and charge transport properties of epitaxial Ba0.6La0.4TiO3−δ thin films
Cite as: Appl. Phys. Lett. 114, 202902 (2019); https://doi.org/10.1063/1.5093749 Submitted: 25 February 2019 . Accepted: 05 May 2019 . Published Online: 22 May 2019
Qiang Li, Aihua Zhang, Dong Gao, Min Guo, Jiajun Feng, Min Zeng , Zhen Fan , Deyang Chen, Xingsen Gao, Guofu Zhou, Xubing Lu, and J.-M. Liu
Oxygen vacancy mediated conductivity and charge transport properties of epitaxial
Ba 0.6 La 0.4 TiO 3d thin films
Cite as: Appl. Phys. Lett.114, 202902 (2019);doi: 10.1063/1.5093749 Submitted: 25 February 2019
.
Accepted: 5 May 2019.
Published Online: 22 May 2019
QiangLi,1AihuaZhang,1DongGao,1MinGuo,1JiajunFeng,2MinZeng,1 ZhenFan,1 DeyangChen,1 XingsenGao,1GuofuZhou,3XubingLu,1,a)and J.-M.Liu1,2
AFFILIATIONS
1Guangdong Provincial Key Laboratory of Quantum Engineering and Quantum Materials and Institute for Advanced Materials, South China Academy of Advanced Optoelectronics, South China Normal University, Guangzhou 510006, China
2Laboratory of Solid State Microstructures and Innovation Center of Advanced Microstructures, Nanjing University, Nanjing 21009, China
3Guangdong Provincial Key Laboratory of Optical Information Materials, South China Academy of Advanced Optoelectronics and National Center for International Research on Green Optoelectronics, South China Normal University, Guangzhou 510006, China
a)Author to whom correspondence should be addressed:[email protected]
ABSTRACT
We report on the effects of the oxygen vacancy (VO) regarding the microstructure, conductivity, and charge transport mechanisms of epitaxial Ba0.6La0.4TiO3d(BLTO) films. The VOconcentration can be largely regulated from 21.5% to 37.8% by varying the oxygen pressure (PO2) during film deposition. Resistivity-temperature and Hall effect measurements demonstrate that the BLTO films can be tuned remark- ably from an insulator to a semiconductor, and even to a metallic conductor by regulating the VOconcentration. The role of VOconcentra- tion in the charge transport mechanism is clarified. For films with low VOconcentration, the charge transport is dominated by variable range hopping (VRH) at low temperatures, and it shows small polaron (SP) hopping at high temperatures. For films with high VOconcentration, the carrier transport remains VRH at low temperatures, while it changes to SP hopping at moderate temperatures, and is dominated by thermal phonon scattering at high temperatures. Furthermore, the lower starting temperature of SP hopping for films with higher VOcon- centrations indicates that VOfavors electron-phonon coupling. Different charge transport mechanisms are assumed to be due to different VO-induced defect energy levels in the BLTO films, which has been verified by their soft x-ray absorption spectroscopy results.
