Mam Thesis PDF

Investigations on Certain Tellurium based Bulk
Chalcogenide Glasses and Amorphous Chalcogenide

Films having Phase Change Memory (PCM)
Applications
A Thesis
Submitted for the Degree of
Doctor of Philosophy
in the Faculty of Engineering
By
Chandasree Das
Department of Instrumentation and Applied Physics

Indian Institute of Science
Bangalore - 560 012
India
September 2011
dedicated to
my father mr. mihir ranjan das,
my mother mrs. ilora das
&
My Teachers
Declaration
I hereby declare that the work reported in this thesis entitled Investigations on
Certain Tellurium based Bulk Chalcogenide Glasses and Amorphous Chalcogenide
Films having Phase Change Memory (PCM) Applications is entirely original and has
been carried out by me independently in the department of Instrumentation and Applied
Physics, Indian Institute of Science, Bangalore, 560 012 (INDIA), under the supervision of
Professor S. Asokan and Professor G. Mohan Rao.
I further declare that this work has not formed the basis for the award of any degree,
diploma, fellowship, associateship or similar title of any University or Institution.
September 2011 (Chandasree Das)

Acknowledgements
It is the time to thank all the people who made this thesis a successful one.
First of all I would like to express my deep and sincere gratitude to my
supervisors, Professor S. Asokan and Professor G. Mohan Rao for their immense
support, guidance, encouragement, help and valuable discussion, without which it
would have been impossible for me to finish this work. Joining IISc for my
doctoral work was a turning point of my life and also a wonderful experience.
My sincere thanks goes to Prof. K. Rajanna, Chairman, Department of

Instrumentation and Applied Physics and also ex Chairman Prof. R.M. Vasu, for
their encouragement and support and also allowing me to use the departmental
facilities for my work.
I would like to thank Prof. A.G. Menon, Prof. S. Mohan, Dr. M. Chandran,
Dr. Gunashekhar, Dr. N.C. Shivaprakash and all other faculty members of the
department who have been encouraging and helpful at various times and on
various occasions.
I sincerely thank Dr. K. Narasimha Rao for allowing me to use optical

spectroscopy in his lab without which my thin film studies would have been
incomplete.
I wish to express my warm and sincere thanks to Professor U.

Ramamurty, Department of Materials Engineering for his support to use Nano-
indentation facility in his lab.
I am very thankful to Prof. Vikram Jayaram, Department of Materials

Engineering for allowing me to use the electron microscopy facility at Nano
Centre independently.
I would like to thank Dr. Kiran and Eswar Prasad for helping me to
carryout Nano-indentation experiments in their lab.
I am heartily thankful to Kesab Barai, Somnath, Amit of Nano Centre for

helping me in electron microscopy training and operation during my early days
to Nano Centre.
I deeply acknowledge the support and help I received from Tamil Selvan,
Rakesh, Servanan, Yuvaraj from Narasimha Rao Sir’s lab for optical
spectroscopy.
I owe my deepest gratitude to Dr. S.V. Jagadeesh Chandra for his support
during my visit to South Korea to attend a conference. The care and affection I
received from Bindu and Nani is ever memorable.
It’s my immense pleasure to acknowledge the support I received from Office

Staff, Udaya Madam and Natarajan Sir for dealing with all official work, during
my entire stay here in this department.
My Sincere thanks goes to all the members of Glass blowing lab and
mechanical workshop for their cooperation through out my work. Particularly
Anandan Sir deserve a special mention for helping me with quartz ampoule as
well as all other glass blowing related work.
Thanks are due for Vijaya Madam and Rajum Madam for supporting me
and helping me during my entire stay in this lab.
I amply acknowledge the invaluable help and ever memorable company

that I got from my senior lab mates Pulok da, Anbu, Bhanu Sir, Manikandan,
Aashia, Madhav, Ramya, Damayanti, Sumeet, Lokesh, Madhu, Deepthi Madam,
Sreekumar Sir during my initial days in this lab.
I am very much thankful to my lab mates specially Srinivas, Tamil and

Arun for supporting me and assisting me in my experiments as well as taking
care of my samples during preparation in my absence.
I always enjoyed the company of my present lab mates Srividya, Lakshmi,

Guruprasad, Shivananju, Sarad ,Dinar, Vanitha, Reshmi, Ravi Sir, Chethna. I
cherish the time I spent with Hembram, Ambily, Rajesh, Eswar, Tulsi, Habib,
Venu in Mohan Rao Sir’s Lab.
I especially thank Dr. M.G. Mahesha, Department of Physics, Manipal

