Growth and structure of sputtered gallium nitride films
Brajesh S. Yadav, S. S. Major, and R. S. Srinivasa
Citation: J. Appl. Phys. 102, 073516 (2007); doi: 10.1063/1.2786100
View online: http://dx.doi.org/10.1063/1.2786100
View Table of Contents: http://jap.aip.org/resource/1/JAPIAU/v102/i7
Published by the American Institute of Physics.
Related Articles
L10 ordered phase formation in FePt, FePd, CoPt, and CoPd alloy thin films epitaxially grown on MgO(001)
single-crystal substrates
J. Appl. Phys. 111, 07A708 (2012)
Comment on “Forming-free resistive switching behavior in Nd2O3, Dy2O3, and Er2O3 films fabricated in full
room temperature” [Appl. Phys. Lett. 99, 113509 (2011)]
Appl. Phys. Lett. 100, 076101 (2012)
Epitaxial NiO gate dielectric on AlGaN/GaN heterostructures
Appl. Phys. Lett. 100, 063511 (2012)
Coexistence of bi-stable memory and mono-stable threshold resistance switching phenomena in amorphous
NbOx films
Appl. Phys. Lett. 100, 062902 (2012)
Investigation of intermolecular interactions in perylene films on Au(111) by infrared spectroscopy
J. Chem. Phys. 136, 054503 (2012)
Additional information on J. Appl. Phys.
Journal Homepage: http://jap.aip.org/
Journal Information: http://jap.aip.org/about/about_the_journal
Top downloads: http://jap.aip.org/features/most_downloaded
Information for Authors: http://jap.aip.org/authors
Downloaded 24 Feb 2012 to 14.139.97.76. Redistribution subject to AIP license or copyright; see http://jap.aip.org/about/rights_and_permissions
JOURNAL OF APPLIED PHYSICS 102, 073516 共2007兲
Growth and structure of sputtered gallium nitride films
Brajesh S. Yadav and S. S. Major
Department of Physics, Indian Institute of Technology Bombay, Mumbai 400076, India
R. S. Srinivasaa兲
Department of Metallurgical Engineering and Materials Science, Indian Institute of Technology Bombay,
Mumbai 400076, India
共Received 1 July 2007; accepted 9 August 2007; published online 10 October 2007兲
GaN films have been deposited by radio frequency sputtering of a GaAs target with pure nitrogen.
The growth, composition, and structure of the films deposited on quartz substrates have been studied
by x-ray diffraction, transmission electron microscopy, and Raman spectroscopy. Films deposited
below 300 ° C are amorphous and As rich. Above 300 ° C, polycrystalline, hexagonal GaN is
formed, along with As rich amorphous phase, which reduces with increasing substrate temperature.
At a substrate temperature of 700 ° C, GaN films, practically free of amorphous phase, and As
共⬍0.5 at. % 兲 are formed. The preferred orientation depends strongly on the substrate temperature
and is controlled by surface diffusion of adatoms during growth stage. Below 500 ° C, the surface
diffusion between planes dominates and results in the 共101̄1兲 preferred orientation. Above 500 ° C,
the surface diffusion between grains takes over and results in 共0002兲 preferred orientation. © 2007
American Institute of Physics. 关DOI: 10.1063/1.2786100兴
I. INTRODUCTION
The commercialization of gallium nitride based high
brightness light emitting diodes and long life, short wavelength laser diodes has been followed by the realization of
high frequency power transistors, solar blind photodiodes,
ultraviolet light emitting diodes and detectors, high electron
mobility transistors, and numerous other electronic
devices.1–5 The large direct band gap of 3.4 eV, high breakdown voltage with added advantage of high thermal conductivity, and thermal stability of GaN have led to electronic
devices that can operate at high temperatures.6 GaN has also
received enormous attention due to its applications in white
light sources.7 Applications in spintronic devices for magnetic sensors, high-density memories, and spin polarized
light emitters for optical encoding have been proposed, based
on ferromagnetic doping of GaN.8 Piezoelectric properties of
GaN and its negative electron affinity have also opened
newer avenues of research.1–3 Further, its low electron affinity and the ability of high n-type doping make it a promising
material for field emitters.9,10
Device quality GaN films are usually grown epitaxially,
using metal organic chemical vapor deposition 共MOCVD兲 or
molecular beam epitaxy 共MBE兲 techniques.1,2 Polycrystalline
GaN films have also been grown by MBE and MOCVD for
field electron and light emitter applications.10–12 In view of
the wide ranging, potential applications of polycrystalline
GaN, sputtering technique has recently become attractive,
owing to its versatility and scalability. There have been several early reports on the growth of GaN films by reactive
sputtering of Ga target with nitrogen or nitrogen-argon
mixture,13–15 though problems of reproducibility arising out
of the low melting temperature of gallium have been
a兲
Electronic mail: [email protected]
0021-8979/2007/102共7兲/073516/9/$23.00
reported.16 However, during the last few years, GaN films
have been deposited using both Ga 共Refs. 17–20兲 and GaN
共Refs. 21–24兲 targets. Magnetron sputtering has also been
utilized to deposit epitaxial GaN films on a sapphire
substrate.25–27
The inherent ability of sputtering to deposit large area
films on a variety of substrates at low temperatures opens up
enormous opportunities for a robust semiconductor, such as
GaN and III-nitrides in general, for potential applications in
optoelectronics, photonics, white lighting, and photovoltaics.
