Highly conductive ZnO grown by pulsed laser deposition in pure Ar
Robin C. Scott, Kevin D. Leedy, Burhan Bayraktaroglu, David C. Look, and Yong-Hang Zhang
Citation: Applied Physics Letters 97, 072113 (2010); doi: 10.1063/1.3481372
View online: http://dx.doi.org/10.1063/1.3481372
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/97/7?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Stable highly conductive ZnO via reduction of Zn vacancies
Appl. Phys. Lett. 101, 102101 (2012); 10.1063/1.4748869
Optical and electrical properties of transparent conducting B-doped ZnO thin films prepared by various
deposition methodsa)
J. Vac. Sci. Technol. A 29, 041504 (2011); 10.1116/1.3591348
Ga-doped ZnO grown by pulsed laser deposition in H 2 : The roles of Ga and H
J. Vac. Sci. Technol. A 29, 03A102 (2011); 10.1116/1.3523296
Effects of ZnO buffer layers on the fabrication of GaN films using pulsed laser deposition
J. Appl. Phys. 101, 093519 (2007); 10.1063/1.2730573
Synthesis and characterization of ZnO thin film grown by electron beam evaporation
J. Appl. Phys. 99, 123105 (2006); 10.1063/1.2204333

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
209.147.144.10 On: Fri, 06 Feb 2015 17:16:20

APPLIED PHYSICS LETTERS 97, 072113 共2010兲

Highly conductive ZnO grown by pulsed laser deposition in pure Ar
Robin C. Scott,1 Kevin D. Leedy,2 Burhan Bayraktaroglu,2 David C. Look,3,a兲 and
Yong-Hang Zhang4
1

School of Mechanical, Aerospace, Chemical and Materials Engineering, Arizona State University,
Tempe, Arizona 85287, USA
2
Air Force Research Laboratory, Wright-Patterson AFB, Ohio 45433, USA
3
Semiconductor Research Center, Wright State University, Dayton, Ohio 45435, USA
4
School of Electrical, Computer, and Energy Engineering, Arizona State University, Tempe,
Arizona 85287, USA

共Received 21 July 2010; accepted 29 July 2010; published online 19 August 2010兲
Ga-doped ZnO was deposited by pulsed laser deposition at 200 ° C on SiO2 / Si, Al2O3, or quartz in
10 mTorr of pure Ar. The as-grown, bulk resistivity at 300 K is 1.8⫻ 10−4 ⍀ cm, three-times lower
than that of films deposited at 200 ° C in 10 mTorr of O2 followed by an anneal at 400 ° C in
forming gas. Furthermore, depth uniformity of the electrical properties is much improved. Mobility
analysis shows that this excellent resistivity is mostly due to an increase in donor concentration,
rather than a decrease in acceptor concentration. Optical transmittance is approximately 90% in the
visible and near-IR spectral regions. © 2010 American Institute of Physics.
关doi:10.1063/1.3481372兴
Forming-gas 共FG兲 共5% H2 in Ar兲 annealing of ZnO films
doped with donor impurities such as Al, Ga, or In, has been
shown to increase conductivity1,2 and is often used in preparing transparent conductive oxides 共TCOs兲. The improved
conductivity has often been attributed to the ability of H to
directly provide donors and also to passivate negatively
charged acceptors, including those causing potential barriers
in grain boundaries.3,4 However, as shown recently, the bulk
electrical properties after FG anneals are not uniform with
depth, as evidenced by the thickness dependences of resistivity ␳, mobility ␮, and carrier concentration n.5,6 Furthermore, H is known to be quite mobile, potentially impacting
the reliability of certain devices.7 Thus, it is preferable to
obtain the desired high film conductivity without the assistance of H.
Ga-doped ZnO films are often grown using pulsed laser
deposition 共PLD兲, and generally in an oxygen-rich environment, presumably to ensure stoichiometric films. However,
as the partial pressure of oxygen is increased, the film resistivity increases and electron mobility decreases.8,9 This high
resistivity resulting from excess oxygen is likely related to
共1兲 an increase in compensating point-defect-related acceptors, such as the Zn vacancy VZn and 共2兲 diffusion of O2
共forming O2−兲 into grain boundaries 共GBs兲 and other macroscopic defects. The O2 centers at GBs capture electrons that
can induce high potential barriers; these barriers can further
impede transport by reducing mobility.
In the present work, Ga-doped ZnO was deposited at
200 ° C in Ar instead of O2. There are reasons to believe that
growth in Ar might enhance conductivity by inhibiting the
formation of VZn acceptors in the bulk and O2 acceptors at
grain boundaries. The Ga-doped ZnO films were deposited
using a Neocera Pulsed Energy deposition tool. The target,
consisting a ZnO ceramic sintered with 3 wt % Ga2O3, was
ablated with a 248 nm KrF pulsed laser operating at a frequency of 30 Hz. The energy density at the target surface
was 2.6 J / cm2. During deposition, the target was rotated at
a兲

