Dark current suppression of amorphous selenium based photosensors
by the ZnO hole blocking layer
Tung-Yuan Yu
a, Fu-Ming Pan
a,*, Cheng-Yi Chang
a, Tien Hu
a, Jenn-Fang Chen
b,
Jia-Feng Wang
b, Cheng-Lu Lin
b, Tsung-Han Chen
c, Te-Ming Chen
caDepartment of Materials Science and Engineering, National Chiao-Tung University, 1001 Ta Hsueh Road, Hsinchu 30010, Taiwan, ROC bDepartment of Electrophysics, National Chiao-Tung University, Hsinchu 30010, Taiwan, ROC
cAU Optronics Corporation, Hsinchu Science Park, Hsinchu 30078, Taiwan, ROC
a r t i c l e i n f o
Article history:
Received 26 November 2013 Received in revised form 7 February 2014 Accepted 13 February 2014 Available online 4 March 2014
Keywords: Zinc oxide Amorphous Se Hole blocking layer Oxygen vacancy DLTS
a b s t r a c t
To study the influence of defects in the hole blocking layer (HBL) on the dark current of amorphous selenium (a-Se) based photosensors, we prepared ZnO thinfilms by reactive sputter deposition (RSD) for the use as the HBL of the photosensors. The ZnO HBL layers prepared with different oxygenflow rates were characterized by X-ray photoelectron spectroscopy, Raman scattering analysis and photo-luminescence, indicating that the density of oxygen vacancies in the ZnO thin films is significantly affected by the oxygenflow rate. The deep level transient spectroscopy measurement reveals two hole trap levels present in the RSD deposited ZnO thinfilms; one is at 0.94 eV and the other at 0.24 eV above the valence band edge. The electrical performance of the a-Se photosensor is largely influenced by the amount of oxygen vacancies in the ZnO thinfilm. The a-Se photosensor with the ZnO HBL of the most oxygen vacancies has the lowest dark current and demonstrates the highest breakdownfield.
Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction
Zinc oxide is a direct wide band gap oxide semiconductor with the band gap in the near ultraviolet (UV) spectral region and a large exciton binding energy of 60 meV[1e3]. As-grown ZnO is usually an n-type semiconductor because of the presence of intrinsic de-fects that introduce donor levels in the band gap, such as interstitial zinc and oxygen vacancies. Owing to many advantageous material and optoelectronic properties, ZnO has received extensive research on chemical and optoelectronic applications in the past few de-cades, such as transparent thin-film transistors, photosensors, blue/ UV light-emitting diodes and chemical sensors. Among the wide variety of promising applications, the use of ZnO as the hole blocking layer (HBL) for organic photovoltaic devices has been proposed based on its n-type nature with the wide band gap[4e6]. In this study, we used ZnO thinfilms as the HBL for amorphous selenium (a-Se) based photosensors, which have recently received much attention for X-ray imaging applications[7e9].
The photoconversion quantum efficiency of a-Se is drastically increased when the electric bias reaches afield of >70 V/
m
m. Thiseffect is due to the avalanche multiplication of impact ionization kindled by hole carriers drifting at the high electricfield[10,11]. The avalanche effect makes a-Se a prominent photoconductors for im-aging applications that require operations under an illumination condition of very low photon dose, such as medical X-ray imaging [7,12,13]. However, the high photoconversion gain at the highfield is usually accompanied by a high dark current, resulting in the degradation in the image contrast. The dark current is mainly contributed from the hole injection from the anode to the a-Se active layer [14,15]. In order to suppress the dark current, a-Se based photoconductors require an HBL to impede the hole injec-tion. In contemporary a-Se photosensors, such as high-gain avalanche rushing photoconductor (HARP), the HBL is generally made of a CeO2 thin film. Based on the observation of time
dependence of the dark current at different bias fields, the hole blocking mechanism is generally proposed to be associated with hole trapping in the CeO2HBL[16e18]. The hole trapping results in
charge accumulation in the blocking layer and a potential barrier is thus developed by the positive space charges, thereby suppressing hole carrier injection from the anode. Because the hole trapping is closely related to defects present in the HBL, and ZnO is one of the n-type oxide semiconductors that have been extensively studied on the nature of defects, we used ZnO thinfilms as the HBL layer of a-Se photosensors in this study. Fig. 1 shows the interface band
* Corresponding author. Tel.: þ886 3 5131322; fax: þ886 3 5724727. E-mail address:fmpan@faculty.nctu.edu.tw(F.-M. Pan).
