This content has been downloaded from IOPscience. Please scroll down to see the full text.
Download details:
IP Address: 140.113.38.11
This content was downloaded on 25/12/2014 at 03:14
Please note that terms and conditions apply.
Simulation study on electrical characteristic of AlGaN/GaN high electron mobility transistors
with AlN spacer layer
View the table of contents for this issue, or go to the journal homepage for more 2014 Jpn. J. Appl. Phys. 53 04EF08
(http://iopscience.iop.org/1347-4065/53/4S/04EF08)
Niraj Man Shrestha1, Yiming Li2*, and Edward Yi Chang1*
1Compound Semiconductor Device Laboratory, Department of Material Science and Engineering,
National Chiao Tung University, Hsinchu 300, Taiwan
2Parallel and Scientific Computing Laboratory, Department of Electrical and Computer Engineering,
National Chiao Tung University, Hsinchu 300, Taiwan E-mail: ymli@faculty.nctu.edu.tw; edc@mail.nctu.edu.tw
Received September 24, 2013; revised November 20, 2013; accepted November 25, 2013; published online February 24, 2014
Two-dimensional electron gas (2DEG) property is crucial for the performance of GaN-based high electron mobility transistors (HEMTs). The 2DEG-related concentration and mobility can be improved as device’s performance booster. Electrical characteristics of AlGaN/AlN/GaN HEMT are numerically simulated and compared with conventional AlGaN/GaN HEMT. The main findings of this study indicate that 2DEG’s concentration level is increased when a spacer layer of AlN in the interface of AlGaN/AlN/GaN is inserted owing to large conduction band off set, high polarizationfield, and high barrier. Notably, when a thin spacer layer of AlN is introduced, the 2DEG’s distribution virtually shifts away from the interface which reduces the interface scattering. The scattering appearing in conventional AlGaN/GaN HEMT includes alloy and interface roughness scatterings. They are reduced in AlGaN/AlN/GaN HEMT due to binary nature of AlN material. A critical thickness of spacer layer for mobility is 0.5 nm and the maximum drain current and transconductance (Gm) are at 1.5 and 1.2 nm thickness of AlN spacer layer. Increasing
thickness of AlN spacer layer deteriorates the ohmic resistance of source/drain contact and hence degrades the performance of device beyond 1.5-nm-thick AlN spacer layer. ©2014 The Japan Society of Applied Physics
1. Introduction
GaN-based high electron mobility transistors (HEMTs) have attracted a great deal of attention for high-frequency and high-power microwave applications because nitride-based material systems have desirable fundamental physical
proper-ties, such as a large band gap (3.4 eV), high breakdownfield
(about 3© 106V/cm) and strong spontaneous and
piezo-electric polarization fields. A unique feature of GaN-based
HEMTs is high sheet carrier concentration (1© 1013cm¹2), which is achieved in the channel not only due to large band gap discontinuty1) at interface but also due to piezoelectric
and spontaneous polarization effect without intentionally
doping barrier layer.2–4) Because of these advantageous
material properties, GaN and its related alloys have been used in a variety of light-emitting devices and electron devices on a commercial basis. Nevertheless, more techno-logical innovations are needed to realize high-efficiency, highly reliable and low-cost devices. High power density is one of key parameters for diverse power devices. In order to increase the power density, the product of electron
concen-tration and electron mobility should be maximized.5,6) To
improve device’s performance, various barrier and channel
alternatives have been studied in nitride-based HEMTs.7,8)
Carrier’s concentration level can be increased by increasing
Al mole fraction and AlGaN barrier layer thickness.9,10)
However, when the Al content of AlGaN layer increases, surface quality of AlGaN layer is degraded and then reduces
the two-dimentional electron gas (2DEG) mobility.9,11) The
existence of the different scatterings, such as acoustic and optical phonons scattering,12,13) ionized impurity scattering,
interface roughness scattering,14) dislocation scattering, and
alloy disorder scattering15–18) play major role to limit the mobility of 2DEG in conventional AlGaN/GaN HEMT. These scatterings are unique to GaN-based HEMTs due to large dislocation density and strong polarization effects.
