2850 IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 65, NO.

7, JULY 2018

Role of Surface States and Interface Charges

in 2DEG in Sputtered ZnO Heterostructures
Rohit Singh , Md. Arif Khan , Shaibal Mukherjee , Member, IEEE ,
and Abhinav Kranti , Senior Member, IEEE

Abstract — In this paper, we report on the influence of optoelectronic properties MgZnO/ZnO-based heterostructures
surface (donor and acceptor) states and interface charges also show the formation of 2-D electron gas (2DEG)
on the formation of 2-D electron gas (2DEG) in MgZnO/ZnO at the MgZnO/ZnO heterointerface due to spontaneous
heterostructure grown by sputtering. Donor and acceptor
states alone cannot yield an order of magnitude higher value and piezoelectric polarization [9]. The 2DEG density in
of 2DEG as observed by sputtering growth as compared MgZnO/ZnO heterostructure can be obtained by different
to epitaxial techniques. While surface donor states are growth mechanisms such as epitaxial and sputtering. However,
reported to govern the formation of 2DEG in epitaxially the 2DEG density is nearly an order of magnitude different
grown heterostructures, our simulations suggest a non- in heterostructure grown by sputtering [2], [10], [11] and
negligible contribution of surface acceptor states in the
2DEG density. In addition, the existence of interface charges epitaxial [1], [9], [12] techniques. Therefore, the contributing
due to high defect density along with appropriate values factors to the source of electrons in 2DEG in MgZnO/ZnO
of donor and acceptor states is essential to accurately heterostructure in both growth techniques should be different
predict an order of magnitude higher 2DEG density grown by and require insightful analysis. Ibbetson et al. [13] suggested
sputtering. This paper analyzes the distinct roles of donor the presence of surface donor states to be the predominant
and acceptor states along with interface charges in 2DEG
formation and achieved values. origin of electrons in 2DEG for AlGaN/GaN heterostructures,
which might also hold for MgZnO/ZnO heterostructures
Index Terms — 2-D electron gas (2DEG) sheet charge den- grown by molecular beam epitaxy (MBE). However, an exact
sity, epitaxy, sputtering, zinc oxide (ZnO) heterostructure.
physical mechanism of formation of 2DEG, its understanding
and contributions of different affecting parameters is still
I. I NTRODUCTION lacking in the literature mainly for sputtered deposited

Z INC oxide (ZnO)-based heterostructures are being

explored for light-emitting diode (LED), ultra violet-
photo-detectors, solar cells, and high-electron-mobility
MgZnO/ZnO heterostructure [13], [14].
The difference in the values of 2DEG by sputtering and epi-
taxial growth suggests that the formation of 2DEG is affected
transistor (HEMT) [1]–[5]. ZnO has superior optoelectronic by the introduction of an additional source of electrons in
properties, including the free-exciton binding energy case of growth by sputtering. In order to explain the source of
of 60 meV and direct band gap of 3.37 eV at room temperature electrons in 2DEG in MgZnO/ZnO heterostructures grown by
makes the MgZnO/ZnO heterostructures-based LED [6] work different techniques, the roles of surface states and interface
at elevated temperature in comparison to conventional charges in MgZnO/ZnO heterostructure are investigated using
Si-based LED [7]. ZnO-based thin film transistors are an insightful analysis based on electrostatic neutrality across
also used for liquid crystal display panel [8]. Apart from barrier and buffer layer. The proposed theory is compared with
simulations to account for different contributing factors and
Manuscript received March 6, 2018; revised April 24, 2018; accepted
May 16, 2018. Date of publication May 28, 2018; date of current version the effect of each such factor on 2DEG density. The analysis
June 19, 2018. This work was supported by DST SERB Project under is further verified by comparing the results with published
Grant EMR/2016/006847 dated October 13, 2017. The review of this experimental data for epitaxial [9] and sputtering [10] growth
paper was arranged by Editor G. Verzellesi.(Corresponding author:
Md. Arif Khan.) mechanisms. This paper improves the understanding of roles
R. Singh and M. A. Khan are with the Hybrid Nanodevice Research of donor and acceptor states and provides an explanation
Group (HNRG) and Low Power Nanoelectronics Research Group, Disci- for an order of magnitude high 2DEG density in sputtered
pline of Electrical Engineering, IIT Indore, Simrol, Indore 453552, India
(e-mail: [email protected]; phd1501102009 @iiti.ac.in). heterostructures through the contribution of interface charges.
S. Mukherjee is with the Hybrid Nanodevice Research Group (HNRG), This understanding is essential for the estimation of 2DEG
Electrical Engineering, IIT Indore, Simrol, Indore 453552, India (e-mail: under different growth mechanisms, which is crucial to the
[email protected]).
A. Kranti is with the Low Power Nanoelectronics Research Group, performance of HEMTs.
Discipline of Electrical Engineering, IIT Indore, Simrol, Indore 453552,
India (e-mail: [email protected]). II. S IMULATIONS
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org. Fig. 1(a) shows the experimental 2DEG density in
Digital Object Identifier 10.1109/TED.2018.2838546 MgZnO/ZnO heterostructure, available in [9] and [10], grown