Contents lists available atScienceDirect
Current Applied Physics
j o u r n a l h o me p a g e : w w w . e l s e v i e r . c o m/ l o ca t e / c a p
http://dx.doi.org/10.1016/j.cap.2014.02.011
diagram of the heterojunction between ZnO and indium doped tin oxide (ITO) to illustrate hole traps in the ZnO layer. The work function of ITO (4.75 eV) and the electron affinity of ZnO (4.35 eV) are collected from the literature[19,20], and the band gap of ZnO is experimentally obtained in the study as discussed later. Hole car-riers injected from the ITO electrode can be trapped in deep trap centers in the ZnO HBL. Increasing trap centers in the ZnO HBL is beneficial for lowering the dark current of the photosensors. The properties of intrinsic defects in ZnO highly depend on the growth techniques, growth conditions and post-growth thermal treatment [21,22]. We prepared the ZnO HBL by reactive sputter deposition (RSD), and X-ray photoelectron spectroscopy (XPS), photo-luminescence (PL) and Raman scattering spectroscopy were used to study oxygen vacancies present in the ZnO thin films. The mea-surement of hole trap levels was performed by deep level transient spectroscopy (DLTS).
2. Experimental
ZnO thin films were deposited on the ITO glass or the SiO2
substrate at room temperature by reactive sputter deposition in a magnetron sputtering system using a Zn metal target of 99.9% in purity. The ZnO thinfilm was deposited on the thermally oxidized Si wafer for material characterizations, while, for the fabrication of a-Se photosensors, it was deposited on the ITO glass. During the ZnOfilm deposition, the argon flow rate was kept at 20 sccm and the oxygenflow rate varied from 10 to 50 sccm; the sputtering power and the working pressure was 50 W and 1 102torr, respectively. All the ZnO samples prepared with different O2flow
rates at a constant Arflow rate of 20 sccm had a thickness of 50 nm. The major carrier concentration of ZnO thinfilms was analyzed by Hall-effect measurements (Ecopia HMS-3000). The surface morphology of ZnO thinfilms was examined by scanning electron microscopy (SEM, JEOL JSM-6500F) and atomic force microscope (AFM, Bruker Dimension Icon). The microstructure of the ZnO thin films was analyzed by transmission electron microscopy (TEM, JEOL JEM-3000F) and X-ray diffraction spectroscopy (XRD, Bruker D8 Advance) using the Cu K
a
source. The incident angle of the X-ray beam with respect to the sample surface is 1 during the XRD analysis. A Horiba Jobin Yvon LABRAM HR800 spectrometer was used to study Raman scattering of the ZnO thinfilms using anAr-ion laser as the excitatAr-ion source. The surface chemical composi-tion of the ZnO thinfilms was analyzed by X-ray photoelectron spectroscopy (XPS, Thermo VG 350). Photoluminescence (PL) emission at room temperature from the ZnO thinfilms was studied by a PL spectrometer (Horiba Jobin Yvon iHR-320). DLTS (Sula Technologies) was conducted at the temperature range between 110 and 350 K and at a quiescent bias of 0 V with thefilling pulse height of 5 V into the reverse bias and a width of 100 ms. For the DLTS measurement, the ZnO thinfilm with a thickness of 800 nm was deposited on a 500 nm thick SiO2layer thermally grown on the
Si wafer, and Pd and Al were used as the schottky and ohmic contacts, respectively.
