Study On The Semiconducting Properties

Download as pdf or txt
Download as pdf or txt
You are on page 1of 155

HELWAN

University

Study of The Physical Properties of Some


Semiconductor Materials.
Thesis Submitted
In
Partial Fulfillment for the Requirement of the Master
Degree in Science (Solid State Physics)
To
Physics Department, Faculty of Science, Helwan University
By
Amira Ali Abdel-Wahab

B. Sc. in Physics, 2007

2012
Study the Physical Properties of Some
Semiconductor Materials.
Thesis Submitted
In
Partial Fulfillment for the Requirement of the Master
Degree in Science (Solid State Physics)
To
Physics Department, Faculty of Science, Helwan University
By
Amira Ali Abdel-Wahab
B. Sc. in Physics, 2007
Supervisors:

Prof. Dr. Abdel-Rahman-Abdel-El-Mongy.


Head of Physics Department, Faculty of Science,
Helwan University.

Assist.Prof. Dr. Hatem Hassan Amer.


Solid States and Electrons Accelerator Department,National Researches
Center For Radiation and Technology Atomic Energy Authority.

Assist.Prof.Dr.Yasser El-Gendy.
Physics Department, Faculty of Science,
Helwan Univerity.
Study of the Physical Properties of Some
Semiconductor Materials.
Thesis Submitted
In
Partial Fulfillment for the Requirement of the Master
Degree in Science (Solid State Physics)
To
Physics Department, Faculty of Science, Helwan University
By
Amira Ali Abdel-Wahab

B. Sc. in Physics, 2007


Examiners:

Prof. Dr. Sayed Mohamed El-Araby.


Chairman of Atomic Energy Authority.Cairo
Chairman of Nuclear Researches Center-Anshas.
Prof. Dr. Mohmed Mohamed El-Ocker.
Physics Department, Faculty of Science,
El-Azher University.

Assist .Prof. Dr. Hatem Hassan Amer.


Solid States and Electrons Accelerator Department,National Researches
Center For Radiation and Technology Atomic Energy Authority.

Prof. Dr. Prof. Dr. Abdel-Rahman-Abdel-El-Mongy.


Head of Physics Department, Faculty of Science,
Helwan University.
To my beloved husband,

And to my beautiful future baby,

To my great father,

To my loving mother,

To my supportive brothers and sisters,


Contents i

Contents

List of Figures .................................................................................iv


List of Tables...................................................................................vi
List of Symbols................................................................................
viii
Acknowledgment ............................................................................xi
Abstract ...........................................................................................
xiii
Summary .........................................................................................
xv
Chapter I Introduction and Literature Survey ...........................1
1.1. Introduction .............................................................................1
1.2. Literature Survey ................................................................ 4
1.3. Aim of the Work 13
15
Chapter II Theoretical Background .............................................
2.1. Disordered systems ................................................................15
2.2. The Chalcogenide Glasses ......................................................
15
2.3. Electronics Band Structure and Defects ...............................
16
2.3.1. Types of Defects ................................................................ 20
2.4. Thin Film Growth Process .....................................................
23
2.5. Optical Properties of Amorphous Materials ........................
26
2.5.1. Optical Absorption Mechanisms ........................................
27
2.5.2. Absorption Edge ................................................................ 30
2.6. The Electrical Properties of Chalcogenide Glasses..............
33
2.7. Switching in Alloys Glasses ....................................................
35
2.8. Radiation Sources ................................................................ 37
2.8.1. Gamma Radiation ................................................................
37
Contents ii

2.8.2. Interaction Of Gamma Radiation with Matter .................


38
2.9. Radiation Effects on Solids .....................................................
41
2.9.1. Defect Production by Gamma – Rays ................................
43
Chapter III Experimental Techniques .........................................
44
3.1. Preparation of Bulk compositions .........................................
44
3.2. Preparation of Thin Films ......................................................
47
3.3. Methods for Thin Film Thickness Measurment...................
50
3.3.1. Quartz crystal thickness monitor technique ......................
51
3.4. Density Determination ............................................................
52
3.5. Structure Measurements ........................................................
53
3.6. Optical Measurements ............................................................
60
3.7. Electrical Measurements ........................................................
61
3.8. Spectrophotometer Measurements ........................................
68
3.9. Gamma Irradiation Source ....................................................
70
Chapter IV Results & Discussion .................................................
72
4.1. X-ray Diffraction Identification of Bulk Samples ................
72
4.2. Energy dispersive X-ray analysis (EDX)...............................
73
4.2.1. Scanning electron microscope technique ...........................
76
4.3. Differential Thermal Analysis (DTA) ................................ 78
4.4. The Density dependence of (In) content ................................
79
4.5. The Effect of In Content on Conductivity ............................
82
4.6.Switching Characteristics of Amorphous 87
Semiconductor…………………………………………………
4.6.1.Temperature dependence of Switching 89
Characteristics………………………………………………..
Contents iii

4.7. Theoretical Study of I-V Characteristics of Switching ........


92
4.7.1. The Suggested Switching Model .........................................
95
4.8. Optical Band Gap of In-Se-Sn-Bi Thin Films ......................
97
4.9. Thermal and Radiation–induced defects in thin film 102
devices .............................................................................................
4.10. Effect of Gamma-Irradiation on Optical Band Gap .........
102
4.10.1. Effect of Gamma-Irradiation on Threshold 102
Switches…………………………….
4.11. Modern Applications of InxSn20Se60-xBi20 ............................
103
4.12. Trends for Future Work .......................................................
105
Chapter V Summary and Conclusion ..........................................
107
References .......................................................................................
111
List of Publications .........................................................................
120
List of Figures iv

List of Figures
Figure Page
Title
No. No.
(2-1) Showing bonding in (a) Ge and (b) Se………….…….……… 17
Shows various forms proposed for the density of states in
(2-2) amorphous semiconductors. The shaded areas represent 19
localized states ………………………………….……………
Formation of charged defects (valence alternation pairs) in
( 2-3) 21
chalcogenide glasses………………………………………….
Basic growth processes:(a) island, (b) layer-by- layer, and (c)
(2-4) 26
Stranski-Krastanov type …………………………...…………
(2-5) Absorption spectrum of thin film …………………………… 28
(2- 6) Optical inter band transitions……………………...…………. 30
(2-7) Parts A, B, C of the absorption edge……………….………… 31
Current – Voltage characteristic curve (I-V) of switch Ih
(2-8) 35
denote current …………………………………………………
Dynamic (I-V) characteristic curve for thin film of amorphous
(2-9) 36
semiconductor (Memoryswitch)………………………………..
(2-10) The interaction of gamma rays with matter probabilities ……. 40
Mechanism of the interaction of gamma rays with matter
(2-11) 40
probabilities……………………….…………….....................
(3-1) Silica tubes used for bulk amorphoussemiconductors preparation 45
Design flow chart for Preparation of bulk amorphous
(3-2) 46
InxSn20Se(60-x)Bi20 …………………………………………….
(3-3) a) Schematic diagram b) captured photo Vacuum coating unit. 48
( 3-4) Design flow chart for Preparation of InxSn20Se(60-x )Bi20 Films 49
(3-5) The copper mask designed for E-306A ……….……………… 50
(3-6 ) Thicknesses monitor (TM-200)………………………………. 50

(3-7) X-ray diffractmeter, "Shimadzu XRD-6000"……………….. 54


List of Figures v

Figure Page
Title
No. No.
Typical DTA thermo gram illustrating the definition of the
(3-8) 56
different transition temperature Differential thermal analysis…
(3-9) Captured photo of Differential Thermal Analysis…………..… 57
( 3-10) Principle diagram of Differential thermal analysis……………. 58
(3-11) Design flow chart for DTA thin film Preparation……………… 59
(3-12) AJEOL-5400 Scanning Electron Microscope (SEM) with (EDX) 60
(3-13) Shimadzu UV-160ASpectrophotometer……………………..… 61
(3-14) Gold planar electrodes…………………..…………………….. 62
Construction used for controlling the temperature of the sample
(3-15) in the Range from room Block diagram of the circuit used for 62
measuring electrical Conductivity..........................................
Block diagram of the circuit used for measuring electrical
(3-16) 64
Conductivity…………………………………………………….
Design flow for Preparation of thin film amorphous
(3-17) 66
InxSn20Se(60-x)Bi20 for switching…………………….……….
(3-18) A special cell construction for I-V measurements……………... 67
(3-19) A Simple Circuit used for measuring I-V characteristics D.C… 68
(3-20) SP8 -200 Optical Diagrams……………………………………. 69
(3-21) J6500 Irradiator…………………………………..……………. 71
X-Ray Diffraction patterns of bulk sample of the system
(4-1) 72
InxSn20Se(60-x)Bi20, (where x=0,0.1,0.2 and 0.3at.%) …………..
EDX qualitative analysis for InxSn20Se (60-x) Bi20(where x=0%)
(4-2) 74
bulk amorphous materials…….....................
EDX qualitative analysis for InxSn20Se (60-x) Bi20 (where
(4-3) 74
x=0.1%)…………………..……………………………...…….
(4-4) A SEM photograph of Sn20Se(60-x)Bi20………………………… 77
DTA measurements for In XSn20Se 60-x Bi20 glasses heating rate
(4-5) 79
10c/min………………………………………….…….………..
List of Figures vi

Figure Page
Title
No. No.
Dependence of density on In content in the system InxSn20Se60-
(4-6) 80
xBi20 (with x=0, 0.1, 0.2, 0.3)…………………………………

Variation of ln (σ) vs reciprocal absolute temperature for films


(4-7) of Inx Sn20 Se(60-x)Bi20 where x= 0, 0.1, 0.2 and 0.3 %at 84
constant thickness 100nm……………………………………..
(4-8) Variation of σ as a function of In content……………………. 85
(4-9) Variation of of σ as a function of activation energy………….. 85
Variation of activation energy and cohesive energy as a
(4-10) 86
function of In content…………….. ……………………………
I-V Characteristics of D.C switching For the FilmsInxSn20 Se(60-
(4-11) 89
x)Bi20 at thickness 100nm………………………………..……..

Statics I-V Characteristics curves For x=0, 0.3at.% thin film


(4-12) 90
sample at thickness 100 nm at different ambient temperature T
(4-13) Relation between rise time and cohesive energy……………… 96
Transmission spectra of InxSn2Se(6-X)Bi2 thin film (where x=0,
(4-14) 97
0.1 ,0.2 and0.3 at %)……………….…………………………
Absorption coefficient (α) versus photon energy (hυ) for the
(4-15) 98
Inx Sn20 Se(60-x)Bi20thin film…………………………………..
Best fit of (αhυ)1/2 versus photon energy(hυ) plots for the
(4-16) 99
InxSn20Se(60-x)Bi20thin films…………...………………………
The variation in the optical band gap (E0 ) as function of In
(4-17) 100
content of the InxSn20Se(60-x) Bi20thin films…………………….
List of Tables vi

List of Tables
Table Page
Title
No. No.

(4-1) The composition dependence of density……………………… 80

Shows values of the optical band gap, density, coordination


number, heat of atomization (Hs), bond energy and electro
(4-2) 82
negativities of Sn, Bi, Se and In respectively which are used
for calculations…………...……………………………………

The average coordination number (Nco) and the constraints


(4-3) number (Ns) as function of In content of InxSn20Se (60-x)Bi20 82
glasses…………………………...…………………………….

Compositional dependence of the electrical characteristic


quantities for the thin film glasses in the system InxSn20Se(60-
(4-4) 83
x)Bi20 where x= 0 , 0.1 , 0.2 and 0.3 at constant thickness

100nm………………….………………………………………

Switching characteristics of the composition InxSn20 Se(60-


(4-5) x)Bi20 where x=0, 0.1, 0.2 and 0.3% at constant thickness 89
100nm………………………………………………………….

Values of filament temperature at Vs and after switching as


calculated from experimental (I-V) curves and the
(4-6) corresponding values of current for samples of the 92
composition Inx Sn20Se(60-x) Bi20 where x= 0, 0.3 at thickness
100 nm…………………………………………………………
List of Tables vii

Table Page
Title
No. No.

Values of rise time and cohesive energy for samples of the


(4-7) composition InxSn20Se(60-x)Bi20 (where x=0,0.1,0.2 & 96
0.3at.%)………………………………………………………

Bond energy probabilities and relative probabilities of


(4-8) formation of various bonds in Inx Sn20 Se (60-x) Bi20Films, 101
taking the probability of Sn-Se bond as unity…………...…….
Acknowledgment xi

Acknowledgment

First and last thanks to Allah who gives me the power to go


forward in a way illuminated with His merciful guidance.

I would like to express my thanks to Prof. Dr. H. H. Amer,


Solid State Department, National Center For Radiation Research and
Technology, for giving me the chance to be one of his students and
for his generous advices, valuable discussions which helped me
greatly, and the good proof reading of this thesis. He did not only
guide this work and devote time to discuss it with me but also gave
me the confidence to express my ideas freely.

I will always remember how his ideas and suggestions always


work and how he could simply pick the small mistakes. Actually he
was more than a supervisor; he was a teacher who inspired me and
pushed me forward. I am thankful to Prof. Dr. A. Rahman Abdel-
Mongy, Head of Physics Department, Faculty of Science, Helwan
University, for his help, encouragements and valuable advice.I would
also like to express my thanks to Dr. Yasser Mohamed El-Gendy,
Physics Department, Faculty of Science, Helwan University, for his
adequate reading of the thesis, worthy discussions and his kind
assistance.

I am so grateful to Prof. Dr Rizk abdel Moniem Rizk, Dean


of Faculty of Science, Helwan University, for his continuous help,
encouragement and facilities.

I am thankful to all staff members and colleagues, Solid State


Department, National Center For Radiation Research and
Technology, Atomic Energy Authority for their support and
appreciated help.
Acknowledgment xii

I would also like to thank my Colleagues in the Physics


Department, Faculty of Science, Helwan University for their help.

Special thanks to my best friends, Hanaa Selim Ali and


Doaa Gamal and Soraya Mohamed for their warm encouragement,
continuous assistance and advice.
Abstract xxi

Abstract

Chalcogenide glasses are a recognized group of


inorganic glassy materials which always contain one or
more of the chalcogenide elements S, Se or Te but not O, in
conjunction with more electro positive elements as As, Sb,
etc.

Chalcogenide glasses are generally less robust, more


weakly bonded materials than oxide glasses. Glasses were
prepared from Sn, Se, Bi and In elements with purity
99.999 %. These glasses are reactive at high temperature
with oxygen. Therefore, synthesis was accomplished in
evacuated clean silica tubes. The tubes were washed by
distilled water, and then dried in a furnace whose
temperature was about 100°C.

The weighted materials were introduced into the


cleaned silica tubes and then evacuated to about 10-4 torr
and sealed. The sealed tubes were placed inside the furnace
and the temperature of the furnace was raised gradually up
to 900°C within 1 hour and kept constant for 10 hours.
Moreover, shaking of the constituent materials inside the
tube in the furnace was necessary for realizing the
homogeneity of the composition. After synthesis, the tube
was quenched into ice water. The glassy ingots could be
obtained by drastic quenching. Then materials were
removed from the tubes and kept in dry atmosphere. The
proper ingots were confirmed to be completely amorphous
using x-ray diffraction and differential thermal analysis.
Abstract xxii

Thin films of the selected compositions were prepared by


thermal evaporation technique with constant thickness
100nm. The effect of radiation, optical and some other
effects on composition was studied.

The structural properties of Inx Sn20 Se(60-x) Bi20


amorphous semiconductor in the powder and thermally
evaporated thin films have been investigated. Differential
Thermal Analysis, DTA, for InxSn20Se(60-x)Bi20 in the
powder form showed that an endothermic peak in the DTA
curve results from an increase in specific heat at the glass
transition temperature Tg. The absence of any sharp
exothermic peak in the DTA curve is good indicator for
absence of the structural changes. The analysis of X-Rays
Diffraction Patterns (XRD) of Inx Sn20 Se(60-x) Bi20 in the
powder form confirmed amorphous state. Scanning
electron microscope SEM micrographs were made for Se-
Bi-Sn films deposited at room temperature. The film
consisted of individual grains, which are irregular in size
and shape and separated by well-defined inter-grain
boundaries. By adding In, further separation of the
surrounding media gives rise to large grains in size at x=0.1
at%. Then large grains can be seen for partially crystalline
at x=0.2 at % the grain sizes become smaller for x=0.3 at%
and the number of grains become larger.

The density of the as prepared glasses of the system


Inx Sn20 Se(60-x) Bi20 films has been determined by the
hydrostatic method with an accuracy of ± 0.05 %. It has
been noticed that the density decreases by increasing In
from 5.712 gm/cm3 for the composition Inx Sn20 Se(60-x) Bi20
Abstract xxiii

at x=0% to 4.682 gm/cm3 for composition InxSn20Se(60-


x)Bi20 at x=0.03%.

The optical properties of Inx Sn20 Se(60-x) Bi20 have


been characterized by the measurements of the
transmittance and reflectance in the wave length 200 –
1100 nm for the deposited films. The type of the electronic
transition responsible for optical properties is indirect
allowed transition with transport and onset energy gap in
the range 0.79 – 1.7 eV. The values of the optical energy
gap Eopt were found to decrease with increasing In content
which could be due to the fact that In has a metallic
behaviour.

The absorption spectra of Inx Sn20 Se(60-x) Bi20 is


recorded in the UV region. Some important parameters
such as coordination number Nco , the number of constraints
NS, the parameter r determined the deviation of
Stoichiometry. If there is a linear dependence between the
bond strength and the average band gap, and if one allows
their superposition to describe the compounds, then the
addition of In will affect the average band gap. By
increasing the In content, the average bond strength of the
compound decreases, and hence Eg will decrease. In order
to emphasize the relationship between Eg and the average
bond strength more clearly, Eg is compared with HS/Nco
which is the average single-bond energy in the alloy.

The electrical properties of Inx Sn20 Se(60-x) Bi20 alloys


include the measurements of DC conductivity for
InxSn20Se(60-x)Bi20 films and the measurements of switching.
Abstract xxiv

The DC conductivity of InxSn20Se(60-x)Bi20 thin films has


been measured as function of temperature. The
dependence of the electrical DC conductivity on the
temperature showed the existence of two distinct linear
regions with two activation energies ∆E1 and ∆E2.

The switching measurements have been made for


InxSn20Se(60-x)Bi20 thin films and the addition of In has led
to an increase in both the threshold voltage (Vs) and
threshold current (Is) from 1.5 volt and 1.0 µA respectively
at x=0 up to 5.5 volt and 2 µA respectively at x=0.3 for
constant film thickness d=100 nm. As for the holding
voltage (Vh), it was found to increase with the increase of
In content from 0.4 volt at x=0 to 2.0 volt at x=0.3. On the
contrary, the increase of In content has caused a decrease in
the holding current (Ih) from 45 µA at x=0 to 19 µA at
x=0.3 for a constant thickness 100 nm. It was proved that
the threshold power increased by increasing In content.
This means that the quality of switching is reduced by
increasing the In content. The addition of In content
increases the cohesive energy and consequently affects the
switching properties. Raising the film temperature
improved the switching characteristics where the threshold
voltage decreased and the threshold current increases.
Also, the addition of In reduced the filament temperature,
thus reducing the switching ability. Increasing the In
content from x=0 to 0.1, 0.2 and 0.3 led to an increase in
the switching rise time from tr = 25 to 40, 100 and 200
nano second respectively and a decrease in the cohesive
energy from C.E. = 2.8186 to 2.8192, 2.8199 and 2.820 eV
respectively. These results indicate that composition Inx
Abstract xxv

Sn20 Se(60-x) Bi20 shows good electrical threshold switching


results and promises a useful threshold switching device in
computer applications.

Finally, the study of effect of gamma rays on the Inx


Sn20 Se (60-x) Bi20 showed that the gamma radiation did n’t
have a noticeable effect, for a dose of 15MRad showed a
constant value in the transmittance upon the addition of In.
Summary

Summary
The subject of amorphous semiconductors (a-S.C.)
has been of great interest in the recent years and it is
considered a particularly active field in solid state science.
In the last decade, considerable attention has been focused
on a-S.C. especially those known as chalcogenide glasses.
Their structure is investigated by X-ray Diffraction and
their amorphous nature is confirmed by the Differential
Thermal Analysis (DTA). They are characterized by their
sensitivity to light, thus leading to structural or optical
changes. The study of the optical parameters, e.g. the
absorption coefficient, provides information about the band
structure and energy gap in the material. Memory
switching is also a phenomenon that is observed in a-S.C.
The technological importance of the S.C. chalcogenide
glasses is not only due to its valuable technological
applications in modern devices, but also because of its
cheapness in relative with other S.C. materials. Their
applications in modern technology comprise energy
management, thermal fault detection, temperature
monitoring and night vision. They are selected for
switches, memory and computer applications due to their
favorable switching characteristics. Moreover, they are
applied in film transistors and electrographic units. An
interesting application of a-S.C is in the fabrication of
sensors for environmental protection and medical
diagnosis. Also, optical fibers are made from chalcogenide
a-S.C. and these are commonly used in telecommunication
systems, illumination and imaging optics. Optical fibers
have become the focus of researchers due to their potential
use in ultra-fast switching devices and surgical purposes.
Also, such a-S.C. compete favorably with silicon devices
for solar power conversion as they are less expensive, thus
they have recently been used to manufacture solar cells.
Since a-S.C. are characterized by their sensitivity to
Summary

external factors, especially ionizing radiation, they are used


for radiation dosimetry and as radiation detectors.

Chalcogenide glasses are a recognized group of


inorganic glassy materials which always contain one or
more of the chalcogenide elements S, Se or Te but not O, in
conjuction with more electro positive elements as As, Sb,
etc.

Chalcogenide glasses are generally less robust, more


weakly bonded materials than oxide glasses. Glasses were
prepared from Sn, Se, Bi and In elements with purity
99.999%. These glasses are reactive at high temperature
with oxygen. Therefore, synthesis was accomplished in
evacuated clean silica tubes. The tubes were washed by
distilled water, and then dried in a furnace whose
temperature was about 100°C.
The weighted materials were introduced into the
cleaned silica tubes and then evacuated to about 10-4 torr
and sealed. The sealed tubes were placed inside the
furnace and the temperature of the furnace was raised
gradually up to 900°C within 1 hour and kept constant for
10 hours. Moreover, shaking of the constituent materials
inside the tube in the furnace was necessary for realizing
the homogeneity of the composition. After synthesis, the
tube was quenched into ice water. The glassy ingots could
be obtained by drastic quenching. Then materials were
removed from the tubes and kept in dry atmosphere. Thin
films of the selected compositions were prepared by
thermal evaporation technique under vacuum 10-4 torr with
constant thickness 100 nm.