We have earlier reported28 on the reactive sputtering of a
GaAs target in 100% nitrogen, which resulted in practically
arsenic-free gallium nitride films, deposited at about 500 ° C
on quartz substrates and subsequently on nanocrystalline
GaN films deposited by the same approach.29 In the present
work, the growth process of GaN films by the sputtering of a
GaAs target in nitrogen has been studied in detail. The influence of significant growth parameters on the composition
and structure of these films has been investigated using x-ray
fluorescence 共XRF兲, x-ray diffraction 共XRD兲, transmission
electron microscopy 共TEM兲, and Raman spectroscopy.
II. EXPERIMENTAL DETAILS
The films were deposited by radio frequency 共rf兲 magnetron reactive sputtering of a 3 in. GaAs target with 100%
nitrogen as sputtering-cum-reactive gas. A Hartley oscillatortype rf power supply 共13.56 MHz兲 was used and the anode
current was used to monitor the rf power. An rf shielded
heater was used to control the substrate temperature up to
700± 10 ° C. Film deposition was preceded by evacuating the
chamber to a base pressure of 1 ⫻ 10−6 mbar, followed by
presputtering with nitrogen at 8 ⫻ 10−3 mbar for 5 min.
Films were deposited at a nitrogen pressure of 8
⫻ 10−3 mbar and a flow rate of 25 SCCM 共SCCM denotes
cubic centimeter per minute at STP兲. The growth rate of the
102, 073516-1
© 2007 American Institute of Physics
Downloaded 24 Feb 2012 to 14.139.97.76. Redistribution subject to AIP license or copyright; see http://jap.aip.org/about/rights_and_permissions
073516-2
J. Appl. Phys. 102, 073516 共2007兲
Yadav, Major, and Srinivasa
FIG. 1. Growth rate of nitrogen sputtered films deposited at different substrate temperatures and at a constant rf power corresponding to an anode
current of 0.20 A.
films was estimated from the film thickness measured by
Ambios XP-2 surface profilometer. The composition of films
was studied by Phillips 2404 x-ray fluorescence spectrometer
having a 4 kW x-ray tube with Rh target. The structure of the
films was studied using PANalytical X’Pert Pro powder diffractometer using Cu K␣ radiation. TEM studies were carried out using a Philips model CM200 Supertwin microscope
operated at 200 kV. Additional information about the composition and structure of the films was obtained by Raman
spectroscopy, using Jobin Yvon Horiba HR-800 confocal
micro-Raman spectrometer equipped with a 18 mW, Ar+ laser 共514.5 nm兲.
III. RESULTS AND DISCUSSION
Figure 1 shows the variation of the growth rate with the
substrate temperature, at an anode current of 0.20 A. The
growth rate remains ⬃1600 nm/ h for films deposited in the
range of 30– 200 ° C. Subsequently, a drastic fall in the
growth rate to ⬃800 nm/ h is seen at ⬃300 ° C. The growth
rate remains nearly constant up to a substrate temperature of
⬃500 ° C, beyond which it decreases slightly to saturate
⬃700 nm/ h. The growth rate was found to increase monotonically with rf power 共anode current兲. Typically, for a film
deposited at 550 ° C, it changed from ⬃100 nm/ h at an anode current of 0.12 A to⬃ 1100 nm/ h at 0.24 A.
A. Compositional studies
As the films were deposited by sputtering of GaAs with
nitrogen, the possible elements present in the films are Ga,
N, and As. As XRF is not a sensitive technique for nitrogen,
Ga K␣ and As K␣ peaks were recorded. Quantitative estimates of As/ Ga ratios in the films sputtered at different substrate temperatures were obtained by using a sputtered GaAs
film 共with argon兲 as reference. The GaAs film was similar in
morphology and thickness to the nitrogen sputtered films and
hence matrix effects were assumed to be nearly similar. The
variation of As/ Ga ratio in the films with the substrate temperature is shown in Fig. 2. The films deposited below
FIG. 2. Variation of As/ Ga atomic ratio in nitrogen sputtered films with the
substrate temperature.
300 ° C exhibit high As/ Ga ratios 共0.8–0.96兲, which fall
drastically to ⬍0.1 at ⬃350 ° C. The transition from high to
low As content films is concurrent with the sharp fall in the
growth rate. Thus the drastic fall in the growth rate seen
⬃300 ° C in Fig. 1 is attributed to the efficient removal of As
from the growing film. Consequently, the films deposited in
the range of 500– 600 ° C exhibit 共As/ Ga兲 ratios below 0.05,
which reaches a low value of less than ⬃0.01, for the film
deposited at 700 ° C.
B. Structural studies
XRD studies were carried out on 500– 800 nm thick nitrogen sputtered films. Typical XRD results for films deposited in the range of 30– 600 ° C are shown in Fig. 3. The film
deposited at room temperature shows weak and broad
humps. All the other films deposited up to 250 ° C show a
broad hump at 2␪ ⬃ 36° extending towards lower 2␪, as
shown typically in Fig. 3, for the film deposited at 250 ° C.
The value of 2␪ ⬃ 36.8° corresponds to the strongest reflection, 共101̄1兲 of hexagonal GaN. The value of 2␪ ⬃ 32.4° corresponds to 共101̄0兲 reflection of GaN, as well as the strongest
reflection, 共012兲 of As. It is thus inferred that all the films
deposited up to 250 ° C are amorphous but may be multicomponent. The films do not seem to contain a-GaAs, as no
humps are seen around the expected values of 2␪ ⬃ 27.3°,
31.7°, 45.4°, and 53.8° for GaAs. The possible phases in
these films could thus be a-GaN, a-GaAsN or elemental As.
