IAES International Journal of Artificial Intelligence (IJ-AI)

Vol. 8, No. 1, March 2019, pp. 33~43

ISSN: 2252-8938, DOI: 10.11591/ijai.v8.i1.pp33-43 r 33

Comparative between (LiNbO3) and (LiTaO3) in detecting

acoustics microwaves using classification

Hafdaoui Hichem1, Benatia Djamel2

1,2
Electronics Department, Faculty of Technology, University of Batna 2 (Mostefa Benboulaid), Algeria

Article Info ABSTRACT

Article history: Our work is mainly about detecting acoustics microwaves in the type of BAW
(Bulk acoustic waves), where we compared between Lithium Niobate
Received Jun 12, 2018 (LiNbO3) and Lithium Tantalate (LiTaO3), during the propagation of acoustic
Revised Feb 11, 2019 microwaves in a piezoelectric substrate. In this paper, We have used the
Accepted Feb 25, 2019 classification by Probabilistic Neural Network (PNN) as a means of numerical
analysis in which we classify all the values of the real part and the imaginary
part of the coefficient attenuation with the acoustic velocity for conclude
Keywords: whichever is the best in utilization for generating bulk acoustic waves.This
study will be very interesting in modeling and realization of acoustic
Microwaves microwaves devices (ultrasound) based on the propagation of acoustic
Numerical analysis microwaves.
Probabilistic Neural Network
(PNN)
Piezoelectricity Copyright © 2019 Institute of Advanced Engineering and Science.
All rights reserved.

Corresponding Author:
Hafdaoui Hichem,
Electronics Department, Faculty of Technology,
University of Batna 2 (Mostefa Benboulaid),
Algeria.
Email: hichemhafdaoui@yahoo.fr

1. INTRODUCTION
Multiferroic materials exhibiting ferroelectricity and ferromagnetism in the same phase have attracted
much attention due to their basic physics and potential technological applications in recent years [1, 2].
Lithium tantalate (LiTaO3) and lithium niobate (LiNbO3) are well-known materials for ferroelectric,
piezoelectric, acoustooptical, electro-optical and nonlinear optical applications [3, 4]. The interest in the use of
piezoelectric materials as the wave propagation medium lies in the propagation of appearance. These waves in
this case, spread in a resilient part (or acoustic) and power; hence the name electroelastic or electroacoustic [5].
Various types of transducers such as bulk acoustic wave (BAW) (Bulk acoustic waves) transducers, shear wave
transducers and interdigital transducers IDTs (Inter-Digitized transducers) are reported for generation and
reception of acoustic waves in SAW (Surface acoustic waves) devices. IDTs are widely used transducer in
SAW devices and they are metallic comb shaped electrodes fabricated over a piezoelectric substrate [6]. Theory
on generation and propagation of BAW in a SAW device with IDTs is well explained in [5, 7].
Some materials, when deformed, become electrically polarized.This effect is known as piezoelectricity [8, 9].
In this work, we searched for all Acoustic velocity and Attenuation coefficients on the level of Piezoelectric
substrate (LiNbO3) and then we compared it with the results of Piezoelectric substrate (LiTaO3) ,to conclude
wich of them is better in utilization for generating bulk acoustic waves.

Journal homepage: http://iaescore.com/online/index.php/IJAI

34 r ISSN: 2252-8938

Interdigital transducer

Input of signal Z

Output of signal
Surface
waves X
Piezoelectric substrate
(LiNbO3) or (LiTaO3)

Bulk acoustic waves (B.W)

(into the crystal)

Figure 1. LiNbO3 or LiTaO3 Crystals excited by transducer.

2. PHENOMENOLOGICAL TENSORIAL PIEZOELECTRIC EQUATIONS

The electrical induction and the electric field E in the piezoelectric materials are bound by: [8-11]

D = ε .E + eT .S (1)

Where ε is the permittivity tensor (F/m)?

e: piezoelectric tensor (c/m)
S: strain tensor (Sij=1/2[∂Ui/∂Xj+∂Uj/∂Xi])

The electric polarization of the medium under the effect of deformation, also involves the creation of stresses
under the effect of an external electric field. This constraint is:

T = C. S.− e. E (2)

where C : represents the elastic tensor (N/m²)

In one form the stress tensor and the electric induction are defined as follows: [8-11]