Institute of Technology for helping me in thin film coatings. His suggestions and
cooperation are heartily acknowledged.
I would always like to recall with great fondness the memorable moments
spent with other friends of the department especially, Tushar da, Baishali,
Sreejita, Madhurima di, Chintan. My coffee breaks with Mitesh and Nimisha
iv
deserve a special mention. As we stepped in to the IISc in the same day, we had
many things to share and chat. I will always relish those moments.
I will be failing in my duties if I don’t express my gratitude to this premier

institute, IISc for giving me the opportunity and providing me the facilities to
carry out my doctoral work. I will always remain indebted to this institute.
It gives me a great deal of pleasure to acknowledge all my teachers for

making me reach this stage in my life. I would like to thank all my friends for
being with me through various stages of my life.
It would not be complete if I don’t express my gratitude to my parents. This

thesis would have not been possible without their blessings and unconditional
support. I would like to thank my sister Soma and my brother-in law Abhijit for
their support during my work. I express my gratitude to them during this
moment.
Finally and most importantly, I heartily acknowledge my husband Deep for

being supportive, cooperative, helpful at every stage of my doctoral work. The
sweet innocent smile of my daughter Shubhangi is always a source of inspiration
and encouragement for me. I owe my every achievement to both of them.
CHANDASREE DAS
v
Preface
Chalcogenide glass based Phase Change Memories (PCMs) are being considered
recently as promising alternatives to conventional non-volatile Random Access Memories
(NVRAMs). PCMs offer high performance & low power consumption, in addition to other
advantages, such as high scalability, high endurance, compatibility with complementary
metal oxide semiconductors (CMOS) technologies, etc. Basically PCM is a resistance
variable non-volatile memory in which the memory bit state is defined by the resistance of
the material. In this case, the initial ‘OFF’ state (logic zero) corresponds to the high
resistance amorphous state and the logic 1 or ‘ON’ state corresponds to low resistance
crystalline state.
The present thesis work deals with electrical, thermal, mechanical and optical
characterization of certain tellurium based chalcogenide glasses in bulk and thin film form
for phase change memory applications. A comparative study has been done on the
electrical switching behavior of Ge-Te-Se & Ge-Te-Si amorphous thin film samples with
their bulk counterparts. Further, electrical switching and thermal studies have been
undertaken on bulk Ge-Te-Bi and Ge-Te-Sn series of samples. The composition
dependence of switching voltages of bulk and thin film samples studied has been
explained on the basis of different factors responsible for electrical switching. The thesis
contains ten chapters:
Chapter 1 deals with a brief introduction on chalcogenides and their applicability
in phase change memories. The glass transition phenomenon, synthesis of chalcogenide
alloys, different structural models of amorphous semiconductors, electrical switching
behavior etc. are also discussed in detail in this chapter. Further, a brief description of
optical and mechanical properties along with the principles of few characterization
techniques used is discussed. Also, a brief overview on PCM application of chalcogenides
is presented.
The second chapter provides the details of various experimental techniques used to
measure electrical, thermal, optical and mechanical properties of few tellurium based
chalcogenide glassy systems.
In the third chapter, the electrical switching behavior of amorphous Al23Te77 thin
film devices, deposited in co-planar geometry, has been discussed. It is found that these
samples exhibit memory type electrical switching. Scanning Electron Microscopic studies
show the formation of a crystalline filament in the electrode region which is responsible
for switching of the device from high resistance OFF state to low resistance ON state. The
switching behavior of thin film Al-Te samples is found to be similar to that of bulk
samples, with the threshold fields of bulk samples being higher. This has been understood
on the basis of higher thermal conductance in bulk, which reduces the Joule heating and
temperature rise in the electrode region.
Electrical switching and thermal behavior of bulk; melt quenched Ge18Te82-xBix
glasses (1 ≤ x ≤ 4) are presented in chapter 4. Ge-Te-Bi glasses have been found to exhibit
memory type electrical switching behavior, which is in agreement with the lower thermal
diffusivity values of these samples. A linear variation in switching voltages (also known as
threshold voltages) (Vt) has been found with increase in thickness. The switching voltages
have been found to decrease with an increase in temperature which is due to the decrease
in the activation energy for crystallization at higher temperatures. Further, Vt of Ge18Te82-
xBix glasses have been found to decrease with the increase in Bi content, indicating that in
the Ge-Te-Bi system, the resistivity of the additive has a stronger role to play in the
composition dependence of Vt, in comparison with the network connectivity and rigidity
factors. In addition, the composition dependence of crystallization activation energy has
been found to show a decrease with an increase in Bi content. X-ray diffraction studies on
thermally crystallized samples reveal the presence of hexagonal Te, GeTe, Bi2Te3 phases.
The fifth chapter deals with the electrical switching studies and optical band gap
measurements on GexSe35-xTe65 (17 ≤ x ≤ 23) amorphous thin film samples. These thin
film samples coated with sandwich geometry are found to switch with very low voltages
as compared to bulk samples of the same chalcogenide glasses. The switching voltages
and optical band gap are found to increase with the addition of Ge at the expense of Se.
High structural cross linking with progressive addition of 4-fold coordinated Ge atoms
could be the one of the reasons of increasing switching voltage and stronger Ge-Se bond
strength could be the reason of increasing band gap for these chalcogenide glasses.
In chapter 6, electrical switching studies on amorphous Ge15Te85-xSix (1 ≤ x ≤ 6)