An additional thin AlN spacer between AlGaN and GaN
may improve the 2DEG concentration1) and the mobility at
low temperatures.18)It is found that, the carriers’ mobility in
AlGaN/AlN/GaN HEMTs is comparatively higher than that of the conventional device without AlN spacer layer.18,19)A
thin AlN spacer layer in AlGaN/GaN interface causes the reduction of alloy scattering which plays a dominant factor in the carriers’ mobility. Another effect of AlN spacer layer insertion is the reduction in the forward Schottky gate current, which makes it possible to apply a high gate voltage
in the transistor operation.20) The drain current of AlGaN/
AlN/GaN HEMTs is better than that of conventional AlGaN/
GaN HEMTs.21–23) Therefore, the AlN spacer layer is one
of the fascinating ways to achieve high speed, high power
switching device with low specific on resistance.
Addition-ally, the thickness of AlN is important for the mobility.19)
When the thickness is changed, the polarization field and
the conduction band offset are changed which also affects the 2DEG-related concentration and mobility. Therefore, to know the potential thickness of spacer layer for high power device and to know the real cause of reduction of mobility at higher spacer layer thickness will be interesting studies.
In this study, the effect of AlN spacer layer thickness on electrical characteristic in AlGaN/AlN/GaN HEMT is numerically studied. The most potential spacer layer thick-ness for power device is discussed. In addition, effect of the AlN layer thickness on interface roughness scattering will be highlighted. This paper is organized as follows. In Sect. 2, we introduce the device structure and simulation settings used for studying the effect of AlN spacer layer thickness on AlGaN/AlN/GaN HEMT. In Sect. 3, we report our result and examine the transport property as determined by the thickness of AlN spacer layer. Finally, we draw our con-clusions and suggest future works in Sect. 4.
2. Simulation methodology
The epitaxial layer grown step and fabrication step is shown in Figs. 1(a)–1(f ). The physical device studied in this paper is grown on sapphire substrates by using metal organic chemical vapor deposition system at different temperatures.
Entire growth process is conducted by using ammonia (NH3),
trimethylgallium (TMGa), trimethylaluminum (TMAl) as precursors and hydrogen as carrier gas. Following an AlN nucleation layer, a 2-µm-thick undoped GaN buffer is grown. A thin AlN spacer layer of varying thickness (0–2 nm) and a 25-nm-thick undoped AlGaN barrier layer are grown on top. The entire epitaxial layer is grown at high temperature. The Al mole fraction determined by using X-ray diffraction is around 30%. Figure 2 shows a scanning electron microscope (SEM) image of cross section of the fabricated AlGaN/AlN/ GaN HEMT structure.
The simulated device structures of AlGaN/GaN HEMT with and without the AlN spacer layer are shown in Figs. 3(a) and 3(b), where the parameters of simulated devices are tabulated in Table I. Figure 3(a) is the device is the conventional structure without the spacer layer and Fig. 3(b) with the AlN spacer layer and both structures use the same device parameters except the thickness of AlN spacer layer. As shown in Fig. 3(b), the thickness of AlN spacer layer varies from 0 to 2 nm while keeping the total thickness of AlN and AlGaN layers at 25 nm. Notably, an insertion of thin AlN layer between AlGaN and GaN can enhance the confinement of 2DEG. Therefore, improvement of electron transport characteristic could be estimated. The spacer layer affects the transport carrier which is related to 2DEG’s concentration and mobility. The band gaps of the relevant binary compounds are computed, by Eqs. (1) and (2), as a function of temperature T: EgðGaNÞ ¼ 3:507 0:909 10 3T2 T þ 830 ; ð1Þ EgðAlNÞ ¼ 6:23 1:799 103T2 T þ 1462 : ð2Þ
The band gap of ternary compound depends on composition fraction x is given by
EgðAlxGa1xNÞ
¼ EgðAlNÞ þ EgðGaNÞð1 xÞ 1:3xð1 xÞ: ð3Þ
Polarization is one of key mechanisms in controlling the
formation of the 2DEG at the AlGaN/GaN interface. PPZof