The equilibrium solubility of nitrogen in GaAs is known to
be 2%–3% 共Refs. 30 and 31兲 and the introduction of a larger
quantity of nitrogen results in a GaAsN alloy with poor
crystallinity.32 The presence of As either in elemental form or
as amorphous GaAsN is in tune with XRF results. It can thus
be inferred that, at substrate temperatures below 300 ° C,
though nitrogen gets incorporated in the films, it does not
completely replace As in GaAs nor is As efficiently removed
from the films. The incomplete removal of As from the
growing films is thus responsible for the high growth rate, as
seen below 300 ° C.
Downloaded 24 Feb 2012 to 14.139.97.76. Redistribution subject to AIP license or copyright; see http://jap.aip.org/about/rights_and_permissions
073516-3
Yadav, Major, and Srinivasa
FIG. 3. XRD patterns of nitrogen sputtered films deposited at different
substrate temperatures 共as indicated in the figure兲.
Typical XRD results are also shown in Fig. 3 for films
deposited at and above 300 ° C. The XRD pattern of the film
deposited at 300 ° C shows two small peaks at 32.4° and
58.0°, along with a weak broad hump, indicating the beginning of the formation of crystalline phase. The films deposited in the temperature range of 350– 700 ° C showed sharp
J. Appl. Phys. 102, 073516 共2007兲
Bragg peaks, clearly indicating the formation of polycrystalline films. For these films, the d values corresponding to all
the peaks matched with those for hexagonal GaN and were
accordingly indexed. No peaks corresponding to GaAs, elemental Ga, or As were observed. As the 共012兲 peak of As
nearly coincides with the 共101̄0兲 peak of GaN, a marginal
presence of As in these films cannot be completely ruled out.
These results thus show that the films deposited above
300 ° C consist of polycrystalline hexagonal GaN. However,
as indicated by the XRF studies, these films also contain As
as an impurity, typically ⬃1 at. % in the films deposited
above 500 ° C, assuming the Ga/ N ratio to be nearly unity.
The film deposited at 700 ° C, clearly has the lowest As content, well below 0.5 at. %.
It is noticed from Fig. 3 that the preferred orientation of
crystallites changes with the substrate temperature. The film
deposited at 350 ° C showed a 共101̄0兲 preferred orientation,
which changed to 共101̄1兲 at 400 ° C. Similar XRD patterns
共not shown兲 were obtained for films deposited at 450 and
500 ° C. With the increase in substrate temperature to
550 ° C, an enhancement is seen in the intensity of 共0002兲
peak, which strongly dominates the patterns of films deposited at 600 共not shown兲 and 700 ° C. The change in the preferred orientation is accompanied by a small decrease in the
film growth rate, as seen in Fig. 1. The effect of the rf power
on the preferred orientation of films was also investigated,
but was found to be insignificant. The preferred orientation
effects will be discussed in detail later.
TEM studies were carried out on thinner films
共70– 80 nm兲, grown under the same conditions as above.
Figure 4 shows typical TEM results in the form of dark field
images along with the diffraction patterns. The dark field
image of the film deposited at 100 ° C shows dark regions
attributed to large voids or possibly very low density material. It does not indicate the presence of distinct crystallites
and the relatively brighter regions are attributed to the amorphous phase. The electron diffraction pattern of the film
shows only a “diffused diffraction halow,” confirming the
amorphous nature of this film. Interestingly, the dark field
image also shows some white spots, indicating the existence
of small crystallites within the amorphous matrix. Similar
features were seen in the films deposited up to 250 ° C,
which shows a continuous and void-free film, with a larger
density of crystallites. The film deposited at 250 ° C shows
some broad diffraction rings superimposed on a diffused pattern. As the diffraction rings correspond to reflections from
hexagonal GaN, the small crystallites formed within the
amorphous matrix may be identified as those of GaN. These
observations are in general agreement with XRD studies,
which showed that the films deposited up to 250 ° C were
amorphous. The small crystallites within the amorphous matrix seen in the dark field images were not evidenced from
XRD patterns because of low crystalline content. For the
film deposited at a substrate temperature of 300 ° C, a uniform distribution of crystallites is seen in the dark field images. The corresponding diffraction pattern clearly shows
sharp rings. Thus both XRD and TEM studies show that a
change in the structure of the films from amorphous to poly-
Downloaded 24 Feb 2012 to 14.139.97.76. Redistribution subject to AIP license or copyright; see http://jap.aip.org/about/rights_and_permissions
073516-4
Yadav, Major, and Srinivasa
J. Appl. Phys. 102, 073516 共2007兲
FIG. 4. Dark field TEM images of nitrogen sputtered
films deposited at different substrate temperatures 共as
indicated in figure兲. Insets show the corresponding electron diffraction patterns.
crystalline takes place in the substrate temperature range of
250– 300 ° C. It may be reiterated that the change in structure
of the films is concurrent with the drastic decrease in the
growth rate and efficient removal of As from the films.
With the further increase in the substrate temperature,
the diffraction rings become sharper and an increase in the
crystallite size is also seen in the dark field images. A typical
case is shown in Fig. 4 for the film deposited at 550 ° C. The
d values were calculated for all the reflections seen in the
films as deposited above 250 ° C. All the d values matched
well with those of hexagonal GaN and no diffraction rings
corresponding to any other phase such as GaAs, elemental
Ga, or As were observed.
