Porous Si(1 1 1) and Si(1 0 0) as an intermediate buffer layer for nanocrystalline InN films

Article abstract of DOI:10.1016/j.jallcom.2009.01.085

We report preliminary results on the indium nitride (InN) films grown on porous silicon (PSi) substrates by reactive magnetron sputtering using an indium target. PSi is used as an intermediate layer between silicon and InN and it provides a large area com

Full text of DOI:10.1016/j.jallcom.2009.01.085

Journal of Alloys and Compounds 479 (2009) L54–L58
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jallcom
Letter
Porous Si(1 1 1) and Si(1 0 0) as an intermediate buffer layer for nanocrystalline
InN films
∗
L.S. Chuah , Z. Hassan, S.S. Ng, H. Abu Hassan
Nano-Optoelectronics Research and Technology Laboratory, School of Physics, Universiti Sains Malaysia, 11800 Minden, Penang, Malaysia
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 7 October 2008
Received in revised form 16 January 2009
Accepted 21 January 2009
Available online 6 February 2009
We report preliminary results on the indium nitride (InN) films grown on porous silicon (PSi) substrates
by reactive magnetron sputtering using an indium target. PSi is used as an intermediate layer between
silicon and InN and it provides a large area composed of an array of voids. The PSi samples were pre-
pared using photoelectrochemical method on n-type silicon wafer with (1 1 1) and (1 0 0) orientation. To
prepare porous structures, the samples were dipped into a mixture of HF:ethanol (1:1) for 5 min with
−2
current densities of 50 mA cm , and subjected to external illumination with a 500 W UV lamp. The sur-
face morphology and the crystalline structure of the InN films were characterized by scanning electron
microscope (SEM) and X-ray diffraction (XRD). Structural analysis revealed nanocrystalline structure with
crystallite size of 32 ± 0.05 nm and 20 ± 0.05 nm for InN films grown on Si(1 1 1) and Si(1 0 0) orientation,
respectively. The optical properties of InN layers were examined by micro-Raman and Fourier transform
infrared (FTIR) reflectance spectroscopy at room temperature. Micro-Raman results showed that A1(LO) of
Keywords:
Porous silicon
InN
Reactive sputtering
XRD
Raman
SEM
FTIR
−1
−1
hexagonal InN/Si(1 1 1) and InN/Si(1 0 0) have been observed at 588.4 cm and 587.0 cm , respectively.
From the results of FTIR spectroscopy, the TO [E1(TO)] phonon mode of the InN/Si(1 1 1) and InN/Si(1 0 0)
are clearly visible at 472 cm and 470 cm , respectively.
−1
−1
©
2009 Elsevier B.V. All rights reserved.
1
. Introduction
InN and relatively low melting point of indium metal. It is difficult
to directly grow InN films from indium and nitrogen. Various meth-
ods have been used to fabricate InN films, such as MBE, MOCVD, and
pulse laser deposition (PLD).
The group-III nitrides have become the focus of materials
research because their wide-range bandgaps make then suitable
for optoelectronics applications such as near-IR/near-UV devices.
Among the III nitrides, indium nitride (InN) has the smallest
electron effective mass, largest mobility, highest peak, saturation
electron drift velocities, and smallest direct bandgaps [1]. With
improvement in material quality, studies have suggested that the
InN bandgap is close to 0.7 eV, much lower than the previously
accepted value of 1.9 eV. Thus, it is a promising material for micro-
electronic and optoelectronic devices such as solar cells with high
efficiency, optical coatings, and various types of sensors.
Recently, InN has received more attention for the properties
of its alloys with GaN and AlN which permit the fabrication of
InGaAlN-based short-wavelength semiconductor laser diodes [2,3].
Moreover, InN has distinct advantages in high frequency centimeter
and millimeter wave devices because the transport characteristics
of InN are superior to those of GaN and GaAs [4]. The main obstacles
of producing InN thin films are the low dissociation temperature of
Since bulk InN substrates are not available, the InN films must be
grown on a foreign substrate. Usually, sapphire is used as a substrate
material for the heteroepitaxy of InN because of its wide availabil-
ity, hexagonal symmetry, and its ease in handling and pre-growth
cleaning. However, from the perspective of device technology, sap-
phire possesses severe problems as a substrate material because it
is difficult to cleave. Moreover, since the sapphire substrate is not
conductive, it is impossible to create optical devices which operate
between the films and the substrates. Therefore, it is highly desir-
able to deposit the InN films on other semiconductor substrates
[5–7].