Eq. (4) is piezoelectric polarization which is related to strain tensor: PPZ¼ 2as a0 a0 E31 C13 C33E33 ; ð4Þ
where E31 and E33 are piezoelectric constants, and C13 and
C33 are elastic constants. When an AlxGa1¹xN grown on
GaN, the piezoelectric polarization is expressed as
PPZðxÞ ¼ ð3:2x 1=9x2Þ106cm2: ð5Þ
In addition to strain induced polarization, cations and anions are spontaneously displaced with respect to each other, producing spontaneous polarization. Then total polarization is given by
Ptot¼ PPZþ PSP: ð6Þ
The Schottky barrier height estimated is 1.4 V. A 2D quantum mechanically corrected device simulation is per-formed, where the quantum mechanical correction is considered with a density-gradient approach. Performance
of device mainly depends on the carriers’ mobility which is
affected by various types of scattering mechanism; such as, the phonon scattering (at room temperature), the alloy disorder scattering (potential disorder from ternary alloy, at low and room temperatures), the surface roughness scattering (at low temperature), the ionized impurities scattering, the dislocation scattering, and the dipole scattering. Among these scattering mechanisms, the alloy disorder scattering is the
limiting factor14) at low and room temperatures when the
carrier’s concentration is high. Binary nature of AlN makes comparatively low alloy disorder scattering. Notably, the
Fig. 1. (Color online) Epilayer grown and processflow of the studied AlGaN/AlN/GaN HEMT.
GaN AlGaN
AlN
1 μm
Fig. 2. (Color online) A SEM image of the AlGaN/GaN structure with the AlN spacer layer.
(a) (b)
Fig. 3. (Color online) Schematic plots of the AlGaN/GaN HEMT. (a) Conventional device is without the AlN spacer layer and (b) the explored device which is with the AlN spacer layer.
Table I. Parameters used for the simulated devices.
Gate width (µm) 15
AlGaN layer thickness (nm) 25
GaN layer thickness (µm) 2
Al composition in AlGaN (%) 30
Gate length (µm) 1
AlN thickness (nm) 0, 0.3, 0.5, 1, 1.2, 1.5, 2
Jpn. J. Appl. Phys. 53, 04EF08 (2014) N. M. Shrestha et al.
one is the lowfield mobility model and the other is the nitride
specific high field dependent mobility model. The first low
field mobility model consists of the interface roughness scattering and alloy scattering. The expression for the interface roughness scattering is given by24)
1 I
0ðND; TÞ
¼ aðND=1017Þ lnð1 þ 2cwÞ=ðT=300Þ1:5
þ bðT=300Þ1:5þ c=expð=T 1Þ; ð7Þ
where ¼ h!LO=kB is related to the phonon scattering
resulting from the scattering of interface roughness, 2cw¼
3ðT=300Þ2ðN
D=1017Þ2=3, ND (cm¹3) is ionized donor
con-centration, T (K) is absolute temperature. Notably, the interface roughness scattering is one of the most important factors to determine the longitudinal optical phonons; in particular, for binary compound quantum wells where the
contribution from the alloy scattering is absent.25,26) The
expression for the alloy scattering, is given by27)
A0ðND; TÞ ¼ min ðT=300Þ1
þ ðmax minÞ ðT=300Þ2
1 þ ½ND=1017 ð300=TÞ3ðT=300Þ4
; ð8Þ
where min and max are the minimum and maximum
mobility for the materials.
Therefore, the nitride specific high field dependent mobility model is27)
ðEÞ ¼
A
0ðND; TÞI0ðND; TÞ=½A0ðND; TÞ þ I0ðND; TÞ þ vsatðEn11=En01Þ
1 þ ðE=E0Þn2þ ðE=E0Þn1 ; ð9Þ
where I0ðND; TÞ and A0ðND; TÞ are calculated from Eqs. (7)
and (8), vsat is the saturation velocity, E is electric field, and the other parameters used in Eqs. (7)–(9) are tabulated in Tables II and III. The current characteristics of device can be evaluated through Schrödinger equation into a self-consistent computation with the Poisson equation.28–31)
3. Results and discussion
Simulation study of electron transport and DC characteristic of the purposed AlGaN/AlN/GaN HEMT is performed in this section. We compare the results with that of conventional AlGaN/GaN HEMT to show the advantage of the spacer layer of AlN in the interface of the AlGaN/AlN/GaN of the explored device. In Sect. 3.1, we show and discuss the physical characteristic of the device. In Sect. 3.2, we report the electrical characteristics including the transport current and the transconductance.