Tij = C ijkl . S kl − e kij . E k (3)

D i = e jkl . S kl + ε ik . E k (4)

with i,j,k,l =1,2,3

In the quasi-static approximation, the Maxwell equations reduce to the Poisson equation:

! ∂D i
div. D = =0 (5)
∂X i

The movement of the particles under the action of stress is described by Newton's equation:

∂u ²
∇T = ρ. (6)
∂t ²
Substituting (3) and (4) in (5) and (6) we obtain the piezoelectric tensor phenomenological equations:

IJ-AI Vol. 8, No. 1, March 2019: 33 – 43

IJ-AI ISSN: 2252-8938 r 35

2 2 2
∂ Uk ∂ U ∂ Uj
C ijkl + e lij 4 =ρ
(7)
∂ .Xi ∂ .X l ∂ .X k ∂ .Xi 2
∂ .t

2 2
∂ Uk ∂ U
e ikl − ε ik 4 =0
(8)
∂ .Xi ∂ .X l ∂ .X k ∂ .Xi

3. GENERAL FORM OF THE SOLUTION

The piezoelectric wave is the solution of the elastic and electric equations that satisfy elastic and
electric boundary conditions. When these waves attenuate exponentially inside the piezoelectric material as
shown in Figure 1, they are known as surface waves and if this is not the case they are known as bulk waves.
Consider the following form of the surface wave (partial wave).

U i = u i exp(jβ. α i X ) (9)
3
where ui(i=1,2,3) are displacement amplitudes, ui(i=4) represents the amplitude of the electric
potential, β is the propagation constant, αi are the attenuation coefficients of the wave within the piezoelectric
crystal (axis Y show in Figure 1) and ω is the angular pulsation.We will be interested in the coefficient αi. With
a chosen LiNbO3 (Lithium Niobate) or LiTaO3 (Lithium Tantalate). Solutions of this type of wave correspond
to waves propagating with or without attenuation along the direction X. The elastic displacements Ui and the
electric potential U4 may vary from the normal direction to the flat surface (Y), but following invariant Z as
show in Figure 1. Substituting (11) into (9) and (10) we obtain the system:

[A].[U] = [0] (10)

Where [A] is a symmetric matrix 4x4

T
[U] = [u1, u2, u3, u4] Components determine

3.1. For LiNbO3 (Lithium Niobate)

Development of the determinant of [A] of the general, an eighth order polynomial:

8 i
Det ( A) = ∑ Bi .α = 0 (11)
i=0
This polynomial is called dispersion equation or secular equation.
The solution of (11) gives for each β (β = 2.π.f / Vs: constant propagation) eight roots. These roots are based
on Vs (acoustic velocity). Each root generates three displacement components of the particle and an electric
potential. Thus, the general solution is a combination of eight roots (8 secondary wave) given by the expression:

n =1
{
U i = ∑ C n . D in .exp jβ. α n . X 3 + jβX 1 (1 + j. γ ) − j. β. V. t} (12)

i=1,4
where D in are the components of the eigenvector of the system (10) associated with the eigenvalue α n
Cn: constant to be determined by the boundary conditions.

3.2. For LiTaO3 (Lithium Tantalate)

Development of the determinant of [A] of the general, an sixth order polynomial:

Comparative between (LiNbO3) and (LiTaO3) in detecting acoustics microwaves… (Hafdaoui Hichem)
36 r ISSN: 2252-8938

6 i
∑ β i .α = 0 (13)
i =0
This polynomial is called dispersion equation or secular equation.
The solution of (13) gives for each β (β = 2.π.f/Vs: constant propagation) six roots. These roots are based on
Vs (acoustic velocity). Each root generates three displacement components of the particle and an electric
potential. Thus, the general solution is a combination of six roots (6 secondary wave) given by the expression:

n =1
{
U i = ∑ C n . D in .exp jβ. α n . X 3 + jβX 1 (1 + j. γ ) − j. β. V. t} (14)

i=1,3
where D in are the components of the eigenvector of the system (10) associated with the eigenvalue α n
, Cn: constant to be determined by the boundary conditions.