Silicon (Si) is a desirable substrate due to its large role in the
electro-optic industry. If high quality InN film can be grown on
Si substrate, the production of optoelectronic integrated circuits
will be possible using present Si-based IC processing technologies.
Moreover, it has a potential for the development of high efficiency
and low cost solar cells by the formation of InN/Si two-junction tan-
dem structures [8]. They can be easily cleaved, etched and removed
when necessary. These substrates can serve as a backside ohmic
contact. Indium nitride films have been prepared by reactive sput-
tering methods [9] and vapor phase deposition methods (MOCVD
∗ Corresponding author.
E-mail addresses: chuahleesiang@yahoo.com (L.S. Chuah), zai@usm.my
(
Z. Hassan).
0925-8388/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2009.01.085

L.S. Chuah et al. / Journal of Alloys and Compounds 479 (2009) L54–L58
L55
Table 1
Sputtering condition for InN deposition.
Target
Indium (99.999%)
3 in.
Target diameter
Target–substrate distances
Substrate
Residual pressure
RF power
∼15 cm
Porous silicon
−
6
<4 × 10 Torr
60–70 W
Gas
N2 + Ar mixture
75:25%
N2 (99.995%)
30:10 sccm
Gas ratio (N2:Ar)
Sputtering ambient
Gas flow rate (N2:Ar)
Sputtering pressure
Substrate temperature
Deposition time
−
2
2.0 × 10 Torr
◦
Room temperature (∼25 C)
90 min
(
Ar) and nitrogen (N2). The sputtering system consists of a vacuum chamber with
target and substrate holder, a turbo molecular pump parallel to a rotary pump to
provide high vacuum circumstance, an RF power supply at 13.56 MHz, and mass
flow controllers to maintain steady gas flow rate. The target was cleaned with a 5 min
pre-sputter by Ar plasma prior to each deposition process. The growth conditions
for InN films are summarized in Table 1.
After deposition, the thickness of InN film was determined by a surface step
profile analyzer. Structural characterization using SEM (JOEL JSM-6460LV) and XRD
Fig. 1. Schematic of PEC wet etching apparatus.
(
PANalytical X’pert PRO MRD) were carried out at room temperature. The carrier
concentration and mobility were measured by Hall effect.
All the optical measurements were performed at room temperature. Micro-
Raman scattering experiments were performed by using a Jobin Yvon HR 800 UV
spectrometer system. The excitation light source was the 514.5 nm line of Ar-ion
laser and the incident laser power was 20 mW.
In this work, micro-Raman scattering experiments were carried out in the z(x,
unpolarized) z¯ scattering configuration. Here, the Porto’s notation is used for scat-
tering geometries with z parallel to wurtzite c-axis, and (x, unpolarized) refers
to the polarization of the incident and scattered light. Under this configuration,
the allowed zone-center phonon modes that can be detected for wurtzite struc-
[
10], MOVPE [11]). In the former, there have been several compli-
cated and ambiguous relations between deposition conditions and
properties of deposited films.
In the last decade, porous semiconductors have been widely
studied, primarily due to the potential for intentional engineering
of properties not readily obtained in the corresponding crystalline
precursors as well as the potential applications in optoelectron-
ics, chemical and biochemical sensing [12,13]. When porosity is
formed, these materials exhibit various special optical features,
for instances, the shift of bandgap [14], luminescence intensity
enhancement [15], as well as photoresponse improvement [16,17].
Among porous semiconductors, porous silicon (PSi) is a material
creating great scientific and technological interest because of its
ample range of application; it has recently been used as an active
element and as an anti-reflective film in solar cells.