The lattice constants 共a and c兲 of hexagonal GaN phase
in all the polycrystalline films were determined from the d
values obtained from XRD and TEM. The lattice constants
along with their uncertainties are plotted in Fig. 5 for films
deposited at different temperatures. Within the uncertainty
limits, the XRD and TEM results agree well and both a and
c are found to be nearly independent of the substrate temperature, having average values of 3.205 and 5.217 Å, respectively, which are slightly larger than the respective standard values for hexagonal GaN 共a = 3.190 Å and c
= 5.189 Å兲. The slightly but reproducibly higher values of
lattice constants at all substrate temperatures is a significant
observation. The corresponding strain was estimated to be
⬃5 ⫻ 10−3 共for both a and c兲 for all the films. The expected
thermal strain introduced by differences in the thermal expansion coefficients of the quartz substrate 共␣quartz = 0.5
⫻ 10−6 K−1兲 and GaN film 共␣a,c = ⬃ 5.6⫻ 10−6 K−1兲 at a sub-
FIG. 5. Plots of lattice constants a and c of GaN films against the substrate
temperature. The dashed lines show the standard values of lattice constants.
Downloaded 24 Feb 2012 to 14.139.97.76. Redistribution subject to AIP license or copyright; see http://jap.aip.org/about/rights_and_permissions
073516-5
J. Appl. Phys. 102, 073516 共2007兲
Yadav, Major, and Srinivasa
TABLE I. Crystallite size determined from the strong XRD peak共s兲 and dark
field TEM images. Corresponding planes are indicated within the brackets
for XRD data
Substrate
temperature 共°C兲
Crystallite size 共nm兲
XRD
TEM
400
15 共101̄1兲
8±6
450
27 共101̄1兲
10± 6
500
41 共101̄1兲
14± 8
550
33 共101̄1兲 and 13 共0002兲
12 共0002兲
13 共0002兲
16± 8
20± 12
600
700
13± 9
strate temperature of 600 ° C is ⬃3 ⫻ 10−3 and should be
smaller at lower substrate temperatures. The significantly
larger value of the measured strain and its temperature independence shows that thermal stress may not be significantly
responsible for the strain in these films. Thus, the slightly
larger values of lattice constants of GaN films are attributed
to the incorporation of a small quantity of As in GaN lattice.
However, the substrate temperature-independent increase in
the lattice constant is in contrast to the observed As/ Ga ratios in these films 共Fig. 2兲, which varies from as high as 0.13
共350 ° C兲 to less than 0.01 共700 ° C兲. This indicates that although most of these films may have a significantly larger As
content its actual incorporation in GaN lattice is independent
of the substrate temperature and may be less than 1 at. %, as
seen in the film deposited at 700 ° C. This is in tune with the
reported30,31 observations that it is not possible to incorporate
a significant amount of As into GaN lattice, owing to its
larger ionic radius. Thus most of the As in the films deposited at relatively lower temperatures 共300– 500 ° C兲 may be
present as a-GaAsN, as mentioned earlier. However, at
higher substrate temperatures, the quantity of As incorporated in the films is much less 共艋1 at. % 兲, most of which
may be incorporated only in the GaN lattice. It may thus be
inferred from the above studies that at low substrate temperatures, nitrogen gets incorporated in the films to form amorphous GaN / GaAsN. These films contain a large quantity of
As, which may partly be in elemental form. As the substrate
temperature is increased, hexagonal polycrystalline GaN is
formed and most of the free As released during the formation
of GaN evaporates and is thus efficiently removed from the
film. The efficient removal of excess As from the films deposited at higher substrate temperatures is responsible for the
drastic fall in the growth rate above substrate temperatures
⬃300 ° C.
The crystallite sizes of polycrystalline GaN films deposited at 400 ° C or above, estimated from dark field images as
well as from the width of the most prominent XRD peak共s兲,
are presented in Table I. The average crystallite size estimated from TEM increases monotonically with the increase
of the substrate temperature. For the films deposited above
500 ° C, the average crystallite size estimated from the dark
field image agreed well with the values obtained from the
width of 共0002兲 peak. However, in case of the films deposited below 500 ° C, in which 共101̄1兲 peak was most promi-
FIG. 6. XRD patterns of GaN films 共ⱗ100 nm兲 deposited at different substrate temperatures 共as indicated in the figure兲.
nent and much sharper, the crystallite sizes estimated from
its width are significantly higher than those obtained from
dark field images. This difference is attributed to the fact that
the crystallite size obtained from the XRD pattern is along
具101̄1典 crystallographic direction, perpendicular to the plane
of the film for the 共101̄1兲 preferentially orientated crystallites. In comparison, the crystallite size obtained from TEM
represents the average lateral size of crystallites. The difference in the two crystallite sizes indicates anisotropy in the
shape of crystallites and hence a columnar structure. In contrast, in the films deposited above 500 ° C, the crystallite
sizes measured from XRD and TEM are nearly same, which
suggests that the crystallites are relatively isotropic in shape.
As seen from Fig. 3 and discussed earlier, polycrystalline
GaN films exhibit preferred orientation of crystallites, which
is strongly dependent on the substrate temperature. The films
deposited at 300 and 350 ° C show 共101̄0兲 preferred orientation, but as these films contain a significant amount of amorphous phase as well as As, preferred orientation effects were
analyzed only for the films deposited at and above 400 ° C.
Though the preferred orientation behavior was independent
of the film thickness in the range of 500– 800 nm, it was
found that less than 100 nm thick films showed significantly
different behavior. XRD patterns of thinner films deposited
at three typical temperatures are shown in Fig. 6. All these
XRD patterns are dominated by a strong 共0002兲 peak, indicating that in thinner films the crystallites exhibit a preferred
orientation of 共0002兲 planes, irrespective of the substrate
temperature. However, with the increase in the film thick-
Downloaded 24 Feb 2012 to 14.139.97.76. Redistribution subject to AIP license or copyright; see http://jap.aip.org/about/rights_and_permissions
073516-6
Yadav, Major, and Srinivasa
ness, depending on the substrate temperature, the preferred
orientation either remains the same or changes to 共101̄1兲.