The structural properties of InxSn20Se(60-x)Bi20


amorphous semiconductor in the powder and thermally
evaporated thin films have been investigated. Differential
Thermal Analysis, DTA, for InxSn20Se(60-x)Bi20 in the
Summary

powder form showed that an endothermic peak in the DTA


curve results from an increase in specific heat at the glass
transition temperature Tg. The absence of any sharp
exothermic peak in the DTA curve is good indicator for
absence of the structural changes. The analysis of X-Rays
Diffraction Patterns (XRD) of InxSn20Se(60-x)Bi20 in the
powder form confirmed amorphous state. Scanning
electron microscope SEM micrographs were made for Se-
Bi-Sn films deposited at room temperature. The film
consisted of individual grains, which are irregular in size
and shape and separated by well-defined inter-grain
boundaries. By adding In, further separation of the
surrounding media gives rise to large grains in size at x=0.1
at%. Then large grains can be seen for partially crystalline
at x=0.2 at% the grain sizes become smaller for x=0.3 at%
and the number of grains become larger.

The density of the as prepared glasses of the system


InxSn20Se(60-x)Bi20 films has been determined by the
hydrostatic method with an accuracy of ± 0.05 %. It has
been noticed that the density decreases by increasing In
from 5.712 gm/cm3 for the composition InxSn20Se(60-x)Bi20
at x=0% to 4.682 gm/cm3 for composition InxSn20Se(60-
x)Bi20 at x=0.03%.

The optical properties of InxSn20Se(60-x)Bi20 have


been characterized by the measurements of the
transmittance and reflectance in the wave length 200 –
1100 nm for the deposited films. The type of the electronic
transition responsible for optical properties is indirect
allowed transition with transport and onset energy gap in
the range 0.79 – 1.7 eV. The values of the optical energy
gap Eopt were found to decrease with increasing In content
which could be due to the fact that In has a metallic
behaviour.
Summary

The absorption spectra of InxSn20Se(60-x)Bi20 is


recorded in the UV region. Some important parameters
such as coordination number Nc o , the number of
constraints (NS) , the parameter (r) determined the
deviation of Stoichiometry. If there is a linear dependence
between the bond strength and the average band gap, and if
one allows their superposition to describe the compounds,
then the addition of In will affect the average band gap. By
increasing the In content, the average bond strength of the
compound decreases, and hence Eg will decrease. In order
to emphasize the relationship between Eg and the average
bond strength more clearly, Eg is compared with HS/Nco
which is the average single-bond energy in the alloy.

The electrical properties of InxSn20Se(60-x)Bi20 alloys


include the measurements of DC conductivity for
InxSn20Se(60-x)Bi20 films and the measurements of switching.
The DC conductivity of InxSn20Se(60-x)Bi20 thin films has
been measured as function of temperature. The
dependence of the electrical DC conductivity on the
temperature showed the existence of two distinct linear
regions with two activation energies ∆E1 and ∆E2.

The switching measurements have been made for


InxSn20Se(60-x)Bi20 thin films and the addition of In has led
to an increase in both the threshold voltage (Vs) and
threshold current (Is) from 1.5 volt and 1.0 µA respectively
at x=0 up to 5.5 volt and 2 µA respectively at x=0.3 for
constant film thickness d=100 nm. As for the holding
voltage (Vh), it was found to increase with the increase of
In content from 0.4 volt at x=0 to 2.0 volt at x=0.3. On the
contrary, the increase of In content has caused a decrease in
the holding current (Ih) from 45 µA at x=0 to 19 µA at
x=0.3 for a constant thickness 100 nm. It was proved that
the threshold power increased by increasing In content.
This means that the quality of switching is reduced by
Summary

increasing the In content. The addition of In content


increases the cohesive energy and consequently affects the
switching properties. Raising the film temperature
improved the switching characteristics where the threshold
voltage decreased and the threshold current increases.
Also, the addition of In reduced the filament temperature,
thus reducing the switching ability. Increasing the In
content from x=0 to 0.1, 0.2 and 0.3 led to an increase in
the switching rise time from tr = 25 to 40, 100 and 200
nano second respectively and a decrease in the cohesive
energy from C.E. = 2.8186 to 2.8192, 2.8199 and 2.820 eV
respectively. These results indicate that composition
InxSn20Se(60-x)Bi20 shows good electrical threshold
switching results and promises a useful threshold switching
device in computer applications.

Finally, the study of effect of gamma rays on the


InxSn20Se(60-x)Bi20 showed that the gamma radiation did n’t
have a noticeable effect, for a dose of 15MRad showed a
constant value in the transmittance upon the addition of In.
Summary
Introduction and Literature Survey 1

Chapter 1

Introduction and Literature Survey

1.1 Introduction

Many amorphous materials can be called


Semiconductor in the sense that they are neither good
conductors nor good insulators, but instead they are poor
conductors. Many are also similar to their crystalline
counterparts in that they possess an optical gap.
Amorphous Semiconductor is characterized by:

1. Their electrical properties are similar to intrinsic S.C. or


perfectly compensated S.C.
2. They are partially transparent in the infrared region.
3. Their room temperature conductivities are lower than
103 - 104 Ω-1 cm-1.

A major category of a-S.C. is the chalcogenide


glasses. The first chalcogenide glass to be commercially
developed in 1950s was As2S3, produced for passive bulk
optical component for the mid-IR. During the next two
decades, other sulphide and selenide-telluride glasses have
been used as optical components for the far infrared which
have since then been exploited commercially [1, 2].
Applications of infrared optics include energy
management, thermal fault detection and electronic circuit
detection, temperature monitoring and night vision [3].
The blackbody radiation emitted by room temperature
objects such as the human body in the wavelength 8–12 µm
region is an example of the latter, where Se-Te based
glasses are applicable for thermal imaging. Starting from
Introduction and Literature Survey 2

1970, chalcogenide glasses have been recognized as a


stable and active electronic device component in
photocopying and switching applications [4].

Through the 1980s, attention was focused on the


fabrication of ultra-low loss IR fibres for
telecommunication signal transmission to compete with
silica optical fibres [5, 6].

In the 1990s, the development of the optical glasses


with IR transmittance for infrared purposes has been
proceeding in two main directions: infrared imaging and
wave guide applications.

More advancement has taken place in the last decade


and till present to take advantage of the space applications
of a-S.C. and their application in the field of
nanotechnology.

Chalcogenide glasses are a recognized group of


inorganic glassy materials which always contain one or
more of the chalcogen elements S, Se or Te, in conjunction
with more electropositive elements as As, Sb and Bi [7].
These glasses are band gap S.C. and they are generally less
robust, more weakly bonded materials than oxide glasses.
Chalcogenide glassy S.C. have several useful properties
that can be employed in various solid state devices. They
show a continuous change in physical properties with
change in chemical composition [8].

Chalcogenide a-S.C. materials exhibit a number of


interesting changes when exposed to light having a photon
energy comparable to the band gap. Such changes can be
structural, mechanical, chemical or optical (photodarkening
and photobleaching). The light induced changes can in
general be either irreversible, i.e., the changes are
Introduction and Literature Survey 3

permanent after irradiation, or reversible, in which case the


changes can be removed by annealing to the glass transition
temperature (Tg). These changes are favored in
chalcogenide glasses due to their structural flexibility and
also due to their high-lying lone pair (LP) p states in their
valence bands [9].
Memory switching in chalcogenide glasses has been
widely reported in literature and is fairly well understood.
Important device applications of switching in
chalcogenides are computer memory arrays, display
devices, optical mass memories…, etc. Enough evidence is
available to believe that at a threshold voltage, memory
switching occurs due to formation of a filamentary path
which is crystalline in nature [10].

Amorphous Semiconductor is present in three types


[11]: elemental, covalent alloys and ionic. The first
category contains elements such as S and Se which can be
obtained in the amorphous state. The S and Se are
characterized by chain and ring structure, and a short-range
order which extends over a distance, depending on
temperature and thermal history of the material. The
second category contains covalent bonded alloy glasses.
These alloys possess compositional and translational
disorders; therefore, all atoms locally satisfy their valence
bond requirements. The third category contains ionic
materials such as silicate glasses, which contain at least one
element of the chalcogens (S, Se and Te).

Impurity effects in chalcogenide glasses have


importance in fabrication of glassy semiconductors.
Several workers have reported the impurity effects in
various chalcogenide glasses. They are interesting as core
materials for optical fibers used for transmission especially
when short length and flexibility are required. Since the
advent of electro photography, amorphous Selenium has
Introduction and Literature Survey 4

become a material of commercial importance. Selenium


exhibits the unique property of reversible phase
transformation [12]. Its various device applications like
rectifiers, photocells, xerography, switching and memory,
etc. have made it attractive, but pure selenium has
disadvantages like short lifetime and low sensitivity. This
problem can be overcome by alloying Se with some
impurity atoms (Bi, Te, Ge, Ga, Sb, As, … etc.), which
gives higher sensitivity, higher crystallization temperature
and smaller ageing effects.

Ag-doped chalcogenide glasses and their films have


many current and potential applications in optics,
optoelectronics, chemistry and biology (optical elements,
gratings, memories, micro lenses, waveguides, bio- and
chemical-sensors, solid electrolytes, batteries,…etc.) [13]
The Ag-doped glasses can be used as optical memory
materials and materials for holography. The sensitivity can
be increased by the simultaneous application of an
electrical field with light. The Ag-doped binary and ternary
tellurides are becoming important because some of them
are used as materials for phase-change optical storage
(DVD disks, … etc). They exhibit single glass transition
and single crystallization temperatures, which is important
for rewritable disks.

1.2 Literature Survey

Chalcogenide glasses have been recognized as


promising materials for infrared optical elements [14] and
for the transfer of information [15]. They have also found
application sin Xerography [16], switching and memory
devices [17], in the fabrication of in expensive solar cells
[18], and more recently, for reversible phase change optical
records [19]. Thus, it is important to have an insight into
Introduction and Literature Survey 5

their electronic properties. It has been reported that for any


chalcogenide glassy system, increasing the relative atomic
mass of the chalcogen or its proportion in glass diminishes
the average bond strength and hence decreases the glass
transition temperature [20]. The addition of chalcogen
which act as chains or network terminators tend to decrease
the glass transition temperature and increase the thermal
expansion coefficient [21].

Thin film of Se1-x Tex (x=0.2, 0.4 & 0.6) deposited


on a glass substrate were studied and investigated by H.EL-
Zahed et al [22]. Optical band gap Eg were determined
from the absorbance and transmittance measurements in the
visible and near IR spectral range (500-1100) nm Optical
band gap Eg was found to be decreased with increasing
tellurium concentration.

M. A. Abdel-Rahim [23] reported and discussed the


results of differential thermal analyses (DTA) under non
isothermal conditions for three compositions of the Se(x-
85)Te15SbX (x=0, 3 and 9). The onset crystallization
temperatures (Tc), and the peak temperature of
crystallization (Tp) were found to be dependent on the
compositions and the heating rates. From the dependence
on heating rates of (Tg) and (Tp) the activation energy for
glass transition (Et) and the activation energy for
crystallization (Ec) were calculated and their composition
dependence were discussed. The crystalline phases
resulting from DTA have been identified using X-ray
diffraction and Scanning electron microscope (SEM).

D.C. conductivity measurements at high electric


fields in thermal vacuum evaporated thin films of
amorphous Se80-x Te20Cdx (x=0, 5, 10, 15 at %) systems
have been studied by S. P. Singh et al [24] Current-voltage
Introduction and Literature Survey 6

(I-V) characteristics have been measured at various fixed


temperatures .They observed that, at low electric fields, the
studied samples have ohmic behavior, but at high electric
fields (E~104 V/cm), non ohmic behavior was observed.

A. Dahshan et al [25].reported the effect of


replacement of selenium by antimony on the optical gap
and some other physical parameters of new quaternary
chalcogenide As14 Ge14 Se72-xSbx (where x=3.6.9 and15
at%)thin films. thin films with thickness 200-220 nm of
As14 Ge14 Se72-xSbx were prepared by thermal evaporation
of bulk samples, increasing antimony content was found to
affect the average of atomization, the average coordination
number, number of constraints and cohesive energy of
As14 Ge14 Se72-xSbx alloys. Optical absorption
measurements showed that fundamental absorption edge is
a function of, composition. Optical absorption is due to
allowed non direct transition and the energy gap decreases
with increasing antimony content. The chemical bond
approach has been applied successfully to interpret the
decrease in the optical gap with increasing antimony
content.

Thin films were thermally evaporated from ingot


pieces of the As30 Se70-x Sbx (with 2.5<x<17.5 at.%) glasses
under vacuum of ~10-5Torr. by A. A. Moharram et al [26]
.Increasing Sb content was found to affect the thermal and
optical properties of these films. Non-direct electronic
transition was found to be responsible for the photon
absorption successfully to interpret the decrease of the glass
optical gap with increasing Sb content.

P. Sharma et al [27] were studied the optical


properties of Se substituted by Bi in Ge20Se80 thin films.
Optical reflection and transmission spectra, at normal
Introduction and Literature Survey 7

incidence of Ge20 Se80-x Bix thin films (x=0, 4, 6, 8, 10, 12)


were obtained in the range 200nm-840nm.The optical
energy gap was estimated from the absorption coefficient
values using Tauc's procedure. It is found that the optical
band gap decreases with increasing Bismuth content.

Glasses based on selenium and tellurium were


carefully characterized to establish the interdependence
between chemical composition and the magnitudes of the
physical parameters. Optical measurements were performed
on thin amorphous chalcogenide films from the system of
pure (Se0.8Te0.2) and metal doped (Se0.8Te0.2)0.9M0.1 (M= Cu,
Ag and Sn) by A. F. Maged et al [28]. Parameters
considered are density, molar volume, the concentrations of
metal atoms per unit volume, and optical energy gap. Bulk
Se0.8Te0.2 alloy was prepared by the standard melt
quenching technique and the different metals were added to
the binary system using the same technique. Thin films of
the prepared compositions was grown using thermal
evaporation method. The effects of different metallic
additions and annealing at a temperature below the glass
transition temperature Tg on the optical characteristics was
explored. The bulk and thin films of the samples were
tested by X-ray diffraction, which reveals that they are
amorphous. The optical energy gap E0 was found to
decrease with the addition of metal, the amount of decrease
depends on the chemical character of added metal.
Annealing leads to a decrease of the optical energy gap
with annealing time. The gamma irradiation (up to 100
kGy) has no detectable effect on the optical energy gap.

Amorphous Se90In10−xSnx (x=2, 4, 6 and 8) thin films of


thickness 1000 Å were prepared on glass substrates by the
thermal evaporation technique by Adel A. Shaheen et al
[29]. Optical parameters of the films were investigated, in
Introduction and Literature Survey 8

the wavelength range 400–700 nm, before and after


irradiation by 4, 8, and 12 kGy doses of γ-rays. The optical
absorption coefficient α for as-deposited and gamma
irradiated films was calculated from the reflectance R and
transmittance T measurements, which were recorded at
room temperature. From the knowledge of α, at different
wavelengths, the optical band gap Eg was calculated for all
compositions of Se–In–Sn thin films before and after
gamma irradiation. Results indicate that allowed indirect
optical transition is predominated in as-deposited and
irradiated films. Besides, it is found that the band gap
decreases with increasing Sn concentration and this is
attributed to the corresponding decrease in the average
single bond energy of the films. The band gap, after
irradiation at different doses of γ-rays, was found to
decrease for all compositions of the studied films. This
post-irradiation decrease in the band gap was interpreted in
terms of a bond distribution model.

Gamma radiation is known to induce changes in


physical, optical, and structural properties in chalcogenide
glasses, but previous research has focused on As2S3 and
families of glasses containing Ge. For the first time, we
present composition and dose dependent data on the As–S
binary glass series. Binary AsxS100−x (x = 30, 33, 36, 40, and
42) glasses were irradiated with gamma radiation using a
60
Co source at 2.8 Gy/s to accumulated doses of 1, 2, 3, and
4 MGy. The irradiated samples were characterized at each
dose level for density, refractive index, X-ray diffraction
(XRD), and Raman spectrum. An initial increase has been
reported in density followed by a decrease as a function of
dose that contradicts the expected compositional
dependence of molar volume of these glasses. This unusual
behavior is explained based on micro void formation and
nanoscale phase-separation induced by the irradiation.
XRD, Raman, and electron spin resonance data provide
Introduction and Literature Survey 9

supporting evidence, underscoring the importance of


optimally- or overly constrained structures for stability
under irradiation [30].

A. F. Maged et al [31].have Studied the effects of


addition of tellurium on transition temperature, density,
molar volume and optical properties for Ge-As-Se system
and the effect of γ-irradiation on IR transmission for the
system x=0 and x=40. Oxygen impurities, which increased
after γ-irradiation, produce an absorption between 12 and
16 µm due to Ge-O, As-O and Se-O

M.M.EL-Ocker et al [32]. Investigated the effect of


addition In content on dc electrical conductivity and DTA
of The system (AS2Se3)1-xInx x=0,0.1,0.05 . The electrical
energy gap was found to increase for an In content 0.01%
and decrease for an In content 0.05%.The Samples exhibit
the three conduction mechanisms proposed by Mott and
Davis. The activation energy was calculated for each
mechanism. The effect of heating rate on the transition
temperatures (Tg,Tc,Tm) was studied and variation of the
crystallization-peak position was used to calculate the
activation energy and the order of the crystallization
process.

Chalcogenide glasses are interesting materials due to


their infrared transmitting properties and photo induced
effects exhibited by them. Thin films of the glasses
Sn10Sb20_xBixSe70 prepared by melt quenching technique
were evaporated in a vacuum better than 10_5 mbar. Optical
transmissions spectra of all the deposited films were
obtained in a range (400–2500) nm. The optical band gap
and the absorption coefficient were calculated by Muneer
Ahmad et al [33] from the transmission data. The optical
band gap initially increases with increase in Bi content (for
x = 2) and then decreases sharply for higher Bi
Introduction and Literature Survey 10

concentrations. The refractive index as well as absorption


coefficient decrease with increase in wavelength. The dark
activation energy initially increases with increase in Bi
content and then decreases with further addition

Chalcogenide alloys of Sn10Sb20_xBixSe70 system


were prepared by the melt quenching technique .Thin films
were prepared on well-cleaned glass substrates by the
thermal evaporation technique. The X-ray diffract gram for
System reveals the amorphous nature as no sharp peak is
observed but the sample with shows sharp peaks. The glass
transition, crystallization, melting temperatures and glass
forming tendency of the amorphous samples were
determined from differential scanning calorimetric
measurements. The glass transition activation energies and
the crystallization activation energies were determined
using the Kissinger method. Optical transmission and
reflection spectra of thin films were obtained in the
range400-2500nm. The conductivity activation energy and
optical gap initially increases with increasing Bi
concentration and then decreases sharply for higher Bi
content .The values of the band tailing parameter and the
pre-exponential factor are also reported and discussed by
Muneer Ahmed [34].

Both dynamic and static I–V characteristic curves of


amorphous thin films of Se75Ge25−xAsx for switching and
memory behavior have been studied by Fadel [35]. The
films were prepared by thermal evaporation of high purity
(99.999%) material. X-ray diffraction patterns revealed the
formation of amorphous films. The electrical measurements
are made at room temperature and at elevated temperatures
up to the glass transition temperature (Tg~418 K). The
conduction activation energy, Eσ is determined. The
threshold voltage, Vth is determined. It is found that Vth
Introduction and Literature Survey 11

increases linearly with the film thickness and decreases


with As content. Moreover, Vth decreases exponentially
with temperature. The rapid transition between the highly
resistive and conductive states is attributed to an electro
thermal mechanism from the Joule heating of a current
channel.

Kotkata et al. studied [36] the switching effects in


amorphous GeSe2, GeSe4, GeSe2Tl and GeSe4Tl thin films.
The observed switching phenomenon for these
compositions was of the memory type. The threshold
switching voltage was found to increase linearly with
increasing film thickness (80–740 nm), while it decreased
exponentially with increasing temperature (T < Tg). The
effect of adding Thallium to both amorphous GeSe2 and
amorphous GeSe4 results in decreasing the values of the
threshold electric field, the activation energy of switching,
as well as the thermal activation energy of conduction. The
results obtained are explained in accordance with the
electro thermal model of breakdown.

Observations of memory switching in thin films of


amorphous As2Se3 have been made at various film
thicknesses by (Thornburg, 1972) [37]. The distribution of
threshold voltages for a given thickness shows a strong
peak which is attributed to the intrinsic switching
mechanism. A plot of the most probable threshold voltage
vs film thickness shows the switching process to be field
controlled. Microscopic evidence is presented for strong
Joule heating caused by capacitive discharge upon filament
formation.

Ovshinsky and Fritzsche [38], reported the


performance and reliability of amorphous semiconductor
devices that deal with the handling of information in the
Introduction and Literature Survey 12

form of switching, modulation, storage, and displays.


Structural changes between a disordered and a more
ordered state and the concomitant large change in many
material properties offer the possibility of using amorphous
semiconductors for high-density information storage and
high-resolution display devices. The structural changes can
be initiated by various forms of energy such as an electrical
pulse, a short light pulse, or a brief light exposure. Many
materials show good structural reversibility. The sensitivity
of an amorphous photo structural film is amplified by
several orders of magnitude by first forming a latent image
by photo nucleation and subsequent dry development by
heat or radiation. Examples of optical contrast and
resolution in image formation are given. The major
differences between crystalline and amorphous
semiconductors are briefly outlined.

Rajesh and Philip [39] reported the discovery of


electrical switching in chalcogenide glasses. A complete
understanding of the mechanisms responsible for this
phenomenon is lacking. It is believed that threshold
switching in chalcogenide glasses is electronic in origin
whereas memory switching is of thermal origin. According
to the thermal model, the on state during memory switching
is caused by a thermal breakdown of the steady state of the
material when the heat generated by Joule heating cannot
be removed fast enough by thermal conduction. Joule
heating causes crystallization of the glass into fine
filaments resulting in excess carrier concentration in the
current path due to the large electric field present. Even
though electrical conductivity and related properties of a
number of materials have been investigated during and
after switching, no measurements seem to have been done
to determine the variation in thermal conductivity and
related properties during switching. The results of their
Introduction and Literature Survey 13

measurements of thermal conductivity and heat capacity of


In–Te glasses were reported, which exhibit clear memory
type electrical switching at threshold fields in (80–140) V
cm−1 range. Measurements have been made on bulk
samples using an improved photo thermal technique.
Results show that thermal conductivity of the samples
increase considerably during electrical switching, whereas
heat capacity remains more or less constant. The results are
discussed in the light of the thermal model for memory
switching applicable to Te-based chalcogenide glasses.