0018-9383 © 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

Authorized licensed use limited to: Pt Dwarka Prasad Mishra Indian Inst of Info Tech. Downloaded on January 21,2021 at 05:18:42 UTC from IEEE Xplore. Restrictions apply.
 SINGH et al.: ROLE OF SURFACE STATES AND INTERFACE CHARGES IN 2DEG 2851

TABLE I
PARAMETERS OF M G Z N O C ALCULATED F ROM M OLE F RACTION
(x) AT R OOM T EMPERATURE

Fig. 1. (a) 2DEG density varying with Mg composition for sputtering [10]
and epitaxially [9] grown MgZnO/ZnO heterostructures. (b) Schematic
conduction band diagram for MgZnO/ZnO heterostructure showing vari-
ous space charge components. Inset: device schematic.

by MBE [9] and radio frequency (rf) sputtering [10] tech-

niques, with varying Mg composition (x) in MgZnO barrier
layer at zero bias. Therefore, all the investigations done in this
paper are at zero bias. For a 30-nm barrier layer thickness (t),
the value of 2DEG density in sputtering grown MgZnO/ZnO
heterostructures is nearly one order of magnitude higher than
that grown by MBE. Ibbetson et al. [13] suggested that the
surface donor-like states to be the origin of 2DEG electrons, negative nature of fixed trap charges. The electrons in 2DEG
the same mechanism is not appropriate to predict an order of thus can be obtained as
magnitude higher 2DEG values in rf-sputtered samples which
requires careful reconsideration of the contributing factors and qn s = σsurface + σinterface + σintegrated . (1)
the distinct roles of each for high values of 2DEG density. The device structure has been analyzed through ATLAS
Chin et al. [10] has suggested that high values of 2DEG simulation tool [19]. In order to ensure accuracy of physics-
density (∼1014 cm−2 ) in MgZnO/ZnO samples is due to high based simulations, parameters such as band gap energy, spon-
defect density in their samples which contribute more carriers taneous polarization coefficients, lattice parameters, elastic
to the interface, and thus increases 2DEG density. stiffness constants, piezoelectric constants, and relative per-
As shown in the schematic conduction band diagram mittivity, as outlined in Table I were used in simulation
[Fig. 1(b)] of MgZnO/ZnO heterostructure, the space charge of MgZnO/ZnO heterostructure. Bandgap narrowing model
components consists of the following. is included to account for the variation of band gap with
1) Negative charge due to electrons in 2DEG at the het- temperature, Fermi-Dirac [20] statistics and Shockley–Read–
erointerface (qns , where q is the charge of electron and Hall Recombination [21], [22] model has been included in the
n s is the 2DEG density). simulations. The models of carrier transport are also taken into
2) Polarization-induced sheet charges in barrier and buffer account [23], [24].
layer.
3) Integrated sheet charge (σintegrated ) due to ionized bulk III. R ESULTS AND D ISCUSSION
donors in MgZnO (σMgZnO ) and ZnO (σZnO ) layers. As per the existing theory valid for epitaxially grown
4) charge due to ionized surface states (σsurface ), including heterostructures [8], electrons which compensate the positive
donors (N D ) and acceptors (N A ) surface states (Here, polarization charge difference, come from integrated sheet
N D and N A are not doping densities but surface state charges from bulk donors (σint egrat ed ) and ionized surface
densities.) states (σsur f ace ) for single crystalline MgZnO/ZnO system.
5) Charges due to formation of interface states at the The σint er f ace is neglected because its value is very small for
barrier–buffer interface (σinterface ). single crystalline systems [25]. Therefore, the expression of n s
In this paper, σinterface is considered as fixed charge as the for single crystalline epitaxially grown heterostructures can be
study done in this paper is for zero bias conditions. In case expressed as
of (0001) oriented MgZnO/ZnO heterostructure, the polariza-
qn s = σsurface + σintegrated . (2)
tion difference created at the heterointerface is positive. This
polarization difference at the heterointerface creates an electric Fig. 2(a) shows the comparison of epitaxially grown exper-
field, which attracts free electrons present in the heterostruc- imental results [9] of 2DEG density with our simulation for
ture to nullify the positive polarization charge difference at varying barrier layer thickness at fixed Mg composition of
the heterointerface. If the deposition is not single crystalline x = 0.37. It is observed that the surface donor and accep-
i.e., defective, then there is possibility for fixed trap charges in tor densities combined can attain the experimental 2DEG
barrier layer, which will enhance or decrease the electric field density values. The ionized surface donor states density of
created by polarization difference depending on positive or 7 × 1012 cm−2 at 0.8 eV energy level below barrier layer