The a-Se photosensor fabricated in the study is basically with the HARP device structure without the electron hole blocking layer. Fig. 2schematically shows the cross-sectional structure of the a-Se photosensor. Wefirst deposited the 50 nm-thick ZnO HBL on the ITO glass substrate at room temperature, followed by the deposi-tion of an a-Sefilm of 13
m
m in thickness at 30 C by thermal evaporation deposition. A cellulose acetate (CA) polymerfilm of 2m
m in thickness was then spin-coated on the a-Se layer before the electron beam deposition of the Al electrode pad with an area of 1 mm2. The CA polymerfilm was used as a distributed resistive layer (DRL), which can mitigate electrical discharge at electrode edges due to the high electrical stress during photosensing opera-tion [23,24]. The dark current of the a-Se photosensors was measured with a Keithley 237 source-measure unit.3. Results and discussion
Five ZnO thinfilms were prepared by reactive sputter deposition with different O2flow rates at a constant Ar flow rate of 20 sccm.
We will designate thereafter the ZnO thinfilms by ZnO-X, where X represents the O2flow rate in sccm. The electron density of all the
ZnO samples measured by Hall-effect is in the order of 1013cm3. The ZnO-20 thin film has the highest electron density (8.53 1013cm3) and the ZnO-10 thinfilm has the smallest one
(2.37 1013cm3).Fig. 3(a) and (b)shows the plane-view SEM and
AFM images of the ZnO-20 thinfilm, respectively. The surface of the thinfilm has a root-mean-square roughness of w1.64 nm according to the AFM analysis. The variation in the O2flow rate results in little
difference in the surface topography and the microstructure of the ZnO thinfilms.Fig. 3(c)shows the cross-sectional SEM image of a 50 nm thick ZnO-20 thinfilm capped by a 13
m
m thick a-Se layer,Fig. 1. The interface band diagram for the ITO/ZnO heterojunction. EFandfrepresent
the Fermi level and the work function of ITO, respectively. Labels for the ZnO semi-conductor are follows: EFs: Fermi level,c: electron affinity, EC: conduction band edge
and EV: valence band edge. Hole trap centers in ZnO are illustrated as the short lines
near EV.
indicating that the thin film has a microstructure composed of equiaxed and columnar grains with a size in the order of 20e30 nm. The cross-sectional TEM image of the ZnO-20 thinfilm is shown in Fig. 3(d)with a high resolution image (inset) showing the lattice fringe of the ZnO (0002) plane (w0.26 nm). Fig. 4 shows XRD spectra of ZnO thinfilms deposited with various oxygen flow rates. The distinct peak at 34.4corresponds to the (0002) lattice plane of the ZnO hexagonal wurtzite structure, indicating that the ZnO thin
films have a preferential orientation in the direction of the c-axis. Because the (0002) plane of wurtzite ZnO has the smallest surface free energy compared with other lattice planes, the sputter-deposited ZnO thin films tend to grow along the c-axis orienta-tion[25]. Based on the Scherrer equation, the average grain size of the ZnO thinfilms is around 20 nm, which is consistent with the observation of the cross-sectional SEM image.
Fig. 5(a) shows XPS spectra of the O1s electron for the ZnO samples. The O 1s XPS signal can be curve-fitted with three peak components assuming a GaussianeLorentzian peak shape[26e29]. The Oapeak at 530.1 eV is due to lattice oxygen ions (O2) in the
ZnO wurtzite structure, and the Oc peak at 532.3 eV can be
attributed to adsorbed species on the ZnO surface, such as H2O and
weakly adsorbed O2molecules. The Obpeak at 531.2 eV is generally
assigned to Oions, which are associated with oxygen vacancies (VO) in the surface layer of the ZnO thin films[26,30]. Fig. 5(b)
presents the ratio of the peak area of the individual peaks to the summed area of the three component peaks as a function of the O2
flow rate used for the ZnO deposition. It is obvious from the figure that the Obpeak of the ZnO-20 sample has the largest area ratio,
suggesting that the ZnO-20 thinfilm has the highest oxygen va-cancy density. Raman scattering study discussed below supports the observation by XPS.