Stocker [40] studied the switching phenomena which


take place in thick bulk samples of semiconducting glass,
once a path of devitrified material is established. Potential
probe and infrared micro radiometer measurements reveal
that the switching action takes place in a small region
somewhere along this path. Application of voltage pulses
can move this region to a different position. Evidence of
partial devitrification and melting is also found in thin film
switches made from many different glass compositions.
Memory switching has also been observed in all bulk and
thin film experiments to date. Since the characteristics of
bulk and thin film switching are remarkably similar, doubt
is cast upon the interpretation of switching phenomena as
due to electronic properties of amorphous semiconductors.

1.3 Aim of the work

The theoretical survey made on various


technological applications of semiconductors Known as
chalcogenide glasses indicated that they have different
useful properties, e.g easy to prepare, light in weight and
cheap. These compositions can be used in solid state
devices to produce economical chalcogenide compositions.
Introduction and Literature Survey 14

The study has shown that the selected composition


has not been dealt with in many previous studies.
Therefore, it has been chosen to prepare anew composition
with different percentages of composition that have not
been done before, so it can be used in the future in different
electronic applications.

The goal is then to use cheep and reliable materials


for technological application in deriving scientists to
develop application. Rectifiers, photocells, switching and
memory devices, detectors and sensors, optical imaging or
storage media computer, memory arrays and display
devices.
Theoretical background 15

Chapter 2

Theoretical Background

2.1 Disordered systems

In solid-state physics disordered is very often explained


through the terminology of order. Two aspects of order are
important for this treatise:
• Short-range order is a regular arrangement of the closest
neighboring atoms.

• Long-range order has a strict periodicity and translation


in variance of the crystal lattice. An unperturbed and
infinite lattice is ideal and considered as the zeroth
approximation in the calculation of solid state
properties. Perturbations can be classified as dynamic,
in elementary excitations, or static as in point
imperfections. A lattice is considered ordered when it is
possible to explain its characteristics with an infinite
lattice with ideal long-range order and with addition of
perturbation theory including the dynamics and static
perturbations. An arrangement is considered disordered
when this approximation is not meaningful.

2.2 The Chalcogenide Glasses

The properties of amorphous semiconductor containing


one or more of the chalcogenide elements S, Se or Te are
reviewed [41-55].
Theoretical background 16

It is possible to form glasses by combination with one


or more of the elements As, Ge ,Si ,Tl ,Pb ,P ,Sb and Bi ,
among others. The binary chalcogenide glasses As2Se3,
As2S3 and As2Te3 have been extensively studied. .Mixed
systems such as the As2Se3-As2Te3 binaries and the As2Se3-
As2Te3-TI2Se systems have also been the subjects of
detailed investigations. This is due to the large variety of
such ternary and quaternary systems. Classification of these
materials is difficult, particularly in view of the freedom
that is allowed in amorphous systems to depart from
stoichiometric proportions of the constituents. However,
the use of the stoichiometric compositions allows useful
comparison with the material in its crystalline phase.

2.3 Electronics Band Structure and Defects


The absence of long–range order in amorphous
semiconductors does not have a major effect on the energy
distribution of the electronic levels. The density of states
retains a profile similar to that of the crystalline phase. The
absence of long-range order manifest itself in the form of
band tailing and localization of the electronic states.
Localized states near both conduction and valence bands
are separated by the so–called mobility edges. This is
because conduction through the localized states can only
occur by thermally assisted hopping or tunneling [56]. The
electronic states of a solid may be considered to first
approximation to be a broadened superposition of the
molecular orbital states of the constituent bands. In Ge
3
fourfold coordinated the hybridized Sp orbitals are split
into bonding σ and anti-bonding σ* states, Fig (2-1a) [57].

In the solid phase molecular states are broadened into


bands. Thus, in tetrahedral semiconductors the bonding
band forms the valence band and the anti-bonding band
Theoretical background 17

forms the conduction one. In Se the S states lie well below


the P–states thus no hybridization occurs. Then only two of
the three P states can be utilized for bonding, Se is found in
two fold coordination. This leaves one nonbonding electron
pair, Fig (2-1b)

Figure (2-1) Showing bonding in (a) Ge and (b) Se


A: atomic bonding
B: hybridized states
C: molecular states
D: broadening of states into bands in the solid phase
unshared or lone pair L electrons Form a band near the
original P–states energy.

The σ and σ* are split symmetrically with respect to this


reference energy. Thus the bonding band is no longer the
valence band and this role is played by the LP band in the
tetrahedral materials localized states are produced in the
gap due to the formation of dangling bonds. In the
Theoretical background 18

chalcogenide materials the LP bands lie in the energy


region between bands. The question of states in the gap,
whether of extrinsic or intrinsic nature is of considerable
importance. Cohen, Fritzche and Ovshinsky (CFO) [58]
model supposed that the non crystalline structure would
lead to overlapping of band tails of localized states as in
Fig (2-2a). Those derived from the conduction band would
be neutral when empty and those derived from the valence
band are assumed to be neutral when occupied.

In the overlap region they would be charged leading to


centers with unpaired spins. In this model known as CFO
model they emphasized the existence of mobility edges at
energies in the band tails. These are identified with critical
energies separating localized states from extended states.
The difference between the energies of the mobility edges
in the valence and conduction bands is called the “mobility
gap”. Mott and Davis [59] proposed a band model in which
they made a strong distinction between localized states that
originate from the lack of long–range order and those
which are due to defects in the structure.

The first kind of localized states extend only to


energy EA and EB in the mobility gap, Fig (2-2b) The defect
states form longer tails but of insufficient density to pin the
Fermi level. The authors proposed a band of compensation
near the gap center in order to account for the pinning of
the Fermi level and suggested that if the states of the
compensation band arose from defect centers such as
dangling bonds then they could act as deep donors Ey and
acceptors. This means that the compensation band will be
split into two bands Ey and Ex as shown in Fig (2-2c).
Theoretical background 19

Figure (2-2) Shows various forms proposed for the density


of states in amorphous semiconductors. The shaded areas
represent localized states.

(a) Overlapping of conduction and valence band tails as


proposed by Cohen et. al. (the CFO model)
(b) Real gap in the density of states (Mott-CFO model).
(c) The same as (b) but with compensation bands Ey and
Ex.

The Fermi level will lie between Ey and Ex if they do not


overlap or be pinned within them, if they do, since the
above models were proposed many experimental data have
emerged providing a clearer picture of the density
distribution of states in the gap of amorphous
semiconductors. Spear proposed that the centers
responsible may arise from pairs of dangling bonds at
Theoretical background 20

defects similar in nature to the divacancy in the crystal. The


lower Ey and upper Ex levels associated with this defect
correspond to bonding and anti-bonding states and are
separated by more than energy proposed in the model of
Mott and Davis.

2.3.1 Types of Defects

The importance of defects lies in the fact that many


properties of amorphous materials can be defect controlled
as in the case of crystalline solids. Some of these properties
are magnetic properties, optoelectronic behavior,
vibrational properties and mechanical characteristic. For
certain materials, e.g. chalcogenide glasses, the ideal
amorphous state is impossible to achieve experimentally
since structural defects are present even in thermal
equilibrium in the melt and are consequently frozen on
vitrification. A dangling bond is simply a broken or
unsatisfied bond and it normally contains one electron and
it is electrically neutral. A dangling bond can only be
formed in covalent solid and it has no meaning in a solid
formed from non–directional bonds such as in a metal ionic
salt or rare gas. Structural defects such as dangling bonds
are expected to introduce electron states deep into the gap
which is empty in the ideal case except for band tailing.
The precise position of these states in the gap will depend
on factors such as the electronic character of the states in
the top of the valence band and the bottom of the
conduction band from which the Eigen functions of the
defect states derives. The density of the states in the gap for
an amorphous semiconductor containing dangling bonds
defects might be as shown in Fig (2-3) the dangling bond
level is broadened by disorder into a band. The lower mid
gap band corresponds to dangling bond containing single
electron and is donor–like, i.e. neutral when occupied. The
Theoretical background 21

upper band corresponds to a different charge state of the


same defect namely when an extra electron is placed in it
and is acceptor–like, i.e., neutral when empty.

The essential features of the charged dangling band model


may be understood by considering a monoatomic system of
amorphous Se as in Fig (2-3).

Figure (2-3) Formation of charged defects (valence


alternation pairs) in chalcogenide glasses.
Theoretical background 22

(a)Illustration of the formation of threefold coordination D+


(C+) and single coordinated D- (C-1) defect centers by
exchange of an electron between two Do (Co1) centers.
(b)Configuration – coordinate diagram for the formation of
D+- D- p

Amorphous Se is twofold coordinated and it consists


mainly of chains. Any chain end will be a site of a dangling
bond which contains an unpaired electron. Mott et al
referred to this dangling bond defect as Do where the
superscript indicates the charge state. They postulated
following Anderson [60] electrons residing at Do centers
should experience negative effective correlation energy Ueff
and electron pairing should be energetically favorable as a
result of atomic rearrangements. The transfer of an electron
from one Do center to another produces on site which has
the original dangling bond orbital containing two spin –
paired electrons and which is consequently negatively
Theoretical background 23

charged D- and the other which has an empty orbital which


is then free to form a band with the lone pair of a fully
connected neighbouring atom. The defect now becomes
three fold coordinated and positively charged D+.

The repulsive Hubbard energy U involved in placing


an extra electron on the same site to form a D- center is
postulated to be weighed by the energy gained in forming
the extra bond at the D+ site rendering the reaction
exothermic:
2D → D + +D -

The process is illustrated in Fig (2-4a) and (2-4b) where the


coordinate “q” may be taken to be the sum of the distance
between two Do centers and their respective nearest
neighbor but non- directly bonded atoms. Kastener et al
[60] considered the same process of spin pairing at defects
in amorphous chalcogenide referring to it as “Valence
Alteration” and used the notation C3+ for D+ and C1- for D-
where C stands for chalcogen atom and the superscript
refers to charge state and subscript refers to the
coordination.

2.4 Thin Film Growth Process

Any thin film deposition process involves three main


steps:

(1) Production of appropriate atomic, molecular or ionic


species.
(2) Their transport to the substrate through a medium and

(3) Condensation on the substrate, either directly or via a


chemical and/or electrochemical reaction, to form a solid
deposit. Formation of a thin film takes place via nucleation
Theoretical background 24

and growth processes. The general picture of the step-by-


step growth process emerging out of the various
experimental and theoretical studies can be presented as
follows:

1. The unit species, on impinging the substrate, lose


their velocity component normal to the substrate
(provided the incident energy is not too high) and are
physically adsorbed on the substrate surface.

2. The adsorbed species are not initially in thermal


equilibrium with the substrate and move over the
substrate surface. In this process they interact among
themselves, forming bigger clusters.
3. The clusters or the nuclei, as they are called, are
thermodynamically unstable and tend to desorbed in
a time depending on the deposition parameters. If the
deposition parameters are such that a cluster collides
with other adsorbed species before getting desorbed,
it starts growing in size. After a certain critical size is
reached, the cluster becomes thermodynamically
stable and nucleation barrier is said to have been
overcomes. This step involving the formation of
stable, chemisorbed; critical-sized nuclei is called
nucleation stage.
4. The critical nuclei grow in number as well as in size
until a saturation nucleation density reached. The
nucleation density and the average nucleus size
depend on a number of parameters such as the
energy of the impinging species, the rate of
impingement, the activation energies of adsorption,
desorption and thermal diffusion, and the
temperature, topography, and chemical nature of the
substrate. A nucleus can grow both parallel to the
substrate by surface diffusion of the adsorbed
Theoretical background 25

species, as well as perpendicular to it by direct


impingement of the incident species. In general,
however, the rate of lateral growth at this stage is
much higher than the perpendicular growth. The
grown nuclei are called islands.

5. The next stage in the process of film formation is the


coalescence stage, in which the small islands start
coalescing with each other in an attempt to reduce
the surface area. This tendency to form bigger
islands is termed agglomeration and is enhanced by
increasing the surface mobility of the adsorbed
species, as for example, by increasing the substrate
temperature. In some cases, formation of new nuclei
may occur on the areas freshly exposed as a
consequence of coalescence.

6. Larger islands grow together, leaving channels and


holes of uncovered substrate. The structure of the
films at this stage changes from discontinuous island
type to porous network type. A completely
continuous film is formed by filling of the channels
and holes

The growth process may be summarized as consisting


of a statistical process of nucleation, surface-diffusion
controlled growth of the three-dimensional nuclei and
formation of a network structure then its subsequent filling
to give a continuous film. Depending on the
thermodynamic parameters of the deposit and the substrate
surface, the initial nucleation and growth stages may be
described as of (a) island type, (b) layer type, and (c) mixed
type (called Stranski-Krastanov type). This is illustrated in
Fig (3-3). In almost all practical cases, the growth 3takes
place by island formation.
Theoretical background 26

Figure (2-4) Basic growth processes: (a)


(a)island, (b) layer-
by- layer, and(c) Stranski
Stranski-Krastanov type

2.5 Optical Properties of Amorphous Materials

The distribution in energy of the electron states in an


amorphous material has gross features that resemble those
of the material in its crystalline form. Optical techniques
such as U.V. absorption and phphotoemission can be used to
probe the spectrum of electron levels in non – crystalline
systems. One of the most important properties of glass iis its
transparency in IR region. The implication of this as far as
the energy spectrum of electron states is conce
concerned is that
Theoretical background 27

are empty, just as crystalline semiconductors and


insulators. So, in ordinary window glass this gap must be
larger than the energy of quanta, whereas in In –Sn– Se –
Bi which is opaque the transparency is in the infra-red. The
origin of the gap in semiconducting or insulating
amorphous materials cannot be considered from the
viewpoint of scattering of block waves by periodic lattice
potential, which is the normal approach for crystals. In
glasses there is no translational symmetry. The gross
features of the energy spectra of electron states in many
solids (particularly the density of valence band states) is
now known to depend on the nearest neighbor environment
of a particular atom , with long – range interactions affect
details only. Since in most amorphous materials the forces
that bind atoms together are virtually the same as in the
crystalline state. One frequently finds nearest–neighbor
bond length and angles similar to those in the
corresponding crystals and thus a similar density of states
distribution.

2.5.1 Optical Absorption Mechanisms

Optical absorption in solids can occur by several


mechanisms all of which involve coupling of the electric
vector of the incident radiation to dipole moments in the
material and a consequent of energy.
Semiconductors show all the optical properties of
insulators and metals though not of course to the same
degree. The main features are as follows to Fig (2-5) [61].
Theoretical background 28

Figure (2-5) Absorption spectrum of thin film.

In the ultraviolet, and sometimes extends into the


visible and infra-red, intense absorption due to electronic
transitions between valence and conduction bands can be
observed. Such transitions generate mobile electrons and
holes resulting in photo-conductivity. The absorption
coefficient is typically in the range 105 to 106 cm-1 on the
high energy side of this band (typically around 2 eV).
There is often a smooth fall in absorption over a range of
several electrons volts. On the low energy side, the
absorption coefficient falls more rapidly and may fall as
much as six orders of magnitude within a few tenth of eV.
In semiconductors, this low energy boundary of the
fundamental absorption is often the most striking feature of
the spectrum and is referred to as the “absorption edge”.
• The limit of the absorption edge corresponds to the
photon energy required to promote electrons across the
Theoretical background 29

minimum energy gap Eg. The edge region often shows


some structure in particular that are due to excitons. An
exciton is formed when an electron, having been excited
insufficiently to escape from the influence of the hole it
leaves behind, is able to exist in a stable state in which it
does not recombine with the hole. The electron and hole
pair are held together as hydrogen atom by their mutual
Coulomb attraction and the separated charges can exist
in one of a series of quantized energy states. Exciton
absorption is more pronounced in insulators particularly
ionic crystals than in semiconductors and can leads to
strong narrow – line absorption as in ionic spectra.
• As the wave length is increased beyond the absorption
edge, the absorption starts to rise slowly again. This
increase is due to electronic transitions within the
conduction or intraband transitions and is referred to as
"free carrier absorption" or "intraband transition". It
extends throughout the infrared and microwave region
of the spectrum.
• At photon energies between 0.02 and 0.05 eV (50 to 20
µm wave length) a new set of absorption peaks appears.
These are due to interaction between the incident
photons and the vibrational modes of the lattice. If the
crystal is ionic the absorption coefficient may reach 105
cm-1 and strong reflection occurs.
• Impurities give rise to additional absorption but only at
low temperatures such that thermal energy is less than
the ionization energies of the impurity atoms.
• Absorption may occur in solids due to electron spin
reversal. Solid containing paramagnetic impurities will
show absorption line spectra in the presence of external
magnetic field.
Theoretical background 30

2-5-2 Absorption Edge


Electronic transition between the valence and
conduction bands shown in Fig (2-6) in the crystal start at
the absorption edge which corresponds to the minimum
energy difference, Eg between the lowest minimum of the
conduction and highest maximum of the valence band.

Figure (2-5) Optical interband transitions in


(a) Direct (b) Indirect band gap semiconductor

If these extreme lie at the same point of the K- space,


the transitions are called direct. If this is not the case, the
transitions are possible only when phonon–assisted and are
called indirect transitions.
Theoretical background 31

The rule governing these transitions is the


observation of quasi momentum during transitions either of
the electron alone in direct transitions or the sum of the
electron and phonon quasi momenta in indirect transitions.
The value of the gap Eopt depends in a rather subtle way on
the structure and actual values of the pseudo potential in the
crystal. When the semiconductor becomes amorphous one
observes a shift of the absorption edge either to
towards lower
or higher energies. The he shape of the absorption curve
appears to be similar for many amorphous semiconductors.
In manyy amorphous semiconductors the absorption ed edge
has the shape shown in Fig ((2-6).

Figure (2-6) parts A, B, C of the absorption edge


Theoretical background 32

It can be distinguished by the high absorption region


4
A (α > 10 cm-1). The exponential part B extends over 4
orders of magnitude of α and the weak absorption tail C.
The high absorption region A probably corresponds to
normal one–electron transitions and carriers information on
the energy dependence of the density of states at the band
edges. In general the high absorption region is often
observed in semiconducting glasses that at high enough
absorption levels (α ≥ 104 cm-1) the absorption constant α
has the following frequency dependence:

n0 ℏγα (γ ) = (ℏγ − E gopt ) n (2.2)

Where n is a constant that depends on the type of transition


and no is the refractive index. For direct transitions the
constant n in the above equation can take either the value
1/2 or 3/2 depends on whether the transition is allowed or
forbidden in the quantum mechanical sense. For amorphous
chalcogenides the spectral dependence of α is given by the
following relation:
α = B ( ℏ − E gopt ) 2 ℏ ω (2.3)

where B is a constant. For indirect transition:


(ℏω − Egopt + hγ ph )n (ℏω − Egopt + hγ ph )n
α n0ℏγ = −1
+ −1 (2.4)
 hγ   hγ 
exp   1 − exp  
 KT   KT 

rom transitions involving phonon absorption and


emission respectively. For allowed transitions n=2 and for
forbidden transitions n=3. The exponential region of the
absorption edge (part B), i.e. in the absorption constant
range from 1 cm-1 to about 104 cm-1, the absorption
constant α(ω) is described by the formula:
Theoretical background 33

α (ω) = α0 exp(ℏω E0 ) (2.5)

Where Eo is the width of the band tail. The energy Eo


characterizing the slope is almost temperature independent
at low temperatures. It is interesting to note that Eo values
between 0.05 eV and 0.08 eV.

2.6 The Electrical Properties of Chalcogenide Glasses

The D.C. conductivity σ of most of the chalcogenide


glasses near room temperature obeys the relation:

 1 Eσ 
σ = C exp −  (2.1)
 2 KT 

where Eσ is the activation energy for conduction. The


3 4
constant C is often in the range of 10 - 10 ohm-1 cm-1 but it
-5 8
can be as low as 10 ohm-1 cm-1 and as high as 10 ohm-1
cm-1. The D.C. conductivity, in general , shows that log σ is
a fairly linear function of 1/T and the variable range
hopping conductivity behaving approximately as Aexp(-
1/4
.B/T) is not observed for chalcogenide [62-63]. However
deviation from linearity can occur at low temperatures as a
result of different mechanisms for conduction.

The thermo power for the chalcogenide glasses is


normally positive, early work was frequently done at the
liquid phase. For As2Se3, Edmond [64] has found that the
activation energy in the liquid phase appears greater than in
the solid phase suggesting that the gap decreases with
increasing temperature T. In chalcogenide glasses there are
different conduction mechanisms which can be observed.
Theoretical background 34

The conductivity σ in the chalcogenide glasses can be


written as[65]:
∆E − E1 − E2
σ = σ O EXP ( − ) + σ 1 EXP ( ) + σ 2 EXP ( ) (4.3)
KT KT KT

The three terms arise from three different conduction


mechanisms and they are to be discussed separately.

a) The high temperature region

In the first region, the dominant mechanism is the band


conduction through the extended states. This region is
expressed by the first term of the R.H.S of equation (4-3).
The constant σo for the chalcogenide glasses varies from
10-2 to 5×10-9Ω1cm1 and is found to depend on the
composition [65], where ∆E is the activation energy, K
being the Boltzmann constant and is the absolute
temperature.

b) Hopping Conduction Via Localized states

This is responsible for the conduction in the second region.


Here the conductivity arises from tunneling through
unoccupied levels of the nearest neighbouring centers. The
value of σ1 is approximately (102-104)) times less than σ0
partly because of the smaller density of localized states and
their low mobilities.