In this field, to the best of our knowledge, this is the first time
that PSi layer was used as an intermediate buffer layer between
Si wafer and InN. In this work, we report results of indium nitride
films grown at room temperature (RT) by reactive RF sputtering
method on porous silicon. The surface morphology and the crys-
talline structure of the InN films were characterized by scanning
electron microscope (SEM) and X-ray diffraction (XRD). The opti-
cal properties of InN layers were examined by micro-Raman and
Fourier transform infrared (FTIR) reflectance spectroscopy at room
temperature.
2.
Experimental
Prior to the RF sputtering deposition, the porous samples were prepared using
photoelectrochemical (PEC) method on n-type silicon wafer with (1 1 1) and (1 0 0)
orientation. Before electrochemical etching, RCA chemical cleaning was applied. The
chemical etching reveals atomic terraces and passivates silicon surface atoms by
hydrogen atoms (H-termination).
To prepare porous structures, the samples were dipped into a mixture of
−
2
HF:C2H5OH electrolyte (1:1) for 5 min with current densities of 50 mA cm . Due
to the hydrophobic character of clean silicon surface, the C2H5OH is necessary to
improve the wettability of porous silicon. Ethanol is the surface active agent and its
presence facilitates in removing the bubbles of hydrogen formed during etching.
For the dissolution of n-type silicon, an external source of light is necessary to
obtain sufficient holes flux density. Samples were illuminated with 500 W UV lamp
and subsequently well rinsed in de-ionized water. The anodization is performed
in potentiostatic (voltage-controlled) mode. Typical PEC wet etching apparatus is
schematically shown in Fig. 1. In electrochemical setup, silicon was the anode and
Pt was the counter electrode.
InN films were grown by magnetron reactive sputtering deposition technology
from high purity indium targets with a purity of 99.999% in a mixture gas of argon
Fig. 2. SEM images of porous silicon layers on: (a) Si(1 1 1) and (b) Si(1 0 0) substrates
by PEC etching.

L56
L.S. Chuah et al. / Journal of Alloys and Compounds 479 (2009) L54–L58
ture layer will be A1(LO), E2(low), and E2(high), unless there are some disoriented
microstructures.
Room temperature IR measurements are carried out by using an FTIR spectrome-
ter (Spectrum GX FT-IR, PerkinElmer) with a KBr beam splitter and a mid-IR triglycine
sulfate (TGS) detector. The measurements are taken in the reflection mode at a res-
olution of 4 cm–1 and the number of scans is 128. A PerkinElmer variable angle
reflectance accessory is used to measure the reflectance spectrum at an incident
◦
angle of 15 .
3. Results and discussion
Electrochemical etching results in selective removal of bulk sil-
icon. In the process, the presence of the electronic holes at the
semiconductor surface is necessary. Low holes density in n-type
silicon, where holes are the minority carrier results in low current
flow. To support the electrochemical reaction, an external source
of holes is necessary. Increase in holes quantity result in increase
of the current density and etching. For the holes generation, the
most effective way is illumination. During illumination, electrons
in the valence band take the photons energy and are excited into the
conduction band, leaving the holes in the valence band. Density of
holes is dependent on the light intensity. Positive holes react on the
−
surface with negative solvated ions F , which result in the disso-
lution of the surface. Formation of the porous silicon includes both
oxidation and etching of the silicon atoms. Removal of the silicon
atoms from the surface proceeds through fluoride formation.
Fig. 2 presents the plan view SEM micrographs of PSi prepared on
n-Si. For porous Si(1 1 1) substrate, in the plan view the rigid trian-
gular shaped macroarray network of 300–400 nm sizes are clearly
seen. The characteristic faceted, triangular shaped pores illustrate
an orientation dependence of the photoanodization process of the
Si(1 1 1) material, probably due to the anisotropy of Si(1 1 1) dis-
Fig. 4. XRD phase analysis scan of InN films grown on porous: (a) Si(1 1 1) and (b)
Si(1 0 0) substrates.
solution. For porous Si(1 0 0) substrate, pores formed by the rigid
circular shaped network with diameter up to 400 nm are clearly
seen in the plan view SEM image. The thickness for porous (1 1 1)
and (1 0 0) orientation were around 0.5 ␮m and 1 ␮m.