4. BEHAVIOR OF THE ROOTS OF THE SECULAR EQUATION

The behavior of the roots depends on the acoustic velocity Vs at a velocity less transverse to the low
velocity of volume in the medium, all the roots are complex conjugate pairs:

α (1) = α (re1) + jα (im1) et α ( 2 ) = α (re2 ) + jα (im2 )

where α (rei) = α (rei+1) et α (imi) = α (imi+1) with j=1,3,5,7

This corresponds to the partial wave:

[ ]
U (i 1) ≈ exp j. β. α (re1) . X 3 − β. α (im1) . X 3 .exp( jβ. X 1 ) (15)

[ ]
U (i 2 ) ≈ exp j.β. α (re2 ) . X 3 + β. α (im2 ) . X 3 .exp( jβ. X 1 ) (16)

After development we have:

[ ]
U (i 1) ≈ exp −β. α (im1) . X 3 .exp j.β α (re1) . X 3 + X 1 [ ] (17)

[ ]
U (i 2 ) ≈ exp +β. α (im2 ) . X 3 .exp j.β α (re2 ) . X 3 + X 1[ ] (18)

Remarks
if α im(i ) = 0 and α (1)
re < 0, the bulk waves will be obtained [12-14].

5. PROBABILISTIC NEURAL NETWORK

A Probabilistic Neural Network (PNN) is defined as an implementation of statistical algorithm called
Kernel discriminate analysis in which the operations are organized into multilayer feed forward network with
four layers: input layer, pattern layer, summation layer and output layer. A PNN is predominantly a classifier
since it can map any input pattern to a number of classifications. [15, 16] Among the main advantages that
discriminate PNN is: Fast training process, an inherently parallel structure, guaranteed to converge to an
optimal classifier as the size of the representative training set increases and training samples can be added or
removed without extensive retraining. Accordingly, a PNN learns more quickly than many neural networks
model and have had success on a variety of applications. Based on these facts and advantages,
PNN can be viewed as a supervised neural network that is capable of using it in system classification andpattern
recognition [16]. The PNN consists of nodes allocated in three layers after the inputs show in
Figure 2. [15, 17]

IJ-AI Vol. 8, No. 1, March 2019: 33 – 43

IJ-AI ISSN: 2252-8938 r 37

Figure 2. The architecture of PNN [15,17]

6. APPLICATION ON LiNbO3 (Lithium Niobate)

In this example, we will show the variations of the real and the imaginary parts of attenuation
coefficients (α3) show in Figure 3 and (α4) show in Figure 4, we will indicate the presence of bulk waves (zero
imaginary part and real part is negative) and for a good detection of bulk waves we should have (Imaginary
part equals zero and the minimal negative value of real part). We set the value of the real part for (α3), which
we will look for it in the Classification as shown Figure 3.
Real part

zero imaginary
part and the
minimal negative
value of real part

imaginary part

Figure 3. Plot - Real part versus Imaginary part

We set the value of the real part for (α4), which we will look for it in the Classification show in Figure 4.

Comparative between (LiNbO3) and (LiTaO3) in detecting acoustics microwaves… (Hafdaoui Hichem)
38 r ISSN: 2252-8938

Real part

zero imaginary part and

the minimal negative
value of real part

imaginary part

Figure 4. Plot - Real part versus Imaginary part

7. RESULTS AND DISCUSSION (for LiNbO3 Lithium Niobate)

7.1. For coefficient attenuation α3
To know the accurate values of each of the real part and the imaginary part which represent coefficient
attenuation α3 and Acoustic velocity, Selection variable stays always Acoustic velocity and we change the
Classification factor where we find: show in Figure 5, we have classification where we find classification factor
is imaginary part and selection variable is acoustic velocity and number of cases in training set 51, in the
minimal negative value of real part we find that the imaginary part equals 0. Show in Figure 6, we have
classification where we find classification factor is real part and selection variable is acoustic velocity and
number of cases in training set: 51, We find that the minimal negative value of real part equals -0.5654. Show
in Figure 7, we have classification where we find classification factor acoustic velocity and selection variable
acoustic velocity and number of cases in training set: 51, in the minimal negative value of part part (-0.5654)
We find that the acoustic velocity equals 4050 m/s.

zero imaginary
part and the
minimal negative
value of real part

Figure 5. Neural Network Bayesian Classifier - Imaginary part (Acoustic velocity)