Authorized licensed use limited to: Pt Dwarka Prasad Mishra Indian Inst of Info Tech. Downloaded on January 21,2021 at 05:18:42 UTC from IEEE Xplore. Restrictions apply.
 2852 IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 65, NO. 7, JULY 2018

This may be due to sudden decrease in interface charges at

x = 0.15 onward lower Mg composition because of decrease
in lattice mismatch between barrier and buffer layer [15] which
reduces the chance of formation of high interface charges.
Another probable reason for this sudden transition may be that
the polarization field difference for the Mg0.15 Zn0.85 O/ZnO
interface in RF sputtered film is weaker than those with higher
Mg content. The carriers induced by the polarization field in
this case may not be sufficient to passivate the defects [30].
From Fig. 2(b) the followings are evident.
1) The surface donor and acceptor states alone cannot attain
the experimentally obtained 2DEG values in sputtered
heterostructures [10]. The experimental values, as shown
in Fig. 2(b), are one order of magnitude higher than the
simulated values achieved only by surface donor and
acceptor states, and undoped barrier and buffer layer.
Therefore, the surface donor and acceptor states are not
the sole source of 2DEG electrons in sputtered ZnO
Fig. 2. Comparison of simulated and experimental 2DEG density grown heterostructures.
by (a) MBE, and (b) rf-sputtering. (c) Simulated 2DEG density for different
doping densities in ZnO buffer layer showing that higher values of 2DEG
2) The other possible source of 2DEG electrons is the
density are least likely to be due to the doping of buffer layer. integrated sheet charges in barrier and buffer layer due
to doping. However, these charges were not able to
achieve the experimental values even when the doping
conduction band minima along with surface acceptor states density in buffer ZnO layer was increased excessively
density of 1013 cm−2 at 0.2 eV above barrier layer valence (>1019 cm−3 ), while keeping the constant doping in
band maxima, are used in the simulation. These values of sur- barrier MgZnO layer as 1017 cm−3 , as can be seen
face donor and acceptor energy level as predicted in our work in Fig. 2(c).
agrees very well with the published results of Tampo et al. [26] 3) Experimental values of 2DEG density in sputtered
and Park et al. [27], respectively. The value of N D as predicted MgZnO/ZnO heterostructures [10] were only achieved
in our work is also in reasonable agreement with experimen- when another source of 2DEG electrons was considered
tally obtained value of ∼8 × 1012 cm−2 by Tampo et al. [26] in the analysis. As the surface of ZnO is generally
at x = 0.42. The value of surface acceptor states density not very stable due to presence of oxygen vacan-
of 1013 cm−2 is used to obtain a good fit in the simulations, cies [31] and hydrogen adsorption [32], it is highly likely
and is also in the range of reported values for surface acceptor that the interface of sputtered MgZnO/ZnO has defect
states density [28]. The value of σMgZnO is ∼1017 cm−3 states which contribute toward an enhanced 2DEG den-