thin film samples have been described and the results are compared with their bulk
counterparts. Similar trend has been found for both bulk and film samples when the
threshold field is varied with composition. Optical band gap has been measured as a
function of composition for these films, which also shows a behavior similar to that of
switching voltages. The increasing trend in the variation with composition of electrical
viii
switching voltages and optical band gap are due to the increase in network connectivity
and rigidity as Si atoms are incorporated into the Ge-Te system.
Chapter 7 summarizes the electrical switching and glass forming ability of the Ge-
Te-Sn glasses of two different composition tie-lines, namely Ge15Te85-xSnx and Ge17Te83-
xSnx. Glasses belonging to both the series have been found to exhibit memory type of
electrical switching behavior. The thickness dependence of threshold voltages is also
found to support the memory switching behavior of the system. Further, ADSC studies are
undertaken to explore the thermal behavior of these glasses which indicates that the
crystallization tendency increases as Sn concentration is increased in the Ge-Te network.
XRD studies done on two representative samples from both the series reveal the fact that
Sn atoms do not take part actively to enhance the network connectivity and rigidity. The
composition dependence of crystallization temperature, metallicity factor and results of
XRD studies are put together to explain the variation with composition of threshold
voltages for both the series of samples.
In chapter 8, investigations on the electrical switching behavior of Ge15Te85-xSnx
(1 ≤ x ≤ 5) and Ge17Te83-xSnx (1 ≤ x ≤ 4) amorphous thin films have been discussed. Both
the series of samples have been found to exhibit memory type of electrical switching
behavior. The composition dependence of threshold voltage shows a decreasing trend,
which has been explained on the basis of the Chemically Ordered Network (CON) model,
bond strength and the metallicity factor. The optical band gap variation of both the series
also exhibits a similar decreasing trend with composition. The observed behavior has been
understood on the basis of higher atomic radius of Sn atom than Ge atom, which makes
the energy difference between bonding and anti bonding state less at band edge.
Chapter 9 deals with the nano-indentation studies on Ge15Te85-xSix (0 ≤ x ≤ 9) bulk
glasses. The composition dependence of young’s modulus and hardness is studied
systematically in this glassy system. The density of the samples of different compositions
has also been measured, which strongly supports the variation of Young’s Modulus and
hardness with composition. The composition dependence of mechanical properties of Ge-
Te-Si samples has been understood on the basis of the presence of an intermediate phase
and a thermally reversing window in this glassy system.
A summary of the significant results obtained in the present thesis work is

presented in the last chapter along with the scope for future work.
ix
Contents
Declaration ............................................................................................................................ i
Acknowledgements ............................................................................................................. iii
Preface ................................................................................................................................vii
Chapter 1: Glassy Chalcogenides and Phase Change Memories - A Brief Overview ....... 1
1.1 Introduction ..................................................................................................................... 1
1.2 Glass Transition and Glass Forming Ability................................................................... 2
1.2.1 Glass Transition Phenomenon .................................................................................. 2
1.2.2 Glass Forming Ability .............................................................................................. 3
1.3 Synthesis of Glassy Alloys ............................................................................................. 4

1.4 Structural Criteria and Properties .................................................................................... 5
1.4.1 Structural Models ..................................................................................................... 5
1.4.1.1 The Continuous Random Network (CRN) Model ............................................. 5
1.4.1.2 The Random Covalent Network (RCN) and Chemically Ordered Network
(CON) Models ............................................................................................................... 6
1.4.1.3 Topological Thresholds ..................................................................................... 6
1.4.2 Electrical, Optical and Mechanical Properties ......................................................... 8
1.4.2.1 Electrical Switching ........................................................................................... 9
1.4.2.1.1 Different models proposed for electrical switching .................................. 11
1.4.2.1.2 Different time constants and Switching parameters for chalcogenide

glasses ...................................................................................................................... 12
1.4.2.2 Optical Properties ............................................................................................ 15
1.4.2.3 Mechanical Properties...................................................................................... 16
1.5 Application of Chalcogenides ....................................................................................... 17

1.6 Phase Change Memory and Chalcogenides: A brief Overview .................................... 18
1.7 Objectives and Scope of the Present Work ................................................................... 20
References ........................................................................................................................... 21
Chapter 2: Experimental Techniques ............................................................................... 27
2.1 Introduction ................................................................................................................... 27
2.2 Synthesis of Chalcogenide Glasses ............................................................................... 27
2.2.1 Bulk Glass preparation ........................................................................................... 27
2.2.2 Thin film coating .................................................................................................... 28
2.3 Thickness Measurement................................................................................................ 31

2.4 Electrical Switching Studies ......................................................................................... 32
2.5 Scanning Calorimetry ................................................................................................... 33
2.5.1 Differential Scanning Calorimetry ......................................................................... 33
2.5.2 Alternating Differential Calorimetry ...................................................................... 34
2.6 X-Ray Diffraction Studies ............................................................................................ 35

2.7 Scanning Electron Microscopy (SEM) ......................................................................... 36
2.8 Optical Spectroscopy .................................................................................................... 37
2.9 Nano-indentation........................................................................................................... 37
References ........................................................................................................................... 39
Chapter 3: Electrical Switching Behavior of Amorphous Al23Te77 Thin Film Sample .. 41
3.1 Introduction ................................................................................................................... 41
3.2 Experimental techniques ............................................................................................... 42
3.3 Results and Discussion ................................................................................................. 42
3.4 Conclusions ................................................................................................................... 47
References ........................................................................................................................... 48
Chapter 4: Electrical Switching and Thermal Studies on Bulk Ge-Te-Bi Glasses ......... 51
4.1 Introduction ................................................................................................................... 51
4.2 Experimental procedures .............................................................................................. 51
4.3 Results and Discussion ................................................................................................. 52
4.3.1. Electrical switching behavior of Ge-Te-Bi glasses ............................................... 52
4.3.2. Temperature dependence of threshold voltages .................................................... 55
4.3.3. Thickness dependence of switching voltages........................................................ 55
4.3.4. Activation energy for crystallization ..................................................................... 56
4.3.5. Crystallization behavior of Ge18Te82-xBix glasses ................................................. 57
4.3.6. Electron Microscopic analysis of Ge-Te-Bi .......................................................... 59
4.4 Conclusions ................................................................................................................... 59