During anodic etching, the material between pores was gener-
ally depleted of carriers, and the presence of a depletion layer is
responsible for current localization at pore tips. After reaching a
critical thickness, pore walls are devoid of holes and hence inert to
lateral anodic etching [18]. But the etching rate of the Si(1 0 0) crys-
tal plane is about two orders of magnitude larger than that of the
(1 1 1) crystal plane. The use of a porous layer is also a good alter-
native in lattice-mismatch heteroepitaxy. The PSi layer is obtained
on the Si substrate. Its skeleton is supple, thus reducing the stress
induced at the cooling phase and limiting the formation of disloca-
tion and cracks in InN layers.
The deposited films were semi-transparent and dark red color.
Fig. 3 shows the plan view SEM images of InN films grown on porous
Si(1 1 1) and Si(1 0 0) substrates. It can be seen that the film consists
of agglomerated nanocrystals. The shape of the particles is a conse-
quence of lattice-mismatched materials where the nucleation of the
nitride film is initiated on protruding crystals on the rough under-
lying substrate. The thickness of the InN cross section from the SEM
image is about 201 nm which is in good agreement with the thick-
ness of InN film determined by a surface step profile analyzer which
is about 200 nm.
The unintentionally doped InN films exhibited n-type conduc-
tivity. This has long been attributed to nitrogen vacancies. Such
nitrogen deficiency could arise from the low dissociation energy
of the indium–nitrogen bond. Other possible explanations of the
n-type characteristic include nitrogen on indium sites, oxygen on
Fig. 3. SEM images of InN films grown on porous: (a) Si(1 1 1) and (b) Si(1 0 0)
substrates.

L.S. Chuah et al. / Journal of Alloys and Compounds 479 (2009) L54–L58
L57
nitrogen sites, or interstitial hydrogen. From the Hall measure-
ments, the resulting unintentionally doped InN films with Hall
2
−1
mobility and electron concentration about (200–300) cm Vs and
18
−3
.
(
1–3) × 10 cm
X-ray diffraction is performed to investigate the crystal phase of
the studied samples. Fig. 4(a) and (b) represents typical XRD profile
of the InN film grown on the porous silicon substrates. For Fig. 4(a),
¯
only the peak from (1 0 1 1) InN plane is observed apart from the
(
1 1 1) peak from the porous silicon substrate. For Fig. 4(b), the peak
◦
located at around 69.5 is attributed to the Si family planes (4 0 0).
The c-axis orientation was not observed in our samples. The low
intensity of the InN peak and a broadening centered at the crys-
talline region could be an indication that these films are a mixed
phase of crystalline and amorphous structure. This is probably a
signature of the microcrystalline or nanocrystalline phase for InN.
The average size of the crystallites was determined from the
¯
(
1 0 1 1) peak using the Scherrer formula:
Kꢀ
t =
(1)
ˇ cos ꢁ
where t is the crystallite size, K is a constant, which is 0.9, ꢀ is the
wavelength of the incident beam, ˇ is the half width of the peak,
and ꢁ is the diffraction angle of the peak [19]. Using Eq. (1), the crys-
Fig. 6. IR reflectance spectrum of InN thin films grown on porous: (a) Si(1 1 1) and
(
b) Si(4 0 0) substrates.
tallite (grain) sizes determined by the Scherrer method are about
3
2 ± 0.05 nm and 20 ± 0.05 nm for InN/PSi(1 1 1) and InN/PSi(1 0 0),
respectively, thus confirming the nanocrystalline structure of our
films.
A typical Raman spectrum at room temperature for the grown
−
1
−1
InN/PSi and PSi in the spectrum range from 0 cm to 800 cm is
shown in Fig. 5(a) and (b). The Raman scattering was recorded in
back scattering geometry using an Ar-ion laser. InN has a hexagonal
4
wurtzite structure with four atoms per unit cell and belongs to a c6v
(
P6 mc) space group. From group theory, one would expect one A ,
3
1
one E , two E2 and two B1 optical phonon modes. Phonon modes
1
of A and E symmetry are both Raman and IR active; the E mode
1
1
2
is only Raman active; and the B mode is silent (forbidden for both
1
Raman and IR excitations). The polar character of the A1 and E1
vibrational modes gives rise to both longitudinal and transverse
components.