Fig. 6shows Raman spectra ranging from 350 cm1to 650 cm1 for the ZnO thin films. The fluorescence-background has been removed from the Raman spectra. A broad peak is observed in the range between 375 and 475 cm1, which can be assigned to the superposition of three Raman peaks, A1(TO) (380 cm1), E1(TO)
(407 cm1) and E2(high) (473 cm1)[31]. The E2(high) mode is Fig. 3. The Plane-view SEM image (a) and the AFM image (b) of the ZnO-20 thinfilm; (c) cross-sectional SEM image of the ZnO-20 thin film overlaid by a 13mm thick a-Se layer; (d) the cross-sectional TEM image of the ZnO-20 thinfilm with the inset showing a high-resolution image.
associated with oxygen vibration and its frequency shifts if lattice distortion occurs in the ZnOfilm. According to the Raman spectra, all the ZnO samples have the E2(high) peak situated at 437 cm1,
indicating that the variation of the O2flow rate results in little effect
on the lattice strain of the ZnO samples. On the other hand, all samples except the ZnO-20 has the E1(LO) peak centered at
583 cm1. Because the A1(LO) mode, which is situated at 574 cm1
for wurtzite ZnO, is obviously absent in the Raman spectra of all the ZnO samples other than the ZnO-20, we attribute the lower peak position (w4 cm1) of the ZnO-20 thinfilm to a redshift of the E1(LO) peak but not an interference from the A1(LO) mode. A small
redshift of the E1(LO) mode can be ascribed to the presence of point
defects, such as O-vacancy and Zn-interstitial, in the ZnO wurtzite lattice[31]. The observed redshift of the E1(LO) mode is in
agree-ment with the XPS study, which shows that the ZnO-20 has a higher concentration of oxygen vacancy than the other ZnO sam-ples. Therefore, we believe that the ZnO-20 thinfilm has a larger VO
density than other ZnO samples.
The room temperature PL spectra of the ZnO thin films are shown inFig. 7. The sharp luminescence peak situated at 377 nm is due to the near-band-edge emission. The broad band emission ranging from 450 nm to 700 nm is centered at 550 nm, which is generally attributed to defect-related deep level emissions in the
ZnO thinfilm[32e34]. The peak position and shape of the green band has little dependence on the oxygenflow rate for the ZnO deposition. Therefore, the green band emission of all the ZnO thin films must be induced from the same electronic transition process involving a specific defect level. The relative concentration of the defect centers resulting in the green PL thinfilm can be estimated by the PL intensity (peak height) ratio of the green emission (Igr) to
the edge emission (Iuv). The inset ofFig. 7shows the Igr/Iuvratio as a
function of the oxygenflow rate. The ZnO-20 sample has the largest PL ratio suggesting that it has the highest concentration of the defect producing the green emission. Although the PL emission of ZnO has been extensively studied, the origin of the green band emission is still under much debate. Among the various defects proposed to be responsible for the green PL emission of ZnO, oxy-gen vacancies are the most often suggested defect inducing the emission, which is considered to result from electronehole recombination in singly ionized VOcenters situated about 0.9 eV
above the valence band maximum (VBM) of ZnO [21,35e39]. Because the XPS and Raman analyses reveal that the ZnO-20 thin film has a larger VOdensity than other ZnO samples, the green
emission is likely associated with VOrelated defect centers. We do
find by DLTS analysis that a deep level trap is present at 0.9 eV above the VBM.
Fig. 8(a) presents the DLTS spectrum of the ZnO-20 sample showing two signal dips. The two hole-trap related signals have the minimum at 216 K (L1) and at 287 K (L2). The thermal emission activation energy of the two trap levels can be determined by the Arrhenius plot as shown inFig. 8(b), in which ln(
s
T2) is plotted against 1000/T, wheres
is the capacitance transient time constant deduced from the DLTS window setting and T is the absolute temperature[40]. The activation energy of the thermal emission of the L1 and L2 traps are thus calculated to be 0.246 eV and 0.942 eV, respectively, with respect to the VBM. The trap density derived from the DLTS measurement is 1.50 1011 cm3 and2.93 1011 cm3 for L1 and L2, respectively. A defect level at
w0.9 eV above VBM has been previously reported in the defect study of ZnO by DLTS[41,42]. Much research, including theoretical and experimental studies, has proposed that the energy level
Fig. 5. (a) O 1s XPS spectra of the ZnO thinfilms deposited with various O2flow rates;
(b) the ratio of the peak area of the three component peaks to the summed area as a function of the O2flow rate.
around 0.9 eV is due to oxygen vacancies[41,43e46]. The measured activation energy for the L2 hole trap is in very good agreement with the theoretical value obtained by Lany et al. for the VO(þ/0)
transition level[44]. The reactive sputter-deposited ZnO thinfilms have an optical band gap of 3.21 eV as determined by the Tauc plot based on the ultravioletevisible absorption spectroscopy analysis.