and σZnO is ∼1016 cm−3 is used throughout in the simula- sity [10]. The defect density is high in sputtered samples,
tions unless its changed value is explicitly mentioned. Our which leads to high trapped charges, and therefore,
simulation results as shown in Fig. 2(a) suggests that the higher values of effective interface charges [10].
existing theory [13] for the source of 2DEG electrons in single This would imply that for sputtering grown samples the charge
crystalline AlGaN/GaN system partially holds true for single balance equation is same as (1).
crystalline MgZnO/ZnO system. Published results have also While the exact mechanism for the formation of high defect
reported the presence of both surface donor and acceptor states density in MgZnO/ZnO samples is still under research for
in ZnO surface [28], [29]. It can also be seen in Fig. 2(a) sputtered deposited samples, our work has shown that it has
that as the surface acceptor states density increases, 2DEG significant impact on the formation of 2DEG in sputtered
density value decreases because it pulls the Fermi energy MgZnO/ZnO heterostructure [27], [33]. In MBE, the depo-
level down and decreases the triangular potential well depth, sition is nearly defect free and, therefore, there are negligible
which is formed at MgZnO/ZnO heterointerface. It is also to charges at MgZnO/ZnO interface [25].
be noted that without surface acceptor states, the simulated Fig. 3(a) shows that the conduction band for epitaxially
2DEG density values are higher than the experimental values, grown heterostructures (without interface charges), that sug-
and vary from 8.56 × 1012 cm−2 to 8.73 × 1012 cm−2 for gests an abrupt change at the interface and the triangular
corresponding change in thickness of MgZnO layer from 3 to potential well is formed only in buffer layer side. This is
20 nm, not shown in the graph. because the source of 2DEG electrons, to be collected in the
Fig. 2(b) shows the comparison of 2DEG density in exper- triangular potential well, is surface states which pin the Fermi
imental results of rf-sputtering [10] with our simulated 2DEG energy level at a specific energy level below conduction band
density, with and without interface charges for varying Mg minima, i.e., fixes the position of conduction band minima
composition at fixed barrier layer thickness of 30 nm. Our as well. In sputtering grown heterostructure (with interface
simulated 2DEG density is unable to attain the experimental charges, ∼1013 cm−2 ), we observe that the conduction band
2DEG density for lower Mg composition of x = 0.15. shifts downward, thus, more 2-D density of states are available

Authorized licensed use limited to: Pt Dwarka Prasad Mishra Indian Inst of Info Tech. Downloaded on January 21,2021 at 05:18:42 UTC from IEEE Xplore. Restrictions apply.
 SINGH et al.: ROLE OF SURFACE STATES AND INTERFACE CHARGES IN 2DEG 2853

of the research and development work undertaken in the

project under the Visvesvaraya PhD Scheme of Ministry of
Electronics and Information Technology, Government of India,
being implemented by Digital India Corporation (formerly
Media Lab Asia).