(c) Hopping conduction near the Fermi level

This third contribution to conductivity in an amorphous


semiconductor is analogous to impurity conduction in
heavily doped semiconductors. In this case the conductivity
Theoretical background 35

is given by the third term on the R.H.S. of eq.(4


eq.(4-3). In the
present
ent study results on the D C conductivity of thin film
samples of amorphous InxSn20Se(60-x)Bi20 semiconductor

2.7 Switching in Alloys Glasses


The realization that films of chalcogenide alloys
show fast and reversible switching from a hi high to a low
resistance state [66]] was one reason for the rapid growing
interest in these materials from year 1968 onwards. There
are, of course, many forms of switching which can occur in
a wide variety of materials and even in lliquid alloys of S,
Se and Te [67]; ]; it is unlikely that the same mechanism is
responsible in all cases. The current
current–voltage characteristics
of a typical glass switching
tching device is shown in Fig (2-7)

Figure (2-7) Current – Voltage characteristic curve (I


(I-V) of
switch Ih denote current.
Theoretical background 36

The current in on state depends on temperature or


voltage, and the current is maintai
maintained unless the “holding
current”drops
drops below some critical value.
In the memory switch, constructed from a less stable
alloy (e.g. Ge17 Te19 Sb2 S2), partial crystallization of a
conducting channel occurs in some milliseconds after
threshold switching, memory switching occurs due to
formation of a filamentary path which is crystalline in
nature soon
oon after the formation of a crystalline filament, the
D.C. conductance increase by many orders of magnitude
and the device gets locked in the on state even after
removal of the applied DC voltage
voltage. Fig (2-9) shows the
dynamic I-V V characteristic curve ffor thin film of
amorphous (memory switch). A forming process may occur
during the initial switching
witching event, but switches can be
constructed in which no forming occurs. The main
contraversy about the mechanism of switching in these
devices has been whether itt is thermal, (a hot conducting
channel being formed leading to negative resistance), or
whether somee electronic process [68
[68] like double injection
is involved.

Figure (2-8) Dynamic (I-V)


V) characteristic curve for thin
film of amorphous semiconductor (Memory switch)
Theoretical background 37

A system in which switching is probably thermal is the


vandate glass switching investigated by Higgins [69] and
by earlier workers. However, the evidence, reviewed by
Adler et al [70], suggests strongly that thermal mechanism
is not the correct model for the chalcogenide glasses.
2-8 Radiation Sources
The sources of radiation, which are used in radiation
studies and applications, can be divided into two groups,
those employing natural and artificial radioactive isotopes,
and those that employ some form of particle accelerator.
The first group consists of the classical radiation sources,
radium and radon, and such artificial radioisotopes as
cobalt-60 cesium-137, and of various types, and
accelerators such as van de Graaff accelerator and
cyclotron used to generate beams of positive ions. Nuclear
reactors have also currently, the most widely used radiation
sources are cobalt-60 (γ-radiation) and electron
accelerators (electron beam).
Choice of a particular radiation source is generally
dependent on the nature and size of the object to be
irradiated. Gaseous materials can be irradiated successfully
using any type of ionizing radiation, but irradiation of bulk
liquid or solid samples requires one of the more penetrating
and ionizing radiations such as (γ-radiation) or a beam of
energetic electrons in MeV range. Less penetrating
radiations such as α or β radiation or lower-energy
electrons can be used if irradiation is to be restricted to the
surface layers of the sample.
2-8-1 Gamma Radiation
Gamma rays are electromagnetic radiation of nuclear
origin with wave lengths in the region of 3×10-11 m to
3×10-13 m. It is more convenient to describe the radiation in
Theoretical background 38

terms of energy than in terms of wavelength since it is the


energy absorbed from the radiation is basically of interest.
The relationship between wavelength and energy is
E = hc / λ (2.6)
Where h is Plank's constant, c is the velocity of light,
and λ is the wave length. Substituting for the constants
gives
E (eV ) = (1.24 ×10−6 ) / λ (m) (2.7)
In terms of energy the wave length range 3×10-11 m
to 3×10-13m becomes approximately 40KeV to 4 MeV. The
γ-rays emitted by radioactive isotopes are either
monoenergetic or have a small number of discrete energies.
Cobalt-60, for example, gives equal numbers of gamma
photons of energy 1.332 and 1.173 MeV.
2-8-2 Interaction Of Gamma Radiation with Matter
In passing through matter, γ-radiation interacts with
matter in a variety of processes. The three main processes
are the photo electric effect, Compton scattering and pair
production. The probability of each of these interactions
depends on the energy of the incident photon. At low
energies, the Compton Effect dominates, and at high
energies pair production dominates. as shown in fig (2-9)
(i) photo electric effect
When γ-rays are incident on a solid, a single electron
absorbs the incident photon and becomes exited to the
conduction band or ejected from the atom. This ejected
electron will collide with other electrons, sharing the
electron. This will result in many electrons exited to the
conduction band, each with roughly the same energy.
Eventually, these electrons will fall back to the more stable
Theoretical background 39

ground state. When this occurs, each will emit a photon


with energy approximately equal to the band gap. Because
this process happens so quickly, all the electrons will fall
back to the ground state at roughly the same time. The
photoelectric effect is the dominant energy transfer
mechanism for X and γ-ray photons with energies below 50
KeV.
(ii) Compton scattering
When a γ-ray collide elastically with an electron, the
electron absorbs some of the energy, and the photon
continues in a new direction with less energy and a longer
wavelength. The amount of energy absorbed by the
electron is dependent on the scattering angle of the γ-
photon after collision takes place. When maximum energy
is transferred to the electron, the rebound photon has
maximum wavelength and minimum energy. Compton
scattering is thought to be the principal mechanism for γ-
rays in the intermediate energy range from 100 keV to
10MeV.
(iii) Pair production
A gamma ray may spontaneously change into an electron
and positron pair, in the vicinity of the nucleus. A positron
is the anti-matter equivalent of an electron. It has the same
mass as an electron, but it has a positive charge equal in
strength to the negative charge of an electron. Energy in
excess of the equivalent rest mass of the two particles (1.02
MeV) appears as the kinetic energy of the pair and the
recoil nucleus. The positron has a very short lifetime (if
immersed in matter) (about 10-8seconds).At the end of its
range, it combines with a free electron .The entire mass of
these two particles is then converted into two gamma
photons of 0.51 MeV energy each. The secondary electrons
(or positrons) produced in any of these three processes
Theoretical background 40

frequently have enough energy to produce many ionization


up to the end of range.

Figure (2-10) The interaction of gamma rays with matter


probabilities

Figure (2-11) Mechanism of the interaction of gamma rays


with matter probabilities.
Theoretical background 41

2-9 Radiation Effects on Solids


Historically, one of the earliest examples of the action
of radiation on solids was the production of pleochroic
holes in mica by radiation from inclusions of radioactive
substances such as uranium or thorium [71]. Early studies
of the action of nuclear radiations on solids showed, among
their effects, that colorless glass becomes colored up on
exposure to radiation and that the coloration can be
removed by the action of heat or light. Clearly, a variety of
effects may be observed, depending on the nature of the
radiation and the solids. Much of the information available
is concerned with physical rather than chemical effects [72]
and no attempt is made here at an exhaustive treatment.
All types of ionizing radiation are able to produce
ionized and excited atoms in the solid and, when ionizing
radiation is absorbed in semiconductor materials a
temporary change in electrical conductivity is induced by
the production of electrons or positive holes having
sufficient energy to be free to move through the material. In
certain materials a permanent change in conductivity may
be produced by radiation damage. This change in
conductivity can be measured as a function of the absorbed
dose in the semiconductor detector, so that the system is
then an integrating dosimeter [73-75]. The materials most
commonly used are Silicon, Germanium, Cadmium Sulfied
and recently porcelain [76-77] and Oxide Glass [78]
The theory of radiation damage is based on the
assumption that the simplest defect arising in a solid is a
vacancy or displaced atom with a more or less stable
position in the interstitial. That is, if the energy transferred
to the lattice atom is greater than a certain amount, the
lattice atom will be displaced to an interstitial position. The
displaced atom will leave behind a vacancy in the lattice.
Theoretical background 42

The result vacancy–interstitial pairs is called a “Frenkel


pair” [79-86]. Frenkel defects are the simplest point defects
in lattice and differ from dislocation and from the more
complex and extended imperfection.
Frenekel defects are produced uniformly through
o
lattice. At finite temperature, particularly above -173 C,
the vacancies are mobile in the solid, and they unite with
themselves or with impurities in the solid, and from defect
complex in order to obtain lower energy levels in the
forbidden gap. Theoretical analysis of the problem is
difficult especially due to the lack of data concerning the
exact values of the capture cross–sections. The number of
defects produced by hard radiation and the energy levels of
such defects can be studied by measuring the electrical
conductivity and the Hall Effect.
Two aspects for the interaction of radiation with
matter are recognized. A large fraction of the energy of an
incident energetic goes in electronic processes (excitation
and ionization), and this produces, temporary or transient
disturbance in the material which disappear shortly after the
removal of radiation source. The remainder of the incident
optical energy goes into atomic processes and produce
displacement of atoms within the solid lattice. The fraction
of these displacements which remain after long time, (103)
seconds at room temperature will be called damage.
There are many factors influencing the amount and
type of damage produced by radiation. Some of these
factors can change the results by a factor of 2 or 3, and
some others change the results by more than an order of
magnitude for a given radiation. The amount of permanent
radiation damage is given by:
Damage = f ( A, B, C, D, E)
Theoretical background 43

Where A, B, C….etc. are all parameters, each depend on:


i. The temperature of sample during irradiation.
ii. Inactive dissolved impurity concentration.
iii. P-Type or n–type.
iv. History after irradiation.
v. Rate of irradiation.
So, in studying the damage produced by radiation, these
parameters must be known, or at least they must be kept
constant. Electrons, gamma–rays, neutrons and heavy
particles, are most interesting types of radiation commonly
used in solid state studies.
2-9-1 Defect Production by Gamma–Rays
Gamma rays, like fast electrons produce point
defects. The probability of an atom being displaced by a
direct interaction between gamma–rays and atomic nuclei
in crystal is very small.
Most defects are formed by the secondary fast
electrons produced as a result of the photo electric effect,
the gamma rays Compton scattering or/and the electrons–
positions pairs creation formed at high gamma–ray
energies. The total absorption cross–section of gamma–
rays is determined as a result of the above three processes.
Evans and some others could calculate the probabilities of
occurrence for each of these events.
Experimental Techniques 44

Chapter 3

Experimental Techniques
3.1 Preparation of Bulk compositions

There are at least a dozen of different techniques that


can be used to prepare materials in the amorphous phase.
Of these, two normal ways are commonly used in one form
or another to produce most non-crystalline material; by
cooling from a melt or by condensation from the vapour.
The first method forms bulk materials, while the second
yields thin films as in thermal evaporation, sputtering or
glow discharge techniques. During the preparation of the
amorphous material, the faster the rate of cooling or
deposition the farther the amorphous solid lies from
equilibrium .The present glasses were prepared from Sn,
Se, Bi and In elements with purity 99.999 %. These glasses
are reactive at high temperature with oxygen. Therefore,
synthesis was accomplished in evacuated clean silica tubes.
The tubes were washed by distilled water, and then dried in
°
a furnace whose temperature was about100 C. For each
composition the proper amounts of materials were
weighted using an electrical balance type (Sartorius) with
-4
accuracy ± 10 gm. The weighted materials were
introduced into the cleaned silica tubes and then evacuated
-6
to about 10 torr and sealed. The sealed tubes were placed
inside the furnace and the temperature of the furnace was
raised gradually up to 900°C within 1 hour and kept
constant for 10 hours. Moreover, shaking of the constituent
materials inside the tube in the furnace was necessary for
realizing the homogeneity of the composition.
Experimental Techniques 45

After synthesis, the tube was quenched into ice


water. The glassy ingots could be obtained by drastic
quenching. Then materials were removed from the tubes
and kept in dry atmosphere. The ingot materials were
identified as glass due to their bright features. The proper
ingots were confirmed to be completely amorphous using
x-ray diffraction and differential thermal analysis.
Homogeneity of the prepared samples was proved by
determination of density of different parts. Silica tubes used
for bulk amorphous is illustrated in Fig (3-1).

Figure (3-1) Silica tubes used for bulk amorphous


semiconductors preparation.
Experimental Techniques 46

Washing silica tubes with distilled water and


dried in a furnace whose temperature about 100
°C.

Weighting of each component in composition.

Weighted materials introduced into the cleaned


silica tubes.

Cleaned silica tubes evacuated to about 10-6 torr


and sealed.

The sealed tubes placed inside the furnace and the


temperature of the furnace was raised gradually
up to 900°C within 1 hour and kept constant for
10hours.

Shaking the constituent materials inside the tube


in the furnace is necessary for realizing the
homogeneity of the composition.

After synthesis the tube was quenched into ice


water.

Materials removed from the tubes and kept in dry


atmosphere.
Figure (3-2) Design flow chart for Preparation of bulk
amorphous InxSn20Se(60-x)Bi20.
Experimental Techniques 47

3.2 Preparation of Thin Films

Thin films of the selected compositions were


prepared by the most widely used technique known as
thermal evaporation technique of bulk samples on to glass
substrates for both optical and dc conductivity
measurements. In order to prepare well-formed and
homogeneous films onto glass substrates and homogenous
glass substrate with low surface roughness the following
steps were carried out:

i) Cleaning the substrate

• The substrates were washed several times


using hot distilled water and soap.
• Then substrates were exposed to ultrasonic
waves using Branson-120 device for 15
minutes in a solution of distilled water and
ethyl alcohol.
• Finally the substrates were washed with
distilled water and ethyl alcohol separately and
then, dried in an oven.

ii) Thermal deposition of thin film

An Edward 306E coating unit shown in Fig (3-3)


was used for thin film deposition. The vacuum system
consists mainly of a rotary pump, diffuse on pump, penning
bridge for measuring vacuum, high A.C. current source and
the bell jar. The heating of the tungestun boat was achieved
by spiral tungsten wire. The Tungetun boat had to be
cleaned every time before evaporation. This was
accomplished by using hydrochloric acid, then washing
Experimental Techniques 48

several times with boiled distilled water and finally it was


dried in a furnace, whose temperature was about 100 °C.
The Fig (3-4) illustrated a flow chart for Preparation of Inx
Sn20Se(60-x )Bi20 Films and Fig (3-5) showed that the
copper mask designed for E-306A.

(a) (b)

Figure (3-3) a) schematic diagram b) captured photo

Vacuum coating unit


Experimental Techniques 49

Cleaned tungestun boat Washing several times


every time before with boiled distilled water
evaporation by using and dried in furnace
hydrochloric acid. whose temperature was
about 100 °C.

Edward 306E coating unit The heating of the silica


was used for thin film boat was achieved by
deposition at a pressure of spiral tungsten wire ,
10-6 torr.

The distance between the The vacuum system was


substrate and upper ends turned ON until the
of the boat was adjusted vacuum inside the bell jar
to be 14cm. reached 10-6 torr

During the deposition The thickness of the


process (at normal prepared thin film
incidence), the samples was measured by
substrates were suitably using thickness monitor
rotated in order to obtain
films of uniform

Figure (3-4) Design flow chart for Preparation of


InxSn20Se(60-x )Bi20 Films.
Experimental Techniques 50

Figure (3-5) The


he copper mask designed for E
E-306A.

3.3 Methods for thin film thickness measurement

The film thickness is one of the most important


parameters. The thin film properties show a high degree of
dependence on it, so it should be measured accurately. In
the present work,, the film thickness was measured by
Applying
ing common techniques:

• A monitoring
toring technique using a quartz crystal
thickness monitor.

Figure (3-6) Thickness monitor (TM


(TM-200).
Experimental Techniques 51

3.3.1 Quartz crystal thickness monitor technique

In this method, the film was deposited onto one


electrode of a quartz crystal connected to an oscillator to an
oscillator circuit. The crystal thickness (dq) has the
fundamental resonance frequency (f) [87]

f = V p / 2d q = C / d q (3.1)

Where VP is the velocity of transverse elastic wave in the


direction of the thickness dq and C is the frequency constant
when a certain amount of the material is deposited on the
quartz crystal; the thickness will be increased by δdq, where

δ d q = δ m / Aρq (3.2)
where δm is the mass of the deposited film, ρq is the density
of quartz and A is the film area on the crystal.
Due to increase in the mass, the frequency will change by
δƒ, where
δ f = ( f 2 / C ).(δ m / Aρq ) (3.3)

The linear dependence of δƒ on δm was given by the


above equation, which has a limited range of validity for
film thickness of approximately 2µ and with thicker film
the crystal becomes non-linear, The film thickness ,d ,is
determined by

D = G(∆f / ρm ) (3.4)

Where G is a constant in gm/Hz.cm2, ∆ƒ is the frequency


changein Hz, ρm is the material density in gm.cm-3.
Experimental Techniques 52

3.4 Density Determination

Density is an important physical parameter which is


related to other physical properties of the material and also
is used to examine the homogeneity of the as-prepared
materials. The density of the considered samples was
determined using the method of hydrostatic weight using
toluene. A single crystal of germanium was used as a
reference material for determining the toluene density. The
latter has been determined from the formula:

'
Wair − Wtoluene
'
dtoluene = '
× dGe (3.5)
Wair

Where, w' is the weight of single Ge crystal. Then,


the sample density was calculated from the formula:

Wair
dsample = ×d (3.6)
Wair −Wtoluene toluene

Where w is the weight of the sample


.
While the density dth of the prepared compositions
was also calculated theoretically using Myuller,s formula
[88].
Pi −1
d th = [ ∑ ] (3.7)
i di

Where Pi is the fraction of weight of the ith element and di is


its density.
Experimental Techniques 53

3.5 Structure Measurements

The Structure characteristics of the investigated In-Sn-


Se-Bi system in both powder and thin film forms have been
investigated by:

1. X-Ray analysis.
i) X-Ray Diffraction techniques (XRD).
ii) Energy Dispersive X-ray (EDX).
2. Diffraction Thermal Analysis (DTA) Technique.
3. Scanning Electron Microscope (SEM).

A brief description of these methods will be given below.

1- X-Ray analysis

X-ray analysis for both bulk and thin films include two
analyses for getting compositional and structural
information about prepared samples.

i) X-ray diffraction techniques (XRD).

X-Ray diffraction has been used to investigate and


characterize the structure of the as prepared samples. This
has been done with the aid of a “Shimadzu XRD-6000’’ as
shown in Fig (3-7) diffractmeter which consists of
αpw1400/90 stabilized X-raygenerator,αpw1050/70 vertical
goniometer, αpw1995/60 proportional counter and
αpw1930 electronic panel. Nickel–filtered copper radiation
with λ=1.542 Å was used in the present investigation.

The choice of a particular X-ray radiation source of a


certain wavelength depends on the fact that if the radiation
Experimental Techniques 54

wave length λ is less than that of the absorption edge of the


examined sample, large absorption will cause a faint
reflection. The filter was located between the scatter slit
and the count tube. The X- ray has been operated at 50 kv
and 25 mA throughout the measurements.

A sample of the considered material was ground to


fine powder flat rectangular specimen holders, made of
aluminum, and is used, each with an aperture of 2 x 1 x 0.2
cm3. The rear of aperture was closed with metal plates. A
smooth specimen surface was prepared by backing with a
glass slide. By an appropriate choice of the slit system, at
least half of the specimen surface area was examined in
each angular range from 10 to 80 degrees (2θ) X–ray
diffraction pattern was recorded automatically by the
diffract meter, is essentially a plot of intensities as a
function of the angle of reflection given by the material.

The resulting diffraction pattern of crystal comprises


both positions and intensities of the diffraction effects.

Figure (3-7) X-ray diffractmeter, “Shimadzu XRD-


6000’’
Experimental Techniques 55

Analysis of the position of the diffraction effects


leads immediately to a knowledge of the size, shape, and
orientation of the unit cell, while the position of the
individual atoms in the cell may be found from the peak
intensity measurements. The measured angle (2θ) of the
crystalline samples and the inter planar (d') are related
through Bragg’s law.
nλ = 2dhkI
'
sin θ hkI (3.8)
Where λ is the wave length of the X–ray source, θ is
the Bragg’s angle (angle of the incidence or reflection),
d’hkI refers to the orientation of the plane, and n is an
integer equal to 1, 2, 3 ….etc.

ii) Energy Dispersive X-ray:

The elemental analysis of the prepared samples in


the bulk and thin film form was performed by a technique
known as Energy Dispersive X-ray (EDX) analysis In this
analysis the specimen is subjected to an energetic electron
beam(20KeV) resulting in production of characteristic X-
rays from the specimen surface. The emitted X-ray
spectrum was used in determining qualitatively and
quantitatively the composition of each sample and to ensure
the sample homogeneity as well. The electron beam energy
20keV was selected for two reasons; the first is because,
the maximum X-ray energy collected by the device is
20keV on the X-ray scale. The Second reason is to prevent
high penetration of the film, since energy greater than
20keV may occur. “AJEOL-5400” Scanning Electron
Microscope (SEM) with (EDX) attachment was used in
determining the accurate composition of both bulk and thin
film samples.
Experimental Techniques 56

2- Differential Thermal Analysis (DTA)

Differential thermal analysis (DTA) is the technique


of measuring the heat effects associated with physical or
chemical changes that take place as a substance is heated at
a constant rate.. The appearance of many papers in the last
decades in which quantitative DTA results showed that
under controlled conditions, fairly good quantitative results
can be obtained also, the availability of better equipment
has assisted in getting reproducible conditions. The basic
principle of DTA is the measurement of the temperature
difference between the sample and a reference material as
they are heated simultaneously under uniform rate.
A solid state reaction causes an evolution of heat,
which is shown as temperature difference ∆T between the
sample and the reference material [89]. As this heat is
dissipated to the surroundings, ∆T reduces to zero again.
Measurements of the temperature T and ∆T over a suitable
range gives a thermo gram characteristic of the reaction
occurred. Accordingly, the DTA curve of Fig (3-8)
continues in an approximately rectilinear manner, until the
test material undergoes some physical or chemical change.

Figure (3-8) Typical DTA thermo gram illustrating the


definition of the different transition temperature.
Experimental Techniques 57

The curve begins to deviate from the baseline; the


first deviation is more representative of the start of a
transition then the peak. V.G. Hill and Roy [90] stated that
the nearest approach to transition temperature is the point at
which the curve leaves the baseline. This point is difficult
to be determined, so the intersection of the baseline and the
extrapolation of the straight part of the adjacent side of the
peak is the glass transition temperature Tg, then the curve
begins to deviate from the base line forming the
crystallization peak at temperature Tc. After that, the curve
returns to the new baseline. The onset of an endothermic
reaction is indicated by the downwards deflection of the
baseline giving rise to endothermic peak, melting
temperature Tm.
A micro–DTA apparatus, Shimadzu DTA-50 model,
was used for the measurement of DTA. Fig (3-8), Fig (3-9)
showed the captured photo and a block diagram illustrating
the main units of the used DTA.

Figure (3-9) Differential thermal analysis


Experimental Techniques 58

Figure (3-10) Principle diagram of Differential thermal


analysis.