3.1 Physical characteristic
Figure 4 illustrates the patterns of conduction band profile of
the proposed AlGaN/AlN/GaN HEMT and the conventional
AlGaN/GaN device. A thin AlN spacer layer in the hetero-interface of AlGaN/GaN produces large effective conduction band offset because of the potential drops across the spacer layer due to large piezoelectric and spontaneous polarization
field.32) The increase in the conduction band offset when
using the thin AlN spacer layer inside the structure of AlGaN/GaN is explained by EAlN c EcAlGaN¼ exp AlN N2D "AlN dAlN; ð10Þ
whereEAlNc andEAlGaNc are the effective conduction band offsets between the AlGaN/AlN/GaN and the AlGaN/GaN,
respectively; N2D is the sheet carrier concentration of the
AlGaN/AlN/GaN structure, ¾AlN is dielectric constant,·AlN
is the polarization induced charge at the AlN/GaN inferface, and dAlNis the thickness of AlGaN.10)Effect of the thickness
of AlN spacer layer on the conduction band offset was analyzed by means of device simulation and results are plotted as shown in Fig. 5(a). The conduction band offset becomes large when the thickness of AlN spacer layer increases. Numerically calculated values of the quantum well depth with respect to the thickness of AlN spacer layer is tabulated in Table IV. Figure 5(b) shows the quantum well depth for the device with respect to different thickness of AlN
c (V s¹1cm¹2) 1.7© 10¹2 Vsat(cm/s) 1.9064© 107 ¡ 6.12 n1 7.2 n2 0.786 E0(kV/cm) 220.89
Table III. The adopted parameters in Eq. (8).27)
¢1 ¢2 ¢3 ¢4 £
GaN ¹1.02 ¹3.84 3.02 0.81 0.66
AlGaN ¹1.33 ¹1.75 6.02 1.44 0.29
AlN ¹1.82 ¹3.43 3.78 0.86 1.16
Fig. 4. (Color online) Illustration of the pattern of band diagram for the (a) AlGaN/GaN and (b) AlGaN/AlN/GaN HEMTs.
spacer layer. The results show that 2DEG’s depth has 50% increases when the AlN spacr layer is 1.5 nm thick in the
AlGaN/AlN/GaN HEMT.
2DEG, created at the hetero-interface on growing high band gap material Al0.3Ga0.7N on comparatively small band
gap material GaN, is crucial mechanism to confine electrons
in quantum well which leads to high electron concentration
and high mobility. It is found that large polarization fields
in the AlN barrier layer in the AlN/GaN
heterostruc-ture results in high values of the 2DEG sheet density. The 2DEG’s concentration in the AlGaN/AlN/GaN structure is given by10) N2D¼ AlNþ "AlN "AlGaN dAlGaN dAlN AlGaN1 e "AlN dAlN
½Efþ ’sþ EcðAlGaN=AlNÞ EcðAlN=GaNÞ
1 þ "AlN "AlGaN dAlGaN dAlN ; ð11Þ
where·AlGaNis the polarization charge at AlGaN surface,¤s
is the depth of the Fermi level at the AlGaN surface with respect to its conduction band edge;¾AlGaNand dAlGaNare the
dielectric constant and the thickness of AlGaN, respectively.