R EFERENCES
Fig. 3. (a) Conduction energy band diagram of MgZnO/ZnO heterostruc- [1] H. Tampo et al., “Two-dimensional electron gas in Zn polar
ture for without interface charges and with interface charges. (b) Charge ZnMgO/ZnO heterostructures grown by radical source molecular
distribution profile of 2DEG across MgZnO/ZnO heterostructure for beam epitaxy,” Appl. Phys. Lett., vol. 89, no. 13, p. 132113, 2006,
without interface charges and with interface charges. doi: 10.1063/1.2357588.
[2] I-C. Cheng, B.-S. Wang, H.-H. Hou, and J.-Z. Chen, “MgZnO/ZnO
heterostructure field-effect transistors fabricated by RF-sputtering,” ECS
Trans., vol. 50, no. 8, pp. 83–93, 2013, doi: 10.1149/05008.0083ecst.
for 2DEG. This is due to the defect generated charge carri- [3] B. S. Wang, Y. S. Li, and I. C. Cheng, “Mobility enhancement
ers [34] that pull down the conduction band minima further in RF-sputtered MgZnO/ZnO heterostructure thin-film transistors,”
below in buffer layer, until the positive polarization difference IEEE Trans. Electron Devices, vol. 63, no. 4, pp. 1545–1549,
Apr. 2016, doi: 10.1109/TED.2016.2526649.
at the barrier–buffer interface is completely abolished. For fur- [4] M. A. Khan, R. Singh, S. Mukherjee, and A. Kranti, “Buffer layer
ther increment of positive interface charges [35], the induced engineering for high (≥1013 cm−2 ) 2-DEG density in ZnO-based
free carriers accumulate at the barrier–buffer interface under heterostructures,” IEEE Trans. Electron Devices, vol. 64, no. 3,
pp. 1015–1019, Mar. 2017, doi: 10.1109/TED.2016.2647258.
additional positive columbic attraction and since, the band-
[5] R. Singh, M. A. Khan, S. Mukherjee, and A. Kranti, “Analytical model
offset is below the Fermi energy level, the free carriers are for 2DEG density in graded MgZnO/ZnO heterostructures with cap
accumulated on the barrier side as well. This further increases layer,” IEEE Trans. Electron Devices, vol. 64, no. 9, pp. 3661–3667,
the depth of potential well, and hence, 2DEG increases. Sep. 2017, doi: 10.1109/TED.2017.2721437.
[6] P.-N. Ni, C.-X. Shan, B.-H. Li, and D.-Z. Shen, “High Mg-content
Fig. 3(b) shows the charge distribution profile of 2DEG wurtzite MgZnO alloys and their application in deep-ultraviolet light-
across MgZnO/ZnO heterostructure for a 30-nm barrier layer emitters pumped by accelerated electrons,” Appl. Phys. Lett., vol. 104,
thickness over buffer layer ZnO for without interface charges no. 3, p. 32107, 2014, doi: 10.1063/1.4862789.
[7] K. Xu, “Integrated silicon directly modulated light source using p-
and with interface charges. The peak in electron carrier con- well in standard CMOS technology,” IEEE Sensors J., vol. 16, no. 16,
centration at 30 nm from surface, in Fig. 3(b), shows the con- pp. 6184–6191, Aug. 2016, doi: 10.1109/JSEN.2016.2582840.
finement of 2-D electrons at the MgZnO/ZnO heterointerface. [8] L.-Y. Kim and O.-K. Kwon, “A design method for flicker-free liquid
crystal display panels based on indium-gallium-zinc-oxide thin-film
It can be noticed that the electron concentration in barrier transistors,” J. Soc. Inf. Display, vol. 25, no. 9, pp. 544–553, 2017,
layer is significantly lower than the bulk electron concentration doi: 10.1002/jsid.604.
(in buffer layer), suggesting the depletion of barrier electrons [9] H. Tampo et al., “Polarization-induced two-dimensional electron gases