Electronic mail: david.look@wright.edu.

40 deg/s and rastered by + / −2 deg to prevent preferential
ablation at the center of the target, and the susceptor was
rotated at 20 deg/s The target to wafer distance was fixed at
9.5 cm. The substrate was heated by means of a coiled resistive element located approximately 3 mm from the substrate
wafer. The pressure was controlled for both O2 and Ar processes by flowing the appropriate volume of gas and setting
the turbo pump to 250 Hz. The number of laser pulses was
chosen to achieve film thicknesses of 30 to 300 nm.
To determine donor ND and acceptor NA concentrations,
temperature-dependent Hall-effect measurements were carried out over a range of 15–320 K using a Lakeshore 7507
apparatus. The values of n were independent of temperature,
and those of ␮ and ␳, only slightly dependent. It is instructive to compare n for a sample grown in Ar 共called the “Ar
layer”兲 with that of one grown in O2 共the “O2 layer”兲. Since
the O2 layer had earlier been grown on SiO2 / Si, the Ar layer
was also grown on this material; however, up to this point we
have noticed no great dependence of n on choice of substrate. The Ar-layer thickness was 82 nm, as measured by
spectroscopic ellipsometry 共SE兲, and the O2-layer, 99 nm. In
the as-grown O2 layer, n was low, about 6 ⫻ 1019 cm−3. As
mentioned earlier, such a low value of n is probably due to
high concentrations of VZn–related acceptors in the bulk and
O2-related acceptors at the GBs. However, when the O2 layer
was annealed in FG at 400 ° C for 10 min, n increased
greatly, up to about 4.3⫻ 1020 cm−3. This increase is likely
due to the H-induced reduction of O2 from the GBs and the
H passivation of VZn, i.e., VZn + 2H → VZn − 2H. On the other
hand, the Ar layer has a high value of n, as deposited, about
8.1⫻ 1020 cm−3. Upon annealing in FG, n actually decreases
slightly, while the mobility ␮ increases slightly, a scenario
that can be explained only by a decrease in the donor concentration ND. This observation is not in line with conventional wisdom which would have predicted an increase in
ND, due to the addition of H donors, and a decrease in NA,
due to the passivation of acceptors. Thus, the role of H is
unclear in this case.

0003-6951/2010/97共7兲/072113/3/$30.00
97,is072113-1
© 2010 American InstituteDownloaded
of Physics to IP:
This article
is copyrighted as indicated in the article. Reuse of AIP content
subject to the terms at: http://scitation.aip.org/termsconditions.
209.147.144.10 On: Fri, 06 Feb 2015 17:16:20

072113-2

Appl. Phys. Lett. 97, 072113 共2010兲

Scott et al.