High level of 2DEG’s concentration is owing to the large
conduction band offset and reduced wave propagation into
the AlGaN layer33)caused by large band gap of the inserted
AlN spacer layer. Device simulation results listed in Table IV and plotted in Fig. 6(a) show dependency of the conduction band offset on the thickness of AlN spacer layer improves the level of 2DEG’s concentration. As shown in Fig. 6(a), the level of 2DEG’s concentration increases with respect to the thickness of the AlN spacer layer. The level of 2DEG’s concentration is improved by 63% when the thickness of the AlN spacer layer is 1.5 nm thick in the AlGaN/AlN/GaN
interface. The electron concentration distribution for the device with respect to different thickness of the AlN spacer
layer at Vg= 0 V, plotted in Fig. 6(b), shows the 2DEG
distribution. Using the simulated data of electron density,
the averaged position of 2DEG’s distribution is further
calculated by hxi ¼ Z n x dx Z n dx ; ð12Þ
where n is the 2DEG’s concentration and x is respective position.
Figure 7 shows the average position of 2DEG distribution. The results show that the averaged position of 2DEG’s
AlN Spacer Layer Thickness (nm)
0.0 0.5 1.0 1.5 2.0 2.5
Quantum W
ell Depth (eV)
0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.17 (a) Distance (μm) 0.295 0.300 0.305 0.310 Quantum W
ell Depth (eV)
-0.15 -0.10 -0.05 0.00 0.05 0.10 (b)
Fig. 5. (Color online) (a) Variation of the quantum well depth for the device at zero bias with respect to the different thickness of the AlN spacer layer. (b) Plot of the quantum well depth for the device at zero bias varying with the thickness of the AlN spacer layer.
Table IV. Simulated results of transport and DC characteristics for the device with respect to different thickness of AlN spacer layer. AlN spacer layer
thickness (nm) 2DEG depth (eV) Electron concentration (©1019cm¹3) Mobility
(cm2v1s¹1) Drain current(mA/mm)
Maximum transconductance (mS/mm) 0.0 0.095 0.89 1495 417 170 0.3 0.125 1.02 1673 451 180 0.5 0.128 1.11 1675 473 188 1.0 0.1346 1.31 1670 518 194 1.2 0.1378 1.37 1665 533 195 1.5 0.14213 1.45 1653 551 194 2.0 0.14614 1.58 1637 550 163
Jpn. J. Appl. Phys. 53, 04EF08 (2014) N. M. Shrestha et al.
concentration shift away from the interface on inserting AlN spacer layer, which indicates that the interface scattering is reduced. It has been known that the interface roughness is the main scattering mechanism when the level of 2DEG’s concentration closes to the hetero-interface with the increase
of the level of 2DEG’s concentration.34) As summarized in
Table V, the average position of 2DEG’s distribution shift
away from interface attains on inserting thin AlN spacer layer in AlGaN/GaN interface which implies that the low
interface roughness scattering on the AlGaN/AlN/GaN
HEMTs in comparison with the conventional HEMTs. For
the first time, the averaged position of 2DEG’s distribution
with respect to different thickness of the AlN spacer layer is calculated. Distance of averaged position of 2DEG’s distribution attains its maximum when AlN thickness is 0.5 nm. Beyond this thickness, sufficiently high electron concentration leads the 2DEG position back toward the interface. Notably, the interface scattering reaches the minimum when the thickness of the AlN spacer layer is 0.5 nm thick and increase again on further increase the thickness. Therefore, mobility increase on increasing the AlN thickness, become maximum at the critical thickness and decrease beyond this thickness. The device simulation result depending on the electron mobility is affected by the thickness of the AlN spacer layer, as shown in Fig. 8.
Simulation result shows that the 2DEG’s mobility increases
when the thickness increases from 0 to 0.5 nm. Our result shows that the mobility is increased by 12% when the 0.5-nm-thick AlN spacer layer is introduced in the AlGaN/GaN heterointerface. Low interface roughness and low alloy scattering mechanism increase the mobility. However, high carrier concentration at a thicker AlN spacer layer leads to moving 2DEG distribution toward the interface. It results in increase of the interface roughness scattering and coulomb scattering in 2DEG. Consequently, the mobility is decreased beyond 0.5-nm-thick AlN space layer.