in ZnMgO/ZnO heterostructures,” Appl. Phys. Lett., vol. 93, no. 20,
in the formation of 2DEG. pp. 202104-1–202104-4, Nov. 2008, doi: 10.1063/1.3028338.
[10] H.-A. Chin et al., “Two dimensional electron gases in polycrystalline
IV. C ONCLUSION MgZnO/ZnO heterostructures grown by RF-sputtering process,” J. Appl.
Phys., vol. 108, no. 5, p. 054503, 2010, doi: 10.1063/1.3475500.
The epitaxially grown MgZnO/ZnO heterostructures show [11] H.-A. Chin et al., “Electrical properties of modulation-doped RF-
a non-negligible role of acceptor states toward the formation sputtered polycrystalline MgZnO/ZnO heterostructures,” J. Phys. D,
Appl. Phys., vol. 44, no. 45, pp. 455101-1–455101-4, Nov. 2011, doi: 10.
of 2DEG density. The surface donor density shows a causal- 1088/0022-3727/44/45/455101.
effect relationship in presence of acceptor states. In contrast, [12] H. Tampo et al., “High electron mobility Zn polar ZnMgO/ZnO het-
the 2DEG density in sputtered MgZnO/ZnO samples is an erostructures grown by molecular beam epitaxy,” J. Cryst. Growth,
vols. 301–302, pp. 358–361, Apr. 2007, doi: 10.1016/j.jcrysgro.
order high, and exceeds the barrier-buffer polarization differ- 2006.11.169.
ence limit. This is due to the interface states present at the [13] J. P. Ibbetson, P. T. Fini, K. D. Ness, S. P. DenBaars, J. S. Speck, and
MgZnO/ZnO interface due to sputtering growth technique and U. K. Mishra, “Polarization effects, surface states, and the source of
electrons in AlGaN/GaN heterostructure field effect transistors,” Appl.
leads to formation of interface trap charges which contribute Phys. Lett., vol. 77, no. 2, pp. 250–252, 2000, doi: 10.1063/1.126940.
toward enhancement of 2DEG density. This paper improves [14] G. Koley and M. G. Spencer, “On the origin of the two-dimensional
the understanding of underlying physical mechanisms and electron gas at the AlGaN/GaN heterostructure interface,” Appl. Phys.
contributing factors in the formation of 2DEG in MgZnO/ZnO Lett., vol. 86, no. 4, p. 042107, 2005, doi: 10.1063/1.1850600.
[15] S.-H. Park and D. Ahn, “Spontaneous and piezoelectric polarization
heterostructure fabricated by different growth mechanisms. effects in wurtzite ZnO/MgZnO quantum well lasers,” Appl. Phys. Lett.,
vol. 87, no. 25, p. 253509, Dec. 2005, doi: 10.1063/1.2149294.
[16] S.-H. Jang and S. F. Chichibu, “Structural, elastic, and polarization
ACKNOWLEDGMENT parameters and band structures of wurtzite ZnO and MgO,” J. Appl.
M. A. Khan would like to thank the Ministry of Phys., vol. 112, no. 7, p. 073503, 2012, doi: 10.1063/1.4757023.
Electronics and Information Technology (MeitY), Government [17] M. Yano et al., “Polarization-induced two-dimensional electron gas at
Zn1−x Mgx O/ZnO heterointerface,” J. Cryst. Growth, vols. 301–302,
of India, for providing fellowship grant under Visvesvaraya pp. 353–357, Apr. 2007, doi: 10.1016/j.jcrysgro.2006.11.114.
PhD scheme for Electronics and IT. Prof. S. Mukherjee [18] J. D. Ye et al., “Two-dimensional electron gas in Zn-polar ZnMgO/ZnO
would like to thank MeitY, Government of India for Young heterostructure grown by metal-organic vapor phase epitaxy,” Appl.
Phys. Lett., vol. 97, no. 11, pp. 111908-1–111908-3, Sep. 2010,
Faculty Research Fellowship (YFRF) under Visvesvaraya PhD doi: 10.1063/1.3489101.
Scheme for Electronics and IT. This publication is an outcome [19] ATLAS User Manual. Santa Clara, CA: Silvaco International, 2010.

