Optik 127 (2016) 9212–9221

Contents lists available at ScienceDirect

Optik
journal homepage: www.elsevier.de/ijleo

Theoretical study of electronic and optical properties of BN,

GaN and Bx Ga1−x N in zinc blende and wurtzite structures
A. Said ∗ , M. Debbichi, M. Said
Laboratoire de la matière condensée et nanosciences, Faculté des Sciences, Université de Monastir, 5019 Monastir, Tunisia

a r t i c l e i n f o a b s t r a c t

Article history: Using density-functional theory (DFT) within local density approximation (LDA), the
Received 28 April 2016 electronic and optical properties of GaN, BN and BGaN alloy have been investigated in
Accepted 28 June 2016 zinc-blende (ZB) and wurtzite (WZ) structures. Our calculations show a strong nonlinearity
of lattice constants and band gap depending on the boron content. By increasing the boron
Keywords: concentration, the crossover compositions where the alloy switches from direct to indi-
Density functional theory rect band gap material are predicted. The band gap is tunneled from the blue to ultraviolet
BGaN
(UV) spectral region and the obtained bowing parameter is rather different for the both
Optical properties
structures. The calculated dielectric functions and absorption spectra demonstrate that the
increasing band gap enhances the optical absorption and strong absorptions have been
found in ultraviolet region. In order to overcome the underestimation of the energy band
gap, a scissors operation with a rigid upward shift of the conduction band has been used to
correct the LDA.
© 2016 Elsevier GmbH. All rights reserved.

1. Introduction

The III-nitride compounds and their alloys have received a significant attention due to their potential applications in
optoelectronic devices operating in the infrared to the ultraviolet (UV) spectral region. Owing to their larger band gaps and
stronger bonds, the groups gallium nitride and related materials have been attractive for high frequency, high-power and
high-temperature optoelectronic applications [1]. Due to its large ionicity, high thermal conductivity, high heat capacity,
and its wide band gap (3.5 eV) GaN semiconductor is a potential candidate for several micro- and optoelectronics devices
[2,3]. It can exist in two different structures: wurtzite which is the most stable at room temperature and zinc blende [2].
The necessity for compact ultraviolet laser devices with even shorter wavelengths seems to be increasing day by day,
and this has motivated many scientists to search for materials with larger band gaps than that of GaN [3,4].
Therefore, Boron represents a potential candidate for band-gap modification in nitride-based semiconductor materials.
In fact the addition of boron to GaN can be tunneled the band-gap from the blue to ultraviolet (UV) spectral region [5,6],
and reduce the lattice constant mismatch between GaN epilayers and 6H-SiC substrates [7]. Boron nitride (BN) can be found
in cubic (C-BN), hexagonal (h-BN) or wurtzite (WZ-BN) forms. C-BN has been used in a different microelectronic devices
working at high temperatures, however, WZ-BN has been used in vacuum technology and nuclear energy [8,9]. Wurtzite
BN which has a smaller lattice constant compared to that of AlN it has a potential along with the other materials in the
III-nitrides family either as substrate or as part of a layer in a strain balanced heterostructure. Hexagonal BN is classified

∗ Corresponding author.
E-mail address: [email protected] (A. Said).

http://dx.doi.org/10.1016/j.ijleo.2016.06.103
0030-4026/© 2016 Elsevier GmbH. All rights reserved.
 A. Said et al. / Optik 127 (2016) 9212–9221 9213