This scattering geometry allowed the observation of E (high)
2
and A (LO) (longitudinal optical) phonons of InN. The A1 modes
1
give atomic displacements along the c-axis, while the others, the
E gives atomic displacements perpendicular to the c-axis. In addi-
2
−
1
tion, a peak from the silicon substrate is also seen at ∼520 cm
,
−
1
and a band at ∼300 cm due to the acoustic phonons of Si. The InN
−
1
Raman mode of E (high), which is located at 488 cm , is hardly
2
detected from the spectrum because it has been masked by the
−
1
stronger Si peak (at 520 cm ). The A (LO) Raman line shape is
1
very sensitive to lattice disorder in InN crystals, because the non-
stoichiometric defects will disrupt the long-range ionic ordering.
Fig. 5. (a and b) Room temperature micro-Raman spectrum of InN/PSi substrate as
well as the Raman spectrum of porous Si substrate.

L58
L.S. Chuah et al. / Journal of Alloys and Compounds 479 (2009) L54–L58
The disorder will induce the phonon modes near the A (LO) on the
Micro-Raman results showed that A (LO) of hexagonal InN/Si(1 1 1)
1
1
phonon dispersion curve to be Raman active. For InN thin films, only
and InN/Si(1 0 0) have been observed. From the results of FTIR spec-
the allowed Raman phonon modes of A (LO) are clearly visible at
troscopy, the TO [E (TO)] phonon mode of the porous InN/Si(1 1 1)
and porous InN/Si(1 0 0) is clearly visible.
1
1
−
1
−1
5
88.4 cm and 587.0 cm for InN film grown on porous Si(1 1 1)
and Si(1 0 0) substrates, respectively.
In comparison to the result obtained from Davydov et al. [20],
Acknowledgements
−
1
which has A (LO) mode of wurtzite InN at 586 cm , it is found
1
that our A (LO) mode of wurtzite InN is shifted towards higher fre-
Financial support from Science Fund, MOSTI, and fellowship
from Universiti Sains Malaysia are gratefully acknowledged. The
authors thank Mr. Mohtar bin Sabdin for FTIR measurements.
1
quency. This discrepancy can be attributed to the quality of the InN
thin film as well as the coupling effects between the LO and the free
carrier concentrations, which is, known as the LO plasmon–phonon
(
8
LPP) coupling effects. The Raman mode of E (low), located at
7 cm [20], is invisible from the spectrum because of the cut-
References
2
−
1
[
1] C.B. Yang, X.L. Wang, H.L. Xiao, X.B. Zhang, G.X. Hu, J.X. Ran, C.M. Wang, J.P. Li,
J.M. Li, Z.G. Wang, Appl. Surf. Sci. 255 (2008) 3149–3152.
2] M. Jamil, R.A. Arif, Y.-K. Ee, H. Tong, J.B. Higgins, N. Tansu, Phys. Stat. Sol. (a) 205
(2008) 1619–1624.
off limit of the 514.55 nm notch filter used in this work. The Raman
peaks exhibit broad and asymmetric line shapes. The broadening
in the spectra is due to the small phonon correlation length in the
samples.
Fig. 6 shows the IR reflectance spectrum of InN thin films grown
on the porous Si substrates. The IR reflectance spectrum of porous
Si substrates is also shown for comparison. In general, the InN Rest-
[
[3] I. Akasaki, S. Sota, H. Sakai, T. Tanaka, M. Koike, H. Amano, Electron. Lett. 32
1996) 1105–1106.
4] S.K. O’Leary, B.E. Foutz, M.S. Shur, U.V. Bhapkar, L.F. Eastman, J. Appl. Phys. 83
1998) 826–829.
(
[
(
[5] A. Kadir, M.R. Gokhale, A. Bhattacharya, A. Pretorius, A. Rosenauer, J. Cryst.
Growth 311 (2008) 95–98.