It is mainly a combination of a sample holder, DA-


30 amplifier, control unit and two pen recorder R–22T. The
control unit permits temperature control for the sample
holder, also it permits the temperature to be displayed
digitally either in C or K . The pen of the recorder provides
a mark at every 1 degree or 10 degree on the chart to ensure
more precise temperature measurement.
Also, the temperature is simultaneously recorded as a
solid line on the chart. The heating rate can be adjusted to
ten different rates ranging from 1 degree/min up to 100
degree/min. The experimental conditions adopted were in
accordance with the recommendations of Machenzi [91].

i)Preparation of Samples for DTA Measurements

The studied samples were 25mg powder of each of


the compositions under investigation. They were heated up
to about 500 °C in aluminum cells. All measurements were
referenced to a α-Al2O3 powder in Al cell. However, in
order to obtain good and reproducible results, the following
precautions were taken during measurements:
Experimental Techniques 59

• The temperature distribution in the sample was made as


uniform as possible.
• The sample must be in good contact with the cell, and the
cell must be in good contact with the detector. This means
that the bottom of the cell must be flat.
• Nitrogen flow must be allowed at a constant rate of about
50ml/min to prevent oxidation of the sample.

25mg powder of each of the composition under


investigation

Heated up to 500 °C in aluminum cell

Measurements were referenced to α –Al2O3 powder


in Al cell

To obtain good results take into consideration the


following Temperature distribution in the sample
was made as uniform as possible.

1- Bottom of the cell must be flat to have good


contact with the sample.

2- Nitrogen flow must be at constant rate


50ml/min

Figure (3.11) Design flow chart for DTA thin film


Preparation.
Experimental Techniques 60

3) Scanning Electron Microscope technique (SEM)

The morphology of the surface of In-Sn-Se-Bi thin


films could be investigated by using a Scanning Electron
Microscope.

Figure (3-12) AJEOL-5400 Scanning Electron


Microscope (SEM) with (EDX).

3.6 Optical Measurements

Optical transmittance and reflectance for thin


film have been measured using a double beam
(Shimadzu UV-160A) Spectrophotometer. Values of
the absorption coefficient (α) of the absorption
coefficient (α) for the InxSn20Se(60-x)Bi20 (where x=0
,0.1 ,0.2 & 0.3at.%) thin film were calculated from
the Transmittance T and Reflectance R. According
to Tauc's relation for the allowed non-direct
Experimental Techniques 61

transition, the photon energy dependence of the


absorption coefficient can be described by

(α hυ )1 2 = B (h − E0 )
n
(3.9)

Where is n a parameter that depend on the


transition probability and Eo is the optical energy gap.
Values of Eo were taken as the intercept of the
straight (αhυ)1/2 versus hυ line with the energy axis
where hυ is the photon energy .

Figure (3-13) Shimadzu UV-160A)


Spectrophotometer

3.7 Electrical Measurements

In order to measure the electrical conductivity of the


InxSn20Se(60-x)Bi20 (where x=0 ,0.1 ,0.2 & 0.3at.%) thin
films two gold planar electrode with spacing 0.1 cm and
length1.4 cm as shown in Fig (3-13)
Experimental Techniques 62

Figure (3-14) Gold planar electrodes.

The D.C conductivity has been measured for thin


film samples. By measuring Sheet resistance for these
films. The resistance value is given by equation

ρL
R= (3.10)
db

Where d is the thickness, b is the length of electrode and L


is the distance between electrodes. If (L=b), then equation
(3.4) becomes

ρ
R= = Rs (3.11)
d

So that the resistance Rs of one square of film is


independent of the size of the square but depend only on
resistivity ρ and thickness. The quantity Rs of the film and
is expressed in ohms per square cm. It is a very useful
quantity that is widely used for comparing films,
particularly those of the same material deposited under
similar conditions. If the thickness is known, the resistivity,
is readily obtained from

ρ = dRs (3.12)
Experimental Techniques 63

The electrical measurements were carried out in a dry


nitrogen atmosphere using a specially designed system
capable of measuring resistance, Fig (3-14) shows a block
diagram of the circuit used for measuring electrical
conductivity.

Figure (3-15) Construction used for controlling the


temperature of the sample in the Range from room
temperature to liquid nitrogen.
Experimental Techniques 64

Figure (3-16) Block diagram of the circuit used for


measuring electrical Conductivity

i) Preparation of Thin Film for Switching

After testing many types of substrate materials,


pyrographite was chosen as a suitable substrate for thin
film evaporation for switching measurements. One surface
of each substrate is highly polished using three different
grades of A1203, finishing with the finest. The polished
substrates were washed several times with boiled distilled
water followed by alcohol to remove all traces connected to
them during the polishing process. After cleaning the
substrates, they were dried in a furnace at moderate
temperature.
Experimental Techniques 65

Preparation of thin film samples was performed


according to the following procedure:

 Chalcogenide glass of the considered composition


was first crushed into small grains and introduced
into the cleaned dry silica boat which was put in its
place inside the electrical spiral heater.

 The highly polished pyrographite plane substrates


were placed horizontally on the copper mask, inside
the holes with the polished surfaces facing tungestun
boat.

 The distance between the substrate and the upper end


of the boat was adjusted to be 14cm.

 The bell jar was fixed in its place and the vacuum
system was turned ON until the vacuum inside the
bell jar reached (10-6) torr. At such an evacuation
value, the tungestun boat which was heated to a
sufficiently high temperature ensures that
evaporation occurs at a suitable rate.

 The thickness of the prepared thin film samples was


measured by using the thickness monitor type “TM-
200”.
Experimental Techniques 66

Cleaned tungestun boat every Washing several times with


time before evaporation by using boiled distilled water and dried
hydrochloric acid in furnace whose temperature
was about 100 °C

Edward 306E coating unit was The heating of the tungestun


used for thin film deposition at a boat was achieved by spiral
pressure of 10-6 torr tungsten wire

The distance between the The vacuum system was turned


substrate and upper ends of the ON until the vacuum inside the
boat was adjusted to be 14cm bell jar reached 10-6 torr

During the deposition process (at The thickness of the prepared


normal incidence), the substrate thin film samples was measured
(Pyrographite) were suitably by using thickness monitor
rotated in order to obtain films of
uniform thickness

Figure (3-17) Design flow for Preparation of thin film


amorphous InxSn20Se(60-x)Bi20 for switching

A special cell has been constructed for switching


measurements. The cell is made of Teflon block in order to
Experimental Techniques 67

give high insulation. The complete assembly is shown in


Fig (3-18). A longitudinal groove is made to accommodate
the upper copper rod electrode with a spring around the rod
so that it could freely move up or down against the spring.
A carbon needle is fixed at the lower end of the copper
electrode. A copper disc is used as a lower electrode which
is embedded in the Teflon block and it is centered with the
longitudinal axis of the upper electrode. The thin film
sample on the pyrographite substrate is placed between the
carbon needle and copper disc electrode of the measuring
cell so that the spring will ensure good contact all the time.
Fig (3-19) illustrated the measuring circuits which are
connected across the cell.

Figure (3-18) A special cell construction for I-V


measurements
1-Spring 2-Cylinder copper
3-Terminal 4-Copper nut
5-Copper rod 6-Carbon needle
7-Sample 8-Lower electrode
9-Teflon
Experimental Techniques 68

Figure (3-19) A Simple Circuit used for measuring I-V


characteristics D.C

The resistance R in the circuit is a limiting current one,


used to protect the thin film against over current during the
ON state. Increasing the voltage applied to the switch
device. Leads to a slow increase of the current passing
through it.

At a certain value of the applied voltage, the current


increases while the voltage across the switch decreases.
This value of the applied voltage is the switching voltage
vs. the reading of the voltmeter and micrometer were
recorded and plotted for each sample.

3.8 Spectrophotometer Measurements

UV/VIS UVIKON 860 a double beam spectrometer


was used to measure the optical density (absorbance). It
mainly consists of a mono chromator that covers the range
Experimental Techniques 69

of wavelength from 190–800800 µm, beam splitter, detection


and recording system, the whole system is computer
controlled. The light source is a quartz halogen lamp
integrated with a system of mirrors, a filter wheel; the beam

Figure (3-20) SP8 -200 Optical Diagrams.

is then splitted into two beams to pass through the samples


and the reference. Fig (3-20)) shows the optics layout of the
used spectrophotometer.

The infrared absorption spectra of the studied glass


samples were measured using Fourier transform (FTIR)
Unicam Mattson–10001000 spectrophotometer in conjunction
with KBr disc technique over the spectral range of 400
400–
-1
6000 cm . All infrared measurements were carried out in
room ambiance. Glass samples of 2 mg were thoroughly
mixed and ground with 200 mg KBr, after which the
mixture was pressed at 7.5 tons/cm2 for 60 minutes under
vacuum to yield transparent discs for mounting and
absorption spectra measurement
measurements.
Experimental Techniques 70

3.9 Gamma Irradiation Source

The gamma irradiator is housed in a shielding building,


Fig (3-21), constructed upon a ground of standard density
concrete (2.36 g/cc), having thickness about 120cm, so that
no one receives more than 10 mr of radiation during 40
hours/week, or the maximum dose rate would not exceed 0.2
mr on all accessible when 1,000,000 curie cobalt radiation
source is utilized. The new irradiation facility, Egypt's
MEGA, gamma I, supplied as type J-6500 by the Atomic
Energy of Canada Ltd., at the National Center for Radiation
Research and Technology, Cairo, has been furnished with a
cobalt–60 source. The source consists of a large planar array
of encapsulated Co60 rods 0.18 cm in diameter and 20.96 cm
in length. Two of such source rod elements are inserted each
double–walled stainless steel pencil of external dimension
1.11 cm in diameter, and 45.2 cm in length. The pencils are
inserted in planar modules 47.5 x 49.5 cm. The source rack
is composed of six modules each having 42 source pencils
and the maximum plaque capacity 252 pencils.

The samples were fixed facing the source at a


distance of 15 cm. At each position, the dose rate was 5.56
Gray / sec. At the time of the present measurements. The
thin film samples were irradiated at room temperature, and
the absorbed dose was measured by using standard
calibrated Perspex dosimeter.
Experimental Techniques 71

Figure (3-21)) J6500 Irradiator


Results & Discussions 72

Chapter 4

Results & Discussions

This chapter deals with the experimental results, of


In xSn20 Se(60-x)Bi20,investigated by X-ray diffraction
analysis (XRD) and differential thermal analysis (DTA)
for the powder and thin film at constant thickness 100 nm
.The surface morphology was investigated by scanning
electron microscope (SEM).

4.1 X-ray Diffraction Identification of Bulk Samples.

The X-ray diffraction technique was used to


investigate the structure and character of our system In x
Sn20 Se(60-x)Bi20 ,(where x=0,0.1,0.2 and 0.3%) .The X-ray
diffraction patterns of the amorphous prepared Samples are
shown in Fig (4-1)

Figure (4-1) X -Ray Diffraction patterns of bulk sample of


the system InxSn20Se(60-x)Bi20, (where x=0,0.1,0.2 and
0.3at.%)
Results & Discussions 73

The patterns show the characteristic diffraction peaks


which reflect the absence of regular crystalline structure.
Such a diffraction pattern characterizes the pure amorphous
state of the investigated samples.
The X-ray diffraction technique is considered to be
as complementary to differential thermal analysis (DTA)
technique and ensures that all the prepared samples of our
system are amorphous.

The diffraction patterns show a two stepped humps


around the Bragg's angles; 4-8 and15-28 degrees (2θ),
respectively. The second hump has the highest intensity
and occupies the largest angular range. This means that this
second step possesses the highest contribution of
diffracting atomic planes in the present domains.

4.2 Energy dispersive X-ray analysis (EDX)

Samples prepared in both powder and thin film


forms; have to be checked out compositionally from the
following three points (i) the type of constituent elements
forming the samples, (ii) the atomic percentage of each
constituent element and finally(iii)the sample homogeneity.
Results & Discussions 74

Figure (4-2) EDX qualitative analysis for InxSn20Se (60-


x)Bi20 (where x=0%) bulk amorphous materials

Figure (4-3) EDX qualitative analysis for InxSn20Se (60-x)


Bi20 (where x=0.
x=0.3 %)
Results & Discussions 75

The adequate analysis for checking out the former


three points is the EDX analysis, because of many reasons,
the most important are:

• When the specimen is subjected to the energetic


electron beam (20keV), the resulting characteristic
X-rays from the specimen are emitted from the
transitions of the inner shells electrons, as any of
such electrons is dislodged from its shell. So, the
emitted X-ray energy is independent of any type of
bonds existing between the outer shells-sharing-
electrons. Therefore, characteristic X-rays spectrum
is the element fingerprint which confirms the type of
the constituent element regardless of its chemical
bonding.

• The collected X-ray spectrum is in the form of X-ray


energy in keV against the number of counts
collected in a certain acquisition time, as shown in
the illustrative example Fig.(4-2,3),where it can be
noticed that the controlled number of counts is not
a sharp line at certain X-ray energy, but rather it is a
symmetrical Gaussian–shaped curve about the
energy value where the area under curve can be
used statistically to determine the atomic percentage
of each element and thus its concentration.

• The EDX allows checking of the sample


homogeneity, by collecting X-ray from sample at
random areas and comparing their results.
Results & Discussions 76

4.2.1 Scanning electron microscope technique


The morphology of the samples can be investigated
by Scanning Electron Microscope (SEM).The sample was
fractured and gold coated before SEM examination to study
the internal morphology as well as the surface. The
scanning micrographs of specimens of different
compositions are shown in Fig (4.4.a) shows which the
SEM micrographs of Se-Bi-Sn film deposited at room
temperature .The film consisted of individual grains, which
are irregular in size and shape and separated by well-
defined inter-grain boundaries. A similar image contrast
has also been observed in other thin films of amorphous
elemental and compound semiconductor.

By adding Indium in Fig (4-4b) further separation of the


surrounding media gives rise to large grains in size then
large grains can be seen in fig(4.4c) for partially crystallite
at x=0.2 at %. In Fig (4-4d) the grain sizes become smaller
for x=0.3 at % and the number of grains become larger
Results & Discussions 77

(a) (b)

(c) (d)

Figure (4-4) A SEM photograph of Sn20Se(60-x)Bi20


(a) bulk, (b) where x=0.1%,
(c) where x=0.2 % , (d) where x= 0.3 %.
Results & Discussions 78

4.3 Differential Thermal Analysis (DTA)

The two categories of materials useful for threshold


and memory devices, respectively, are (A) those whose
structure does not change during the device operation and
(B) those whose structure can be changed in a controlled
and reproducible manner.

Real materials only approximate these idealized


situations, and other factors besides composition affect the
stability and reproducibility of non- crystalline solids in
devices, such as contact, the substrate materials, and
preparation and surface conditions. Stable vitreous
semiconductors of type (A) are found among the three-
dimensionally cross-linked chalcogenide alloy glasses.
There appear to be specific compositions in each glass-
forming system which are particularly stable by having an
optimum number of stable bonds and cross linked
chalcogenide alloy glasses. There appear to be specific
compositions in each glass- forming system which are
particularly stable by having an optimum number of stable
bonds and crosslink. They can be heated to the molten state
and slowly cooled without devitrication or phase
separation. Increasing the number of component of similar
bond strength in a glass stabilizes its structure. At the same
time, however, the numbers of possible compositions
increases, which may phase separate.

An extensive study of stability and the extent of glass


forming regions chalcogenide alloys have been performed
by Hilton, Joes. Brau (1964,1966a,1966b) [92].
Results & Discussions 79

Figure (4-5) DTA measurements for In XSn20Se 60-x Bi20


glasses heating rate 10c/min.

An endothermic peak in the DTA curve results from


an increase in specific heat at the glass transition
temperature Tg. The absence of any sharp exothermic peak
in the DTA curve is good indicator for the absence of the
structural changes as shown in Fig(4-5).

4.4 The density dependence of (In) content

The density of the as prepared glasses of the system


InxSn20Se(60-x)Bi20 films has been determined by the
hydrostatic method with an accuracy of ±0.05%. The
obtained results are given in table (4-1) where it is noticed
that the density decreases by increasing In from 5.712
gm/cm3 for the composition InxSn20Se(60-x)Bi20 at x=0%
down to 4.682 gm/cm3 for composition InxSn20Se(60-x)Bi20 at
Results & Discussions 80

x=0.03%. The atomic volume of In is greater than that of


Se which means that the molecular weight of a given
volume increases by increasing In content.

In at% Composition dTh gm/cm3 dExp(gm/cm3)

0 Sn20Se60Bi20 5.7868 5.712


0.1 In0.1Sn20Se59.9Bi20 5.415 5.343
0.2 In0.2Sn20Se59.8Bi20 5.0955 4.955
0.3 In0.2Sn20Se59.7Bi20 4.811 4.682

Table (4-1) the composition dependence of density.

As it is known, the change of density is related to the


change in the atomic weight and the atomic volume of the
elements constituting the system. The atomic weights of
Sn, Se, Bi and In are 118.69 , 78.96 , 208.980 and 114.82
and their respectively, atomic radii are1.62 , 1.40,1.70,1.66
respectively. Fig (4-2) represents the relation between the
density and In content. It is clear that there is a linear
dependence up to a volume of almost 30%In. It does not
change appreciably.

5.8

5.6 dTh
dexp
Density(gm/cm )

5.4
3

5.2

5.0

4.8

4.6

0.00 0.05 0.10 0.15 0.20 0.25 0.30

In content at.%

Figure (4-6) Dependence of density on In content in the


system InxSn20Se60-xBi20 (with x=0, 0.1, 0.2, 0.3at%).
Results & Discussions 81

Some physical parameters of Sn, Se, Bi and In


elements are given in Table (4-2).

Ioffe and Regel [93] have suggested that the bonding


character in the nearest neighbor region, which is the
coordination number, characterizes the electronic properties
of the semiconducting materials. The coordination number
obeys the so-called 8-N rule, where N is the valency of an
atom; the number of the nearest-neighbour atoms for Sn ,Bi
,Se and In are calculated and listed in table(4-2). The
average coordination number in the quaternary compounds
AαBβCγDλ is as:

α N co ( A) + β N co ( B ) + γ N co (C ) + λ N co ( D )
N co = (4.1)
(α + β + γ + λ )

where α, β ,γ ,λ are the valencies of the elements of


compound

The determination of Nco allows the estimation of the


number constraints (Ns). This parameter is closely related to
the glass-transition temperature and associated properties.
For a material with coordination number Nco , Ns can be
expressed as the sum of the radial and angular valence force
constraints.[94],
N CO
NS = + (2 N CO − 3) (4.2)
2

The calculated values of Nco and Ns for the


InxSn20Se(60-x)Bi20system ,using the elemental coordination
number of Sn, Se ,Bi and In given in Table 1.are listed in
Table( 4-3).
Results & Discussions 82

Physical Sn Se Bi In
characteristics
Energy gap (eV) 0.15 1.95 0.407 1.74
Density(g/Cm3) 7.30 4.79 9.8 7.31
coordination no. 3 2 3 3
Bond energy(kcal/mol) 44.65 30.9 44.02 24.2
Hs(kcal/mol) 72.07 54.17 49.40 58.23
Electro negativity 1.8 2.4 1.9 1.7
Radius(pm) 145 115 160 155
C.E(eV/atom) 3.13 2.45 2.17 2.51

Table (4-2) Shows values of the optical band gap, density,


coordination number, heat of atomization (Hs), bond energy and
electro negativities of Sn ,Bi, Se and In respectively which are
used for calculations.

In Nco NS
content at.%
0 2.6 3.5
0.1 2.601 3.5025
0.2 2.602 3.5050
0.3 2.603 3.5075

Table (4-3) The average coordination number (Nco) and


the constraints number (Ns) as function of In content of
InxSn20Se (60-x)Bi20 glasses.

4.5 The Effect of In Content on Conductivity.

The dependence of ln σ on the reciprocal


temperature in the range (-110oC to 85oC) is illustrated in
Fig (4-7). All samples exhibit common patterns, where two
regions of conductivity are observed. The activation
energy, pre-exponential factors of the two regions were
estimated and listed in Table (4-4)
Results & Discussions 83

∆Eσ σ0 σ1 σ2
Ratio ln σ
−1 −1 −1 −1
%
(ev ) (Ω Cm ) (Ω Cm ) (Ω −1Cm−1 )

x=0 0.800 -26.3 45×10-2 2.44×10-8 2.4×10-10


x=0.1 0.685 -24.3 42.2×10-2 1.35×10-10 4.16×10-13
x=0.2 0.490 -24.12 29.7×10-2 3.74×10-11 8×10-13
x=0.3 0.395 -22.70 27×10-2 3.3×10-12 9.72×10-19

σ R .T ln σ R.T E1 E2 C.E Tg
(Ω −1Cm −1 ) (ev) (ev) (ev) (K )

2.29×10-11 -24.5 0.86 0.57 2.818 240


1.39×10-11 -24.99 0.23 0.06 2.8192 253
1.14×10-11 -25.19 0.22 0.09 2.8199 335
1.7×10-10 -24.79 0.14 0.03 2.820 330

Table (4-4) Compositional dependence of the electrical


characteristic quantities for the thin film glasses in the
system InxSn20Se(60-x)Bi20 where x= 0 , 0.1 , 0.2 and 0.3 at
constant thickness 100 nm.
Results & Discussions 84

-1 6

-1 8

-2 0 X = 0 .0 %
x = 0 .1 %
-2 2 x = 0 .2 %
-2 4
x = 0 .3 %

-2 6
Ln σ

-2 8

-3 0

-3 2

-3 4

-3 6
2 .0 2 .5 3 .0 3 .5 4 .0 4 .5 5 .0 5 .5 6 .0 6 .5 7 .0 7 .5 8 .0
1 0 0 0 /T K

Figure (4-7) Variation of ln (σ) vs reciprocal absolute


temperature for films of Inx Sn20 Se(60-x)Bi20 where x= 0 ,
0.1 , 0.2 and 0.3 %at constant thickness 100nm.

The values obtained for the first region σ0×10-2 suggest


that the conduction mechanism in the high temperature
region is band like conduction through extended states [95].
It is also observed that σ decreases with the decrease of
temperature which gives an indication for the increase of
density of localized states. Similar behavior has also been
observed for Sn-Se-Bi-In composition by Majid [96].

Figure (4-8) illustrates the effect of In content on the value


of σ .It is clear that σ has a linear inverse relation with In
content. This indicates that addition of In leads to an
increase of both the density of localized states. To check
the validity of compensation law, the pre-exponential factor
σ against the activation energy ∆E is shown in Figure (4-9).
.
Results & Discussions 85

2 .5

2 .0

1 .5
-2
σ (Ω cm )*10
-1

1 .0
−1 -

0 .5

0 .0
0 .0 0 0 .0 5 0 .1 0 0 .1 5 0 .2 0 0 .2 5 0 .3 0
In C ontent %

Figure (4-8) Variation of σ as a function of In content.