as wide bandgap material owing to its partially ionic character of B-N bond. Its band gap energy value ranging from 3.6 to
7.1 eV is not clear at present and represents a subject of debate [8,9].
GaN, BN and their alloys Ga1-x Bx N, grown under ambient conditions and assume the hexagonal WZ crystal structure
[10]. On the other hand, the grown of thin films Bx Ga1-x N with the cubic ZB structure is also possible [6]. The incorporation
of boron into semiconductor GaN crystals has been the subject of many theoretical and experimental works [6,11]. On
the other hand, the BGaN crystal growth techniques for ultraviolet LED have been studied. In the low B composition (less
than 3.6%) region, the fabrication of BGaN alloy crystal, by the substitution of B with Ga in the GaN, was realized. Bx Ga1-x N
were grown on AlN/sapphire substrates using metal-organic vapor phase epitaxy and characterized by optical transmission,
photoluminescence, and x-ray diffraction for a composition x ≤ 1.8% [11]. Very recently K. Atsumiet al. [12] indicated that
BGaN semiconductor material can be used in neutron-detection technology, and neutron-based imaging techniques.
Preceding calculations of BGaN alloy largely dedicated to structural and electronic properties in ZB phase predict that
large content boron alloys are possible [14]. Similarly, Kanoun et al. [15] studied the structural parameter and energy gap
of WZ-BGaN using all-electron FPLAPW method. These results demonstrated that BGaN admit a large bowing parameter.
Although, to some extent, the structural and electronic properties of BGaN are demonstrated in both WZ and ZB phases, we
are not aware of any theoretical study of optical properties of this compound. Optical properties are of great importance in
the design and analysis of optoelectronic devices such as light sources and detectors, thus, for a complete understanding
of these materials, a theoretical description of physical properties is very important.We have performed first-principles
calculations to investigate the optical properties of BGaN and give an important guideline on the material design of short-
wavelength optical devices using these nitrides. We report a theoretical study of B-doped GaN systems in, zinc-blende and
wurtzite structure. The electronic band structure, the density of states (DOS), and the optical properties of the B-doped GaN
systems were systematically studied. All calculations are done by using the DFT.
The paper is organized as follows: In Section 2, we outline the computational details. Section 3 is devoted to the main
results obtained from this study. The conclusions are summarized in Section 4.

2. Computational details

Our calculations were performed using the plane wave based density functional theory as implemented in QUANTUM-
ESPRESSO package [16], together with norm-conserving pseudopotentials and Trouiller–Martins form [17] to describe the
interactions between ionic cores.The exchange-correlation interaction was treated in the framework of LDA as parameterized
by Perdew and Zunger [18]. In order to overcome the underestimation of the energy band gap, a scissors operation with a
rigid upward shift of the conduction band has been used.
Cut-off energies set to 80 and 60 Ry and a grid of 4 × 4 × 2 and 3 × 3 × 3 k-points in the reciprocal space is chosenfor, the
zinc-Blende and Wurtzite structure, respectively. The total energy dependence on the cell volume is fitted to the Murnaghan
equation of state (EOS) [19] by:
 
B0 V
 V B’0  V0

Etot =   0
+ B’0 1− − 1 + E0 (1)
B’0 B’0 − 1 V V

Where V0 is the equilibrium volume, B0’ is the bulk modulus derivative andB0 is the bulk modulus.
Optical properties can be described by the complex dielectric constant ε defined as ε = ε1 + iε2 . The imaginary part ε2 (ω)
of the dielectric function can be written as:

e2  
␧2 (ω) = |Mcv (k) |2 ␦ (ωcv (k) − ␻) dk3 (2)
␲m2 ␻2 c,v
BZ

Where Mcv (k) = ϕc,k |u.∇ |ϕv,k  are the transition moments elements between valence ␸v,k and conduction band ␸c,k states,
u is the polarization vector of the incident electric field, k is the reciprocal lattice vector, ωcv (k) = Eck − Evk is the excitation
energy. The real part of dielectric function can be evaluated from the imaginary part by the famous Kramer-Kronig relations
[20]. All of the other optical properties, such as absorption coefficient ˛ (ω), refractive index n (ω), and energy-loss L (ω) can
be evaluated according to the following equations:

 1/2
√
␣ (␻) = 2␻ (␧21 (␻) + ␧22 (␻) − ␧1 (␻) (3)

 1/2
1
N (␻) = √ (␧21 (␻) + ␧22 (␻) + ␧1 (␻) (4)
2
 −1  ε2
L (␻) = Im = (5)
ε (ω) ε21 + ε22
9214 A. Said et al. / Optik 127 (2016) 9212–9221

Table 1
The calculated lattice parameters and bulk modulus for BN and GaN compounds. Available experimental and theoretical data are also given for comparison.

a(Å) c (Å) B (GPa)

GaN (wz) Our Calc. 3.090 5.068 202.300

Other cal. – – 168.000 [31]
Exp. 3.189 [32] 5.181 [32] 188.000 [33]
END>[5pt] (zb) Our Calc. 4.480 – 207.400
Other cal. 4.460 [34] – 201.000 [37]
4.470 [35] – 202.000 [38]
Exp. 4.500 [39] – 190.000 [36]