IJ-AI Vol. 8, No. 1, March 2019: 33 – 43

IJ-AI ISSN: 2252-8938 r 39

zero imaginary part and

real part= -0.5654

Figure 6. Neural Network Bayesian Classifier - Real part (Acoustic velocity)

zero imaginary part

and real part= -0.5654
and Vs=4050 m/s

Figure 7. Neural Network Bayesian Classifier - Acoustic velocity (Acoustic velocity)

7.2. For coefficient attenuation α4

To know the accurate values of each of the real part and the imaginary part which represent coefficient
attenuation α4 and Acoustic velocity, Selection variable stays always Acoustic velocity and we change the
Classification factor where we find: show in Figure 8, we have classification where we find classification factor
is imaginary part and selection variable is acoustic velocity and number of cases in training set 51, in the
minimal negative value of real part we find that the imaginary part equals 0. Show in Figure 9, we have
classification where we find classification factor is real part and selection variable is acoustic velocity and
number of cases in training set: 51, We find that the minimal negative value of real part equals - 0,57212. Show
in Figure 10, we have classification where we find classification factor acoustic velocity and selection variable
acoustic velocity and number of cases in training set: 51, in the minimal negative value of part part (- 0,57212)
We find that the acoustic velocity equals 4000 m/s.

Comparative between (LiNbO3) and (LiTaO3) in detecting acoustics microwaves… (Hafdaoui Hichem)
40 r ISSN: 2252-8938

zero imaginary
part and the
minimal negative
value of real part

Figure 8. Neural Network Bayesian Classifier - imaginary part (Acoustic velocity)

zero imaginary part and

real part= - 0,57212

Figure 9. Neural Network Bayesian Classifier - Real part (Acoustic velocity)

zero imaginary part and

real part= - 0,57212
and Vs=4000 m/s

Figure 10. Neural Network Bayesian Classifier - Acoustic velocity (Acoustic velocity)

IJ-AI Vol. 8, No. 1, March 2019: 33 – 43

IJ-AI ISSN: 2252-8938 r 41

Table 1. Summarizes the full work where we can notice bulk waves detection.
Coefficient Attenuation Imaginary part Real part Acoustic velocity Bulk waves (BAW)
α1 All values All values All values No Detection
α2 All values All values All values No Detection
α3 0 -0,5654 4050 m/s Good Detection
α4 0 -0,57212 4000 m/s Good Detection
α5 All values All values All values No Detection
α6 All values All values All values No Detection
α7 All values All values All values No Detection
α8 All values All values All values No Detection

8. APPLICATION ON LiTaO3 (Lithium Tantalate)

In this example, we will show the variations of the real and the imaginary parts of attenuation
coefficients (α2) show in Figure 11, we will indicate the presence of bulk waves (zero imaginary part and real
part is negative) and for a good detection of bulk waves we should have (Imaginary part equals zero and the
minimal negative value of real part). We set the value of the real part for (α2), which we will look for it in the
Classification show in Figure 11.
Plot of Real part vs Imaginary part

5,4

3,4
Real part

1,4

-0,6
zero imaginary
part and the
-2,6 minimal negative
value of real part

-4,6
-2 0 2 4 6
Imaginary part

Figure 11. Plot - Real part versus Imaginary part

9. RESULTS AND DISCUSSION (for LiTaO3 Lithium Tantalate)

9.1. For Coefficient Attenuation α2
To know the accurate values of each of the real part and the imaginary part which represent coefficient
attenuation α3 and Acoustic velocity, Selection variable stays always Acoustic velocity and we change the
Classification factor where we find: show in Figure 12, we have classification where we find classification
factor is imaginary part and selection variable is acoustic velocity and number of cases in training set 51, in the
minimal negative value of real part we find that the imaginary part equals 0. Show in Figure 13, we have
classification where we find classification factor is real part and selection variable is acoustic velocity and
number of cases in training set: 51, We find that the minimal negative value of real part equals -0.5654. Show
in Figure 14, we have classification where we find classification factor acoustic velocity and selection variable
acoustic velocity and number of cases in training set: 51, in the minimal negative value of part part (-0.5654)
We find that the acoustic velocity equals 4050 m/s.