References ........................................................................................................................... 61
Chapter 5: Electrical Switching and Optical Studies on Amorphous GexSe35-xTe65 (17≤ x
≤ 23) Thin Films................................................................................................................. 63
5.1 Introduction ................................................................................................................... 63
5.2 Experimental Techniques .............................................................................................. 64
5.3. Results and Discussion................................................................................................. 65
5.3.1. Electrical Switching .............................................................................................. 65
5.3.2. Optical band gap .................................................................................................... 69
5.4. Conclusions .................................................................................................................. 72

References ........................................................................................................................... 73
Chapter 6: Electrical Switching and Optical Band gap studies on Ge15Te85-xSix (1 ≤ x ≤
6)Amorphous Thin Film Samples ..................................................................................... 75
6.1 Introduction ................................................................................................................... 75
6.2 Experimental Details ..................................................................................................... 76
6.3 Results and Discussion.................................................................................................. 77
6.3.1. Electrical switching behavior of amorphous Ge15Te85-xSix thin films .................. 77
6.3.2. Optical Band gap ................................................................................................... 81
6.3.3. Electron Microscopic Studies ................................................................................ 83
6.4 Conclusions ................................................................................................................... 84

References ........................................................................................................................... 85
Chapter 7: The Effect of Sn Doping on the Electrical and Thermal Properties of Bulk
Ge-Te Glasses ..................................................................................................................... 89
7.1 Introduction ................................................................................................................... 89
7.2 Experimental Procedure ................................................................................................ 90
7.3 Results and Discussion.................................................................................................. 90
7.3.1. Memory switching behavior of Ge-Te-Sn glasses ................................................ 90
7.3.2 Composition dependence of switching voltages .................................................... 92
7.3.3 Variation of threshold voltage with thickness ........................................................ 97
7.4 Conclusions ................................................................................................................... 98

References ......................................................................................................................... 100
Chapter 8: Studies on Electrical Switching Behavior and Optical Band gap of
Amorphous Ge-Te-Sn Thin Films ................................................................................... 103
8.1 Introduction ................................................................................................................. 103
8.2 Experimental Details ................................................................................................... 103
8.3 Results and Discussion ............................................................................................... 104
8.3.1. Electrical Switching behavior of amorphous Ge15Te85-xSnx and Ge17Te83-xSnx Thin
Films .............................................................................................................................. 104
8.3.2. Optical band gap of amorphous Ge15Te85-xSnx and Ge17Te83-xSnx Thin Films ... 108
8.4. Conclusions ................................................................................................................ 111

References ......................................................................................................................... 112
Chapter 9: Nano-Indentation Studies on Bulk Ge-Te-Si Chalcogenide Glasses .......... 115
9.1 Introduction ................................................................................................................. 115
9.2 Experimental techniques ............................................................................................. 115
9.3 Results and Discussion ............................................................................................... 116
9.4 Conclusions ................................................................................................................. 121
References ......................................................................................................................... 123
Chapter 10: Summary and Future Scope ....................................................................... 125
Scope for Future Work...................................................................................................... 127
List of Publications ......................................................................................................... 127
Chapter 1
Glassy Chalcogenides and Phase Change Memories - A

Brief Overview
1.1 Introduction
Amorphous or non-crystalline or glassy materials are structurally disordered and
they lack long range periodic order in the arrangement of atoms. There are innumerous
materials that constitute the category of glassy materials; however, certain unique
properties make chalcogenide glassy semiconductors containing one or more chalcogen
elements, S, Se and Te, very interesting. Chalcogenide glasses have been in the forefront
of research for over a century.
The earliest report by Schulz-Sellack on an oxygen-free glass dates back to 1870

[1]. Subsequently, a detailed study on the optical properties of the vitreous selenium was
reported by Wood in 1902 [2] and Meier in 1910 [3]. However, intensive investigations on
chalcogenide glasses began only after the discovery that many of these materials transmit
light well in the infrared, up to 12 µm [4-6]. Furthermore, multi-component glassy
chalcogenides attracted attention after it was found that amorphous Tl2Se·As2Te3
exhibited a room temperature conductivity of 10-3 Ω-1cm-1, which is larger than many
crystalline semiconductors [7]. All these interesting aspects led to intensive investigations
on the regions of glass formation and properties of multi component chalcogenide alloys
[8-14]. In addition, halogen containing chalcogenide glasses were also investigated [15-
17] after the discovery that they have desirable features for use as high refractive index
immersion media [18,19].
The research on chalcogenide glasses took a new direction when the

phenomenon of electrical switching was first observed by Ovshinsky in 1968 [20]. The
unusual electrical effects exhibited amorphous chalcogenides such as switching stimulated
from time to time, the development of theory and applications of these materials. The
recent discovery of optical non-linear properties of several chalcogenide glasses, opened
new horizons for optoelectronic applications. Today, the chalcogenide glasses rank among
the most investigated class of semiconductor solids.
Ch.1: Glassy Chalcogenides and Phase Change Memories…
1.2 Glass Transition and Glass Forming Ability
1.2.1 Glass Transition Phenomenon
Even though several terms including amorphous, vitreous, non-crystalline, etc.,