AlN Spacer Layer Thickness (nm)
0.0 0.5 1.0 1.5 2.0 2.5 Electron Concentration (x10 0.0 0.2 0.4 0.6 0.8 1.0 (a) Distance (μm) 0.520 0.525 0.530 0.535 0.540 Electron Concentration (/cm 3 ) 0.0 2.0e+18 4.0e+18 6.0e+18 8.0e+18 1.0e+19 1.2e+19 1.4e+19 1.6e+19 1.8e+19 (b) 2 nm 0 nm Vg= 0 V Vd= 10 V
Fig. 6. (Color online) (a) Electron concentration for the device with different thickness of the AlN spacer layer at Vg= 0 V. (b) Electron
concentration distribution for the device with respect to the thickness of the AlN spacer layer at Vg= 0 V.
0.0 0.5227 0.3 0.5245 1.8 0.5 0.5261 3.4 1.0 0.5251 2.4 1.2 0.5251 2.4 1.5 0.5251 2.4 2.0 0.5250 2.3
AlN Spacer Layer Thickness (nm)
0.0 0.5 1.0 1.5 2.0 2.5
A
v
eraged 2DEG Position (
μ m) 0.5225 0.5230 0.5235 0.5240 0.5245 0.5250 0.5255 0.5260 0.5265 Vg= 0 V Vd= 10 V ∫ ∫ ndx ndx x. Averaged 2DEG position 〈x〉 =
Fig. 7. (Color online) Averaged position of 2DEG’s distribution calculated by using Eq. (11) with respect to the thickness of the AlN spacer layer.
AlN Spacer Layer Thickness (nm)
0.0 0.5 1.0 1.5 2.0 2.5 Mobility (cm 2/Vs) 1480 1500 1520 1540 1560 1580 1600 1620 1640 1660 1680 1700 Vg= 0 V Vd= 10 V
Fig. 8. (Color online) Plot of the variation of mobility versus the thickness of the AlN spacer layer.
3.2 DC characteristics
Note that the AlGaN/AlN/GaN HEMTs show better DC
characteristic in comparison with the conventional AlGaN/
GaN devices21,35) owing to excellent 2DEG property
appearing in the AlGaN/AlN/GaN structures. Figure 9
shows that the simulated Ids–Vds curves for the AlGaN/
AlN/GaN HEMT varing with the thickess of the AlN spacer layer at zero gate bias. We know that drain current is the result of the combined effect of mobility and electron concentration. Result shows that the drain current increases up to 0.5 nm spacer layer thickness due to increase of both electron concentration and mobility with spacer layer thickness. On increasing the spacer layer thickness from 0.5 to 1.5 nm, the drain current enhances due to the carrier concentration. Beyond the thickness of 1.5 nm, the drain
current is lowered owing to decrease in the carrier’s
mobility. Our result shows that the drain current increases by 32% on using 1.5-nm-thick AlN spacer layer. Corre-spondingly, Fig. 10(a) shows the distribution of the trans-conductance with respect to the gate voltage. Figure 10(b) is the variation of the maximum transconductance with
respect to the thickness of the AlN spacer layer.
Figure 10(b) shows that the transconductance increases on the insertion of the AlN spacer layer in the AlGaN/ GaN interface; it increases with the thickness is increased to 1.2 nm thickness of the AlN spacer layer. It slightly decreases from 1.2 to 1.5 nm thickness of the AlN spacer layer and rapidly decreases beyond this thickness. The maximal transconductance of 195 mS/mm is observed at the 1.2-nm-thick AlN space layer which is 15% increase,
compared with the conventional the AlGaN/GaN HEMT.
Beyond the 1.5-nm-thick AlN spacer layer, there is sizeable degradation of ohmic contact resistance because an AlN spacer layer with an extremely wide band gap decreases
the contact resistance significantly.20) The extremely high
ohmic resistance in drain and source contacts degrade the device performance.