BN (wz) Our Calc. 2.500 4.169 394.000

Other cal. 2.560 [15] 4.230 [15] 358.000 [31]
Exp. 2.553 [32] 4.215 [32] 390.000 [55]
END>[5pt] (zb) Our Calc. 3.610 – 381.900
Other cal. 3.580 [37] – 401.700 [38]
3.630 [32] – 371.000 [32]
Exp. 3.620 [32] – 382.000 [40]

Fig. 1. The calculated lattice parameters of the GaBN alloy as a function of Bore composition.

3. Results and discussion

3.1. Structural and electronic properties

Table 1 summarizes the calculated structural parameters of the binary compounds GaN and BN both in WZ and ZB
structures compared to available experimental values and other theoretical predictions. As can be seen from Table 1, our
results are in reasonably agreement with the available ones. Therefore, it is realistic to assume a similar accuracy in our
calculations of the lattice parameters of the alloys.
The lattice parameters of the GaBN alloy computed as a function of Bore composition together with the results from
Vegard’s law are shown in Fig. 1.
Our computational results show a nonlinear characteristic in disagreement with Vegard’s law. The nonlinearity of the
lattice parameters of zinc blend and wurtzite Bx Ga1−x N are described as:
a (x) = aBN x + aGaN (1 − x) − ba x (1 − x) (6)
and c (x) = cBN x + cGaN (1 − x) − bc x (1 − x) (7)
Nonlinear least square fits by Eqs. (6) and (7) to our results for the lattice parameters a and c generate ba = − 0.48 for ZB
in good agreement with that obtained by Riane et al. [21], whereas ba = −1.61 and bc = −3.11 for WZ, where aBN , aGaN , cBN ,
and cGaN are our obtained lattice parameters of BN and GaN listed in Table 1. These nonlinear characteristics of the lattice
parameters are considered to arise from the difference in the chemical bonding property between GaN and BN.
The calculated band structures and the densities of states for the WZ and ZB phases are plotted in Fig. 2.
First inspection from Fig. 2 is that the gap of GaN is direct and that of BN is indirect, between the  and X points for the ZB
and  and K points for WZ. This description is in accordance with calculations performed on both WZ and ZB phases, using
the DFT and the empirical pseudopotential method [1–4]. It is known that the use of the LDA to the density functional theory
 A. Said et al. / Optik 127 (2016) 9212–9221 9215

Fig. 2. Band structures of binaries BN and GaN for the WZ and ZB phases.

Table 2
The calculated band gaps of GaN and BN binaries. Available experimental and theoretical data are also given for comparison.

Energy band gaps

Compounds Our Calc. Other Calc. Experimental Predicted

ZB WZ ZB WZ ZB WZ ZB WZ

GaN 3.04 2.95 2.42 [31,43] 2.92–2.60 [27] 3.20–3.40 [41,42] 3.39–3.50 [32,42] 3.4 3.5
BN 4.85 4.50 5.81–4.35 [32,43] 5.00–6.1 [43,44] 5.70–6.00 [41,42] 5.50–7.00 [42] 6.00 6.00

usually leads to underestimate the energy band gap in semiconductors.In general, the use of nonlocal hybrid-functional DFT
calculations or alternatively, GW quasiparticle calculations are required to reproduce prominent features of the electronic
structure. Nevertheless the gain of accuracy increases the computational cost by two orders of magnitude or even more
compared to conventional local DFT. For this reason and in order to overcome this shortcoming a scissors operation [22]
with a rigid upward shift of the conduction band has been used to correct the LDA energy band gap. The values obtained are
denoted “Predicted”.The results achieved are comparable to those obtained experimentally. These results as well as those
obtained by others calculations are presented in Table 2.
Remarkable progress has been made in the development of optical and electronic devices based on group III–V nitrides.
Important examples are active optoelectronic devices operating in the green, blue and ultraviolet spectral regions. A way
of increasing the flexibility of nitride compounds is to alloy them with boron materials with larger band gaps than that of
GaN. In fact the addition of boron to GaN can be tunneled the band-gap from the blue to ultraviolet (UV) spectral region.
We investigate the Boron concentrations x dependence of the electronic band structure for Bx Ga1-x N. For that, we use the
approximation outlined by Levine et al. [22] to estimate the bandgap:

Eg (x) = (1 − x) [Eg,GaN (Predict) − Eg,GaN (LDA)] + x[Eg,BN (Predict) − Eg,BN (LDA)] + Eg,Bx Ga1−x N (LDA) (8)
9216 A. Said et al. / Optik 127 (2016) 9212–9221

Fig. 3. Bx Ga1-x N band gaps versus Boron content.