Comparative between (LiNbO3) and (LiTaO3) in detecting acoustics microwaves… (Hafdaoui Hichem)
 Real part
-1,6
-4,594
-1,59999
-4,19979
-1,58
-4,18011
-1,56001
-4,14465
-1,49231
-4,12408
Scatterplot -1,48454
42 r -4,08499
ISSN:-1,47654
2252-8938
-4,08339
-1,42987
-4,07547
-1,41232
5,4 Imaginary
Imaginary
-4,00947
part part -1,4
Scatterplot -2 -2 -2,34345
Scatterplot -1,39765
-1,9321
-1,9321 -1,41232
-1,34345
-1,9 -1,4
3,4 -1,9
-1,8 -1,32454
5,4 5,4 -1,8 -1,32454
-1,7 -0,01997
-1,69001
-1,7 -0,4
-0,01996
Real part

-1,68011 -0,39901
-0,01988
1,4 3,4
3,4 -1,69001
-1,66
Real part 0 -0,39878
-1,68011
-1,6
-0,38765
-4,594
-1,59999 1,16874
-1,66

Real part
1,4 -1,58 -4,19979 -0,37878
1,18432

Real part
-0,6 1,4 -1,6
-1,56001
-4,18011 -0,35672
-1,49231
-4,14465
-1,59999 1,19723
-0,6 -1,48454
-4,12408 -0,34769
1,2
zero imaginary -1,58
-1,47654
-4,08499 -0,28999
1,41232
-2,6 -0,6 part and the -1,56001
-1,42987
-4,08339 -0,0399
2,24656
-2,6 -1,41232
-4,07547
minimal negative -1,49231
-1,4 -4,00947
-0,038
2,28876
value of real part -1,48454
-1,39765
-2,34345 -0,035
2,32454
-4,6 -4,6 -2,6 -1,47654
-1,34345
-1,41232 -0,03454
-2 0 2 4 6 -1,32454
-1,4
4,34345
-2 0 2 Imaginary 4
part 6 -1,42987
-0,01997
-0,0328
5,39765
-1,32454
Imaginary part -1,41232
-0,01996
-0,4
-0,02987
5,4
-4,6 -0,01988 -0,02888
0
-1,4
-0,39901
-2 0 2 4 6
-0,39878
-1,39765 -0,0276
Figure 12. Neural Network Bayesian Classifier - Imaginary
Imaginary part part (Acoustic
1,16874 velocity)
-0,38765 -0,02448
Real part -1,34345
1,18432
-0,37878
1,19723
-0,35672 -0,02001
-4,5941,2-1,32454
-0,34769 -0,0128
5,4 -4,19979 -0,01997
1,41232
-0,28999
Real part -0,01001
5,4 2,24656
-4,18011
-4,594 -0,01996
-0,0399
2,28876
-0,038
0
-4,14465
-4,19979 -0,01988
2,32454
-0,035 0,02
5,4 3,4 -4,18011
-4,12408
-4,14465
0Acoustic
4,34345
-0,03454 Velocity(Vs)
0,0343
3,4 -4,08499
5,39765
5,4 -4,12408 5,41,16874
-0,0328
2400 0,0432
Acoustic Velocity(Vs)
3,4 -4,08339
-4,08499
-0,02987
1,18432 2400
0,0723
Real part

1,4 -4,07547
-4,08339
2450
-0,02888 2450
1,19723
Real part

1,4 Scatterplot 3,4 -0,0276

-4,00947 -0,02448
-4,07547 2500 0,0821
2500
Real part

Scatterplot 1,4 -2,34345

-4,00947 1,2 2550 0,092550
Scatterplot -0,02001 2600
-0,6 -2,34345
-1,41232 1,41232 0,09899
2600
Real part