Published under license by AIP Publishing.https://doi.org/10.1063/1.5093749
Barium titanate (BaTiO3, BTO) is an important ferroelectric and dielectric material that has broad applications in ferroelectric memory devices, multilayer ceramic capacitors, piezoelectric devices, etc.1–3 Stoichiometric BTO is an insulator with a high resistivity of about 1010Xcm and a large bandgap of3.2 eV.4Conductive BTO can be achieved by doping pentavalent ions at the B-site or trivalent rare- earth ions at the A-site.5–9In the case of B-site doping, one of the widely studied materials systems is BaTi1xNbxO3(BNTO). Nagono et al.observed a critical doping concentration in BNTO films grown using metal-organic chemical-vapor deposition, at which the film obtained a minimum resistivity value.5However, Shaoet al.reported that BNTO films grown using pulsed laser deposition (PLD) did not exhibit a critical doping concentration. They found that the conductiv- ity of the films increased with the doping concentration when the
latter varied from 0.2 at. % to 100 at. %.6In a recent work, we obtained similar results that Nb-doping can tune the conventional insulating BTO films from an insulating to a highly semiconductive or even to a metallic state.7
In the case of A-site doping of BTO films, much less work has been done compared with B-site doping. The trivalent rare-earth ions of Y, Nd, and La have been reported to regulate the conductivity and charge transport behaviors of RxBa1xTiO3 (R¼Y, Nd, and La) films.8,9Takahashiet al.reported that the resistivity of La-doped BTO semiconducting films tend to decrease with the increase of the temper- ature due to the effective thermal excitation of the small polarons (SPs).9
The oxygen vacancy (VO) has been observed to play significant roles in the structural, electrical, and optical properties of perovskite
oxides.10–12It has also been considered as one of the important strate- gies to regulate the conductivity of BTO films.6,13The VOs belong to n-type defects that can bring additional electrons to the oxide films and improve their conductivity.11,13Compared with the widely investi- gated cation-doping effect on the conductive BTO films, the VOeffect on the microstructure, conductivity, and charge transport characteris- tics of BTO films has not yet been systematically investigated. In par- ticular, the understanding of the charge transport mechanisms and electronic structures in oxygen-deficient BaTiO3d films remains poorly developed and unclear.
In the present work, by controlling the oxygen pressure during the PLD process, we deposited A-site doped Ba0.6La0.4TiO3d(BLTO) films with different VO concentrations. We studied on their micro- structure, conductivity, and charge transport mechanisms systemati- cally, and found that BLTO films with varying VO concentrations exhibit distinctly different microstructures with altered conduction behaviors. In particular, we clarified the charge transport mechanisms in and related electronic structures of the BLTO films with various VO concentrations.
We fabricated BLTO films on (001) MgO single crystal substrates at 650C using the PLD technique with four oxygen pressures varying from 110 to 5104Pa. During the deposition, a KrF excimer laser with a wavelength of 248 nm was operated at 2 Hz, and the laser fluence was fixed at 1.7 J cm2. The thickness of each film was deter- mined by X-ray reflectivity and was in the range of 51–57 nm. The thickness results are shown in Fig. S1 of thesupplementary material.
The crystal structures of the films were studied using X-ray diffraction (XRD) with a PANalytical X’Pert Pro diffractometer, including 2h scans and reciprocal space mapping (RSM). The VO concentration and valence states of elements in the films were analyzed by X-ray photoelectron spectroscopy (XPS, Thermo Fisher Scientific). The tem- perature dependence of the electrical resistivity and the Hall effect of the thin films were measured by the van der Pauw method using a physical property measurement system (PPMS9, Quantum Design).
The O K-edge soft X-ray absorption spectra (SXAS) were collected to investigate the electronic structures of the BLTO films.
Figure 1(a)shows the XRD 2h-scan data for the BLTO films fab- ricated at different oxygen pressures (PO2) of 110, 1101, 1102, and 5104Pa. Only (00l) diffraction peaks of the films are observed and no other phases or randomly oriented peaks appear in the scanning curves. It can be concluded that all the films are grown epitaxially on the MgO (001) substrates. To show more clearly the influence of oxygen vacancy on the crystal structure, a zoom-in view of the (002) peaks for all the films is shown in the right part ofFig.
1(a). For the film grown in high vacuum (PO2¼5104Pa), a clear low angular shift of the (002) peak is found, indicating an increase in the out-of-plane lattice constant. This increase is thought to be due to the larger VOconcentration, which has also been observed in other perovskite oxides.14
Figure 1(b)shows the typical RSM results of the film with a PO2
of 5104Pa, in which the reflection spots from BLTO (113) and MgO (113) are observed. According to the RSM results shown inFig.
1(b), the center H value of the film deviates from that of the MgO sub- strate, indicating that the strain in the film has been relaxed. The lattice constants calculated from the measured RSM results are shown inFig.
1(c), and demonstrate that all the films have a tetragonal structure.
Furthermore, the lattice constants of the film grown under high
vacuum are significantly different from those of the other samples, which is assumed to be due to the higher VO concentration in the BLTO film grown at 5104Pa.