Fig. 4. the total densities of states (TDOS) of BGaN in wurtzite and blende phases.

We took five discrete compositions namely x = 0, 0.25, 0.5, 0.75, 1 and we mainly focus on the band gap bowing parameters
and the crossover compositions where the alloy switches from direct to indirect band gap material.
Fig. 3 illustrates the band gap of ZB and WZ Ga1−x Bx N over the whole alloy composition range in a joint theoretical and
experimental study. Because GaN is a direct semiconductor and BN shows an indirect fundamental band gap, the ternary
alloy exhibits a concentration-dependent direct-indirect band gap crossing point. A crossover from a direct to an indirect
band gap is found to be occurring around x = 0.75 corresponding to the band-gap energies of 4.7 and 3.1 eV for ZB and WZ
structure, respectively. Owing to its crystal-field splitting and lower symmetry, WZ structure has a lower band gap energy
compared to that of ZB.
As can be seen, the band gap increases nonlinearly with increasing B content. The nonlinearity Bx Ga1−x N band gap is
expressed as:

Eg (x) = (1 − x) Eg,GaN + xEg,BN − bx (1 − x) (9)

where b is the bowing parameter.

For zinc blende structure, we notice that these band gap energies increase nearly monotonically. This result was also
observed experimentally [23]. This character appears to be ‘normal’ behavior in Bx Ga1−x N alloy as indicated in the case
of Al-based compounds [24]. However, for wurtzite structure the behavior is nonlinear, in accordance with the results of
Baaziz et al. [25] and Tsai et al. [26]. The direct gap versus concentration has an upward bowing with a value of b = 3.61 and
a downward bowing of b = −1.356 for WZ and ZB, respectively.
In order to get a better understanding of the electronic structure of the investigated compounds, we display in Fig. 4 the
total densities of states (TDOS) of BGaN in wurtzite and blende phases. For pure GaN and BN, the TDOS presents different
regions: two valence regions below the top of the valence band (VB) and one conduction band (CB) above EF . The top of
valence band region is dominated by N p states, and the conduction band by Ga p and B p states ranging from GaN and BN
respectively, whereas the lower valence band is from the electronic transition of N s state in agree well with other theoretical
works [27]. In contrast to wurtzite crystals The DOS curve for ZB GaN has a sequence of singlet peaks located, and this may
 A. Said et al. / Optik 127 (2016) 9212–9221 9217

Fig. 5. The imaginary part ␧2 (␻) of the dielectric function for B-doped GaN.

be due to the existence of resonance interactions of the Ga s, p, and d orbitals with the N s and p orbitals of covalent type
bond.
The VB of Bx Ga1-x N mainly consists of N2p mixed with a small contribution of B2p orbitals. The CB mainly consists of
levels determined by B s and B 2p with few contributions of Ga 2p orbitals. As seen also from Fig. 4, the width of valence band
increases with x. This increase is mainly due to the interaction between p-orbitals from Ga and N elements, which differs
from that between the B and N. By increasing the B concentration, the N 2p states shifts to the lower energy compared to
that from the pure GaN.