1,4 -0,0128 2650

-0,6 Scatterplot -1,41232
-1,4 2,24656
-0,01001 0,3 2700
-0,6 -1,4 0 2650 0,32454
2750
-1,32454 2,28876
-0,6 -1,32454 0,022700 2800
1,0576
-2,6 zero imaginary part and -0,4-0,4 2,324542750
0,0343 2850
-2,6 -2,6 real part= - 4,594 -0,39901
-0,39901 4,34345
0,0432 1,42987
2900
-0,39878
-0,39878 2800
0,0723 2,06783
2950
-2,6 5,39765 3000
-4,6
-0,38765
-0,38765 2850
0,0821 2,09321
-4,6
-4,6 -0,37878 5,40,09 3050
-2 0 2 0 4 6 -0,37878 0,09899
-0,35672
2900 3,083100
-4,6 -2 2 4 6
-2 0 2
Imaginary part 4 Imaginary part 6 -0,35672 0,3 2950
-0,34769 4,0789
3150
-2Imaginary 0part 2 4 6 3200
Imaginary part -0,34769 0,32454
-0,28999 3000 5,06987
3250
-0,28999 1,0576
-0,0399
3050 3300
-0,038 1,42987
Figure 13. Neural Network Bayesian Classifier - Real part (Acoustic -0,0399 velocity)3100
2,06783
3350
Scatterplot -0,035 3400
-0,038 2,09321
-0,03454
-0,035 3,08
3150 3450
-0,0328 3500
-0,03454 4,0789
-0,02987
3200 3550
-0,0328 5,06987
-0,02888 3250 3600
5,4 -0,0276 Acoustic Velocity(Vs)
-0,02987 3300 3650
Scatterplot Scatterplot
-0,02448 2400 3700
-0,02888
-0,02001 2450
3350 3750
-0,02762500 3400
-0,0128
3800
3,4 5,4 5,4 -0,02448 2550 3850
2600 3450
-0,01001 3900
0 -0,02001
2650 3500 3950
3,4 0,02-0,01282700 4000
3,4 3550
Real part

0,0343
-0,01001 2750 4050
1,4 0,0432 2800 3600 4100
0
Real part

1,4 0,0723 4150

Real part

1,4 0,02
0,0821
2850 3650 4200
2900
0,090,03432950 3700 4250
-0,6 -0,6 0,04323000 3750
0,09899 4300
-0,6 4350
0,3 0,07233050
0,32454 3100
3800 4400
-2,6 0,0821 3850 4450
-2,6 zero imaginary part and 1,0576 3150 4500
-2,6 0,09 3200
real part= - 4,594 1,42987 3900 4550
0,09899
2,06783 3250 4600
and Vs=3450 m/s-4,6 0,3
2,09321 3300 3950 4650
-4,6 -2 0 2 4 6
-2 0 2 4 Imaginary60,32454
3,08 part 3350 4000 4700
-4,6 Imaginary part 4,0789 3400 4050
1,05763450 4750
5,06987 4800
-2 0 2 4 6 1,42987 3500 4100 4850
Imaginary part 2,06783 3550 4150 4900
2,09321 3600
3650 4200
Figure 14. Neural Network Bayesian Classifier - Acoustic velocity3,08 (Acoustic velocity)
3700 4250
4,07893750
5,06987
4300
3800
IJ-AI Vol. 8, No. 1, March 2019: 33 – 43 3850 4350
3900 4400
3950
4000 4450
4050 4500
4100
4150
4550
4200 4600
IJ-AI ISSN: 2252-8938 r 43

Table 2. Summarizes the full work where we can notice bulk waves detection.
Coefficient Attenuation Imaginary part Real part Acoustic velocity Bulk waves (BAW)
α1 All values All values All values No Detection
α2 0 -4,594 3450 m/s Good Detection
α3 All values All values All values No Detection
α4 All values All values All values No Detection
α5 All values All values All values No Detection
α6 All values All values All values No Detection

10. CONCLUSION
In this article, we explained that the phenomenon of bulk waves are relying on numerical results at
the level of attenuation coefficients. Changes in real and imaginary parts of the coefficients based on the
acoustic velocity to detect these waves. Since the bulk waves were detected in Lithium Tantalate (LiTaO3) in
one coefficient attenuation α2 and in Lithium Niobate (LiNbO3) were detected in two coefficients attenuation
α3 and α4 so we conclude that (LiNbO3) is the better to use in generating bulk acoustic waves. To compare
between many piezoelectric substrates in detecting acoustic micro waves using PNN classification you just
have to change the characteristics of the material parameters, and those results we need them in realization of
acoustic microwaves devices.