are routinely used to designate structurally disordered solids, the term glass is specifically
used to denote those solids which are obtained by cooling a melt very rapidly below its
freezing temperature, bypassing crystallization. A glass is a substance which is in meta-
stable equilibrium, having the disordered structure of a liquid and the mechanical rigidity
of a solid. When reheated, the glassy solids exhibit the phenomenon of glass transition
which is one of the most fascinating and well-studied aspects of condensed matter physics
[21-24]. The other terms such as amorphous are conventionally used to describe those
solids which are non-crystalline but do not exhibit glass transition while reheating.
However, in many contexts, the terms glassy, amorphous, non-crystalline are used
synonymous.
Figure 1.1: The volume versus temperature diagram showing two different ways of
solidifying a melt: (a) slow cooling to the crystalline state and (b) rapid quenching to the
glassy state [25].
The glass transition phenomenon is best understood by considering a liquid being
cooled from its melt as shown in figure 1.1 [25]. Depending on the viscosity of the liquid
at its freezing temperature, it may crystallize or pass through a meta-stable super cooled
liquid state. The viscosity of the super cooled liquid increases continuously, with a
decrease in temperature below its freezing temperature (Tf). At a temperature called glass
2
transition temperature (Tg) and around a viscosity of 1013 poise, the super cooled liquid
transforms into a glassy solid with no long range structural order.
The transformation of the melt into the glassy state does not involve any latent
heat or volume change as in the case of formation of crystals [25]. The liquid to glass
transition is characterized by a gradual change in volume, entropy and enthalpy [26].
As mentioned earlier, the reverse transformation from a glass to a super cooled

liquid also takes place if a glass is heated to a temperature above glass transition
temperature. The glass transition reaction is a rate dependent process and the thermal
history of the sample determines the magnitude of glass transition temperature. Several
models have been developed to explain glass transition phenomenon in super cooled
liquids. Few of them are configurational entropy model, free volume model, cluster model
etc. [27].
1.2.2 Glass Forming Ability
The term glass has conventionally been reserved for an amorphous solid
prepared by quenching the melt. Earlier it was thought that only a restricted number of
materials such as oxides and organic polymers could be prepared in the form of a glassy
solid. But now it is realized that the ‘glass forming ability’ is almost an universal property
of condensed matter. According to Turnbull [28], “Nearly all materials can be prepared as
glassy solids”, if cooled fast enough and far enough below the normal freezing
temperature. Few of the important issues which influence the glass forming ability in
chalcogenide systems are atomic structure, viscosity of the melt; speed of quenching and
the frustration in a multi component melt [29]. To understand the origin of glass
formation, several models based on structural, thermodynamic and kinetic factors have
been developed:
A macroscopic model of glass formation proposed by Zachariasen predicts that

the substance can readily form an extended three dimensional network lacking periodicity
if the coordination polyhedra in the glass and the energy content in the glass are similar to
that of crystal [30]. Few models also propose that the number of nearest neighbors within
an irregular structural network must be two, three or four only, for good glass formation
[31]. Phillips has tried to explain the composition dependence of glass forming ability,
using the constraints theory, on the basis of percolation in network rigidity [32]. Many
empirical prescriptions have been developed to account for glass forming tendency in
3
certain specific systems [33]. However, there is no definite rule, which can be used
universally to predict glass forming ability in a wide variety of systems.
1.3 Synthesis of Glassy Alloys

Glass formation is a matter of bypassing crystallization. The channel to the
crystalline state is evaded by quickly crossing the dangerous regime of temperature
between Tf and Tg and achieving the amorphous solid state below Tg. The synthesis of
glasses does not require a carefully controlled growth technique, which is a vital economic
advantage in many applications. There are various methods which are used to prepare
amorphous/glassy materials. Few are shown schematically in figure 1.2.
Figure 1.2: Different methods of preparation of amorphous/glassy solids: (a) Slow

cooling, (b) Moderate quenching, (c) Splat quenching, (d) Vapor condensation technique.
The slow-cooling technique, as shown in figure 1.2 (a) is normally adopted for
the materials with very high glass forming tendency. In this technique, the melt is allowed
to cool slowly by simply turning off the furnace or by bringing its temperature down in a
programmed manner. Typical cooling rates for the slow-cooling technique are in the range
from 10-4 to 10-1 0K/sec. Few examples of the glasses which can be prepared by slow
cooling are SiO2, As2S3, etc. Glasses like chalcogenides require somewhat faster rate to
quench, in the range of 101 to 102 0K/sec. An ice-water bath is used to quench the modest
volume of the melt as shown in figure 1.2 (b), known as moderate quenching technique.
The technique sketched in figure 1.2 (c) is another melt quenching method developed
especially for metallic glasses, called the splat quenching technique. In this method, the
liquid droplet is cooled from two sides at once, and is used to produce milligram-size
laboratory specimens of metallic glasses such as Au-Si, etc. The typical rates of cooling in
this technique are in the range of 105 to 108 0K/sec. Vapor condensation technique as
shown in figure 1.2 (d) is one which bypasses the liquid completely and constructs an
amorphous solid in atom by atom fashion from the gas phase on a cold substrate. This
4
technique gives the highest effective quench rates and amorphous films of Si, Ge, etc., are
obtained by this technique [25].
1.4 Structural Criteria and Properties