−1
−1
strahlen region lies between 400 cm and 600 cm whereas the
Si is inactive in IR region. For InN thin films, the features corre-
[6] E.S. Hwang, S.B. Che, H. Saito, X. Wang, Y. Ishitani, A. Yoshikawa, J. Electron.
Mater. 37 (2008) 597–602.
sponding to the TO [E (TO)] phonon modes of the InN are clearly
[7] E. Dimakis, A.Yu. Nikiforov, C. Thomidis, L. Zhou, D.J. Smith, J. Abell, C.K. Kao,
T.D. Moustakas, Phys. Stat. Sol. (a) 205 (2008) 1070–1073.
1
−1
−1
seen at 472 cm and 470 cm for porous Si(1 1 1) and Si(1 0 0) sub-
strate, respectively, which is comparable to the reported values for
wurtzite InN [17]. However, we did not observe the feature (dip)
[
[
8] C.F. Shih, N.C. Chen, C.Y. Tseng, Thin Solid Film 516 (2008) 5016–5019.
9] T.J. Kistenmacher, W.A. Bryden, Appl. Phys. Lett. 62 (1993) 1221–1223.
[10] N. Miura, H. Ishii, A. Yamada, M. Konagai, Y. Yamauchi, A. Yamamoto, Jpn. J. Appl.
Phys. 36 (1997) L259–L265.
due to the LO [A (LO)] phonon mode of the InN (located at around
1
[
11] W.K. Chen, Y.C. Pan, H.C. Lin, J. Ou, W.H. Chen, M.C. Lee, Jpn. J. Appl. Phys. 36
1997) L1625–L1627.
[12] H. Sohn, S. Letant, M.J. Sailor, W.C. Trogler, J. Am. Chem. Soc. 122 (2000) 5399.
−1
5
80–590 cm ). This is mainly due to the weak nature of its spec-
(
tral strength as compared to that of the E (TO) mode. For porous
1
[
[
[
13] L.S. Chuah, Z. Hassan, H. Abu Hassan, Mater. Sci. Poland 26 (2008) 609–615.
14] L.S. Chuah, Z. Hassan, H. Abu Hassan, Adv. Mater. Res. 31 (2008) 39–41.
15] M.A. Steven Kalceff, I.M. Tiginyanu, S. Langa, H. Foll, H.L. Hartnagel, J. Appl. Phys.
89 (2001) 2560–2565.
Si, the fringes are most probably due to interference between the
radiations reflected from the top surface of the Si and that reflected
from the pores.
[
16] M. Mynbaeva, N. Bazhenov, K. Mynbaev, E. Evstropov, S.E. Saddow, Y. Koshka, Y.
Melnik, Phys. Stat. Sol. B 228 (2001) 589–592.
4
. Conclusion
[
17] P.M. Fauchet, L. Tsybeskov, C. Peng, S.P. Duttagupta, J. Von Behren, Y. Kostoulas,
J.M.V. Vandyshev, K.D. Hirschman, IEEE J. Sel. Top. Quantum Electron. 1 (1995)
1126–1139.
Nanocrystalline InN films have been successfully grown at
[
18] L. Koker, A. Wellner, P.A.J. Sherratt, R. Neuendorf, K.W. Kolasinski, Phys. Stat. Sol.
room temperature by reactive sputtering method on PSi(1 1 1) and
PSi(1 0 0) substrates. We have used SEM, XRD, FTIR and Raman scat-
tering to characterize the films. The crystallite (grain) sizes of the
InN film determined by the Scherrer method are about 32 ± 0.05 nm
and 20 ± 0.05 nm for InN/PSi(1 1 1) and InN/PSi(1 0 0), respectively.
(
a) 197 (2003) 117–122.
[
19] D.Y. Lee, I.S. Kim, J.S. Song, Jpn. J. Appl. Phys. 41 (2002) 4659–4662.
[20] V.Y. Davydov, V.V. Emtsev, I.N. Goncharuk, A.N. Smirnov, V.D. Petrikov, V.V.
Mamutin, V.A. Vekshin, S.V. Ivanov, M.B. Smirnov, T. Inushima, Appl. Phys. Lett.
75 (1999) 3297–3299.