If electronic transition from the experimentally determined con-duction band minimum to the L2 level takes place, a green band emission can occur at 2.27 eV (w546 nm), which is consistent with the observation in the PL spectra in Fig. 7. The defect identity associated with the L1 level is unclear to us presently. If radiative electronic transition between the conduction band minimum and the L1 level is allowed, it can produce a violet emission at 2.96 eV (w419 nm). However, the violet emission may be obscured by the band edge emission in this study as a result of the small concen-tration of the L1-related defect. Because the ZnO-20 sample has the largest concentration of VOas revealed by the PL spectra than the
other samples, it is expected to produce the lowest dark current in the a-Se photosensor when used as the HBL.
Fig. 9shows the dark current density of the a-Se photosensors with the ZnO HBL as a function of the electricfield. The dark current density of the a-Se device without the ZnO HBL is about 2 1011A/ mm2 at afield of 10 V/
m
m. When the 50 nm thick ZnO HBL is present in the a-Se photosensor, the dark current density at 10 V/m
m is reduced by nearly one order of magnitude for all the devices with different oxygenflow rates. The dark current density increases with the electric field and demonstrates a sudden increase in a certain field range, indicating the breakdown of the device. The device with the ZnO-20 HBL shows the smallest increase rate of the dark current density (w1 1011 A/mm2 at 40 V/m
m) and thehighest breakdownfield (>50 V/
m
m). The better electrical perfor-mance of the ZnO-20 device can be ascribed to a higher concen-tration of the VO, which may form the deep hole trap center at0.942 eV as discussed above, and thus has a better suppression effect on hole injection from the ITO anode into the a-Se active layer. It has been reported that more defect levels in the CeO2HBL
of a-Se photosensors will lower the potential barrier for holes be-tween the anode and the HBL, thus leading to an increase in the dark current[47]. However, our result seems to indicate that the variation in the VOconcentration in the ZnO HBLs affects
insignif-icantly the potential barrier height. FromFig. 9, a larger dark cur-rent triggers the device breakdown at a lower electricfield; this may be a consequence of the local Joule heating effect. Because the a-Se layer was prepared without the addition of any impurity, such as arsenic, it has a very low glass transition temperature (w41C
according to differential scanning calorimetry analysis). The Joule effect can induce crystallization of the a-Se layer at the low tem-perature and thereby increases the conductivity of the photo-conductor leading to the breakdown [12,48]. The ZnO-20 photoconductor device has the highest breakdownfield as a result of the smallest dark current density.
Fig. 7. PL spectra of the ZnO thinfilms deposited with various O2flow rates. Inset is the
peak height ratios of the green band emission to the band edge emission as a function of the O2flow rate. (For interpretation of the references to color in this figure legend,
the reader is referred to the web version of this article.)
Fig. 8. (a) The DLTS spectrum of the ZnO-20 thinfilm (rate window time constant
s ¼ 0.3743 ms); (b) the corresponding Arrhenius plots for the hole trap levels measured by DLTS analysis.
Fig. 9. The dark current of the a-Se photoconductors with the HBL prepared using different O2flow rates as a function of the electric field.
4. Conclusion
We prepared ZnO thinfilms by reactive sputter deposition, and used as the HBL for a-Se based photoconductor devices. The Ar/O2
flow rate ratio used for the ZnO deposition has a significant effect on material and electrical properties of the ZnO HBL. We correlated oxygen vacancies in the ZnO thinfilm with the dark current sup-pression capability of the HBL by XPS, Raman, PL and DLTS analyses. The XPS, Raman and PL measurements indicate that the ZnO thin film prepared with the Ar/O2flow rate ratio of 20 sccm/20 sccm has
the highest concentration of oxygen vacancies, which create a hole trap level at 0.94 eV above the valence band edge according to the DLTS measurement. The electronic transition from the conduction band edge to the hole trap level results in the green emission centered at 2.72 eV. The a-Se photosensor with the ZnO-20 HBL has the best photoconduction performance, including the lowest dark current and the highest breakdownfield. We believe that oxygen vacancies in the HBL layer play as hole trap centers to suppress the hole injection from the ITO anode of the a-Se photosensor as pro-posed by many previous reports.