REFERENCES
[1] Spaldin NA, Cheong S W and Ramesh R, Multiferroics : Past, present, and future, Phys. Today, Vol 63, 2010, pp 38-
43.
[2] Tao Yan, Ning Ye, Liuwei Xu, Yuanhua Sang, Yanxue Chen, Wei Song, Xifa Long, Jiyang Wang and Hong Liu,
Ferromagnetism in chemically reduced LiNbO3 and LiTaO3 crystals, Journal of Physics D : Applied Physics, 2012,
Vol 49, 195005.
[3] T. Hatanaka, K. Nakamura, T. Taniuchi, H. Ito, Y. Furukawa, K. Kitamura, Quasi-phase-matched optical parametric
oscillation with periodically poled stoichiometric LiTaO 3, Optics Letters, Vol 25,2000, pp 651-653.
[4] J.C. Wu, Z.B. Chen, R.K. Choubey, C.W. Lan, On the study of zinc doping in congruent LiTaO3 crystals, Materials
Chemistry and Physics, Vol 133, 2012, pp 813-817.
[5] Daniel Royer and Eugene Dieulesaint, Elastic Waves in Solids II – Generation, Acoustic-optic Interaction,
Applications, Springer-Verlag, New York (1999).
[6] Mingxi Deng, Simulation of Generation of Bulk Acoustic Waves by Interdigital Transducers, IEEE Ultrasonics
Symposium, Vol 1, 2001, pp 855-858.
[7] Ken-ya Hashimoto, Masatsune Yamaguchi, Gunter Kovacs, Karl Christian Wagner, Werner Ruile, and Robert
Weigel, Effects of the Bulk Wave Radiation on IDT Admittance on 42o YX-LiTaO3, IEEE Transactions on
Ultrasonics, Ferroelectrics, and Frequency Control, Vol 18, No 5, pp1419-1425, 2001.
[8] D. Benatia and M. Benslama, “Analysis of Leaky and Bulk Acoustic Microwaves by Wavelet Technique,” Journal
of Communications in Numerical Methods in Engineering, U.K, Vol 16, pp165-175, 2000.
[9] H. Hafdaoui and D. Benatia, Identification of Acoustics Microwaves (Bulk Acoustic Waves) in Piezoelectric
Substrate (LiNbO3 Cut Y-Z) by Classification Using Neural Network, Journal of Nanoelectronics and
Optoelectronics, Vol 10, No 03, pp 314-319, 2015.
[10] D. Benatia and M. Benslama, Identification of bulk microwaves and detection of Pseudo singularities by wavelet
technique in piezoelectric medium, International Journal of Electronics, Vol 93, No 8, August 2006, pp 567–579.
[11] D.Benatia, Modélisation des ondes de volume rampantes sous la surface (SSBW) par une technique mixte ondelettes
et fractales, Thèse de Doctorat d’Etat, Université de Constantine, institut d’électronique, 1999.
[12] I. D. Avramov: "High Q metal strip SSBW resonator using a SAW design", IEEE Transactions on ferroelectrics and
frequency control, Vol 37, pp 530-534, 1990.
[13] D. Benatia et M. Benslama - Identification des Ondes de fuite et de volume dans les Matériaux piézoélectriques. Cas
du Niobate de Lithium (LiNbO3 Coupe Y-Z) ‘’ 3rd International conference: Sciences of electronic, technologies of
Information and telecommunications March 27-31, 2005 – TUNISIA’’
[14] D. Benatia, T. Fortaki et M. Benslama -Détection des ondes de fuite dans les matériaux piézoélectriques ‘’ 4th
International conference: Sciences of electronic, technologies of information and telecommunications March 25-29,
2007 – TUNISIA’’
[15] H.Hafdaoui, Détection et génération des micro-ondes acoustiques de volume dans les structures piézoélectriques,
Thèse de Doctorat, Université de Batna 2, 2016.
[16] Pv Nageswara Rao, T Uma Devi, Dsvgk Kaladhar, Gr Sridhar, Allam Appa Rao, “A Probabilistic Neural Network
Approach for Protein Superfamily Classification ‘‘Journal of Theoretical and Applied Information Technology, Vol
6, No1, pp 101-105, 2009.
[17] L.Abdelhadi, R.Tlemsani, A. Benyettou. “Un Nouvel Algorithme d‘Apprentissage pour les Réseaux Probabilistes”,
11th African Conference on Research in Computer Science and Applied Mathematics. pp 208-215. 2012.

Comparative between (LiNbO3) and (LiTaO3) in detecting acoustics microwaves… (Hafdaoui Hichem)