0 .5

0 .4
-2

0 .3
σ (Ω cm )*10
-1
−1 -

0 .2

0 .1

0 .0
0.0 0.5 1.0 1.5 2.0

∆ Ε ( ev)

Figure (4-9) Variation of σ as a function of activation


energy.

The observed linear dependence indicates that Sn-


Se-Bi-In obeys the compensation law [97-98]. Such
behavior was also observed by Majid [96]. However, it is
worth mentioning that Mott and Davis argued that there is
Results & Discussions 86

no definite correlation between the intercept σ and the


activation energy ∆E as shown in Fig (4-9).

The values of ∆E decrease linearly by increasing In


content as shown in Fig (4-10). This effect is most likely
due to the reduction of the average binding energy by In
addition. To check this argument, the cohesive energy C.E.
was calculated for the investigated four compositions,
Table 4.4. It is clear that C.E. increases by decreasing In
content which is in good agreement with the above
assumptions.

1.0

2.8200

0.8
∆Ε (eV)
2.8195
C.E (eV)
0.6

C.E (ev)
∆Ε (ev)

2.8190

0.4

2.8185

0.2

2.8180

0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30
In -Content%

Figure (4-10) Variation of activation energy and cohesive


energy as a function of In content.

On the other hand, in the low temperature range of Fig


(4-7), it seems that the dominant mechanism is the hopping
around Fermi level. This assumption is based on the values
of σ2 obtained. The values obtained for the pre-exponential
factor of the second region ensure the fact that such
Results & Discussions 87

systems behave as heavily doped semiconductors. It is


worth mentioning that the variable range hopping is not
favorable in chalcogenide glasses based on ESR and
magnetic susceptibility.

4.6 Switching Characteristics of Amorphous


Semiconductor

It is now clear that the threshold switching which


characterizes thin films of multi component chalcogenide
glasses, first reported by Ovshinsky [99], represents
qualitatively different phenomena with unique properties.
When voltage is first applied to a switching device the
resistance is very high, this is called the off state. If an
applied voltage, VS, known as the threshold voltage, is
exceeded, the device switches into a low resistance state.
The ON state current-voltage characteristic is non ohmic,
the dynamic resistance being of the order of 10 ohms. As
long as a minimum voltage, Vh, known as holding voltage
is maintained through the device, it stays in the on state, if
the voltage falls below Vh, corresponding to the holding
current Ih, the device returns to the off state.

Since the switching process takes place in a very short


time (10-10 sec), it was difficult to record any reading
during this time. We could only record current and voltage
before switching and their values after the switching
process.

Adler [100] has reviewed the experimental results on


threshold switching in thin films of chalcogenide glass.
Peterson and Adler [101] have presented a model for the
recovery of the off state after the holding voltage is
removed from a chalcogenide glass threshold switch. Also,
this model is used to predict quantitatively the resistance of
the device as a function of time for several different values
Results & Discussions 88

of the operating current. Marshall and Owen [102] have


developed a model for carrier transport on the basis of
accumulation-regime field-effect data and transient photo
conductivity measurements for composition InxSn20Se(60-
x)Bi20. Marshall [103] has employed the field effect data to
calculate the screening length of composition InxSn20Se(60-
xBi20. Homma, Henisch and Ovshinsky [104] have made
systematic and detailed comparison between the threshold
switching properties of a typical ovonic threshold material
(InxSn20Se(60-x)Bi20) and a non–chalcogenide alloy (Cd23
Ge12 As65) of nearly equal band gap but higher
conductivity. Their results show great similarities and, in
this way, suggest that no special role is played by lone pair
electrons in the mechanism of threshold switching as such.
On the other hand, the lone pair band in the chalcogenide
alloy may be responsible for its much higher degree of
stability, i.e., its ability to withstand the intense excitation
levels prevailing in the on state. Also, their results provide
further evidence against thermal interpretations and thereby
support electronic models of threshold switching for both
materials.

The specimens employed in the present study were


prepared by evaporation of chalcogenide material onto
substrate of pyrographite. The details of specimen
preparation and the measuring mechanism are represented
in Chapter 3.

Fig (4-11) shows the current-voltage characteristics


for four compositions in the system (Inx Sn20Se(60-x)Bi20 )
where x=0, 0.1, 0.2 and 0.3. From this Figure, it is clear
that composition (Inx Sn20 Se(60-x) Bi20 ) where x=0.3% has
the highest threshold voltage Vs =5.5 V
Results & Discussions 89

Figure (4-11) (I-V) Characteristics of D.C switching for


the films InxSn20 Se(60-x)Bi20 at constant thickness 100nm.

Ratio% Composition Vs(volt) Is Vh(volt) Ih pt(µWatt)


(µΑ) (µA)
0% Sn20Se60Bi20 1.5 1 0.4 45 1.5
0.1% In0.1Sn20Se59.9Bi20 2.5 1.3 1.3 25 3.5
0.2% In0.3Sn20Se59.9Bi20 2.7 1.9 1.5 23 5.13
0.3% In0.2Sn20Se59.8Bi20 5.5 2 4.5 19 11.0

Table 4.5 Switching characteristics of the composition


InxSn20 Se(60-x)Bi20 where x=0, 0.1,0.2 and0.3%at constant
thickness 100nm.

4.6.1 Temperature dependence of Switching


Characteristics

Current-voltage measurements have been carried out


on the two extreme films x=0 and x= 0.3 with a thickness
of 100 nm in a temperature range between 295 and 305 K.
The obtained results for these two compositions are given
in Fig (4-12) and Table (4-6). Since the switching process
Results & Discussions 90

takes place in a very short time (∼ 10-9 sec), it is impossible


to record any reading during this time. We could only
record current and voltage at switching and their respective
values after the switching process had been accomplished.

Figure (4-12) Static I-V Characteristics curves for x=0,


0.3at.% thin film sample at thickness 100 nm at different
ambient temperature T.

Films of the two compositions (x= 0, 0.3) exhibit the


same features. In both films it is observed that Vs
decreases and the switching current increases by increasing
the film temperature. This indicates the thermal nature of
the switching process in the studied compositions of the
chalcogenide glass. In such case one should consider the
concept of filament. The latter is simply the path of the
current inside the films which is considered to be branched
paths. The temperature of such filament path is higher than
that of the film. It was impossible to record any reading
Results & Discussions 91

during this time. We could only record current and voltage


at switching and their respective values after the switching
process had been accomplished.

Temperature inside the filament in the active part of


the switch is calculated at the threshold voltage and after
switching for I-V characteristics of both compositions at
room temperature using the following relation:

Qd
T = TO + (4.4)
2πλ

where
Q = Switching power=ISVS.
Tο = ambient temperature.
λ = thermal conductivity for the substrate, taken to be
0.009 cal/m.sec.°C
d = thin film thickness.

Values of temperature calculated from experimental


static I-V characteristics at room temperature at the
moment of switching and after switching as well as the
corresponding values of current passing through the switch
are given in Table (4-6) for the two compositions x=0, 0.3
of thickness 100 nm. It is found that the addition of In
causes decrease of switching current and switching voltage
with the increase of ambient temperature.
Results & Discussions 92

Comp- Filament Current (µI) Room


osition Tempe-
rature
After
Ratio% At(vs) After(vs) At(vs) switching Co

X=0 331K 420K 11.5 100 298


X=0.3 322K 380K 7.00 70 301

Table 4.6 Values of filament temperature at Vs and after


switching as calculated from experimental (I-V) curves and
the corresponding values of current for samples of the
composition Inx Sn20Se(60-x) Bi20 where x= 0, 0.3 at
thickness 100 nm.

The obtained results can be summarized as follows:


Raising the temperature improves the switching
characteristics as Vs decreases and Is increases. In other
words, raising temperature facilitates switching. On the
other hand, the addition of In reduces the filament
temperature. This in turn reduces the switching ability.
Such effect is consistent with the previously obtained and
above mentioned result.

4.7 Theoretical Study of I-V Characteristics of


Switching

The switching process observed in amorphous


semiconductor may be explained by various models. These
models are categorized into homogenous and heterogenous
models. In the former, the semiconducting film is assumed
to remain essentially homogenous and amorphous during
switching. In the former, the semiconducting film is
assumed to remain essentially homogenous and amorphous
during switching [105]. In the later, a structure change of
the amorphous material takes place in the region of high
Results & Discussions 93

current filament and it does not return to its original


amorphous state if the device is switched off. All evidences
suggest that the forming process creates in the filament
region a new material whose properties have little relation
to the original material [106]. According to the
homogenous model, several trials in the available literature
can explain, the switching process according to thermal
theory [107]. As it was mentioned resistance of
chalcogenide glasses is given by equation (4.5):


R = Const .EXP ( ) (4.5)
2 KT

This equation can be obtained in the following form:

V E 1 1
= R (T ) = R (TO ) EXP ( − ) (4.6)
I 2 K T TO

Where R(TO ) is the resistance at the ambient


temperature TO .

If we take active region of the material under


investigation in the form of a sphere of diameter ''d''
embedded in a spherical infinite medium of thermal
conductivity ''λ''.

The relation between the temperature increase


produced in the active region and the power(IV) generated
within it can be obtained in the form[108]

Qd
T = TO + (4.7)
2πλ
Results & Discussions 94

where

Q =IV, The power generated inside the active region


of the device.
λ =the thermal conductivity of substrate material.
d =is the thickness of thin film sample.

V E 1 1
Ln( ) = LnR(TO ) + σ ( − ) (4.8)
I 2 K T TO

Eσ 1 1
LnV − LnI − LnR (TO ) = ( − ) (4.9)
2 K T TO

1 1
LnV − LnI − LnR (TO ) = B ( − ) (4.10)
T TO

1 1 1
[ LnV − LnI − LnR (TO )] + ( ) = (4.11)
B TO T

where B = Eσ . Temperature inside the active region


2K
equation (4-4) can be obtained in the form:

2πλ d IV (2πλ d + IV )
T =( )To + ( )= = ( MTO + IV ) M (4.12)
2πλ d 2πλ d 2πλ d

Where M=2πλd. By substituting for T in equation (4-11)


we obtain

1 1
[ LnV − LnI − LnR (TO )] + ( ) = ( MTO + IV ) M (4.13)
B TO
Or
1 1
Ψ= [ LnV − LnI − LnR (TO )] + ( ) = ( MTO + IV ) M = 0 (4.14)
B TO
Results & Discussions 95

Using iteration method, a numerical solution for equation


was carried out to obtain values of current corresponding to
values of the applied voltage for all samples prepared from
the composition InxSn20Se(60-x)Bi20 . To solve equation (4-
4), we used experimental "λ'' and the activation energy of
the considered composition. The used for λ the value
0.009cal/cm.sec co.

4.7.1 The Suggested switching model

The obtained results concerning switching properties


allow us to suggest a model, if we consider the atoms in
chalcogenide glass as balls of different sizes, which are
interacted in different ways. In other words, they can be
considered as set of coupled oscillators corresponding to a
set of force constants. One should keep in mind that such
force constants are distributed randomly. In such case one
cannot treat every bond (force constant) separately. Then
a statistical approach by calculating the cohesive energy
C.E.

If C.E. is high, the incoming electric pulse will not


affect the atomic position .In such case the incoming pulse
will affect electrons. The latter is characterized by fast
(short life time) transitions giving rise to switching
phenomena. On the other hand, in case of low C.E., the
incoming pulse leads to atomic displacement to another
metastable position giving rise to memory effect.

Such suggestions agree with the experimentally


observed data where the switching rise time increases by
decreasing C.E .i.e., approach memory behavior are shown
in Figure (4-13).
.
Results & Discussions 96

Table (4-7) illustrates the relation between cohesive energy


C.E and rise time tr For samples of the composition Inx
Sn20Se(60-x)Bi20 at constant thickness 100nm.

InxSn20Se(60-x)Bi20 Cohesive energy (eV) Rise time tr


X=0% 2.560 25
X=0.1% 2.369 40
X=0.2% 2.250 100
X=0.3% 2.134 200

Table 4.7 Values of rise time and cohesive energy


for samples of the composition InxSn20Se(60-x)Bi20 (where
x=0,0.1,0.2&0.3at.%)

220

200

180

160
Rise time (ns)

140

120

100

80

60

40

20

2.1 2.2 2.3 2.4 2.5 2.6


Cohesive energy (e.v)

Figure (4-13) Relation between rise time and cohesive


energy.
Results & Discussions 97

4.8 Optical Band Gap of In-Se-Sn-Bi Thin Films

Transmittance spectra corresponding to the system


Inx Sn20 Se(60-x)Bi20 where x=0, 0.1 ,0.2 , and 0.3 at % thin
film are plotted in Fig (4-14), showing a clear red shift of
the interference-free region with increasing In content.

100

90
X=0.0%
80
X=0.1%
70 X=0.2%
X=0.3%
Transmittance

60

50

40

30

20

10

0
400 600 800 1000

Wavelength(nm)

Figure (4-14) Transmission spectra of InxSn2Se(6-X)Bi2 thin


film (where x=0 ,0.1 ,0.2 and0.3 at%).

Values of the absorption coefficient(α)for the studied films


were calculated from the transmittance T and reflectance R
according to the equation(4-15).

1 2T
α= (4.15)
d ln[(1 − R ) + (1 − R ) 4 + 4 R 2T 2 ]
2

Where
Results & Discussions 98

d is the film thickness. Figure(4-14) shows that the


calculated valueof the absorption coefficient (α) for the
investigated InxSn20Se(60-x)Bi20 (Where x=0,0.1,0.2 and
0.3%) thin films.

0.006

0.005
x=0.0 %
x=0.1%
x=0.2%
0.004
x=0.3%

0.003
α( cm )
-1

0.002

0.001

0.000

0.5 1.0 1.5 2.0 2.5 3.0


hυ(ev)

Figure 4.15 Absorption coefficient (α) versus photon


energy (hυ) for the Inx Sn20 Se(60-x)Bi20thin films.

Analysis of the optical absorption data indicates that


the optical band gap (E0) of the InxSn20Se(60-x)Bi20 thin films
obeys Tauc’s relation for the allowed non-direct transitions.

(α hυ )1 2 = B(hυ − E0 ) (4.16)

where B is a parameter that depend on the transition


probability and E0 is the optical energy gap.
Results & Discussions 99

0.14

0.12 x=0.0%
x=0.1 %
2
/
1
0.10 x=0.2 %
v
e x=0.3 %
))))

1
-
m 0.08
c
2
/
(((( α )))) ((((

1
0.06
u
h
0.04

0.02

0.00
1 2 3

hυ(ev)

Figure (4-16) Shows a typical best fit of (αhυ)1/2Vs.photon


energy(hυ)for the InxSn20Se(60-x)Bi20thin film.

The intercepts of the straight lines with the photon energy


axis at (αhυ) 1/2=0 give the values of the optical band gap
(E0).

The variation of Eg as function of In content is depicted in


Figure (4-17).From this figure, it is clear that Eg decreases
with increasing In content of the investigated glasses.
Results & Discussions 100

7
Eg,TH
6 Eg,EXP

5
Eg(ev)

0
0.00 0.05 0.10 0.15 0.20 0.25 0.30
In content%

Figure (4-17) The variation in the optical band gap (E0 ) as


function of In content of the InxSn20Se(60-x)Bi20 thin films.

The bond energies D(A-B) for heteronuclear bonds


have been calculated by using the empirical relation:

D( A − B) = D( A − A).D( B − B)1 2 + 30(γ A − γ B )2 (4.17)

Where D(A-A) and D(B-B) are the energies of the


homonuclear bonds (in units of Kcal/mol); χA and χB are
the electro negativities for the involved atoms.
Results & Discussions 101

Bond Bond energy Probability Relative probability


(at T=298.15)
Sn-Se 55.13 3.27*1040 1
In-Se 47.33 6.06*1034 1.85*10-6
Sn-Sn 44.65 6.47*1032 1.97*10-8
Bi-Se 44.38 4.12*1032 1.25*10-8
Se-Se 44.02 2.24*1032 6.85*10-9
Bi-Sn 37.44 3.25*1027 9.9*10-17
Sn-In 33.17 2.38*1024 6.97*10-17
Bi-Bi 30.9 5.11*1022 1.56*10-18
Bi-In 28.54 9.37*1020 2.86*10-20
In-In 24.2 6.08*1017 1.85*10-23

Table 4.8 Bond energy probabilities and relative


probabilities of formation of various bonds in Inx Sn20 Se
(60-x) Bi20 Films, taking the probability of Sn-Se bond as
unity.

Knowing the bond energies, we can estimate the


cohesive energy (CE) i.e., the stabilization energy of an
infinitely large cluster of the materials per atom, by
summing the bond energies over all the bonds expected in
the system under test . The (CE) of the prepared samples is
evaluated from the following equation:

CE = ∑ (Ci Di /100) (4.18)

Where Ci and Di are the number of the expected


chemical bonds and the energy of each corresponding
bond. The calculated values of CE for all compositions are
presented in Table 4.4 with the exception of InxSn20Se(60-
x)Bi20 glass, CE increases with increasing In content.
Increasing the In content leads to an increase in the average
molecular weight.
Results & Discussions 102

4.9 Thermal and Radiation–induced defects in thin film


devices

Turnbull [110] and Rudee [111] had pointed out that


the mode of the amorphous to crystalline transformation
yields information about the amorphous phase structure.
Two techniques are used to study the phase transition in
chalcogenide glasses, the isothermal technique and non
isothermal technique. The methods of differential thermal
analysis (DTA) has been universally accepted by
mineralogical laboratories as a rapid and convenient means
for recording the thermal effects that occur as a sample is
heated.

4.10 Effect of Gamma-Irradiation on Optical Band


Gap

The γ- radiation had no a noticeable effect, for a dose


of 15 µRad showed an constant value in the transmittance
upon the addition of In.

4.10.1 Effect of Gamma-Irradiation on Threshold


Switches

Threshold switches devices were exposed to


different doses up to 15 Mrad for glasses x=0 and x=0.3 at
%. Ionization radiation did not cause any significant effects
for I-V characteristics at room temperature. This work
agrees with Nicolaedes [112] which had measured
switching and memory characteristics of doped and
undoped devices, prepared by flash evaporation onto a
glass substrate of a mixture premelted Sn20 Se(60-x)Bi20
subjected to neutron irradiation up to1015 m/cm2(>10KeV).
Results & Discussions 103

They had found that no changes in the electrical


characteristics of the threshold switching devices were
noted after neutron irradiation.

4.11 Modern Applications of InxSn20Se60-xBi20

Since amorphous chalcogenide s.c. are characterized


by their favourable electrical properties, low room
temperature conductivities and suitable optical properties,
the investigated system InxSn20Se60-xBi20 is advantageous in
application to modern devices.

The compound InxSn20Se60-xBi20 has a suitable


optical band gap that lies between 0.7 – 1.7 eV and this
range corresponds to the visible optical range. The optical
band gap is either low or high. When it is low, the
composition is used as an absorption layer in solar cell
construction and photocells. If it is high, the composition
is used as a window in the IR range, therefore, it is used as
a window in IR detection devices.

Moreover, the investigated composition has proved


to have a high absorption coefficient 104 cm-1 which
enables it to absorb the highest amount of light falling on it.
It is also characterized by its low thermal expansion
coefficient with thermal stability, thus it neither expands
nor shrinks which makes it convenient for application in
solar cells that require high thermal stability.

In the investigated composition, the increase of In


content caused an increase in the switching rise time that
reached 200 nsec and a decrease in the cohesive energy,
thus improving the switching characteristics of the system.
The composition InxSn20Se60-xBi20 showed good electrical
threshold switching characteristics and the power threshold
Results & Discussions 104

switching did not exceed 11 watt, thus encouraging its


application as a threshold switching device in computer
memory arrays, display devices, optical mass memories,
thin film transistors, solar cells, and electrophotographic
units.
Such results satisfy the need for cheap and reliable
materials for different technological applications which
scientists seek to apply in rectifiers, photocells, switching
and memory devices, detectors and sensors, optical fibres,
solar cells, and nanotechnology.

To meet the challenge of the need for chemical


information in the fields of environmental protection,
medical diagnosis as well as chemical process control, food
inspection,…etc., a great variety of chemical microsensors
are currently being developed. The composition
InxSn20Se60-xBi20 here with under study can be applied in
the production of different types of sensors for
environmental applications.

Figure (4-18) Array of Sensors of the composition


InxSn20Se60-xBi20
Results & Discussions 105

Chalcogenide semiconductors have been advanced


recently to the point where they compete favorably with
silicon-based devices which are impractically expensive for
solar power conversion. Chalcogenide semiconductors for
solar cells are made by anodizing Bi in a S, Se, Te
electrolyte. They are relatively inexpensive.

Figure (4-19) Solar Cell Array

It has also been found that the composition


InxSn20Se60-xBi20 was almost stable against gamma
radiation with doses up to 15 MRad, which makes it useful
as a shield in the components of devices used in gamma
radiation.

4.12 Trends for Future Work

The present work can be extended to implement


amorphous chalcogenide semiconductors by different gases
with the same the same components i.e., Sn, Se, Bi and In.
Results & Discussions 106

 Such compositions can find some applications as


laser power transmission for industrial welding
operations and also for microsurgery.

 Co2 surgical laser operating at 10.6µm have relied on


bulky articulated arms to deliver the beam to a
micromanipulator and on to the site for surgery.

 Chalcogenide fibers such as Sn-Se-Bi may replace


them as a compact fibre optic competitor.

 Produce some economical chalcogenide composition


for solar cells.

 Production of wide IR windows of chalcogenide with


high resolution of Fourier transform infrared
spectrometry as a remote sensing of gases/liquids is
another.
Summary & conclusion

Chapter 5

Summary & Conclusion

The concept of point defects in amorphous materials is


now well established and they are known to influence the
electrical and optical properties in much the same way as
they do in crystalline semiconductors. Also, the states
deeper in the energy gap can be classified as extrinsic, in
the sense that they arise from point defects whose density is
a strong function of preparation conditions or post-
deposition treatments.

Hence, we have made examination of the structure,


micro structure and related properties of their amorphous
film devices of the compositions under examination.

Four glasses of the system Inx Sn20 Se(60-x) Bi20 where


x=0,0.1,0.2& 0.3 had been prepared from highly pure
Indium, Tin, Selenium and Bithmus (99.9999%purity)
elements by melting the constituents together under
vacuum(10-6torr) in precleaned silica tubes at900C for
about 10 hours and subsequently quenching in liquid
nitrogen. The product ingots were confirmed to be
amorphous and homogenous by X-ray diffraction pattern
(Co-Kα source) , and by differential thermal analysis
(DTA).