Authorized licensed use limited to: Pt Dwarka Prasad Mishra Indian Inst of Info Tech. Downloaded on January 21,2021 at 05:18:42 UTC from IEEE Xplore. Restrictions apply.
 2854 IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 65, NO. 7, JULY 2018

[20] W. B. Joyce and R. W. Dixon, “Analytic approximations for the Fermi [30] S.-T. Lien, H.-C. Li, Y.-J. Yang, C.-C. Hsu, I.-C. Cheng, and J.-Z. Chen,
energy of an ideal Fermi gas,” Appl. Phys. Lett., vol. 31, pp. 354–356, “Atmospheric pressure plasma jet annealed ZnO films for MgZnO/ZnO
Jun. 1977, doi: 10.1063/1.89697. heterojunctions,” J. Phys. D, Appl. Phys., vol. 46, no. 7, p. 075202, 2013,
[21] W. Shockley and W. T. Read, Jr., “Statistics of the recombinations of doi: 10.1088/0022-3727/46/7/075202.
holes and electrons,” Phys. Rev., vol. 87, no. 5, pp. 835–842, 1952, [31] X. Bai, L. Wang, R. Zong, Y. Lv, Y. Sun, and Y. Zhu, “Performance
doi: 10.1103/PhysRev.87.835. enhancement of ZnO photocatalyst via synergic effect of surface oxy-
[22] R. N. Hall, “Electron-hole recombination in germanium,” Phys. Rev., gen defect and graphene hybridization,” Langmuir, vol. 29, no. 9,
vol. 87, no. 2, p. 387, 1952, doi: 10.1103/PhysRev.87.387. pp. 3097–3105, 2013, doi: 10.1021/la4001768.
[23] K. Xu, “Electro-optical modulation processes in Si-PMOSFET LEDs [32] E. V. Lavrov, J. Weber, F. Börrnert, C. G. Van de Walle, and R. Helbig,
operating in the avalanche light emission mode,” IEEE Trans. Electron “Hydrogen-related defects in ZnO studied by infrared absorption spec-
Devices, vol. 61, no. 6, pp. 2085–2092, Jun. 2014, doi: 10.1109/ troscopy,” Phys. Rev. B, Condens. Matter, vol. 66, no. 16, p. 165205,
TED.2014.2318277. 2002, doi: 10.1103/PhysRevB.66.165205.
[24] Y. Taur and T. H. Ning, “Basic device physics,” in Fundamentals of [33] J. Fallert et al., “Surface-state related luminescence in ZnO nanocrys-
Modern VLSI Devices, 2nd ed. Cambridge, U.K.: Cambridge Univ. tals,” J. Appl. Phys., vol. 101, no. 7, p. 073506, 2007, doi: 10.1063/1.
Press, 2015, pp. 11–34. 2718290.
[25] Y. Nishimoto et al., “Plasma-assisted molecular beam epitaxy of high [34] A. Janotti and C. G. Van de Walle, “Fundamentals of zinc oxide as
optical quality MgZnO films on Zn-polar ZnO substrates,” Appl. Phys. a semiconductor,” Rep. Prog. Phys., vol. 72, no. 12, p. 126501, 2009,
Exp., vol. 1, no. 9, p. 091202, 2008, doi: 10.1143/APEX.1.091202. doi: 10.1088/0034-4885/72/12/126501.
[26] H. Tampo et al., “Band profiles of ZnMgO/ZnO heterostructures con- [35] M. Delagebeaudeuf and N. T. Linh, “Metal-(n) AlGaAs-GaAs two-
firmed by Kelvin probe force microscopy,” Appl. Phys. Lett., vol. 94, dimensional electron gas FET,” IEEE Trans. Electron Devices,
no. 24, p. 242107, Jun. 2009, doi: 10.1063/1.3157149. vol. TED-29, no. 6, pp. 955–960, Jun. 1982, doi: 10.1109/T-ED.1982.
[27] C. H. Park, S. B. Zhang, and S.-H. Wei, “Origin of p-type doping diffi- 20813.
culty in ZnO: The impurity perspective,” Phys. Rev. B, Condens. Matter,
vol. 66, no. 7, p. 073202, 2002, doi: 10.1103/PhysRevB.66.073202.
[28] D. J. Leary, J. O. Barnes, and A. G. Jordan, “Calculation of carrier
concentration in polycrystalline films as a function of surface acceptor
state density: Application for ZnO gas sensors,” J. Electrochem. Soc.,
vol. 129, no. 6, p. 1382–1386, 1982, doi: 10.1149/1.2124169.
[29] H. von Wenckstern et al., “Deep acceptor states in ZnO single crystals,”
Appl. Phys. Lett., vol. 89, no. 9, p. 092122, 2006, doi: 10.1063/1. Authors’ photographs and biographies not available at the time of
2335798. publication.

Authorized licensed use limited to: Pt Dwarka Prasad Mishra Indian Inst of Info Tech. Downloaded on January 21,2021 at 05:18:42 UTC from IEEE Xplore. Restrictions apply.