It can be inferred from Fig. 6 that in the nucleation and
initial stage of growth the crystallites grow with a preferred
orientation of 共0002兲 planes parallel to the film surface,
which is referred to as c-axis orientation in hexagonal systems. The 共0002兲 plane in hexagonal systems is usually considered to possess the lowest surface energy,33,34 though in
the case of GaN, there are limited data35,36 available on surface energies of crystallographic planes, indicating nearly
equal values for different planes.35 Also, in the case of AlN,37
共0002兲 plane has been reported as the lowest surface energy
plane. Thus assuming that in GaN also, 共0002兲 plane is the
lowest surface energy plane, 共0002兲 preferred orientation in
thinner GaN films can be attributed to preferential nucleation, which is known33,38,39 to be driven by minimization of
the surface energy. However, the main issue to be addressed
is the occurrence of different preferred orientations in thicker
GaN films deposited below and above 500 ° C. It is interesting to note that the strong c-axis preferred orientation, commonly observed in hexagonal systems such as ZnO,39 is not
seen in the case of sputtered GaN films. This is another important and possibly related observation, which suggests that
the preferred orientation in sputtered GaN films is not solely
determined at nucleation stage but is significantly affected by
the growth process.
Recent reviews by Kajikawa et al.38 and Kajikawa39
have analyzed the preferred orientation in sputtered ZnO and
some nitrides by collecting the experimental observations
and explanations. It is agreed that the preferred orientation at
the growth stage is significantly affected by two processes,
namely, sticking and surface diffusion of adatoms. Fujimira
et al.25 have proposed that the differences in the sticking
probabilities of growing species on different planes determine the preferred orientation. Accordingly, the preferred
orientation shows a strong dependence on the nature and
availability of growing species. In the present case of sputtering of GaAs target with nitrogen, it has been seen that the
preferred orientation effects are independent of the rf power.
This indicates that the higher availability of Ga 共at higher
power兲 relative to the abundantly available nitrogen is not a
significant factor in determining the nature of the preferred
orientation. Further, the nature of the film forming species is
also not expected to drastically vary with the substrate temperature, on which the preferred orientation so strongly and
solely depends. Hence, the nature and sticking coefficient of
the available species do not appear to control the preferred
orientation in GaN films at the growth stage.
The preferred orientation effects seen in GaN films at the
growth stage have thus been explained on the basis of surface diffusion. The consideration of surface diffusion necessarily involves the comparison of surface energies of different planes, which can in general be uncertain. However, as
discussed above, it is assumed that the surface energy of
共0002兲 plane in GaN is lower than that of 共101̄1兲 plane, as in
most other hexagonal systems. The observation made on
thinner GaN films, which exhibit 共0002兲 preferred orienta-
J. Appl. Phys. 102, 073516 共2007兲
FIG. 7. Schematic of the preferred orientation determined by the surface
diffusion process between planes.
tion at the nucleation and early growth stages, independent
of substrate temperature, also supports this assumption.
As mentioned above, at lower substrate temperatures
共⬍500 ° C兲 the thicker films exhibit 共101̄1兲 preferred orientation. This indicates that 共101̄1兲 oriented nuclei 共though
smaller in number in the early stage兲 grow faster along the
growth direction than 共0002兲 oriented nuclei and thus begin
to dominate the preferred orientation in the film. This can be
explained on the basis of surface diffusion, which can promote two kinds of preferred orientation, depending upon
whether it takes place among grains or among planes within
a grain.38,39 When surface diffusion takes place among
planes, the higher surface energy plane becomes the preferred orientation plane because surface diffusion essentially
occurs to conceal the plane with a higher surface energy.
This explains the 共101̄1兲 preferred orientation seen in the
films grown at lower substrate temperatures 共⬍500 ° C兲,
since the smaller surface diffusion coefficient at lower temperatures may limit the surface diffusion to be within the
grains only. This can be understood by referring to Fig. 7, in
which two neighboring crystallites with two different planes
are shown perpendicular to the growth direction. For an obvious relevance to the present case, in one of the crystallites
共A兲, the lower surface energy 共0002兲 plane is perpendicular
to the growth direction, while in the other 共B兲 the higher
surface energy plane 共101̄1兲 is perpendicular to the growth
direction. Since the driving force for surface diffusion in this
case is the difference between surface energies of these two
planes, the adatoms will diffuse to a lower surface energy
plane 共0002兲 and both crystallites will grow in a manner, so
as to maximize the surface areas corresponding to 共0002兲
plane. Notice that in case of 共A兲, the growth will be on the
top surface and hence lateral, while in case of 共B兲 the growth
will be on the inclined surface and hence will contribute
more to the vertical growth of the crystallite. It is obvious
that crystallite 共B兲 will determine the growth rate in vertical
direction, that is, in the direction of the growth of the film. It
is also obvious that in this case the higher surface energy
共101̄1兲 plane will appear as preferentially oriented with respect to the film surface, as has been observed in the present
case.
At higher substrate temperatures 共⬎500 ° C兲 increased
surface diffusion is expected to take the growth process to a
Downloaded 24 Feb 2012 to 14.139.97.76. Redistribution subject to AIP license or copyright; see http://jap.aip.org/about/rights_and_permissions
073516-7
Yadav, Major, and Srinivasa
FIG. 8. Raman spectra of nitrogen sputtered films deposited at different
substrate temperatures 共as indicated in the figure兲.
regime where surface diffusion among grains becomes dominant. In this case, the crystallites with a higher energy surface 共B兲 shrink and those with lower energy surface 共A兲
grow, though more laterally. Consequently, the lower energy
plane 共0002兲 becomes preferentially oriented, as has been
observed at higher substrate temperatures. The dominantly
lateral growth in this case may also be the reason for the
slightly lower film growth rates and the absence of a columnar structure in the films grown above 500 ° C.