The main results can be summarized as follows:

1. The thin film Samples proved to have an amorphous


structure.

90
Summary & conclusion

2. The density decreasing by increasing In content from


5.712 gm/cm3 for the composition at x=0% down to
4.682 gm/cm3 for composition InxSn20Se(60-x)Bi20 at
x=0.03%but does not lead to any change in the
competence of atom.

3. It was found that the electrical conductivity σo


decreases with the decreases of temperature which is
an indication for the increase of density of localized
states. it was also observed that there is an inverse
linear relation between the conductivity and In
content.

4. The electrical activation energy(∆E) was found to


decrease linearly from1.60 ev to 0.79 ev by
increasing the In content from x=0 in the
composition InxSn20Se(60-x)Bi20.On the other hand,
addition of In has almost no effect on Se bonds .It is
worth noting that the cohesive energy increases
linearly with increasing the In content from2.8186 at
x=0% to 2.82 at x=0.3% at const thickness100nm.

5. The addition of In has led to an increase in both the


threshold voltage (Vs) and threshold current (Is) from
1.5 volt and 1.0 µA respectively at x=0 up to 5.5 volt
and 2 µA respectively at x=0.3 for constant film
thickness d=100nm.

6. As for the holding voltage (Vh), it was found to


increase with the increase of either In content from 0.4
volt at x=0 to 2.0 volt at x=0.3.On the contrary, the
increase of In content has caused a decrease in the
holding current (Ih) from 45µA at x=0 to19 µA at x=0.3
for a constant thickness 100nm.

91
Summary & conclusion

7. It was proved that the threshold power increased by


increasing either In content, This mean that the quality
of switching is reduced by increasing the In content.
The addition of In content increase the cohesive energy
and consequently affects the switching properties.

8. Raising the film temperature improved the switching


characteristics where the threshold voltage decreased
and the threshold current increases. Also, the addition of
In reduced the filament temperature, thus reducing the
switching ability.

9. Increasing the In content from x=0 to 0.1, 0.2 and 0.3


led to an increase in the switching rise time from tr =25
to 40, 100 and 200 and 200 nano second respectively
and a decrease in the cohesive energy from C.E=2.8186
to 2.8192 ,2.8199 and 2.820ev respectively. These
results indicate that composition Inx Sn20 Se(60-x) Bi20
shows good electrical threshold switching results and
promises a useful threshold switching device in
computer applications.

10. In the same manner, the X-ray diffraction technique


proved the absence of any regular crystalline structure
which ensured the composition Inx Sn20Se(60-x)Bi20 to be
pure amorphous.

11. The differential thermal analysis (DTA)


Measurements showed that the glass transition
temperature Tg increased with increasing In content. The
absence of any sharp exothermic peak proved the
absence of structural changes.

92
Summary & conclusion

12. The values of the optical energy gap Eopt were found
to decrease with increasing In content which could be
due to the fact that In has a metallic behavior.

13. The γ- radiation had a noticeable effect, for a dose of


15 µRad showed an constant value in the transmittance
upon the addition of In.

14. The investigated system proved to be advantageous


in application to infrared imaging objectives.

93
References

References

[1]J.A. Savage., “IR Optical Materials and their


Antireflection Coating”, Adam Higher, Bristol (1985) 79.

[2]A.R. Hilton, C.E. Jones and M. Brain, Physics and


chemistry of Glasses, 7 (1966) 1051.

[3]R.B. Johnson, Proc. Soc. Photo-optical Instrumen-tation


Engineers, SPIE 915 (1988) 106.

[4]K. Nandakuma and J. Philip, J. Non- Cryst. Solids, 144


(1992) 247.

[5]T. Katsuyama and Matsumura, Infrared Optical Fibers,


Adam Higher, Bristol (1989).

[6]J.M. Parker and A.B. Seddon,High Performance Glasses,


ed. M. Cable and J. M. Parker, Blackie, London (1992)
Ch.13.

[7]A.B. Seddon, “Chalcogenide Glasses: A Review of Their


Preparation , Properties and Applications” , Journal of Non-
Crystalline Solids184 (1995) 44-55.

[8]S.A. Fayek and S.ery. Fouad, Vacuum 52 (1999)359.

[9]A. Zakery, S.R. Elliott, Journal of Non-Crystalline Solids,


330 (2003) 1.

[10]S.A. Fayek and Maged A.F., “Optical Absorption of


Glasses in the system (As2Se3)1-XTIX”, Materials Science &
Engineering B 32 (1995) 47-49. U.K.

90
References

[11]H. Fritzsche and S. R. Ovshinsky, J.Non-Crystalline


Solids, Vol.2, PP. 393 (1970).

[12]A.A. AL-Ghamdi, “Optical Band and Optical Constants


in amorphous Se6-xTe4 Agx Thin Films”, Surface Engineering,
Surface Instrumentation & Vacuum 80 (2006) 400-405.

[13]Miloslav Frumer, Tomas Wagner, “Ag-doped


Chalcogenide Glasses and their application”, Solid State
and Materials Science 7 (2003) 117-126, Elsevier, Holland.

[14]A.R. Ilibton, D. Hayes, M. Rechtin, J. Noncrystalline


Solids 17 (1965) 319.

[15]E. Marquez, P.Villars, R. G. Jimenez, J. Mater. Res. 3


(1988) 314.

[16]J.H. Dessaur, H. E. Clarke, Xerography and Related


Processes Local, London, 1965.

[17]R. Ovshinsky, phys Rev Lett. 21 (1986) 1450.

[18]D.E. Carlson, C. R. Wronski, Appl. phys. Lett. 28 (1976)


671.

[19]J. Fusong, M. Okuda, J. Appl. phys. 30 (1991) 97.

[20]A.B. Seddon, J. Non-Cryst. Solids 184 (1995) 44.

[21]A.B. Seddon, M.A. Hemingway, J. Therm. Anal. 37


(1991) 2189.

[22]H. ELZahed, A. EL-Korashy and M .Dongol, thin solid


films, 259 (1995) 203.

[23]M.A. Abdel–Rahim, J. Noncrystal. Solid 241 (1998) 12.

91
References

[24]S.P. Singh, A. kumar, S.kumar, materials letter, 60


(2006) 1640.

[25]K.A. Aly and A.Dahshan, Philosophical magazine, vol


88, No. 3 (2008) 361-372.

[26]A.A. Moharram, A.A. Othman, H.H. Amer and A.


Dahshan, Thin Solid Films, 457, 235 (2004).

[27]p. Sharma, M. vashistha and I.P. Jain, Journal of


optoelectronics and advanced materials Vol. 7, No. 5 (2005),
p. 2647-2654.

[28]A.F. Maged, G.A.M. Amin, M. Semary and E. Borham,


Journal of Non-Oxide Glasses Volume 1 Number 1, 2009, p.
53-60.

[29]Adel Shaheen, Mousa M.A. Imran, Omar A. Lafi, Moh'd


I. A wadallah, Ma'rouf K. Abdullah, Radiation physics and
chemistry, volume 79, Issue 9,2010, page 923-928.

[30]K. Shanmugavelayutham Sundaram, John S. McCloy,


Brian J. Riley, Mark K. Murphy, Hong A. Qiao, Charles F.
Windisch Jr., Eric D. Walter, Jarrod V. Crum, Roman
Golovchak,, Oleh Shpotyuk, journal of the American society,
2011.

[31]A.F. Maged, K.L. Montasser and H. H.Amer, Materials


Chemistry and physics 56 (1998) 184-188.

[32]M.M. EL-Ocker, M.H. EL-Fouly, S.A. Fayekand G.A.M.


Amin, APPl. Phys. a.60, 233-238 (1995).

[33]Muneer Ahmad, J. Kumar, R. Thangaraj, Journal of


Non-Crystalline Solids 355 (2009) 2345.

92
References

[34]Muneer Ahmad, R.Thangaraj Physica B 404 (2009)


1553–1557.

[35]M. Fadel Vacuum 44 (1993) 851.

[36]M.F. Kotkata, M.A. AFiFi, H.H. Labib, Hegab, M.M.


Abel-Aziz Thin Solid Films 240 (1994) 143.

[37]D. David Thonburg, Journal of Non-Crystalline Solids


11 (1972) 113.

[38]S.R. Ovshinsky., H. Fritzsche, Electron Devices, IEEE


Transactions 20 (1973) 91.

[39]R. Rajesh, J. Philip, Semicond. Sci, Technol. 18 (2003)


133.

[40]H.J. Stocker., (1969) 55.

[41]A.B. Seddon, “Chalcogenide Glasses: A Review of Their


Preparation , properties and Applications” , Journal of Non-
Crystalline Solids184 (1995) 44-50.

[42]A.E. Owen, Firth A.P. and Ewen P.J.S., Phil. Mag. B 52


(1985) 347.

[43]J.A. Savage., ‘‘IR Optical Materials and their


Antireflection Coating”, Adam Higher, Bristol (1985) 79.

[44]J.A. Savage et al, Infared Phys. 20, 313 (1980).

[45]H. Oyanagi, Tanako K., Hosoya S.&Minomura S., J. de


phys. 42, 221 (1981).

[46]D.E. Sayers, Amorphous and liquid semiconductors,


University of Edinburgh (1977).

93
References

[47]H. Maeda,Terauchi H.,Tanabe. K. & Kamijo. N., Jpn J.


Appl. Phys 21, 134 (1982).

[48]R. Parthasarath, RaoK. J. & C. R., J. Phys. Solid State


phys.15, 3646 (1982).

[49]S.R. Elliott Physics of Amorphous Materials, 2nd Ed.


(Longman, Harlow, U.K.) 1990.

[50] K. Homma, Henisch H.K. and Ovshinsky S.R.,J.Non-


cryst. Solid. 35, 36 (1980) 1105-1110.

[51]A. Vohra and K.K. Srivastava: Microwave Switching and


A.C. Conductivity, Phys. Stat. Sol. (a) 121, 341 (1990).

[52]R.B. Johnson in: Proc. Soc. Photo-Optical Instrumen-


tation Engineers, SPIE 915 (1988)106.

[53]F. Demichelis, Kaniadakis G., Tagliaferro A. and


Tresso E., “New Approach to optical Analysis of Absorbing
Thin Solid Films”, Applied Optics, Vol 26 No 9, pp. 1737
(1987).

[54]S.A. Fayek and A.F. Maged, “Optical Absorption of


glasses in the system (As2Se3)1-xTIx)”, Materials Science &
Engineering B32 (1995) 47-49. U.K.

[55]N.F. Mott Sir Neivl, Electronic processes In Non-


Crystalline Materials (1979).

[56]N.F. Mott., Phil. Mag. 19 (1969) 835.

[57]N.F. Mott and Davis E.A., “Electronic in Non-crystalline


Materials”, 2nd edition, Oxford University press (1979).

[58]J.T. Edmond, Br. J. Appl. Phys (1966) 17,979.

94
References

[59]E.A. Davis, J. of Non Cryst.sol. (1985) 71,113.

[60]M. Kastner, Phys. Rev. Lett (1972), 355.

[61]M.H. Cohen, Fritzche H. and Ovshinsky S.R. (1969)


Phys. Rev. Lett. 22, 1065.

[62]E.A. Davis and Mott N.F., Phil. Mag, 22, 903, (1970).

[63]P.W. Anderson, Phys. Rev. Lett. 1975. 34, 953.

[64]R.J Elliott and Gibson A.F. “ An Introduction to Solid


state Physics and Applications”, 1982, El. B. Edition,
McMillan.

[65]S.R. Ovshinsky, Phys. Rev. Lett. 21 (1968) 1450.

[66] G. Busch, Kunzi H.U. and Schweigher A., Phys. Lett.


A33 (1970), 64.

[67]K.E. Peterson and Adler D.J., J. Appl. Phys. 47 (1976)


256.

[68]J.K. Higgins, Temple B.K. and Lewis J.E., J. Non-Cryst.


Sol. 23 (1977) 187.

[69]D. Adler Henisch H.K. and Mott N.F., Rev. Mod. Phys.
50 (1978) 209.

[70]M. Haissinsky, “Nuclear Chemistry and its


Applications”, (Transl. D.G. Tuck), Addison-Wesley
Reading, MA, (1964).

[71]“Radiation Effect in InOrganic Solids”, Faraday Soc.


Discuss. No. 31 (1961).

[72]G. Dearnoley and Northrop D.C., “Semiconductor


Counters for Nuclear Radiation”, Spon, London (1963).
95
References

[73]P.G. Salman and Lallshorth F., “Silicon Surface Barrier


Radiation Detectors: Some Designs and Applications”,
AERE-R-4122 (1963).

[74]L.S. Smirnov, “A Survey of Semiconductor Techniques”,


Moscow (1983).

[75]W.I. Adel Fattah, Alousi D.S. and Fadel M.A. Intern.


Cermic and Glass Magazine, Sprechseal 10(1980).

[76]M.A. Fadel, Abdel Fattah W.I., Abdula A.A. and Kadum


A.A., Radiation Research (1982) in press.

[77]A.M. Nassar and Fadel M.A., J. Phys., D: Appl. Phys. 16


(1983) 431.

[78]R.D. Evans, “The Atomic Nucleus, McGraw-Hill”, New


York (1955).

[79]R.V. Konakova et.al. “Effect of Gamma Radiation on the


Electrical Characteristics of Diode Structure Based on Ga
As &AlxGa1-xAs”, Sov. Phys. Tech. Phys., 26 (1981) 4.

[80]R.V. Konakova et. al., Radiation Effects, 60 (1982) 143.

[81]A.F. Sanaa, “Study of Effects of Gamma Radiation on


Some Physical Properties of Semiconductor Chalcogenide
Glasses”, M. Sc., NCRRT, Egypt (1981).

[82]V. Stupleman and Filaretove: “Semiconductor Devices”,


Mir Publishers, Moscow (1981).

[83]M.S.I. Rageh, El-Behay A.Z. and Soliman F.A.S., J. Rad.


Res. Sci. and APPl. 1 (1983) 1.

96
References

[84]F.A. Soliman, “Radiation Effects on Some Materials


and Devices”, Ph. D., Egypt (1986).

[85]M.A. Morsy, “Studiy of the Radiation and thermal-


induced Changes in some Glasses of The System As4SexTe6-x
Ph. D. NCRRT, Egypt (1984).

[86]K.L. Chopra, “Thin Film Phenomena”, Mc Graw Hill


New York, (1969) P102.

[87]A.A. EL.Shazly, H.T. EL Shair and M.M .EL. Nahass,


1nd. j. Opt. Soc. 12 (1983) 6.

[88]W.J. Smothers and Y. Chiang, “Hand book of


Differential Thermal Analysis”, (1966).

[89]V.G. Hill and R.J. Roy, Cermaic Soc., (1958), 41, 532.

[90]R.C. Machanzi, “Differential Thermal Analysis”, (1970),


Academic Press.

[91]Hilton, Joes, Brau (1964, 1966a, 1966b).

[92]A.F. Loffeand A.R. Regel. prog. Semicond. 4239 (1960).

[93]J.C. Phillps and M.F. Thrope, Solid State Commun.


53699 (1985).

[94]N.F. Mott. and Davis E.A., Electronic Process in Non


crystalline Materials(1978).

[95]C.A. Majid, Philosphical Magazine B, Vol. 49 (1984).

[96]H. Meir, “Organic Semiconductors”, Verlag Chemi


Ganbh, Weinheim (1974).

[97] K. Miyair,OhtaYandLedaM, J. Phys. D. Appl. Phys. 21


(1988) 1519.
97
References

[98]S.R. Ovshinsky, Phys. Rev. Lett. 21(1968) 1450.

[99]D. Adler, Amorphous and Liquid Sem. Con Edingburgh


(1977) 695.

[100]K.E. Peterson and D. Adler, Amorphous Liquid Sem.


Edin burgh (1977) 707.

[101]J.M. Marshalland A.E. Owen, Phil. Mag. 33 (1976)


457.

[102]J.M. Marshall, Phil. Mag. B. Vol. 38 (1978) 4.

[103]K. Homma, H.K. Henishand, S.R. Ovshinsky, J. Non.


Sol35 and 36 (1980) 1105-1110.

[104]D.F. Weirauch, “The Threshold Switching and Thermal


Filaments in Amorphous Semiconductors”, Applied Physics
Letters”, Vol. 16 No. 2, p 72 (1970).

[105]J.R. Bosnell and C.B. Thomas, “PreSwitching


Electrical Properties Forming and Switching in Amorphous
Chalcogenide Alloy Threshold and Memory Devices”, Solid
State Electronics, Vol. 15, P 1261 (1972).

[106]M. Soji and K.C. Kao, “Some Features Relevent to


Switching Process in Amorphous Semiconductors”, J. Non
Crystalline Solids (1975).

[107]F.M. Collins, “Switching by Thermal Avalanche in


Semiconducting Glass Films”, J. Non-Crytalline Solids, Vol.
2 P 496-503(1970).

[108]J.A. Savage et al, Infrared phys. 20, 313 (1980).

[109]D. Turnbull, Phys. Non-Cryst. Sol. (ED. By J. W. Prins)


p. 41 (1965).

98
References

[110]M.L. Rudeeand A. Howie. phil. Mag. 25, 1001 (1972).

[111]R.V. Nicholaides and L.W. Doremus, J.Non-Cryst. Sol.


8-12 (1972) 857.

99
Effect of In addition on some physical properties of Sn20Se(60-x) Bi20 amor-
phous films
H.H.Amer1,A. Abdel-Mongy2,A.A.Abdel-wahab1

1
Solid State Department ,National Center For Radiation Research and Technology,
Nasr City, Cairo ,Egypt.
2
Physics Department, Faculty of Science, Helwan University ,Ain Helwan ,Cairo,
Egypt.
* Corresponding author e-mail: [email protected]

ABSTRACT
The present paper reports the effect of replacement of selenium by indi-
um on the optical gap and some other physical parameters of new quaternary
chalcogenide InxSn20Se(60-x)Bi20 (x = 0, 0.1, 0.2 and 0.3 at. %) thin films. Thin films
with thickness 100 nm of InxSn20Se(60-x)Bi20 were prepared by thermal evaporation
of the bulk samples. Increasing indium content is found to affect the average heat
of atomization, the average coordination number, the number of constraints and the
cohesive energy of the InxSn20Se(60-x)Bi20 alloys. Optical absorption measurements
showed that the fundamental absorption edge is a function of composition. The op-
tical absorption is due to allowed non-direct transition and the energy gap decreas-
es with the increase of indium content. The chemical bond approach has been ap-
plied successfully to interpret the decrease of the optical gap with increasing indi-
um content. The prepared films were irradiated by gamma rays at doses up to
15Mrad. It was found that the compositions were almost stable against gamma ra-
diation.
Keywords: amorphous, chalcogenide, optical properties.

1
‫‪Summery in Arabic‬‬

‫الملخص العربي‬

‫لقد كان موضوع اشباه الموصالت االمورفية ذو اھمية‬


‫كبيرة في السنوات القليلة الماضيةويعتبر مجال حيوي في علم‬
‫الجوامد‪.‬وفي العشرة سنوات الماضية تم التركيز علي اشباة‬
‫الموصالت االمورفية وبخاصة تلك المعروفة بزجاجيات‬
‫الشالكوجنية والتي يتم فحص تركيبھا باستخدام جھاز حيود االشعة‬
‫السينية ويتم تأكيد طبيعتھا االمورفية بأستخدم جھاز حيود التحليل‬
‫الحراري التفاضلي ھذه المركبات تتميز بحساسيتھا للضوء وذلك‬
‫يؤدي الي تغيرات ضوئيةوتركيبية ‪.‬ان دراسة العوامل الضوئية‬
‫مثل معامل االمتصاص تعطي معلومات عن تركيب الحزمة وفجوة‬
‫الطاقة فيالمادة‪.‬وتعتبر ظاھرة القطع) فتح وقفل( والتي تسمي‬
‫بظاھرة )‪ (Switching‬ملحوظة في اشباة الموصالت االمورفية‬
‫وان االھمية التكنولوجية لزجاجيات اشباة الموصالت الشالكوجنية‬
‫ليست فقط نتيجة تطبيقاتھا التكنولوجية القيمة في االجھزة‬
‫الحديثةولكن ايضا ھذة االھمية ترجع الي ثمنھا بالمقارنةبأ شباة‬
‫الموصالت االخري‪.‬ان تطبيقتھا في التكنولوجيا الحديثة تتضمن‬
‫استخدام الطاقة واكتشاف الخطأ الحراري ومراقبة الحرارة‬
‫والرؤية الليلية‪.‬ويتم اختيارھا لتطبيقات الكمبيوتر و الذاكرة والقطع‬
‫بفضل خواص القطع الجيد لھا وايضا يتم تطبيقتھا في الترانزستور‬
‫وفي وحدات الكھروغرافية ‪.‬‬

‫ان اح‪WW‬د التطبيق‪WW‬ات الش‪WW‬يقة الش‪WW‬باة الموص‪WW‬الت االمورفي‪WW‬ة ھ‪WW‬و ف‪WW‬ي‬


‫تص‪WW‬نيع الحساس‪WW‬ات للتحل‪WW‬يالت الطبي‪WW‬ة وحماي‪WW‬ة البيئ‪WW‬ة ‪.‬وايض‪WW‬ا ي‪WW‬تم‬
‫تص‪WWWWW‬نيع االلي‪WWWWW‬اف الض‪WWWWW‬وئية م‪WWWWW‬ن اش‪WWWWW‬باة الموص‪WWWWW‬الت االمورفي‪WWWWW‬ة‬
‫الش‪WW‬الكوجنية وتل‪WW‬ك االلي‪WW‬اف الض‪WW‬وئية يش‪WW‬يع اس‪WW‬تخدامھا ف‪WW‬ي انظم‪WW‬ة‬
‫االتصاالت الالسلكية واالضاءة والتصوير الضوئي‪.‬لقد اصبح تركي‪W‬ز‬
‫الباحثين علي االلياف الضوئية بسيب اس‪W‬تخدامھا المفي‪W‬د ف‪W‬ي اجھ‪W‬زة‬
‫القط‪WW‬ع فائق‪WW‬ة الس‪WW‬رعة واالغ‪WW‬راض الجراحي‪WW‬ة ‪.‬ان آش‪WW‬باة الموص‪WW‬الت‬
‫االمورفي‪WW‬ة تتن‪WW‬افس بتف‪WW‬وق عل‪WW‬ي االجھزةالمص‪WW‬نوعة م‪WW‬ن الس‪WW‬يلكون‬
‫‪Summery in Arabic‬‬