Acknowledgments
We thank the National Science Council of R.O.C. and AU Optronics Corporation for the financial support. The technical support of the National Nano Device Laboratories is also gratefully acknowledged.
References
[1] A. Janotti, C. G. V. d. Walle, Rep. Prog. Phys. 72 (2009) 126501.
[2] Ü. Özgür, Y.I. Alivov, C. Liu, A. Teke, M.A. Reshchikov, S. Dogan, V. Avrutin,
S.-J. Cho, H. Morkoç, S.-J. Appl. Phys. 98 (2005) 041301.
[3] G.W. Tomlins, J.L. Routbort, T.O. Mason, J. Appl. Phys. 87 (2000) 117. [4] S. Schumann, R. Da Campo, B. Illy, A.C. Cruickshank, M.A. McLachlan,
M.P. Ryan, D.J. Riley, D.W. McComb, T.S. Jones, J. Mater. Chem. 21 (2011) 2381. [5] J. Boucle, P. Ravirajan, J. Nelson, J. Mater. Chem. 17 (2007) 3141.
[6] T. Yang, W. Cai, D. Qin, E. Wang, L. Lan, X. Gong, J. Peng, Y. Cao, J. Phys. Chem. C. 114 (2010) 6849.
[7] S. Kasap, J.B. Frey, G. Belev, O. Tousignant, H. Mani, J. Greenspan, L. Laperriere, O. Bubon, A. Reznik, G. DeCrescenzo, K.S. Karim, J.A. Rowlands, Sensors 11 (2011) 5112.
[8] A.R. Cowen, S.M. Kengyelics, A.G. Davies, Clin. Radiol. 63 (2008) 487. [9] J.A. Seibert, Pediatr. Radiol. 36 (2006) 173.
[10] A. Reznik, S.D. Baranovskii, O. Rubel, G. Juska, S.O. Kasap, Y. Ohkawa, K. Tanioka, J.A. Rowlands, J. Appl. Phys. 102 (2007) 053711.
[11] O. Rubel, A. Potvin, D. Laughton, J. Phys.-Condes. Matter 23 (2011) 055802. [12] M.M. Wronski, W. Zhao, A. Reznik, K. Tanioka, G. DeCrescenzo, J.A. Rowlands,
Med. Phys. 37 (2010) 4982.
[13] K. Tanioka, J. Mater. Sci.-Mater. Electron 18 (2007) 321. [14] G. Belev, S.O. Kasap, J. Non-Cryst. Solids 352 (2006) 1616.
[15] J.B. Frey, G. Belev, O. Tousignant, H. Mani, S.O. Kasap, Phys. status solidi C. 6 (2009) S251.
[16] S.A. Mahmood, M.Z. Kabir, O. Tousignant, H. Mani, J. Greenspan, P. Botka, Appl. Phys. Lett. 92 (2008) 223506.
[17] S. Kasap, J.B. Frey, G. Belev, O. Tousignant, H. Mani, L. Laperriere, A. Reznik, J.A. Rowlands, Phys. status solidi B 246 (2009) 1794.
[18] S. Kasap, G. Belev, J. Optoelectron. Adv. Mater. 9 (2007) 1. [19] K. Sugiyama, H. Ishii, Y. Ouchi, K. Seki, J. Appl. Phys. 87 (2000) 295. [20] Z. Guo, D. Zhao, Y. Liu, D. Shen, J. Zhang, B. Li, Appl. Phys. Lett. 93 (2008)
163501.
[21] C.H. Ahn, Y.Y. Kim, D.C. Kim, S.K. Mohanta, H.K. Cho, J. Appl. Phys. 105 (2009) 013502.
[22] Q.P. Wang, D.H. Zhang, Z.Y. Xue, X.J. Zhang, Opt. Mater. 26 (2004) 23. [23] O. Bubon, G. DeCrescenzo, J.A. Rowlands, A. Reznik, J. Non-Cryst. Solids 358
(2012) 2431.