C. Raman scattering studies
Raman spectra of films deposited on quartz substrates
are shown in Fig. 8, along with the spectrum of quartz substrate as reference. Raman spectra of films deposited in the
range of 30– 300 ° C show nearly similar features. Typical
spectra for films deposited at 100 and 300 ° C are shown. A
J. Appl. Phys. 102, 073516 共2007兲
high scattering intensity is seen below 400 cm−1 in both
cases. A broad hump is also seen at ⬃500 cm−1 followed by
another broad hump extending from 700 to 900 cm−1. All the
phonon modes for GaAs 共Ref. 40兲 exist below 300 cm−1 with
a high density of states between 250– 300 cm−1, while for
GaN 共Refs. 41 and 42兲 the acoustic modes lie below
300 cm−1 and the optical modes are in the range of
550– 750 cm−1. In this case too, a high density of states exists in the range of 200– 300 cm−1. It may be noted that As
also exhibits A1g mode at 257 cm−1.43 Further, Raman studies
of dilute GaAs1−xNx alloys44–46 have shown two mode behaviors in these alloys, exhibiting GaAs-like first order
modes below 300 cm−1, as well as second order modes between 500 and 600 cm−1 along with a GaN-like mode
⬃470 cm−1. Thus the significantly large scattering intensity
below 300 cm−1 seen in all the films deposited below 300 ° C
may be attributed to disorder activated Raman scattering,
with contributions from various amorphous phases present in
these films. Though XRD studies do not indicate the presence of amorphous GaAs, it cannot also be completely ruled
out on the basis of Raman studies, since a-GaAs is known47
to exhibit a broad peak around 250 cm−1. The broad hump
centered around 500 cm−1 may also have the same origin,
though the contribution from the quartz substrate is also possible. Figure 8 also shows the presence of another broad
hump extending from 700 to 900 cm−1, which may be partly
influenced by contributions from the quartz hump
⬃800 cm−1. However, in this case, the scattering intensity
appears to be contributed by a broad peak centered around
730 cm−1 followed by a significant scattering intensity extending up to ⬃900 cm−1. This observation is significant because the A1 共LO兲 mode 共LO denotes longitudinal optical兲 of
GaN appears as the most prominent Raman peak at
⬃730 cm−1 in all the films deposited above 300 ° C. Further,
higher order GaN modes due to acoustic-optical combination
have been reported42 at 855 and 915 cm−1, which may contribute to the scattering intensity seen above 800 cm−1. These
features on the higher frequency side of the Raman spectra
are clearly indicative of the formation of GaN, which could
be partly amorphous. It may be recalled that dark field TEM
images indicated the formation of nanocrystallites within a
dominantly amorphous matrix in these films. Based on Raman studies, it may thus be inferred that nanocrystallites of
GaN embedded in an amorphous matrix are formed at substrate temperatures of 100– 300 ° C. Raman studies thus
strongly support the earlier drawn inference from structural
observations that at substrate temperatures below 300 ° C nitrogen does replace As in GaAs, but As is not completely
removed from the films and remains in the form of amorphous GaAsN or elemental As.
In significant contrast to above, the film deposited at
350 ° C shows small peaks ⬃250 and ⬃728 cm−1, with the
latter being identified as the A1 共LO兲 mode of GaN. Two
humps also begin to appear ⬃320 and ⬃560 cm−1, which are
identified as acoustic overtone and E2 共high兲 mode of GaN,
respectively.42 With increase in the substrate temperature, A1
共LO兲 peak becomes sharper and stronger with the other features of Raman spectrum remaining nearly the same. Typical
spectra for the films deposited at 500 and 700 ° C are shown
Downloaded 24 Feb 2012 to 14.139.97.76. Redistribution subject to AIP license or copyright; see http://jap.aip.org/about/rights_and_permissions
073516-8
J. Appl. Phys. 102, 073516 共2007兲
Yadav, Major, and Srinivasa
in Fig. 8. The broad humps around 500 cm−1 and the extended hump from 700– 900 cm−1 seen in the films deposited
at substrate temperatures below 300 ° C are absent in all
these films. It may be mentioned that the positions of both A1
共LO兲 and E2 共high兲 modes are redshifted by 6 – 7 cm−1, with
respect to the respective reported values of 734 and
567.6 cm−1 for epitaxially grown films,33 which may be attributed to the presence of a small quantity of As in the
lattice, as indicated by XRD results. It is also noticed that
beyond 750 cm−1 the background scattering intensity is contributed mainly by the quartz hump and hence no significant
contribution from higher order Raman peaks is seen, as in
the case of films deposited below 300 ° C. All these features
clearly indicate a significant improvement in the crystalline
quality of the films with the increase in the substrate temperature, and thus corroborate well with XRD and TEM results.
Interestingly, the small low frequency peak ⬃250 cm−1
seen in the film deposited at 350 ° C shifts to 258– 260 cm−1
and becomes weaker with the increase of the substrate temperature. It can also be seen that in the films deposited above
500 ° C, apart form this small peak, this part of the spectrum
is determined by Raman scattering from the quartz substrate.
It is difficult to attribute this peak to a GaN mode, because of
the existence of complex phonon dispersion and a high density of states in GaN,42 below 300 cm−1. In order to understand the origin of this peak, it is recalled that the A1g mode
of bulk As is reported at 257 cm−1.43 The formation of As
clusters in GaAs 共Ref. 48兲 is also known to exhibit a Raman
mode around 256– 258 cm−1. Further, a strongly confined
GaAs optical mode has been detected ⬃250 cm−1 in dilute
GaAs1−xNx, and attributed to ordering of As and N atoms.44
Considering the positioning of this peak between
258– 260 cm−1 in the case of films deposited above 500 ° C,
it is attributed to the presence of As clusters in GaN films
deposited above 500 ° C, which contain ⱗ1 at. % As, as discussed above.