Figures 2(a)–2(d)show the results of the O 1s XPS spectra of the BLTO films fabricated at four PO2values. The spectra are carefully decomposed into three peak shapes using Gaussian fits, which reflect different types of O species in the samples.15–17 The low binding energy peak around 529.3 eV corresponds to the oxygen in the BLTO lattice (lattice O). The peak at the intermediate binding energy around 531.0 eV is attributed to oxygen vacancies (nonlattice O). The higher binding energy peak around 532.4 eV is correlated to the absorbed oxygen on the surface, such as hydroxyl groups and CO2(adsorbed O).
FIG. 1.(a) 2h scan patterns of the BLTO films with PO2 from 110 to 5104Pa; (b) the reciprocal space map of the BLTO films with a PO2 of 5104Pa; and (c) the lattice parameters of the BLTO films with PO2of 110 to 5104Pa.
FIG. 2.The O 1s XPS spectra of the BLTO films with (a) PO2of 110Pa, (b) PO2of 1101Pa, (c) PO2of 1102Pa, and (d) PO2of 5104Pa. (e) The relative oxygen vacancy concentration of the BLTO films. (f) Ti 3d XPS spectra of the BLTO films.
Our fitting results on O 1s XPS are well consistent with those of the O 1s XPS spectra of reported BTO films.17
The relative VOconcentration is calculated from the integrated area of these peaks, and the results are shown inFig. 2(e). It can be found that the nonlattice oxygen ratio (VOconcentration) increases from 21.5% to 37.8% as the oxygen pressure decreases from 110to 5104Pa. Figure 2(f)shows the Ti 3d XPS spectra of the BLTO films grown at different PO2. The binding energies of Ti (2p 1/2) and Ti (2p 3/2) are located at 463.7 eV and 457.9 eV, respectively, and nei- ther peak shows further splitting, indicating that only the Ti4þstate exists in the sample.18As a result, the conductive behavior and charge transport mechanisms of the BLTO films are dominated by variations in the VOconcentration.
Figure 3(a) shows the temperature dependent resistivity of the BLTO films with PO2varying from 110to 5104Pa. Clearly, the film resistivity differs considerably when the oxygen pressure is varied during growth, which decreases with decreasing oxygen pressure. The film fabricated at a PO2of 110Pa is an insulator and exhibits the highest resistivity over the entire temperature measurement range. It should be noted that the resistivity around 250 K shows an abrupt change, which is attributed to theT-O structural phase change of BaTiO3dfilms.19,20 For films with PO2 values of 1101 and 1102Pa, the resistivity decreases with the increase of temperature over the entire measured temperature range, demonstrating typical semiconductive behavior.
When the PO2is reduced to 5104Pa, the resistivity decreases further and shows a more complicated conductive behavior. As shown inFig. 3(b), a semiconductor-metal transition is observed at 330 K.
Figure 3(c)shows the temperature-dependent carrier concentration of the films with PO2values of 1102and 5104Pa (the results of the other two samples are unavailable due to their high resistivity).
The film deposited under high vacuum (PO2¼5104Pa) has a much higher carrier concentration than the film grown at 1102Pa, which is due to the higher VOconcentration. Oxygen vacancies have been considered to be n-type defects that can each introduce two elec- trons into the film.11,13
The electron concentrations of the two presented samples increase with the temperature until finally saturating at high tempera- tures. It is assumed that the doped electrons are likely localized at low temperatures. As the temperature increases, the localized electrons are gradually released, resulting in an increase in the electron concentra- tion at higher temperatures. This temperature effect for the carrier concentration is seen more clearly in the 5104Pa film, since it has a much higher VOconcentration than that of the 1102Pa film.
Figure 3(d)shows the corresponding mobility values of the two films. The mobility remains nearly constant for the film at a PO2of 1102Pa over the entire measurement temperature range. For the film with a PO2of 5104Pa, the carrier mobility increases slightly at low temperatures until eventually it gets saturated. The film with a higher carrier concentration clearly exhibits greater carrier mobility, which implies that Coulomb scattering may not be the dominant scat- tering mechanism during the carrier transport. A further discussion about the carrier transport will be given next.