‫لتحوي‪WWW‬ل الطاق‪WWW‬ة الشمس‪WWW‬ية وذل‪WWW‬ك الن اش‪WWW‬باه الموص‪WWW‬الت االمورفي‪WWW‬ة‬


‫ارخ‪WW‬ص ثمن‪WW‬ا وب‪WW‬ذلك ش‪WW‬اع اس‪WW‬تخدامھم ح‪WW‬ديثا ف‪WW‬ي تص‪WW‬نيع الخلي‪WW‬ة‬
‫الشمس‪WW‬ية‪.‬بم‪WW‬ا ان اش‪WW‬باة الموص‪WW‬الت االمورفي‪WW‬ة تتمي‪WW‬ز بحساس‪WW‬يتھا‬
‫للعوام‪WW‬ل الخارجي‪WW‬ة وبخاص‪WW‬ة اآلش‪WW‬عاع االي‪WW‬وني ل‪WW‬ذلك تس‪WW‬تخدم ف‪WW‬ي‬
‫جرعات االشعاع وف‪W‬ي كواش‪W‬ف االش‪W‬عاع‪.‬ان الزجاج‪W‬ات الش‪W‬الكوجنيد‬
‫مجموعه متميزه من المواد الزجاجي‪W‬ة الغي‪W‬ر العض‪W‬ويه الت‪W‬ي تحت‪W‬وي‬
‫عل‪WW‬ي عنص‪WW‬ر ش‪WW‬الكوجنيد واح‪WW‬د او اكث‪WW‬ر مث‪WW‬ل الكبري‪WW‬ت والس‪WW‬يلينوم‬
‫والتريلي‪WW‬وم االكس‪WW‬جين باالتح‪WW‬اد م‪WW‬ع م‪WW‬واد اكث‪WW‬ر ايجابي‪WW‬ة كھربي‪WW‬ة مث‪WW‬ل‬
‫الزرنيخ واالنتموني‪ .‬ان زجاجيات الشالكوجنيد تعتبر عامه اقل شده‬
‫ومواد اضعف في الرابطه ع‪W‬ن زجاجي‪W‬ات االكس‪W‬يد‪ .‬ولق‪W‬د ت‪W‬م تحض‪W‬ير‬
‫الزجاجيات م‪W‬ن العناص‪W‬ر القص‪W‬دير والس‪W‬يلينوم والبزم‪W‬وث واالن‪W‬ديوم‬
‫عالي بنقاء ‪ . %99.999‬ان ھذه الزجاجيات تتفاع‪W‬ل م‪W‬ع االكس‪W‬جين‬
‫عند درجات الح‪W‬راره المرتفع‪W‬ه وذل‪W‬ك ت‪W‬م تص‪W‬نيعھا ف‪W‬ي انابي‪W‬ب س‪W‬يلكا‬
‫نظيفة ومفرغه من الھ‪W‬واء وت‪W‬م غس‪W‬يل ھ‪W‬ذه االنابي‪W‬ب بالم‪W‬اء المقط‪W‬ر‬
‫ثم تجفيفھا في فرن درجه حرارته حوالي ‪ 100‬درج‪W‬ه مئوي‪W‬ه‪ .‬ولق‪W‬د‬
‫تم ادخال الم‪W‬واد بع‪W‬د وزنھ‪W‬ا ف‪W‬ي انابي‪W‬ب الس‪W‬ليكا النظيف‪W‬ه ث‪W‬م تفريغھ‪W‬ا‬
‫حتي حوالي ‪ 10-4‬ضغط جوي وتم غلقھ‪W‬ا ‪ .‬و لق‪W‬د ت‪W‬م وض‪W‬ع االنابي‪W‬ب‬
‫المغلق‪WW‬ة داخ‪WW‬ل الف‪WW‬رن ال‪WW‬ذي ت‪WW‬م رف‪WW‬ع درج‪WW‬ة حرارت‪WW‬ه ال‪WW‬ي ‪ 900‬درج‪WW‬ه‬
‫مئويه في خالل س‪W‬اعه ث‪W‬م تثبيتھ‪W‬ا لم‪W‬ده ‪ 10‬س‪W‬اعات‪ .‬ولق‪W‬د ك‪W‬ان م‪W‬ن‬
‫الض‪WW‬روري رج الم‪WW‬واد المكون‪WW‬ه داخ‪WW‬ل االنبوب‪WW‬ة ف‪WW‬ي الف‪WW‬رن لتحقي‪WW‬ق‬
‫تجانس التركيب‪ .‬بعد التصنيع تم تبريد االنبوبة تبريد فجائ‪W‬ا ف‪W‬ي م‪W‬اء‬
‫مثلج ‪ .‬ولقد تم الحصول علي الزجاجي‪W‬ات الناتج‪W‬ه بالتبري‪W‬د الس‪W‬ريع ‪.‬‬
‫وتم اخراج المواد من االنابيب وتم االحتف‪W‬اظ بھ‪W‬ا ف‪W‬ي المج‪W‬ال ج‪W‬اف ‪.‬‬
‫وت‪WWW‬م تحض‪WWW‬ير الش‪WWW‬رائح الرقيق‪WWW‬ة م‪WWW‬ن المركب‪WWW‬ات بطريق‪WWW‬ة التبخي‪WWW‬ر‬
‫الح‪WWWWW‬راري تح‪WWWWW‬ت ض‪WWWWW‬غط ‪ 10-4‬ض‪WWWWW‬غط ج‪WWWWW‬وي بس‪WWWWW‬مك ثاب‪WWWWW‬ت ‪100‬‬
‫ن‪WW‬انوميتر‪.‬ولق‪WW‬دتمت دراس‪WW‬ة الخ‪WW‬واص التركيبي‪WW‬ة للمرك‪WW‬ب‪Inx Bi20‬‬
‫‪ Se20 Se60-x‬التش‪WW‬بيه الموص‪WW‬ل االم‪WW‬ورفي ولق‪WW‬د اوض‪WW‬ح التركي‪WW‬ب‬
‫الح‪WW‬راري التفاض‪WW‬لي للمرك‪WW‬ب عل‪WW‬ي ھيئ‪WW‬ة ب‪WW‬ودرة ولق‪WW‬د وج‪WW‬د ان قم‪WW‬ة‬
‫الحرارة الداخلية في المنحني التحليل الحراري التفاض‪W‬لي ين‪W‬تج م‪W‬ن‬
‫الح‪WW‬راره النوعي‪WW‬ة عن‪WW‬د انتق‪WW‬ال ح‪WW‬رارة الداخلي‪WW‬ة ف‪WW‬ي منحن‪WW‬ي التحلي‪WW‬ل‬
‫الح‪WW‬راري التفاض‪WW‬لي ين‪WW‬تج م‪WW‬ن الح‪WW‬راره النوعي‪WW‬ة عن‪WW‬د انتق‪WW‬ال ح‪WW‬رارة‬
‫‪Summery in Arabic‬‬

‫الزجاج ‪. Tg‬ان غياب اي قمة حاده للحرارة الخارجية يعتبر موش‪W‬ر‬


‫جيد لغياب اي تغيرات في التركي‪W‬ب ‪ .‬ولق‪W‬د اك‪W‬د تحلي‪W‬ل نم‪W‬اذج الحي‪W‬ود‬
‫لألشعه السينية للمركب ‪ Inx Sn20 Se60-x Bi20‬حالتة االمورفي‪W‬ة‪.‬‬
‫ولقد تم اجراء المسح االلكتروني الميكروسكوبي لشرائح الس‪W‬يليوم‪-‬‬
‫بزموث‪ -‬قصدير تم ترسيبھا عند درجة حرارة الغرفة ‪ .‬ولق‪W‬د تكون‪W‬ت‬
‫االفالم من حبيبات منفصلة غير منتظمة في الش‪W‬كل والحج‪W‬م وتتص‪W‬ل‬
‫بينھا حدود ما بين الحبيبات مح‪W‬دده جي‪W‬د"ا‪ .‬بآض‪W‬افة االن‪W‬ديوم يح‪W‬دث‬
‫انفص‪WWW‬ال اكث‪WWW‬ر ف‪WWW‬ي الوس‪WWW‬ط المح‪WWW‬يط مؤدي‪WWW‬ا" ال‪WWW‬ي زي‪WWW‬اده ف‪WWW‬ي حج‪WWW‬م‬
‫الحبيب‪WW‬ات عن‪WW‬د )‪ (x=0.1%‬ث‪WW‬م يمك‪WW‬ن رؤي‪WW‬ة حبيب‪WW‬ات اكب‪WW‬ر عن‪WW‬د‬
‫)‪ (x=0.2%‬وتص‪WW‬بح حبيب‪WW‬ات اص‪WW‬غر عن‪WW‬د )‪ . (x=0.3%‬ولق‪WW‬د ت‪WW‬م‬
‫تعي‪WWWWW‬ين كثاف‪WWWWW‬ة الزجاجي‪WWWWW‬ات المحض‪WWWWWW‬رة م‪WWWWW‬ن المرك‪WWWWW‬ب بطريق‪WWWWWW‬ة‬
‫الھيدروستاتيك ولقد لح‪W‬ظ ان الكثاف‪W‬ه تتن‪W‬اقص بزي‪W‬ادة االن‪W‬ديوم م‪W‬ن‬
‫‪ (x=0%) Inx Sn20 Se60-x Bi20 (gm/cm3)5.712‬ال‪WW‬ي‬
‫‪ (gm/cm3) 4.682‬للمرك‪WWWW‬ب ‪Inx Sn20 Se60-20 Bi20‬‬
‫حي‪WW‬ث)‪. (x=0.3%‬ولق‪WW‬د تمي‪WW‬زت الخ‪WW‬واص الض‪WW‬وئية للمرك‪WW‬ب ‪Inx‬‬
‫‪ Sn20 Se60-x Bi20‬بقياسات النفاذية واالنعكاس لطول الموجي من‬
‫‪ 200‬الي ‪ 1100‬نانوميتر لالفالم التي تم ترس‪W‬يبھا‪ .‬ان ن‪W‬وع االنتق‪W‬ال‬
‫االلكترون‪WWW‬ي المس‪WWW‬ئول ع‪WWW‬ن الخ‪WWW‬واص الض‪WWW‬وئية ھ‪WWW‬و االنتق‪WWW‬ال الغي‪WWW‬ر‬
‫مباشر المسموح وفجوة الطاق‪W‬ه االبتدائي‪W‬ة ف‪W‬ي المج‪W‬ال‪ 0.79‬ال‪W‬ي‪1.7‬‬
‫الكترون فولت ولقد وجد ان قيم فجوه الطاق‪W‬ة الض‪W‬وئية ‪ Eg‬تتن‪W‬اقص‬
‫بزياده االنديوم وذلك قد يكون بسبب الطبيعة المعدني‪W‬ة لألن‪W‬ديوم‪ .‬لق‪W‬د‬
‫تم تسجيل طي‪W‬ف االمتص‪W‬اص للمرك‪W‬ب ‪ Inx Sn20 Se60-20 Bi20‬ف‪W‬ي‬
‫منطقة االش‪W‬عة الف‪W‬وق بنفس‪W‬جية ولق‪W‬د وج‪W‬د ان بع‪W‬ض العوام‪W‬ل المھ‪W‬م‬
‫دراستھا مثل الرقم االحداثي ‪ , Nco‬وعدد المحددات ‪, Ns‬المعامل )‪(r‬‬
‫تح‪WW‬دد حي‪WW‬ود ‪ .Stoichiometry‬واذا ك‪WW‬ان ھن‪WW‬اك عالق‪WW‬ه خطي‪WW‬ة ب‪WW‬ين‬
‫قوه الرابط‪W‬ة ومتوس‪W‬ط فج‪W‬وة الحزم‪W‬ة ويوض‪W‬عھم ف‪W‬ي اھمي‪W‬ة لوص‪W‬ف‬
‫المركب‪WW‬ات ف‪WW‬أن أض‪WW‬افة االن‪WW‬ديوم س‪WW‬وف يوش‪WW‬ر عل‪WW‬ي متوس‪WW‬ط فج‪WW‬وة‬
‫الحزمة‪ .‬ووجد ان زياده االنديوم فأن متوسط شده الرابطة للمرك‪W‬ب‬
‫تتن‪WW‬اقص وب‪WW‬ذلك ف‪WW‬ان طاق‪WW‬ة‪ Eg‬تتن‪WW‬اقص ولتوكي‪WW‬د العالق‪WW‬ه ب‪WW‬ين ‪Eg‬‬
‫ومتوسط ش‪W‬ده الرابط‪W‬ة بوض‪W‬وح اكث‪W‬ر ولق‪W‬د تم‪W‬ت المقارن‪W‬ة ب‪W‬ين ‪Eg‬‬
‫‪ Hs/Nco,‬والت‪WWW‬ي تعتب‪WWW‬ر متوس‪WWW‬ط طاق‪WWWW‬ة الرابط‪WWW‬ة المفروض‪WWW‬ة ف‪WWWW‬ي‬
‫‪Summery in Arabic‬‬

‫الس‪W‬بيكة‪ .‬ان الخ‪W‬واص الكھربي‪W‬ه لس‪W‬بائك ‪Inx Sn20 Se60-x Bi20‬‬


‫تتضمن قياسات التوصيليه الكھربيه ‪ Dc‬للشرائح الرقيقة من ‪Inx‬‬
‫‪ Sn20 Se(60-x) Bi20‬وقياس‪WW‬ات القط‪WW‬ع )الف‪WW‬تح والقف‪WW‬ل (‪.‬ولق‪WW‬د ت‪WW‬م‬
‫قياس التوص‪W‬يلية الكھربي‪W‬ة ‪Dc‬للش‪W‬رائح الرقيق‪W‬ة كدال‪W‬ة ف‪W‬ي الح‪W‬رارة‬
‫واوضحت النتائج اعتمادالتوصيلية الكھربية ‪ Dc‬ووجود منطقيين‬
‫خطيتين بطاقتي تنشيط ‪.∆E2,∆E1‬‬
‫ولقد تم القيام بدراسات القطع )الفتح والغلق( للشرائح الرقيق‪W‬ة م‪W‬ن‬
‫‪ Inx Sn20 Se60-x Bi20‬ولقد ادي اضافة االنديوم الي زياده الجھ‪W‬د‬
‫االبتدائي)‪ (Vs‬والتيار االبتدائي )‪ (Is‬من ‪ 1.5‬فولت و‪1‬ميك‪W‬رو امبي‪W‬ر‬
‫علي التوالي عن‪W‬د)‪(x=0‬حت‪W‬ي ‪ 5.5‬فول‪W‬ت وتي‪W‬ار ‪ 2‬ميكروامبي‪W‬ر عل‪W‬ي‬
‫التوالي عند )‪ (x=0.3‬لشرائح سمكھا ثابت تساوي ‪ 100‬نانوميتر‪.‬‬
‫ام‪WW‬ا بالنس‪WW‬به جھ‪WW‬د )‪ (Vh‬فق‪WW‬د وجدان‪WW‬ه يزي‪WW‬د بزي‪WW‬اده نس‪WW‬به االن‪WW‬ديوم‬
‫من‪ 0.4‬فولت عند)‪(x =0‬الي ‪ 2‬فولت عن‪W‬د )‪ (x=0.3‬وعل‪W‬ي عك‪W‬س‬
‫ف‪WW‬ان زي‪WW‬اده االن‪WW‬ديوم تس‪WW‬بب نق‪WW‬ص تي‪WW‬ار ‪ Ih‬م‪WW‬ن ‪ 45‬ميكرواميت‪WW‬ر عن‪WW‬د‬
‫)‪ (x=0‬ال‪WW‬ي ‪ 19‬ميكروامبي‪WW‬ر عن‪WW‬د )‪ (x=0.3‬عن‪WW‬د س‪WW‬مك ثاب‪WW‬ت ‪100‬‬
‫ن‪WW‬اموميتر‪ .‬ولق‪WW‬د ثب‪WW‬ت ان الق‪WW‬درة االبتدائيةتزي‪WW‬د بزيادةاالن‪WW‬ديوم وھ‪WW‬ذا‬
‫يعن‪WW‬ي ان خاص‪WW‬ية القط‪WW‬ع تتن‪WW‬اقص بزي‪WW‬ادة االن‪WW‬ديوم‪.‬ووج‪WW‬د ان زي‪WW‬اده‬
‫االنديوم يزيد من طاقة االلتصاق )طاقه التح‪W‬ام( وبالت‪W‬الي ت‪W‬وثر عل‪W‬ي‬
‫خ‪WW‬واص القط‪WW‬ع ‪.‬ووج‪WW‬د ايض‪WW‬ا ان رف‪WW‬ع ح‪WW‬رارة الش‪WW‬ريحة يحس‪WW‬ن م‪WW‬ن‬
‫خ‪WW‬واص القط‪WW‬ع حي‪WW‬ث ان الجھ‪WW‬د االبت‪WW‬دائي يتن‪WW‬اقص والتي‪WW‬ار االيت‪WW‬دائي‬
‫يزيد وايض‪W‬ا اض‪W‬افة االن‪W‬ديوم ي‪W‬نقص ح‪W‬راره الفتيل‪W‬ة وب‪W‬ذلك يقل‪W‬ل م‪W‬ن‬
‫ق‪WWWWWWWWWWWWWWWW‬دره القط‪WWWWWWWWWWWWWWWW‬ع ‪.‬ولق‪WWWWWWWWWWWWWWWW‬د ادي زي‪WWWWWWWWWWWWWWWW‬ادة االن‪WWWWWWWWWWWWWWWW‬ديوم‬
‫من‪x=0‬الي)‪(x=0.1,0.2,0.3‬الي زياده ارتفاع القطع‪.‬زمن ‪ Tr‬م‪W‬ن‬
‫‪ 25‬الي‪ 40,100,200‬نانو ثاني‪W‬ه عل‪W‬ي الت‪W‬والي ‪ .‬ونق‪W‬ص ف‪W‬ي الطاق‪W‬ه‬
‫االلتصاق )‪ (C.E‬من ‪ 2.8186‬ال‪W‬ي ‪ 2.8192‬و ‪ 2.8199‬و‪2.820‬‬
‫الكت‪WW‬رون فول‪WW‬ت عل‪WW‬ي الت‪WW‬والي وتوض‪WW‬ح ھ‪WW‬ذة النت‪WW‬ائج ان المرك‪WW‬ب ل‪WW‬ه‬
‫نت‪WW‬ائج قط‪WW‬ع ابت‪WW‬دائي كھرب‪WW‬ي جي‪WW‬ده وتبش‪WW‬ر باس‪WW‬تخدام مفي‪WW‬د كوس‪WW‬يلة‬
‫قطع ابتدائية في تطبيقات الكمبيوتر‪.‬‬

‫واخيرا فأن دراسة تأثير أشعة جاما علي ش‪W‬رائح ‪Inx Sn20Se60-x‬‬
‫‪ Bi2‬وق‪WW‬د ل‪WW‬وحظ ان التش‪WW‬عيع بالجام‪WW‬ا ل‪WW‬يس ل‪WW‬ه ت‪WW‬أثير واض‪WW‬ح حت‪WW‬ي‬
‫‪Summery in Arabic‬‬

‫الجرعة ‪ 15 MRad‬والتي اوضحت قيمة ثابتة للنفاذية عند اضافة‬


‫االنديوم‪.‬‬
‫المستخلـــــــــص‬

‫الزجاجيات الشالكوجينية ھي مجموعة مميزة من المواد‬


‫الزجاجية الغيرعضوية )‪(Chalcogenideglasses‬والتي‬
‫تحتوي دائما علي واحد او اكثر من العناصر الشالكوجينية‬
‫كالكبريت او السيلينيوم او التليريوم وتتميز ھذه الزجاجيات‬
‫عموما بانھا اقل نشاطا او اقل ترابط من الزجاجيات التي‬
‫تحوى على االكسجين‪.‬‬

‫وقد تم تحضير مركبات من النظام ‪In x Sn20‬‬


‫‪ Se(60-x) Bi20‬حيث ‪ x= 0, 0.1, 0.2, 0.3%‬باستخدام‬
‫عناصر األنديوم و القصدير و السيلينيوم و البيزمس عالية‬
‫النقاوة )‪ (99,999‬وبصھر ھذه المكونات تحت التفريغ ‪4-10‬‬
‫سم زئبق في أنابيب سيلكا سابقة التنظيف عند درجة حرارة‬
‫‪ 900‬سليزيس لمدة ‪ 10‬ساعات ثم تبريدھا تبريدا فجائيا في‬
‫الثلج المجروش‪ .‬وقد تم التاكد من الطبيعة الغير متبلورة‬
‫والمتجانسة للمركبات المحضرة باستخدام جھاز حيود االشعة‬
‫السينية وكذلك استخدام المسح الحراري التفاضلي‪ .‬وقد تم‬
‫تحضير عينات االغشية الرقيقة بسمك ‪ 100‬نانومترباستخدام‬
‫التبخير الحراري للعينات‪.‬‬

‫وسيتم دراسة تاثير التشعيع والضوء وبعض التاثيرات‬


‫االخري علي العينات في ھذه الرسالة‬
‫دراسة الخواص الفيزيائية لبعض مواد اشباة الموصالت‬

‫رسالة مقدمة‬

‫للحصول على درجة الماجستير فى العلوم‬


‫)تخصص فيزياء الجوامد(‬

‫إلى‬

‫كلية العلوم – جامعة حلوان‬

‫من‬

‫أميرة علي عبدالوھاب‬


‫بكالوريوس علوم )‪ (2007‬جامعة حلوان‬

‫‪2012‬‬
‫دراسة الخواص الفيزيائية لبعض مواد اشباة الموصالت‬
‫رسالة مقدمة‬

‫للحصول على درجة الماجستير فى العلوم‬


‫)تخصص فيزياء الجوامد(‬

‫إلى‬
‫كلية العلوم – جامعة حلوان‬

‫من‬
‫أميرة علي عبد الوھاب‬
‫بكالوريوس علوم )‪ -(2007‬جامعة حلوان‬

‫المشرفون‬

‫أ‪ .‬د‪ /.‬عبدالرحمن عبدالمعبود‪ -‬عبدالمنجي‬


‫قسم الفيزياء‪ -‬كلية العلوم – جامعة حلوان‪.‬‬

‫أ‪.‬د‪.‬م‪ /‬حاتم حسن عامر‬


‫قسم الجوامدوالمعجالت االلكترونية‪-‬ھيئة الطاقة الذرية‪.‬‬

‫د‪ /.‬ياسرمحمد الجندي‬


‫قسم الفيزياء‪ -‬كلية العلوم – جامعة حلوان‪.‬‬

You might also like