure. Thisresult was also consistent with the XRD measurement.
To obtain quantitative structural information, the best-fitvalues for the Zn K edge are listed in Table 1. From the results,it can be seen that the undoped ZnO thin films exists at the samelocal structure as the wurtzite ZnO, in which Zn atoms aresurrounded by four O atoms in the first-coordination shell. Thefirst shell Zn�O coordination number NZn�O was 4.018 Å, andthe bond length RZn�O was 1.971 Å. As we know, the bond
Figure 3. Normalized Zn K-edge XANES spectra of undoped ZnO andZnO:Li samples. The inset shows enlargements of the peaks associatedwith the 1s-to-3d transitions.
Figure 4. Fourier transform magnitude of Zn K-edge EXAFS ofundoped ZnO and 3 at. % Li-doped ZnO films.
Table 1. Structural Parameters of ZnO:Li from EXAFS Analyses, where R is the Interatomic Distance, N is the CoordinationNumber, and σ2 is the Debye�Waller Factor
sample interaction type interatomic distance (R) coordination number (N) Debye�Waller (σ2)
undoped ZnO thin film Zn�O 1.971 4.018 0.002
Zn�Zn 3.270 12.09 0.001
3 at. % Li-doped ZnO films Zn�O 1.969 4.013 0.004
Zn�Zn 3.211 11.89 0.006
10255 dx.doi.org/10.1021/jp200815d |J. Phys. Chem. C 2011, 115, 10252–10255
The Journal of Physical Chemistry C ARTICLE
length of RLi�O in ZnO is 1.661 Å, and the RZn�Li bond lengthis 2.703 Å,28 which is much different than the bond length ofRZn�O, excluding the possibility that the interstitial mechanismwas executed. For the 3 at. % Li-doped ZnO films, only thecoordination number and interatomic distance of the secondshell decreased, whereas that of the first shell was similar to theundoped ZnO thin films. The second shell interatomic distanceRZn�Zn of 3 at. % Li-doped ZnO decreased from 3.270 to 3.211Å, which implied a decreased lattice parameter for the Li-dopedZnO films. These results indicate that the substitution of Liatoms for parts of Zn atoms in the ZnO lattice leads to a decreasein the nearest-neighbor bond length between the Zn and Znatoms of Li-doped ZnO films. For the second coordination shell,it can be seen that the Debye�Waller factor (σ2) is larger for the3 at. % Li-doped ZnO as compared with the undoped one. Thisresult is also consistent with the XRDmeasurement because localstructure distortions may occur as a consequence of a latticemismatch induced at a higher doping amount.
4. CONCLUSIONS
The XRD results indicated that all of the ZnO:Li films had(002) preferred orientations with hexagonal wurtzite structures.The electrical measurements by hot probe and the Hall coeffi-cient showed that the lithium-doped ZnO thin films had a p-typeconductive behavior. The optimized doping amount obtained inthis study is at 3 at. % Li-doped ZnO thin films with 0.11Ω 3 cm inelectrical resistivity, 0.22 cm2/V 3 s in Hall mobility, and 3.13 �1018 cm�3 in carrier concentration. The XANES and EXAFSanalyses indicate that the Li substituted for Zn2þ withoutchanging the crystalline structure of ZnO. From the EXAFSresults, it indicates that decreases in the second shell of theZn�Zn coordination number are caused by incorporation oflithium in the substitutional sites rather than in the interstitialsites in p-type ZnO sputtering film.
’AUTHOR INFORMATION
Corresponding Author*Telephone:þ886-6-2606123, ext. 7771. Fax:þ886-6-2602305.E-mail: [email protected] and [email protected].
’ACKNOWLEDGMENT
The authors are grateful to the National Science Council inTaiwan for financially supporting this research under 99-2221-E-024-003 and 98-2221-E-006-075-MY3.
’REFERENCES
(1) Chao, H. Y.; Cheng, J. H.; Lu, J. Y.; Chang, Y. H.; Cheng, C. L.;Chen, Y. F. Superlattices Microstruct. 2010, 47, 160.(2) Lim, M. A.; Lee, Y. W.; Han, S. W.; Park, I. Nanotechnology
2011, 22, 035601.(3) Zheng, K. B.; Gu, L. L.; Sun, D. L.; Mo, X. L.; Chen, G. R.Mater.