To understand the transport mechanism of the BLTO films with PO2< 110Pa, a detailed analysis of theq-Tcurves from low to high temperatures is performed. In particular, the q-Tcurve of the film with a PO2of 5104Pa is divided into metallic and semicon- ducting regimes at the 330 K transition, while the corresponding charge transport mechanism is discussed separately. The fitting results obtained from various transport models are shown inFigs. 3(e)–3(g).
Various transport models and their corresponding temperature ranges are summarized inTable I.
For the films that exhibit semiconducting behavior, the transport model changes from variable range hopping (VRH) to SP hopping when the measurement temperature increases, as shown inFigs. 3(e) and3(f). For the metallic behavior shown in the 5104Pa sample, theq-Tcurve can be fitted well by a thermal phonon scattering model, as indicated byFig. 3(g). In BLTO films, the electron-phonon coupling plays an important role in the transport mechanism.9When the tem- perature is low, the electron-phonon coupling is weak due to the few activated phonons and the high energy barriers in adjacent locations.
In this case, it is difficult for phonons to facilitate electron hopping between the nearest neighbors. However, the electrons can still absorb energy to hop to a remote location with lower potential barriers.
Therefore, the resistivity can be characterized using the VRH model:
ln(q)/T1/4.21
As the temperature increases, the interaction between electrons and phonons becomes enhanced. The local lattice polarization would be induced by the electron movement in crystal lattices. The SP is composed of itinerant electrons and the surrounding polarization field, whose size is ranging from one to several lattices. Consequently, SP hopping can take place between the nearest neighbors. The resistivity FIG. 3.(a) Temperature dependent resistivity of the BLTO films with PO2ranging
from 1100to 5104Pa; (b) temperature dependent resistivity of the BLTO film with a PO2of 5104Pa (300–400 K); temperature dependence of the (c) carrier density and (d) Hall carrier mobility at PO2values of 1102and 5104Pa; (e) fitting results using the VRH model for the BLTO films with PO2<110Pa; (f) fitting results using the SP hopping model for BLTO films with PO2<110Pa;
and (g) fitting results using the thermal phonon scattering mode for the BLTO film with a PO2of 5104Pa.
due to the SP conforms to the model: ln(q)/T3/2exp(WH/KBT).22 Table Ishows that the SP hopping begins at 285, 250, and 190 K for the films with PO2 values of 1101, 1102, and 5104Pa, respectively. The lower starting temperature for the SP hopping in the films with higher VOconcentrations indicates that the oxygen vacan- cies favor electron-phonon coupling. It is assumed that the electrons provided by the oxygen vacancies can remain at a higher energy level than the Fermi level. In addition, the more the oxygen vacancies in the films, the closer the energy level of the electrons from the oxygen vacancies to the conduction band.
The corresponding activation energiesWHcalculated from the respective fitting curves are 0.087, 0.047, and 0.037 eV for the films with PO2values of 1101, 1102, and 5104Pa, respectively.
These values further support the idea that the generated electrons are more likely to be excited to the conduction band at higher VOconcen- trations. In addition, Austin and Mott pointed out that the polaron binding energyWp would be approximately twice the activation energy WH (Wp 2WH).23 According to the results shown in Table I, the film prepared under high vacuum should also have the smallestWpvalue. Therefore, its SPs are most easily dissociated.
When the temperature is increased to 350 K, the SPs in the film with a PO2of 5104Pa are completely dissociated. The conduc- tion behavior of the dissociated electrons is dominated by thermal phonon scattering, and the associated transport mechanism con- forms to the model:q/T3/2.22
Figure 4shows the O K-edge SXAS spectra of films with PO2val- ues of 110, 1101, and 5104Pa. There are five main feature peaks in the spectra, which are labeled A, B, C, D, and E.24The feature peaks A and B are attributed to hybridized states between the O 2p and Ti 3d states. Peaks A and B can be identified ast2gandegorbitals, respectively, produced by crystal field splitting. Peak C is assigned to the O 2p derived states that hybridize with Ba 5d or La 5d, while peaks D and E at higher energy levels belong to hybridized states between O 2p and Ti sp.