3.2. Optical properties

Optical properties are of great importance in the design and analysis of optoelectronic devices such as light sources and
detectors. However, the features of the electronic structure and its relation to the optical properties of GaBN alloy were not
discussed. The optical properties of solids can be described in terms of the absorption, imaginary part of the optical dielectric
function, reflectivity and the refractive index, respectively.
We performed first-principles calculations to investigate the optical properties of Ga1-x Bx N and give an important
guideline on the material design of short-wavelength optical devices using these nitrides. Dielectric function, absorption
coefficient, refractive index, and energy-loss spectrum of B doped GaN systems have been calculated. All the results pre-
sented here are, for a polarization vector perpendicular to the c-axis. To correct the LDA energy band gap, the approximation
of scissors operation is employed.
In Fig. 5, we illustrate the imaginary part ␧2 (␻) of the dielectric function for B-doped GaN. As clearly seen, for GaN blende,
there are three major peaks located at about 4.7, 7 and 7.6 eV. The peak at around 4.7 eV is mainly due to the optical transition
between N p-states in the highest valence band and Ga p-states in the lowest conduction band. The two peaks at about 7
and 7.6 eV correspond to transitions between Ga s- and N-p states. However, for GaN wurtzite two peaks appear at 2.9
and 4.06 eV, caused by the transitions between N-p and Ga-p states. The major peak is located at 7.5 eV and can be due
to the optical transitions between N s and Ga p states. By increasing B concentration x, the peaks move to higher energy,
that might be strongly dependent on the ionic polarization of the BGaN crystal due to the large electronegativity of N. For
pure BN blende, there are also three peaks existing at 10.8, 12.2 and 14.3 eV, the peak at around 10.8 eV originates mainly
from the transition between N p and B s states and its energy value is close to the calculated band gap.The peaks at around
12.2–14.3 eV originate mainly from the transition between N s and B s states. This result is on global agreement with the
calculated and measured data [7]. For BN wurtzite, the major peak is at about 11.07 eV and is mainly derived from the optical
transition between the B p states and N p states.
The real parts of the dielectric function ␧1 (␻) obtained from the imaginary parts by the Krames- kroning conversion [20]
are shown in Fig. 6.
From the ␧1 (␻) curve we can get that BGaN WZ and ZB material display similar trend mainly for higher energy, and
behave like a metallic property above 6.2 eV for GaN and 11 eV for BN. The static dielectric constant ␧(0) given by the low
energy limit of ␧(␻) are listed in Table 3.
The calculated values of ␧(0) agree rather well with experiments [28,29], but they are systematically somewhat smaller.
 2
This is in accordance with the model of Penn [30]ε (0) ≈ 1 + ωp /Eg , showing that a wide band gap yielding a smaller
value of ␧1 (0).
 9218
Table 3
The calculated static dielectric constant ␧(0), n(0) and the positions of Plasmon peaks appearing in L(␻).

␧(0) n(0) ␻p(0)

A. Said et al. / Optik 127 (2016) 9212–9221

Our Calc. Other Calc. Exp. Our Calc. Other Calc. Exp. Our Calc. Other Calc. Exp.

GaN ZB 4.90 4.68, 5.46, 5.7 2.63 2.24[50] 2.5[52], 15.07 – –

5.74[45,47] [46,47] 2.34[51]
WZ 6.15 5.89, 4.62, 5.35 [48] 2.95 2.29[53] 2.6[54] 22.90 – –
5.41,5.60 [45,5]

B0.25 Ga0.75 N ZB 4.02 – – 2.38 – – 15.80 – –

WZ 6.63 – – 3.07 – – 24.14 – –

B0.5 Ga0.5 N ZB 4.08 – – 2.48 – – 17.45 – –

WZ 7.24 – – 3.17 – – 24.45 – –

B0.75 Ga0.25 N ZB 4.13 – – 2.40 – – 19.40 – –

WZ 8.07 – – 3.30 – – 25.12 – –

BN ZB 3.64 3.85, 4.51, 4.14, 4.41, 6.8 [48] 2.27 2.3[55], 1.9[13] 2.1[49] 22.40 25.5[55] –
4.56 [38,49]
WZ 5.37 4.5, 4.19, 7.25 – 2.75 2.2[55] – 30.10 28.5[55], 28.5[55]
[45,47, 28.0[13],33.3
 A. Said et al. / Optik 127 (2016) 9212–9221 9219

Fig. 6. The real parts of the dielectric function ␧1 (␻).

Fig. 7. The absorption edge of Bx Ga1-x N alloys.