Sci. Eng., B 2010, 166, 104.(4) Guo, M. Y.; Fung, M. K.; Fang, F.; Chen, X. Y.; Ng, A. M. C.;
Djurisic, A. B.; Chan, W. K. J. Alloys Compd. 2011, 509, 1328.(5) Chung, J.; Lee, J.; Lim, S. Physica B 2010, 405, 2593.(6) Hongsith, N.; Choopun, S. Chiang Mai J. Sci. 2010, 37, 48.(7) Ozgur, U.; Alivov, Y. I.; Liu, C.; Teke, A.; Reshchikov, M. A.;
Dogan, S.; Avrutin, V.; Cho, S. J.; Morkoc, H. J. Appl. Phys. 2005,98, 041301.
(8) Saw, K. G.; Ibrahim, K.; Lim, Y. T.; Chai, M. K. Thin Solid Films2007, 515, 2879.
(9) Zhang, S. B.; Wei, S. H.; Zunger, A. Phys. Rev. B 2001,63, 075205.
(10) Park, C. H.; Zhang, S. B.; Wei, S. H. Phys. Rev. B 2002,66, 073202.
(11) Lu, J. G.; Zhang, Y. Z.; Ye, Z. Z.; Zeng, Y. J.; He, H. P.; Zhu,L. P.; Huang, J. Y.; Wang, L.; Yuan, J.; Zhao, B. H.; Li, X. H. Appl. Phys.Lett. 2006, 89, 112113.
(12) Wardle, M. G.; Goss, J. P.; Briddon, P. R. Phys. Rev. B 2005,71, 155205.
(13) Liu, X. J.; Song, C.; Zeng, F.; Pan, F. J. Phys.: Condens. Matter2007, 19, 296208.
(14) Norton, D. P.; Pearton, S. J.; Hebard, A. F.; Theodoropoulou,N.; Boatner, L. A.; Wilson, R. G. Appl. Phys. Lett. 2003, 82, 239.
(15) Ktalkherman, M. G.; Emelkin, V. A.; Pozdnyakov, B. A. Theor.Found. Chem. Eng. 2009, 43, 88.
(16) Timoshevskii, A. N.; Ktalkherman, M. G.; Emel’kin, V. A.;Pozdnyakov, B. A.; Zamyatin, A. P. High Temp. 2008, 46, 414.
(17) Koningsberger, D. C.; Prins, R. X-ray Absorption: Principles,Applications, Techniques of EXAFS, SEXAFS, and XANES; Wiley:New York, 1988.
(18) Dimitrov, D. A.; Ankudinov, A. L.; Bishop, A. R.; Conradson,S. D. Phys. Rev. B 1998, 58, 14227.
(19) Ankudinov, A. L.; Ravel, B.; Rehr, J. J.; Conradson, S. D. Phys.Rev. B 1998, 58, 7565.
(20) Wang, D. Y.; Zhou, J.; Liu, G. Z. J. Alloys Compd. 2009, 481, 802.(21) Mohamed, G. A.;Mohamed, E.; Abu El-Fadl, A. Physica B 2001,
308, 949.(22) Onodera, A.; Yoshio, K.; Satoh, H.; Yamashita, H.; Sakagami,
N. Jpn. J. Appl. Phys., Part 1 1998, 37, 5315.(23) Onodera, A.; Tamaki, N.; Kawamura, Y.; Sawada, T.; Yamashita,
H. Jpn. J. Appl. Phys., Part 1 1996, 35, 5160.(24) Yamamoto, T.; Katayama-Yoshida, H. J. Cryst. Growth 2000,
214, 552.(25) Bonasewicz, P.; Hirschwald, W.; Neumann, G. J. Electrochem.
Soc. 1986, 133, 2270.(26) Wang, B.; Tang, L. D.; Qi, J. A.; Du, H. L.; Zhang, Z. B. J. Alloys
Compd. 2010, 503, 436.(27) Kelly, R. A.; Andrews, J. C.; DeWitt, J. G. Microchem. J. 2002,
71, 231.(28) Fu, Z. W.; Zhang, L. N.; Qin, Q. Z.; Zhang, Y. H.; Zeng, X. K.;
Cheng, H.; Huang, R. B.; Zheng, L. S. J. Phys. Chem. A 2000, 104, 2980.