The relative intensities of peaks A, B, and C change with PO2dur- ing film growth. For the 110Pa sample, peak A shows the highest intensity and peak C shows the lowest intensity. For the 1101Pa sample, the intensities of both peaks B and C increase, and peak B has the highest relative intensity. For the 5104Pa sample, peak C has the highest intensity. The relative intensities of peaks D and E for the samples with PO2of 110and 1101Pa show no clear changes.
However, the sample with a PO2of 5104Pa had a much higher D peak intensity than that of peak E.
Changes in the relative peak intensities are assumed to be due to the increased electron concentration occupying the energy levels. The
VOintroduced electrons initially occupy thet2gorbital. As the VOcon- centration gradually increases, the electrons occupy successively higher energy orbitals from theegorbital to the hybridized O 2p-Ti sp state.
The valence bands in the BTO films are formed predominantly by the Ti 3d and O 2p states, and the Fermi level is below the t2gorbital level.25The results shown inFig. 4demonstrate that the increased VO concentration will facilitate more electrons to occupy higher energy levels and be close to the conduction band. This contributes to the decrease in the activation energy of the electrons, increases the electron concentration, and increases the film conductivity.
In summary, the microstructure, electrical conduction, and charge transport of epitaxial BLTO films with different VOconcentra- tions have been investigated. The XPS measurements indicate that the VO concentration in the films increased with the decrease in the TABLE I.List of various transport models and their corresponding temperature ranges andWHfor the BLTO films with PO2<110Pa.
PO2(Pa)
Variable range hopping model temperature range (K)
Small polaron hopping model temperature
range (K)
Thermal phonon scattering model
temperature
range (K) WH(eV)
1101 80–230 285–400 NA 0.087
1102 60–210 250–400 NA 0.047
5104 60–165 190–300 350–400 0.037
FIG. 4.The O K-edge SXAS spectra of the BLTO films with PO2values of 110, 1101, and 5104Pa. [The Fermi level (EF) position is schematically shown by the dotted line.]
oxygen pressure. Theq-Tmeasurements show that increased VOcon- centration leads to reduced electrical resistivity. The film with a PO2of 110Pa behaves as an insulator and exhibits the highest resistivity.
The films grown at PO2values of 1101Pa and 1102Pa exhibit semiconductive behavior. The film grown at a PO2 of 5104Pa exhibits semiconducting behavior at temperatures below 330 K and metallic behavior above 330 K. The role of VOconcentration in the charge transport mechanisms of BLTO films was clarified.
Furthermore, the analysis of the O K-edge SXAS spectra indicates that increases in the VOconcentration facilitate more electrons to occupy higher energy levels, thereby providing reasonable explanations to increased conductivity, lower activation energy, and different charge transport behaviors. These findings not only help to clarify the com- plex charge transport mechanisms in the ferroelectric BTO films, but also provide useful insights for the future study of conductivity and charge transport of other oxygen-deficient perovskite oxides.
Seesupplementary materialfor the details of XRR and thickness results.
This work was supported by the National Natural Science Foundation of China (Contract Nos. 51431006 and 51472093). X.
B. Lu acknowledges the support of the Project for Guangdong Province Universities and Colleges Pearl River Scholar Funded Scheme (2016). This work was also supported by the Guangdong Innovative Research Team Program (No. 2013C102), the Guangdong Provincial Key Laboratory of Optical Information Materials and Technology (Grant No. 2017B030301007), and the 111 Project. The authors also acknowledge the support from the National Synchrotron Radiation Laboratory (NSRL) of Hefei.
REFERENCES
1A. I. Kingon, S. K. Streiffer, C. Basceri, and S. R. Summerfelt,MRS Bull.21, 46 (1996).
2S.-H. Yoon, S.-J. Kim, S.-H. Kim, and D.-Y. Kim,J. Appl. Phys.114, 224103 (2013).
3Z. H. Lin, Y. Yang, J. M. Wu, Y. Liu, F. Zhang, and Z. L. Wang,J. Phys. Chem.
Lett.3, 3599 (2012).
4J. Zhou, X. S. Jing, M. Alexe, J. Y. Dai, M. H. Qin, S. J. Wu, M. Zeng, J. W.