IV. CONCLUSIONS
is practically free of amorphous phase and contains less than
0.5 at. % As, most of which is incorporated in the GaN lattice as an impurity.
These results also throw light on the growth mechanism
involved in the sputter deposition of GaN films, using a
GaAs wafer as target and pure nitrogen as the sputteringcum-reactive gas. The observations that 共i兲 the films grown at
lower temperatures contain a large quantity of arsenic, which
is not completely removed from the film, and 共ii兲 the formation of polycrystalline GaN that takes place only at temperatures above 400 ° C emphasizes the crucial role of the substrate temperature in this unusual growth process and
indicates that the reaction between nitrogen and Ga/ GaAs
leading to the formation of GaN takes place at the substrate
surface.
Less than 100 nm thick GaN films show the prevalence
of preferred nucleation of 共0002兲 plane, irrespective of the
substrate temperature. However, the preferred orientation in
thicker films depends strongly on the substrate temperature
and is explained on the basis of surface diffusion of adatoms
during the growth stage. Below 500 ° C, surface diffusion
between planes dominates and results in the preferred orientation of higher surface energy 共101̄1兲 planes. Above
500 ° C, surface diffusion between grains takes over and results in a preferred orientation of lower surface energy
共0002兲 planes.
Overall, this work shows that polycrystalline, hexagonal
GaN films with strong c-axis preferred orientation and controlled levels of impurity can be deposited by rf sputtering of
a GaAs target with pure nitrogen. This approach can be effectively utilized in applications where the versatility and
scalability of the sputtering technique can be suitably exploited.
ACKNOWLEDGMENTS
The financial support for this work received from the
Board of Research in Nuclear Sciences, Department of
Atomic Energy 共GoI兲 is gratefully acknowledged. SAIF and
CRNTS of IIT Bombay are, respectively, acknowledged for
providing the facilities for TEM and Raman studies.
S. Nakamura and G. Fasol, The Blue Laser Diode 共Springer, Berlin, 1997兲.
M. S. Shur and R. F. Davis, GaN-Based Materials and Devices 共World
Scientific, Singapore, 2004兲.
3
B. Monemar and G. Pozina, Prog. Quantum Electron. 24, 239 共2000兲.
4
H. Morkoc, A. D. Carlo, and R. Cingolani, Solid-State Electron. 46, 157
共2002兲.
5
H. Okumura, Jpn. J. Appl. Phys., Part 1 45, 7565 共2006兲.
6
J. I. Pankove, M. Leksono, S. S. Chang, C. Walker, and B. V. Zeghbroeck,
MRS Internet J. Nitride Semicond. Res. 1, 39 共1996兲. URL: http://
nsr.mij.mrs.org/1/39
7
J. K. Park, M. A. Lim, C. H. Kim, H. D. Park, J. T. Park, and S. Y. Choi,
Appl. Phys. Lett. 82, 683 共2003兲.
8
S. J. Pearton, C. R. Abernathy, D. P. Norton, A. F. Hebard, Y. D. Park, L.
A. Boatner, and J. D. Budai, Mater. Sci. Eng., R. 40, 137 共2003兲.
9
T. Kozawa, M. Suzuki, Y. Taga, Y. Gotoh, and J. Ishikawa, J. Vac. Sci.
Technol. B 16, 833 共1998兲.
10
S. Hasegawa, S. Nishida, T. Yamashita, and H. Asahi, Thin Solid Films
487, 260 共2005兲.
11
D. P. Bour, N. M. Nickel, C. G. Van de Walle, M. S. Kneissl, B. S. Krusor,
P. Mei, and N. M. Johnson, Appl. Phys. Lett. 76, 2182 共2000兲.
12
A. Krost and A. Dadger, Phys. Status Solidi A 194, 361 共2002兲.
13
H. J. Hovel and J. J. Cuomo, Appl. Phys. Lett. 20, 71 共1972兲.
1
The following conclusions may be drawn from the work
presented in this paper on GaN films deposited by sputtering
of GaAs in pure nitrogen. Below 300 ° C, the sputtered films
are dominantly amorphous in nature, though TEM and Raman studies indicate that GaN nanocrystallites begin to form
just above 100 ° C. These films contain a large quantity of
As, and the amorphous phase is likely to consist of GaN
and/or GaAsN. An important inference is that even at these
substrate temperatures nitrogen is incorporated into the film,
though the removal of As from the film is not efficient. This
is attributed to the large quantity of nitrogen available in the
plasma compared to arsenic and the higher strength of Ga–N
bond 共6.81 eV兲 compared to Ga–As bond 共5.55 eV兲.49 As the
substrate temperature is increased to the range of
500– 700 ° C, As released during the formation of gallium
nitride evaporates, leaving behind dominantly single phase
polycrystalline GaN film. The GaN film deposited at 700 ° C
2
Downloaded 24 Feb 2012 to 14.139.97.76. Redistribution subject to AIP license or copyright; see http://jap.aip.org/about/rights_and_permissions
073516-9
14
Yadav, Major, and Srinivasa
E. Lakshmi, B. Mathur, A. B. Bhattacharya, and V. P. Bhargava, Thin
Solid Films 74, 77 共1980兲.
15
T. L. Tansley, R. J. Egan, and E. C. Horrigan, Thin Solid Films 164, 441
共1988兲.