4. Conclusions
For the GaN HEMT with AlN spacer layer, the engineering findings of this study indicate that the 2DEG transport properties and DC characteristic of the AlGaN/AlN/GaN
HEMTs are strongly affected by the thickness of the AlN spacer layer. The electron concentration significantly in-creases on the insertion of AlN spacer layer and is dependent on the thickness of the AlN spacer layer. The level of 2DEG’s concentration is virtually moved away from the interface on inserting the AlN space layer. Distance between the 2DEG concentration and the interface attains its maximum at 0.5-nm-thick AlN space layer. This exhibits that the lowest interface scattering is observed at 0.5-nm-thick AlN space layer. The carrier mobility increases as the thickness of the AlN spacer layer and reaches its maximum at the thickness of 0.5 nm and decreases on further increase of thickness. Furthermore, drain current increases with the thickness is increased to 1.5 nm. Beyond this thickness, the drain current decreases owing to decrease in mobility. As the thickness of the AlN spacer layer is thicker than a critical thickness, the source and drain ohmic contacts are degraded and the drain current decreases. The maximum trans-conductance is observed at 1.2-nm-thick AlN spacer layer. High ohmic resistance makes lower transconductance
beyond 1.5-nm-thick AlN space layer and finally degrades
the device performance. Notably, 1.5-nm-thick AlN spacer layer can be used for high speed, high power switching device.
Drain Voltage (V)
0 2 4 6 8 10 12
Drain Current (mA/mm)
-100 0 100 200 300 400 500 600 Vg= 0 V 2 nm 1.5 nm 1.2 nm 1 nm 0.5 nm 0.3 nm No spacer
Fig. 9. (Color online) Id–Vdcurves for the device with respect to the
different thickness of the AlN spacer layer (they are 0.0, 0.3, 0.5, 1.0, 1.2, 1.5, and 2.0 nm, respectively). Gate Voltage (V) -7 -6 -5 -4 -3 -2 -1 0 T ransconductance (mS/mm) 0 50 100 150 200 250 (a) Vd= 2 V 2 nm 1.5nm 1.2nm 1nm 0.5nm 0.3nm No spacer
AlN Spacer Layer Thickness (nm)
0.0 0.5 1.0 1.5 2.0 2.5 The Maximum T ransconductance (mS/mm) 140 150 160 170 180 190 200 (b) Vd= 2 V
Fig. 10. (Color online) Plots of the (a) distribution of transconductance with respect to the gate voltage and (b) variation of transconductance with respect to the spacer layer thickness.
Jpn. J. Appl. Phys. 53, 04EF08 (2014) N. M. Shrestha et al.
Contracts Nos. NSC 2221-E-009-161 and NSC 102-2911-I-009-302. Mr. N. M. Shrestha would like to thank Mr. T.-T. Luong for the sample grown.
1) L. Guo, X. Wang, C. Wang, H. Xiao, J. Ran, W. Luo, X. Wang, B. Wang, C. Fang, and G. Hu,Microelectron. J.39, 777 (2008).
2) A. Kranti, Rashmi, S. Haldar, and R. S. Gupta,Solid-State Electron.46, 1333 (2002).
3) E. T. Yu, G. J. Sullivan, P. M. Asbeck, C. D. Wang, D. Qiao, and S. S. Lau,
Appl. Phys. Lett.71, 2794 (1997).
4) L. Hsu and W. Walukiewicz,J. Appl. Phys.89, 1783 (2001).
5) Y.-F. Wu, D. Kapolnek, J. P. Ibbetson, P. Parikh, B. P. Keller, and U. K. Mishra,IEEE Trans. Electron Devices48, 586 (2001).
6) Y.-F. Wu, B. P. Keller, P. Fini, S. Keller, T. J. Jenkins, L. T. Kehias, S. P. DenBaars, and U. K. Mishra,IEEE Electron Device Lett.19, 50 (1998).
7) J. Xie, J. H. Leach, X. Ni, M. Wu, R. Shimada, Ü. Özgür, and H. Morkoç,
Appl. Phys. Lett.91, 262102 (2007).
8) M. Miyoshi, T. Egawa, H. Ishikawa, K.-I. Asai, T. Shibata, M. Tanaka, and O. Oda,J. Appl. Phys.98, 063713 (2005).
9) O. Ambacher, J. Smart, J. R. Shealy, N. G. Weimann, K. Chu, M. Murphy, W. J. Schaff, L. F. Eastman, R. Dimitrov, L. Wittmer, M. Stutzmann, W. Rieger, and J. Hilsenbeck,J. Appl. Phys.85, 3222 (1999).