The structure of a crystalline material is understood from its unit cell; however
the absence of long range order creates a difficulty in adopting such a procedure for
amorphous materials. Zachariasen took the first attempt to understand the structure of non-
crystalline solids [30]. According to him, atoms should form an extended three
dimensional network, as the coordination polyhedra in the glass and the corresponding
crystal are similar. Various experimental techniques are used to study the short range order
and medium range order in amorphous materials, like X-ray and neutron diffraction,
Extended X-ray Absorption Fine Structure (EXAFS), Nuclear Magnetic Resonance
(NMR), Raman etc [34-46]. Various models have been proposed to obtain the information
about the amorphous structure in a macroscopic scale as the above mentioned techniques
can give the information only up to the first two coordination shells.
1.4.1 Structural Models
1.4.1.1 The Continuous Random Network (CRN) Model
The CRN model first introduced by Zachariasen is an ideal model to characterize

covalent amorphous structures in which, atoms with coordination numbers less than four
are connected by direct covalent bonds in an united ideal network [30]. The CRN model is
based on few assumptions like:
a) The coordination of the constituent atoms satisfy Mott’s 8-N rule [47], where N
represents the column in the periodic table to which the element belongs.
b) The nearest neighbor distances (bond-lengths) in the glassy and the corresponding
crystalline state are nearly same with an allowed variation of <1% for bond lengths
and ± 10 % for bond angles [25].
c) There are no over-coordinated or under coordinated (dangling bonds) atoms.
According to this model, the local atomic structure in non-crystalline

semiconductors is not completely random; a considerable degree of short range order is
present in these materials.
5
1.4.1.2 The Random Covalent Network (RCN) and Chemically Ordered Network
(CON) Models
The Random Covalent Network (RCN) and the Chemically Ordered Network
(CON) models are two other models proposed to explain the local structure of amorphous
semiconductors. The random covalent network (RCN) model, as its name implies, treats
the distribution of bond types as purely statistical and determined only by the local
coordination and concentration. Thus, any effects leading to preferential ordering like
relative bond energies are neglected [34]. Another approach is the chemically ordered
network (CON) model which assumes that the heteropolar bonds are favored. This model
is based on relative bond energies, with higher bond energies getting precedence over the
lower bond energies during network formation.
The RCN and CON approaches are obviously appropriate only for covalently
bonded system and not for random close-packed structures such as metallic or ionic
glasses. They are also not appropriate for covalent network forming materials which are
alloyed with network modifiers. In this category, two cases may be distinguished:
‘network rupture’ in which the number of fully coordinated network forming anions is
progressively decreased and the connectivity of the network is broken upon addition of
modifier, and ‘network transformation’ in which the coordination of the network forming
cat-ion is progressively increased [34].
1.4.1.3 Topological Thresholds
Network topological thresholds also play an important role in determining the

structural aspects of chalcogenide glasses. Generally the covalent network of chalcogenide
glasses is known to be influenced by two topological effects namely, the rigidity
percolation threshold (also known as stiffness or mechanical threshold) and the chemical
threshold.
(a) Rigidity percolation threshold
The criterion for the mechanical stability of networks, in the context of trusses,
was first introduced by Maxwell in 1864 [48]. Later in 1979, J.C. Phillips proposed a
model, based on the constraint theory [32] to understand the evolution of the local
structure in a covalent network glass and its glass forming tendency as a function of
composition using the Valence Force Fields (VFF) model.
6
According to this theory [32,49-56], bond bending and bond stretching forces are
responsible for the topological constraints in a glassy network. The number of constraints
acting on the atoms due to bond bending and bond stretching forces are r/2 and r (r-1)/2
respectively where r is the coordination number. The total number of constraints Nc acting
on a binary system is given by
2
𝑁𝐶 = r⁄2 + 𝑟(𝑟 − 1)⁄2 = 𝑟 ⁄2 ------------------------------(1)
As per the prediction of Phillips and Thorpe [51], the condition for optimal glass
formation is satisfied only when the total number of constraints equals the number of
degrees of freedom (Nd) i,e.,
𝑁𝑐 = 𝑁𝑑 ------------------------------(2)
Substituting the values for a binary system in equation (2), the average
coordination number becomes
𝑟2
=3
2
𝑟𝑐 = √6 ------------------------------(3)
The addition of every new bond will introduce two more constraints as the angle
between the existing bonds need to be specified. This in turn gives (2r-3) constraints
around an atom with r ≥ 2 [47]. Hence the number of constraints for a ternary system will
be
𝑁𝑐 = 𝑟⁄2 + (2𝑟 − 3) ------------------------------(4)
As mentioned earlier the system will tend to yield a good glass only when
number of constraints is equal to the no of degrees of freedom
𝑁𝑐 = 𝑁𝑑
𝑟⁄2 + (2𝑟 − 3) = 3
𝒓𝒄 = 𝟐. 𝟒 ------------------------------(5)
In literature, rc is commonly known as Rigidity Percolation Threshold (RPT).