Gao, X. B. Lu, and J.-M. Liu,J. Phys. D: Appl. Phys.49, 175302 (2016).
5D. Nagano, H. Funakubo, K. Shinozaki, and N. Mizutani,Appl. Phys. Lett.72, 2017 (1998).
6Y. Shao, R. A. Hughes, A. Dabkowski, G. Radtke, W. H. Gong, J. S. Preston, and G. A. Botton,Appl. Phys. Lett.93, 192114 (2008).
7X. S. Jing, W. C. Xu, C. Yang, J. J. Feng, A. H. Zhang, Y. P. Zeng, M. H. Qin, M. Zeng, Z. Fan, J. W. Gao, X. S. Gao, G. F. Zhou, X. B. Lu, and J.-M. Liu, Appl. Phys. Lett.110, 182903 (2017).
8G. H. Jonker,Solid-State Electron.7, 895 (1964).
9K. S. Takahashi, Y. Matsubara, M. S. Bahramy, N. Ogawa, D. Hashizume, Y.
Tokura, and M. Kawasaki,Sci. Rep.7, 4631 (2017).
10J. E. Hamann-Borrero, S. Macke, W. S. Choi, R. Sutarto, F. Z. He, A. Radi, I.
Elfimov, R. J. Green, M. W. Haverkort, V. B. Zabolotnyy, H. N. Lee, G. A.
Sawatzky, and V. Hinkov,njp Quantum Mater.1, 16013 (2016).
11H. M. I. Jaim, S. Lee, X. H. Zhang, and I. Takeuchi,Appl. Phys. Lett.111, 172102 (2017).
12Y. Kawabea, A. Yamanaka, E. Hanamura, T. Kimura, Y. Takiguchi, H. Kan, and Y. Tokura,J. Appl. Phys.87, 7594 (2000).
13T. Zhao, Z.-H. Chen, F. Chen, H.-B. Lu, G.-Z. Yang, and H.-S. Cheng,Appl.
Phys. Lett.77, 4338 (2000).
14U. S. Alaan, A. T. N’Diaye, P. Shafer, E. Arenholz, and Y. Suzuki,AIP Adv.7, 055716 (2017).
15A. Abliz, C. W. Huang, J. Wang, L. Xu, L. Liao, X. Xiao, W. W. Wu, Z. Fan, C.
Jiang, J. Li, S. Guo, C. Liu, and T. Guo,ACS Appl. Mater. Interfaces8, 7862 (2016).
16P.-T. Hsieh, Y.-C. Chen, K.-S. Kao, and C.-M. Wang,Appl. Phys. A: Mater. Sci.
Process.90, 317 (2007).
17S. Hashimoto, T. Sugie, Z. Zhang, K. Yamashita, and M. Noda,Jpn. J. Appl.
Phys., Part 154, 10NA12 (2015).
18J. Chen, X. R. Xing, R. B. Yu, J. X. Deng, and G. R. Liu,J. Alloys Compd.388, 308 (2005).
19T. Kolodiazhnyi,Phys. Rev. B78, 045107 (2008).
20J. Hwang, T. Kolodiazhnyi, J. Yang, and M. Couillard,Phys. Rev. B82, 214109 (2010).
21Y. Sun, X. J. Xu, and Y. H. Zhang,J. Phys.: Condens. Matter12, 10475 (2000).
22L. F. Liu, H. Z. Guo, H. B. L€u, S. Y. Dai, B. L. Cheng, and Z. H. Chen,J. Appl.
Phys.97, 054102 (2005).
23I. G. Austin and N. F. Mott,Adv. Phys.18, 41 (1969).
24K. Asokan, J. C. Jan, J. W. Chiou, W. F. Pong, M. H. Tsai, Y. K. Chang, Y. Y.
Chen, H. H. Hsieh, H. J. Lin, Y. W. Yang, L. J. Lai, and I. N. Lin,J. Solid State Chem.177, 2639 (2004).
25S. R. Gilbert, L. A. Wills, B. W. Wessels, J. L. Schindler, J. A. Thomas, and C. R.
Kannewurf,J. Appl. Phys.80, 969 (1996).