The dielectric functions are sufficient to derive other physical properties like the absorption coefficient, refractive index
and energy-loss, etc. The absorption coefficient of B-doped GaN systems, which corresponds well to ␧2 (␻) are shown in
Fig. 7.
As can be seen the absorption coefficient increases with the increase of B concentration, which is mainly, ascribed to
the reduction of Ga 4pstates. It is also clearly that the absorption edge of Bx Ga1-x N alloys shifts to a high energy range with
the increasing B doping concentration compared to that of pure GaN and the main absorption part is still located in the UV
region. This red shift phenomenon is in agreement with the electronic structure presented in Fig. 3.
Another parameter necessary to design devices is the refractivity index (n (␻)). We have determined then plotted in
Fig. 8, the refractivity index (n (␻)) of the BGaN systems in the energy range of 0–30 eV.
For B doping systems, the low photon-energy range (<10 eV) is characterized by appreciable refractivity, whereas, the
11–22 eV photon-energy range, the refractivity undergoes an alteration from the strongest to the second weakest. At the
last region (22–30 eV), the refractivity shows an opposite trend. The deduced refractive index n(0) from Fig. 8 is listed in
Table 3 together with the available experimental and theoretical data. For the binary compounds the results are slightly
overestimate compared to the experimental values but the agreement is satisfactory. For the alloys, we note the lack of the
theoretical and experimental data for comparison.
The electron energy-loss functions (ELF), L(␻) for the ZB and WZ crystals calculated according to the Eq. (5) are shown in
Fig. 9.
The sharp maxima are around 15–22 eV and 23–25 for the blende and wurtzite structures, respectively which are related
to the existence of plasma oscillations. The extracted plasma frequency ␻p , corresponding to the major peak in L(␻), are
9220 A. Said et al. / Optik 127 (2016) 9212–9221

Fig. 8. The refractivity index (n(␻)) of the BGaN systems.

Fig. 9. The electron energy-loss functions (ELF) L(␻) for the ZB and WZ crystals.

summarized in Table 3, that indicates an overall agreement of our results for the binary BN. These results seem useful, due
to deficiency of theoretical and experimental results for the alloy.

4. Conclusion

In summary, density functional calculations have been achieved to investigate the structural, electronic and optical prop-
erties of both ZB and WZ structures of Bx Ga1−x N. We found a reasonably agreement of our calculated structural parameters
with both theoretical and experimental results.The electronic band structure and density of states of BGaN reveal that wide
band gap is due to electronegativity difference between B and Ga elements. Furthermore,the absorption edge of Bx Ga1-x N
alloys shifts to a high energy range with the increasing B doping concentration compared to that of pure GaN and the strong
absorption are located in UV region. The relations of the optical properties to the inter-band transitions are also discussed.
An accurate prediction of the band-gap and optical properties is in progress.

References

[1] H. Kim, Y. Hoon Cho, Nano Lett. 6 (2004) 1059–1062.

[2] Z. Usman, C. Cao, J. Phys. Chem. A 115 (2011) 14502–14509.
[3] V.N. Bessolov, Yu. V. Zhilyaev, Tech. Phys. Lett. 28 (2002) 994.
[4] Z. Usman, C. Cao, J. Phys. Chem. A 115 (2011) 6622–6628.
[5] R. Cusco, E. Llado, Phys. Status Solidi (B) 245 (2008) 731–734.
[6] L. Escalanti, G.L. Hart, Appl. Phys. Lett. 84 (2004) 5.
[7] C.H. Wei, Z.Y. Xie, J. Electron. Mater. 29 (2000) 4.
 A. Said et al. / Optik 127 (2016) 9212–9221 9221

[8] L. Liu, Y.P. Feng, Phys. Rev. B 68 (2003) 104102.