Further at this point, generally the glassy system undergoes a percolative transition from a
floppy glass to a rigid solid. The glassy network is said to be under-constrained when the
average coordination number of the network glass is less than rc and over-constrained
when it is more than rc. Several glassy systems such as are Ge-Te [57], Al-Te [58], Al-Ge-
7
Te [59], As-Ge-Te [60], etc., exhibit the signatures of rigidity percolation in the
composition dependence of various properties.
Recently, Boolchand and his co-workers have revealed that, in certain
chalcogenide systems the rigidity transition occurs over a range of compositions and
average coordination numbers. In these systems, an intermediate phase (also known as
Boolchand’s phase or isostatically rigid phase) exists, between the floppy and rigid phase.
Further these systems exhibit two stiffness transitions, from floppy state to an isostatically
rigid state and from isostatically rigid state to stressed rigid state [61-63]. There are a
variety of glassy systems which have shown the presence of intermediate phase in
chalcogenide glasses like, Ge-Se, Se-Si, As-Se, Ge-S, As-Se-Ge, As-Te-Si, Ge-Te-Si, Al-
Te-Si, etc. [64-68].
(b) Chemical Threshold (CT)
The Chemical threshold deals with the chemical ordering in a chalcogenide
glassy system. A chemically ordered glass is one in which only heteropolar bonds are
present, as explained by the Chemically Ordered Covalent Network (COCN) model [69-
71]. Taking an example of AxB1-x glassy system, the composition corresponding to the
chemically ordered glass is given by
𝑅𝑎
𝑥𝑐 = ------------------------------(6)
𝑅𝑎 + 𝑅𝑏
Where the coordination’s of A and B atoms are represented by Ra and Rb

respectively, according to 8-N rule. The glass compositions with x < xc are characterized
by A-B and A-A type of bonds and those with x > xc contain A-B and B-B type of bonds;
whereas at xc the system is chemically ordered comprising of only A-B type of bonds. The
glass formation is found to be very difficult at chemical threshold as the glass is
energetically closest to the crystalline state [72,73]. For the GexSe100-x system the chemical
threshold occurs at rc = 2.67, while it occurs at rc = 2.4 (same as mechanical threshold) for
the AsxSe100-x system [74-77].
1.4.2 Electrical, Optical and Mechanical Properties

Chalcogenides are found to exhibit various interesting properties. A brief
discussion on those properties which have relevance to the present work has been given in
the following subsections:
8
1.4.2.1 Electrical Switching
In insulators, electrical fields in excess of about 106 V/cm lead to a destructive

breakdown; however in certain crystalline or amorphous materials, it can be non
destructive due to switching and negative resistance effect. The negative resistance
phenomenon can be divided into six broad groups namely, Voltage Controlled Negative
Resistance (VCNR) with and without memory, Current Controlled Negative Resistance
(CCNR) with and without memory, and memory and threshold switching as shown in
figure 1.3. A VCNR device as shown in figure 1.3 (A) exhibits an Ohmic behavior up to a
critical voltage Vt, above which the material passes through a negative resistance region to
reach a high resistance state and the path is retraced, when the voltage is reduced. In a
VCNR device with memory (figure 1.3 (B)), the high resistance state is retained even after
the removal of electric field. Al2O3, Ta2O5, ZrO2, etc., are few amorphous oxides
exhibiting the VCNR behavior with or without memory.
In a CCNR behavior, voltage increases initially with current. However, at a
threshold voltage corresponding to threshold current, the voltage starts decreasing with
further increasing current, passing through a negative resistance zone to a low resistance
state. With the reduction in current, the characteristic is retraced as shown in figure 1.3
(C). Figure 1.3 (D) shows the CCNR behavior with memory, in which the low resistance
state once attained is retained under zero bias condition (bi-stable device). Materials like
MoO3, NbO5, etc., have been found to exhibit the CCNR behavior with or without
memory.
9
V
Figure 1.3: Six possible types of negative resistance behavior observed in amorphous
solids. (A) VCNR; (B) VCNR with memory (C) CCNR (D) CCNR with memory (E)
threshold switching (F) memory switching.
Electrical switching is a phenomenon during which a material under an electrical