␮ii共n,K兲 =

24␲3␧20ប3 n
Z2e3mⴱ2 Nii

1
ln关1 + y共n兲兴 −

y共n兲
1 + y共n兲

146.9

=

1/3
ln共1 + 6.46n20
兲−

1

1/3
6.46n20
1/3
+ 6.46n20

冉 冊

1 − K cm2
.
1+K V s
共3兲

Here, dnm is the layer thickness in units of nm, n20 is the
carrier concentration in units of 1020 cm−3, and C is an empirical constant which depends on the details of surface and
interface scattering. For the samples discussed in Ref. 5, a
value of C = 4 best fits the data, while for the three points
presented in Fig. 1, C = 2.5 gives a good fit. The other fitting
FIG. 1. 共Color online兲 Plots of sheet carrier concentration as a function of
layer thickness for samples grown on Al2O3. Inset: mobility as a function of
parameter is K = 0.47, and since K = NA / ND and n = ND − NA,
thickness. The squares are measured data and the solid line is a fit to these
we get finally, ND = 2.3⫻ 1021 cm−3 and NA = 1.1
21
−3
data, with n = 1.2⫻ 10 cm and K = 0.47. The dashed line is the hypotheti⫻ 1021 cm−3. These numbers must be considered tentative
cal mobility for n = 1.2⫻ 1021 cm−3 and K = 0, or equivalently, NA Ⰶ ND.
but clearly the layer has a very high donor concentration,
composed mostly of Ga. Unfortunately, the layer also has a
Thicknesses of Ga-doped ZnO/ SiO2 / Si films deposited
high acceptor concentration, and a key goal of TCO develin oxygen are routinely measured using SE with a wellopment is to get rid of the acceptors. Note that if n were the
established model for the layer stack. However, as the optical
same but NA Ⰶ ND 共i.e., K ⬃ 0兲, then ␮ versus d would folconstants of Ga-doped ZnO films deposited in Ar have not
low the dashed line in the inset of Fig. 1, and the resistivity
been fully characterized, the films were deposited on Al2O3
for thick samples would drop by about a factor three at room
to assure correct thicknesses for ␳ and n calculations. An
temperature. For very thin samples, of course, surface/
abrupt step was obtained for profilometer 共Tencor P-10兲 meainterface scattering, and ultimately quantum effects, will alsurements by partially masking the Ga-doped ZnO film, then
ways limit the mobility.
immersing the wafer into an acid solution known to have a
Unalloyed Ohmic contacts were fabricated using e-beam
high selectivity for ZnO/sapphire, 共1:1000 HCL:deionized
evaporated Ti/Pt/Au:20/30/350nm at room temperature. Exwater兲. In Fig. 1, sheet carrier concentration ns versus thickcellent specific contact resistance values of 2 – 4
ness d is plotted with error bars indicated for the two thickest
⫻ 10−8 ⍀ cm2 were measured using the transmission line
samples. The data were taken at room temperature but there
method 共TLM兲. Temperature stress tests were applied to
is very little change in n from 15–320 K. Within error, the
samples in air at temperatures up to 350 ° C by keeping
points fall on a straight line, and the slope of this line is just
samples at each temperature step for 1 h. No significant
the volume carrier concentration n = 1.2⫻ 1021 cm−3. 共More
change in contact resistivity was observed in this temperapoints will be needed to determine if there is a slight nonlinture range. At higher temperatures, alloyed contacts were
earity in the data.兲 The y-axis intercept, −2 ⫻ 1015 cm−2, can
formed and the contact resistance dropped sharply. The bulk
perhaps be interpreted as a depletion charge due to acceptors
resistivity also was not strongly affected by the stress tests,
in the ZnO/ Al2O3 interface region; however, again more
increasing from 5 ⫻ 10−4 to 7 ⫻ 10−4 ⍀ cm, indicating that
points will be necessary to provide a definite conclusion in
both the conductivity of the films and the unalloyed Ohmic
this matter.
contacts are stable with temperature.
The room-temperature mobilities of the three points are
For transparent electrode applications, the most imporplotted as a function of thickness in the inset of Fig. 1. For
tant parameters are the sheet resistance ␳s and the optical
samples such as these, the low-temperature mobilities are
transmittance OT. For the 277 nm sample in Fig. 1, ␳s
only slightly higher than those at room temperature. This is
= 7.1 ⍀ / sq, a very impressive value. Unfortunately, the
because most of the scattering is due to the ionized donors
backsides of our Al2O3 wafers were not polished, so it was
and acceptors, rather than phonons, even at room temperanecessary to measure OT on other layers, grown on polished
ture. To fit the mobility data we apply a recently developed
quartz. In Fig. 2, the wavelength dependence of the transmitmodel that includes degenerate Brooks–Herring theory along
tance, measured with a Varian Cary 5000 spectrophotometer,
with an empirical formula for boundary scattering, as
is shown for Ar and O2 layers grown on quartz under condifollows:5
tions nearly identical to those pertaining to the samples of
Fig. 1. We note that the OT for as-deposited O2 layers 共not
shown兲 is nearly 90% from the band edge all the way to
−1
−1 −1
␮共d,n,K,C兲 = 关␮ii共n,K兲 + ␮bdry共d,n,C兲 兴 ,
共1兲
2500 nm but unfortunately the conductivity of such layers is
too low for TCO applications. For the highly conductive laywhere
ers, i.e., the O2 layers annealed in FG, and the as-grown Ar
layers, the transmission above 1000 nm decreases due to
10.58 dnm cm2
e d/C
free-carrier absorption, as expected. The decrease in this re␮bdry共d,n,C兲 =
=
,
共2兲
1/3
gion is higher for the Ar layer because its carrier concentraVs
ប 共3␲2n兲1/3
C n20
tion is higher; however, in the visible region, the Ar layer has
and is copyrighted as indicated in the article. Reuse of AIP content is subject
a slightly
better
transmittance.
This article
to the terms
at: http://scitation.aip.org/termsconditions.
Downloaded to IP:

冉 冊

209.147.144.10 On: Fri, 06 Feb 2015 17:16:20

072113-3

Appl. Phys. Lett. 97, 072113 共2010兲

Scott et al.

This work was carried out at the Air Force Research
Laboratory, Wright-Patterson AFB, Dayton, OH. We wish to
thank T. A. Cooper for the Hall-effect measurements. The
work of R.C.S. and Y.H.Z. was partially supported by Science Foundation Arizona, Contract Nos. SRG 0190-07 and
SRG 0339-08, and by the Air Force Research Laboratory/
Space Vehicles Directorate, Contract No. FA9453-08-20228. The work of D.C.L. was partially supported by
AFOSR under Grant No. FA9550-10-1-0079 共K. Reinhardt兲
and NSF under Grant No. DMR0803276 共L. Hess兲.
1

FIG. 2. 共Color online兲 Optical transmittance as a function of wavelength for
Ga-doped ZnO deposited at 200 ° C and 10 mTorr in Ar 共solid line兲 and O2
共dashed line兲.

In conclusion, highly conductive Ga-doped ZnO with
excellent optical transparency has been achieved at low
deposition temperatures using PLD in a pure Ar environment
without the need for postdeposition anneals. The significant
increase in conductivity is attributed mainly to improved Ga
incorporation on donor sites, although a detailed model of
growth dynamics has yet to be worked out. Sheet resistance
versus thickness data suggest that the bulk carrier uniformity
in thick films grown in Ar is much better than that for thick
films grown in O2 and then annealed in FG.

B. Du Ahn, S. H. Oh, C. H. Lee, G. H. Kim, H. J. Kim, and S. Y. Lee, J.
Cryst. Growth 309, 128 共2007兲.
2
K. Yim, H. W. Kim, and C. Lee, Mater. Sci. Technol. 23, 108 共2007兲.
3
E. Millon, J. Perriere, S. Tircot, and C. Boulmer-Leborgne, Proc. SPIE
7005, 70051A 共2008兲.
4
B. G. Choi, I. H. Kim, D. H. Kim, K. S. Lee, T. S. Lee, B. Cheong, Y.-J.
Baik, and W. M. Kim, J. Eur. Ceram. Soc. 25, 2161 共2005兲.
5
D. C. Look, K. D. Leedy, D. H. Tomich, and B. Bayraktaroglu, Appl.
Phys. Lett. 96, 062102 共2010兲.
6
D. C. Look, K. Ghosh, and K. D. Leedy, Zinc Oxide and Related
Materials—2009, MRS Symposia Proceedings No. 1201 共Materials Research Society, Pittsburgh, 2010兲, p. 35.
7
B. Bayraktaroglu, K. Leedy, and R. Bedford, Appl. Phys. Lett. 93, 022104
共2008兲.
8
H. Qi, Q. Li, C. Wang, L. Zhang, and L. Lv, Vacuum 81, 943 共2007兲.
9
S.-M. Park, T. Ikegami, and K. Ebihara, Jpn. J. Appl. Phys., Part 1 44,
8027 共2005兲.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
209.147.144.10 On: Fri, 06 Feb 2015 17:16:20