16
S. Zembutsu and M. Kobayashi, Thin Solid Films 129, 289 共1985兲.
17
T. Miyazaki, T. Fujimaki, S. Adachi, and K. Ohtsuka, J. Appl. Phys. 89,
8316 共2001兲.
18
T. Kikuma, K. Tominaga, K. Furutani, K. Kusaka, T. Hanabusa, and T.
Mukai, Vacuum 66, 233 共2002兲.
19
Q. X. Guo, W. J. Lu, D. Zhang, T. Tanaka, M. Nishio, and H. Ogawa, J.
Vac. Sci. Technol. A 22, 1290 共2004兲.
20
E. C. Knox-Davies, J. M. Shannon, and S. R. P. Silva, J. Appl. Phys. 99,
073503 共2006兲.
21
R. H. Horng, D. S. Wuu, S. C. Wei, S. H. Chan, and C. Y. Kung, Thin
Solid Films 343–344, 642 共1999兲.
22
J. H. Kim, M. R. Davidson, and P. H. Holloway, Appl. Phys. Lett. 83,
4746 共2003兲.
23
S. Shirakata, R. Sasaki, and T. Kataoka, Appl. Phys. Lett. 85, 2247 共2004兲.
24
T. Maruyama and H. Miyake, J. Vac. Sci. Technol. A 24, 1096 共2006兲.
25
M. Park, J.-P. Maria, J. J. Cuomo, Y. C. Chang, J. F. Muth, R. M. Kolbas,
R. J. Nemanich, E. Carlson, and J. Bumgarner, Appl. Phys. Lett. 81, 1797
共2002兲.
26
K. Kusaka, T. Hanabusa, K. Tominaga, and N. Yamauchi, J. Vac. Sci.
Technol. A 22, 1587 共2004兲.
27
Y. Daigo and N. Mutsukura, Thin Solid Films 483, 38 共2005兲.
28
N. Elkashef, R. S. Srinivasa, S. Major, S. C. Sabharwal, and K. P. Muthe,
Thin Solid Films 333, 9 共1998兲.
29
P. A. Nisha, S. Major, N. Kumar, I. Samajdar, and R. S. Srinivasa, Appl.
Phys. Lett. 77, 1861 共2000兲.
30
M. Weyers, M. Sato, and H. Ando, Jpn. J. Appl. Phys., Part 2 31, L853
共1992兲.
31
J. Neugebauer and C. G. Van de Walle, Phys. Rev. B 51, 10568 共1995兲.
32
L. Auvrary, H. Dumont, J. Dazord, Y. Monteil, J. Bouix, C. Bru-
J. Appl. Phys. 102, 073516 共2007兲
Chevallier, and L. Grenouillet, Mater. Sci. Semicond. Process. 3, 505
共2000兲.
33
N. Fujimura, T. Nishihara, S. Goto, J. Xu, and T. Ito, J. Cryst. Growth 130,
269 共1993兲.
34
H. Deng, J. J. Russell, R. N. Lamb, B. Jiang, Y. Li, and X. Y. Zhou, Thin
Solid Films 458, 43 共2004兲.
35
J. E. Northrup and J. Neugebauer, Phys. Rev. B 53, R10477 共1996兲.
36
J. Neugebauer, Phys. Status Solidi B 227, 93 共2001兲.
37
D. R. McKenzie and M. M. M. Bilek, Thin Solid Films 382, 280 共2001兲.
38
Y. Kajikawa, S. Noda, and H. Komiyama, J. Vac. Sci. Technol. A 21, 1943
共2003兲.
39
Y. Kajikawa, J. Cryst. Growth 289, 387 共2006兲.
40
P. Giannozzi, S. D. Gironcoli, P. Pavone, and S. Baroni, Phys. Rev. B 43,
7231 共1991兲.
41
V. Y. Davydov, Y. E. Kitaev, I. N. Goncharuk, A. N. Smirnov, J. Graul, O.
Semchinova, D. Uffmann, M. B. Smirnov, A. P. Mirgorodsky, and R. A.
Evarestov, Phys. Rev. B 58, 12899 共1998兲.
42
H. Siegle, G. Kaczmarczyk, L. Filippidis, A. P. Litvinchuk, A. Hoffmann,
and C. Thomsen, Phys. Rev. B 55, 7000 共1997兲.
43
D. L. Carter and R. T. Bate, The Physics of Semimetals and Narrow Gap
Semiconductors 共Pergamon, New York, 1971兲, p. 285.
44
A. M. Mintairov, P. A. Blagnov, V. G. Melehin, N. N. Faleev, J. L. Merz,
Y. Qiu, S. A. Nikishin, and H. Temkin, Phys. Rev. B 56, 15836 共1997兲.
45
T. Prokofyeva, T. Sauncy, M. Seom, M. Holtz, Y. Qiu, S. Nikishin, and H.
Temkin, Appl. Phys. Lett. 73, 1409 共1998兲.
46
C. Pelosi, G. Attolini, M. Bosi, M. Avella, M. Calicchio, N. Musayeva,
and J. Jimenez, J. Cryst. Growth 287, 625 共2006兲.
47
R. Zallen, M. Holtz, A. E. Geissberger, R. A. Sadler, W. Paul, and M. L.
Theye, J. Non-Cryst. Solids 114, 795 共1989兲.
48
P. S. Pizani, A. Mlayah, J. Groenen, and R. Carles, Appl. Phys. Lett. 66,
1927 共1995兲.
49
CRC Handbook of Chemistry and Physics, edited by R. C. West 共Chemical
Rubber, Boca Raton, FL, 1987兲, p. E102.
Downloaded 24 Feb 2012 to 14.139.97.76. Redistribution subject to AIP license or copyright; see http://jap.aip.org/about/rights_and_permissions