[24] A. Sultana, M.M. Wronski, K.S. Karim, J.A. Rowlands, IEEE Sens. J. 10 (2010) 347.
[25] P.T. Hsieh, Y.C. Chen, C.M. Wang, Y.Z. Tsai, C.C. Hu, Appl. Phys. A-Mater. Sci. Process 84 (2006) 345.
[26] P.T. Hsieh, Y.C. Chen, K.S. Kao, C.M. Wang, Appl. Phys. A-Mater. Sci. Process 90 (2008) 317.
[27] X.Q. Wei, B.Y. Man, M. Liu, C.S. Xue, H.Z. Zhuang, C. Yang, Phys. B 388 (2007) 145.
[28] Z.G. Wang, X.T. Zu, S. Zhu, L.M. Wang, Phys. E 35 (2006) 199.
[29] T.G.G. Maffeis, M.W. Penny, A. Castaing, O.J. Guy, S.P. Wilks, Surf. Sci. 606 (2012) 99.
[30] M. Li, G. Xing, L.F.N. Ah Qune, G. Xing, T. Wu, C.H.A. Huan, X. Zhang, T.C. Sum, Phys. Chem. Chem. Phys. 14 (2012) 3075.
[31] N. Ashkenov, B.N. Mbenkum, C. Bundesmann, V. Riede, M. Lorenz, D. Spemann, E.M. Kaidashev, A. Kasic, M. Schubert, M. Grundmann, G. Wagner, H. Neumann, V. Darakchieva, H. Arwin, B. Monemar, J. Appl. Phys. 93 (2003) 126.
[32] C. Chandrinou, N. Boukos, C. Stogios, A. Travlos, Microelectron. J. 40 (2009) 296.
[33] I.C. Yao, T.-Y. Tseng, P. Lin, Sens. Actuator A-Phys. 178 (2012) 26. [34] B. Lin, Z. Fu, Y. Jia, Appl. Phys. Lett. 79 (2001) 943.
[35] K. Vanheusden, W.L. Warren, C.H. Seager, D.R. Tallant, J.A. Voigt, B.E. Gnade, J. Appl. Phys. 79 (1996) 7983.
[36] B. Cao, W. Cai, H. Zeng, Appl. Phys. Lett. 88 (2006) 161101. [37] L.S. Vlasenko, G.D. Watkins, Phys. Rev. B 71 (2005) 125210.
[38] X. Meng, Z. Shi, X. Chen, X. Zeng, F. Zhuxi, J. Appl. Phys. 107 (2010) 023501. [39] G. Xiong, U. Pal, J.G. Serrano, J. Appl. Phys. 101 (2007) 024317.
[40] Y.-P. Wang, W.-I. Lee, T.-Y. Tseng, Appl. Phys. Lett. 69 (1996) 1807. [41] A.Y. Polyakov, N.B. Smirnov, A.V. Govorkov, E.A. Kozhukhova, V.I. Vdovin, K. Ip,
M.E. Overberg, Y.W. Heo, D.P. Norton, S.J. Pearton, J.M. Zavada, V.A. Dravin, J. Appl. Phys. 94 (2003) 2895.
[42] Z.-Q. Fang, B. Claflin, D.C. Look, L.L. Kerr, X. Li, J. Appl. Phys. 102 (2007) 023714. [43] F. Oba, A. Togo, I. Tanaka, J. Paier, G. Kresse, Phys. Rev. B 77 (2008) 245202. [44] S. Lany, A. Zunger, Phys. Rev. B 72 (2005) 035215.
[45] S.B. Zhang, S.H. Wei, A. Zunger, Phys. Rev. B 63 (2001) 075205. [46] T.R. Paudel, W.R.L. Lambrecht, Phys. Rev. B 77 (2008) 205202.
[47] K. Kikuchi, Y. Ohkawa, K. Miyakawa, T. Matsubara, K. Tanioka, M. Kubota, N. Egami, Phys. status solidi C. 8 (2011) 2800.
[48] S. Abbaszadeh, N. Allec, S. Ghanbarzadeh, U. Shafique, K.S. Karim, IEEE Trans. Electron Devices 59 (2012) 2403.