10) T. R. Lenka and A. K. Panda, Indian J. Pure Appl. Phys.49, 416 (2011). 11) I. P. Smorchkova, C. R. Elsass, J. P. Ibbetson, R. Vetury, B. Heying, P. Fini,
E. Haus, S. P. DenBaars, J. S. Speck, and U. K. Mishra,J. Appl. Phys.86, 4520 (1999).
12) R. Gaska, J. W. Yang, A. Osinsky, Q. Chen, M. Asif Khan, A. O. Orlov, G. L. Snider, and M. S. Shur,Appl. Phys. Lett.72, 707 (1998).
13) L. Wang, W. Hu, X. Chen, and W. Lu,J. Electron. Mater.41, 2130 (2012).
14) Y. Cao and D. Jena,Appl. Phys. Lett.90, 182112 (2007).
15) E. Bellotti, F. Bertazzi, and M. Goano,J. Appl. Phys.101, 123706 (2007).
Solidi B228, 617 (2001).
19) I. P. Smorchkova, L. Chen, T. Mates, L. Shen, S. Heikman, B. Moran, S. Keller, S. P. DenBaars, J. S. Speck, and U. K. Mishra,J. Appl. Phys.90, 5196 (2001).
20) T. Nanjo, T. Motoya, A. Imai, Y. Suzuki, K. Shiozawa, M. Suita, T. Oishi, Y. Abe, E. Yagyu, K. Yoshiara, and Y. Tokuda,Jpn. J. Appl. Phys.50, 064101 (2011).
21) T. Ide, M. Shimizu, S. Hara, D.-H. Cho, K. Jeganathan, X.-Q. Shen, H. Okumura, and T. Nemoto,Jpn. J. Appl. Phys.41, 5563 (2002).
22) M. Miyoshi, A. Imanishi, T. Egawa, H. Ishikawa, K. Asai, T. Shibata, M. Tanaka, and O. Oda,Jpn. J. Appl. Phys.44, 6490 (2005).
23) T. Nanjo, M. Suita, T. Oishi, Y. Abe, E. Yagyu, K. Yoshiara, and Y. Tokuda,
Appl. Phys. Express2, 031003 (2009).
24) J. D. Albrecht, R. P. Wang, P. P. Ruden, M. Farahmand, and K. F. Brennan,
J. Appl. Phys.83, 4777 (1998).
25) R. Gupta and B. K. Ridley,Phys. Rev. B48, 11972 (1993).
26) R. Gupta, Turk. J. Phys.23, 551 (1999).
27) M. Farahmand, C. Garetto, E. Bellotti, K. F. Brennan, M. Goano, E. Ghillino, G. Ghione, J. D. Albrecht, and P. P. Ruden,IEEE Trans. Electron Devices48, 535 (2001).
28) K. C. Sahoo, C.-I. Kuo, Y. Li, and E. Y. Chang,IEEE Trans. Electron Devices57, 2594 (2010).
29) Y. Li, S. M. Sze, and T.-S. Chao,Eng. Comput.18, 124 (2002).
30) T.-W. Tang, X. Wang, and Y. Li,J. Comput. Electron.1, 389 (2002).
31) Y. Li and S.-M. Yu,Comput. Phys. Commun.169, 309 (2005).
32) S. Keller, S. Heikman, L. Shen, I. P. Smorchkova, S. P. DenBaars, and U. K. Mishra,Appl. Phys. Lett.80, 4387 (2002).
33) R. S. Balmer, K. P. Hilton, K. J. Nash, M. J. Uren, D. J. Wallis, D. Lee, A. Wells, M. Missous, and T. Martin,Semicond. Sci. Technol.19, L65 (2004).
34) J. Antoszewski, M. Gracey, J. M. Dell, L. Faraone, T. A. Fisher, G. Parish, Y.-F. Wu, and U. K. Mishra,J. Appl. Phys.87, 3900 (2000).
35) L. Yuan, W. Wang, K. B. Lee, H. Sun, S. L. Selvaraj, S. Todd, and G.-Q. Lo, World Acad. Sci., Eng. Technol.69, 127 (2012).