[9] Y. Cai, et al., Solid State Commun. 141 (2007) 262–266.
[10] K. Watanabe, T. Taniguchi, Nat. Mater. 3 (2004) 404–409.
[11] A. Ougazzaden, S. Gautier, Appl. Phys. Lett. 93 (2008) 083118.
[12] K. Atsumi, Y. Inoue, APL Mater. 2 (2014) 032106.
[13] Yong-Nian Xu, W.Y. Ching, Phys. Rev. B 44 (1991) 7787.
[14] A. Lachebi, et al., Int J. Nanoelectron. Mater. 1 (2008) 81–90.
[15] M.B. Kanoun, S. Goumri-Said, Semicond. Sci. Technol. 23 (2008) 125036.
[16] P. Giannozzi, et al., J. Phys.Condens.Matter 21 (2009) 395502.
[17] N. Troullier, José Luriaas Martins, Phys. Rev. B (1993) 43.
[18] J.P. Perdew, A. Zunger, Phys. Rev. B 23 (1981) 5048.
[19] F.D. Murnaghan, Proc. Natl. Acad. Sci. U. S. A. 30 (1944) 244.
[20] P. Grosse, V. Offermann, Appl. Phys. A 52 (1991) 138–144.
[21] R. Riane a, Z. Boussahl a, S.F. Matar, Solid State Sci. 11 (2009) 200–206.
[22] Z.H. Levine, D.C. Allane, Phys. Rev. B 43 (1991) 4187.
[23] T. Honda, Shibata, J. Appl. Phys. 39 (2000) 2389.
[24] M.B. Kanoun, Goumri, J. Appl. Phys. 98 (2005) 063710.
[25] H. Baaziz, Z. Charifi, Phys. Status Solidi (B) 1 (2006) 10.
[26] M.-H. Tsai, F.C. Peiris, S. Lee, Phys. Rev. B 65 (2002) 235202.
[27] Yong-Mian Xu, W.Y. Ching, Phys. Rev. B: Condens. Matter 48 (1993) 4335.
[28] M.I. Eremets, et al., Phys. Rev. B 52 (1995) 8854.
[29] P. Perlin, I. Gorczyca, Phys. Rev. B 45 (1992) 307.
[30] D.R. Penn, Phys. Rev. 128 (1962) 2093.
[31] V.V. Ilyasov, Phys. Soild State 48 (2006) 4.
[32] J.H. Edgar, Properties of Group III Nitrides, INSPEC, London, 1994.
[33] H. Xia, Q. Xia, A.L. Ruoff, Phys. Rev. B 47 (1993) 12925.
[34] A.F. Wright, J.S. Nelson, Phys. Rev. B 51 (1995) 7866.
[35] M. Fuchs, J.L. Da Silva, Phys. Rev. B 65 (2002) 245212.
[36] M.E. Sherwin, T.J. Drummond, J. Appl. Phys. 69 (1987) 8423.
[37] Kwiseon Kim, Phys. Rev. B 53 (1996) 16316.
[38] M.B. Kanoun, A.E. Merad, Solid-State Electron. 48 (2004) 1601.
[39] I. Vurgaftman, J.R. Meyer, L.R. Ram-Mohan, J. Appl. Phys. 89 (2001) 5815.
[40] M. Rimsditch, E.S. Zouboulis, A. Polian, J. Appl. Phys. 76 (1994) 832.
[41] L.K. Teles, L.M.R. Scolfaro, Appl. Phys. Lett. 80 (2002) 1177.
[42] Y.F. Tsay, A. Vaidyanathan, Phys. Rev. B: Condens. Matter 19 (1979) 5422.
[43] D. Vogel, P. Kruger, Phys. Rev. B 55 (1997) 12836.
[44] V.V. Ilyasov, T.P. Zhdanova, Phys. Solid State 45 (2003) 816.
[45] J. Chen, Z.H. Levine, J.W. Wilkins, Appl. Phys. Lett. 66 (1995) 1129.
[46] U. Kohler, D.J. As, B. Schottker, J. Appl. Phys. 85 (1999) 404.
[47] N.E. Christensen, I. Gorczyca, Phys. Rev. B 50 (1994) 4397.
[48] A.S. Barker, M. Ilegems, Phys. Rev. B 7 (1973) 743.
[49] N. Miyata, K. Mariki, Phys. Rev. B 40 (12) (1989) 028.
[50] R. Riane, A. Zaoui, Physica B 405 (2010) 985–989.
[51] M. Steube, K. Reimann, Phys. Status Solidi (B) 57 (1999) 211.
[52] T. Lei, T.D. Moustakas, J. Appl. Phys. 71 (1992) 10.
[53] E. Ejder, Phys. Status Solidi (A) 6 (1971) 445–448.
[54] T. Egawa, J. watanabe, Appl. Phys. Lett. 70 (1997) 24.
[55] G. Cappelini, Phys. Rev. B 64 (2001) 035104.