field suddenly changes from a low conducting (OFF) state to a high conducting (ON) state
figure 1.3 (E) and (F). Though electrical switching is a CCNR behavior, there is no stable
operating point between high resistance and low resistance state. The electrical switching
is of two types, namely memory switching and threshold switching [20,78,79]. If the ON
state of the material is retained even after the removal of electric field, it is called memory
switching. On the other hand, the threshold switching materials come back to their high
resistance OFF state after the removal of electric field. The threshold materials require a
holding current and a voltage to sustain the ON state. Once the holding current is removed,
they revert back to their original OFF state. It is customary in literature to use the term
“memory switching” in a general way to denote both switching (which is faster) as well as
negative resistance with memory (which is slower). In fact, most of the reports in literature
which talk of memory switching pertain only to negative resistance with memory.
As mentioned earlier, the phenomenon of electrical switching in glassy
chalcogenides, has been first observed by Ovshinsky in 1968 [20], and soon many
amorphous chalcogenide alloys such as Te-Ge-Sb-S, Te-As-Ge-Si, Ge-As-Te, etc. have
10
been found to exhibit threshold and memory switching [80,81]. Apart from these, many
binary glasses like Ge-Te [82,83], Si-Te [57], Al-Te [84] and the ternaries like Al-Ge-Te
[85], As-Ge-Te [60,58], As-Se-Te [86], etc., have been found to exhibit memory
switching. Similarly, glasses like As-Te-In [87], Ge-Te-In [88] etc., have shown threshold
switching characteristics over a wide range of compositions.
It is interesting to note that memory switching is a consequence of a phase
transition from the amorphous to crystalline state [89]. The memory switching occurs in a
material, if the Joule heating caused by the larger current flowing after the initiation of the
switching is sufficient to induce a phase transition in the material. There are various
factors which usually decide whether a chalcogenide glass will exhibit memory switching
or not, such as the ON state current [90], thermal stability of the glass [90], thermal
diffusivity of the material [91], network connectivity and topological effects, etc.[57-
60,85,87]. The origin of memory and threshold switching is discussed in more detail
below:
1.4.2.1.1 Different models proposed for electrical switching
There are several theories and models proposed in literature to explain the array
of complex phenomena associated with the electrical switching effect like thermal,
electronic and electro thermal models:
(a) Thermal Model
The thermal model basically explains the memory switching phenomenon which
is a consequence of phase transition from amorphous to crystalline state [92]. Usually,
memory type electrical switching occurs which de-vitrifies easily. The crystallization of
the chalcogenide glass during the switching process is believed to be caused by the Joule
heating and the resultant excess carrier concentration in the current path due to a high
electric field. There is a formation of crystalline filament between two electrodes in a
memory glass during switching, which has been confirmed by electron microscopy and
optical measurements [89]. As the crystallization is intrinsically an irreversible
phenomenon for a glass and the resistivity of the crystalline material is much smaller than
that of the glassy material, the material remains semi permanently in low resistance ON
state after memory switching.
(b) Electronic Model
The electronic model can be further sub divided into purely electronic and double
injection space charge model. According to the purely electronic model, proposed by
11
Adler and Peterson [92], the conducting state is achieved only after the positively and
negatively charged defect traps existing intrinsically in a chalcogenide glass, are filled by
field injected charge carriers. The lifetime of the charge carriers increases sharply when all
the traps are filled, from a value much lower than the time needed by a carrier to cross the
entire thickness of the material to a value higher than the time needed by the carrier to
cross the film. This leads to a drop in voltage as the current increases leading to a high
conducting state.
The double injection space charge model proposed by Henisch [93], Henisch
et.al [94] and Lucas [95], suggests that both the electrons and holes are injected by the
applied field and they recombine and neutralize the positively and negatively charged
defect states, setting up a negative and positive space charge region adjacent to the anode
and cathode respectively. The space charge region will limit the current flow near
electrode and the field will be redistributed increasing at the centre and decreasing near the
electrodes. With the increase in applied voltage, more charges are injected until the space
charge regions overlap. In the overlapping charge region, the material is neutral but the
conductivity in the centre and the field near the electrode increase. As a result, electrons
and holes are accelerated rapidly across the neutral region and the high field carrier
injection at the leads to an unstable situation. Since the space charge has been neutralized
and the traps are filled, the conductivity increases leading to the ON state characteristics.
(c) Electro-thermal model
According to the electro-thermal model, proposed by Warren to explain
threshold switching, the joule heating due to an applied electric field is responsible to
initiate the formation of a hot narrow conducting channel [96]. By deriving numerical
solutions to the dynamic-heat balance equations and substituting the appropriate
parameters of the material like the material conductivity [97], etc., the non-Ohmic pre-
switching behavior has been predicted. Further, the model suggests that there can be a
temperature rise up to several hundred degrees along the conducting channel before the
current shows any substantial changes.
1.4.2.1.2 Different time constants and Switching parameters for chalcogenide glasses
There are several parameters involved in the switching process of chalcogenide

glasses. Few of them are discussed below:
(a) Switching Voltage (Vt)
12
In both threshold and memory switching, the switching voltage (Vt), also known
as threshold voltage, is defined as the voltage at which the material changes its state from
low conducting OFF state to a high conducting ON state. Different factors affecting the
threshold/switching voltage are thickness of the sample, electrode separation, ambient
temperature, nature of excitation, resistivity of the material, etc.
(b) Holding current (Ih)
The principal difference between a memory and threshold material is the
existence of a holding current (Ih) in a threshold device, below which the ON state cannot
be maintained. When the current falls below this value the threshold switch returns to its
original high resistance OFF state. In contrast, the memory switch retains its conductive
ON state even when the current level approaches zero or reversed.
(c) Switching time (ts)
The switching action in a material takes place when the voltage drop across the
sample exceeds the threshold voltage Vt and the material changes to a low resistive state.
The time taken to attain the low resistance state is commonly referred as the switching
time [98-100].
(d) Delay time (tD)
The time interval between the application of threshold voltage pulse and the
actual initiation of the switching action is known as the delay time (tD). The delay time
decays exponentially with overvoltage (the factor by which the applied voltage exceed the
switching voltage) [101], and attains a value of the order of 10-9 sec when the applied
voltage is 50% higher than the threshold voltage. A typical delay time for a threshold
device is about 10µsec [89,102].
(e) Lock-on time (tlo)
The lock-on time is related to memory switching materials. When a material
switches to the low resistance state, it is required to keep the material in that state for a
minimum period called the lock-on time, to ‘set’ the material in the memory state
[99,100,102]. In case the duration of set pulse is less than ‘td +tlo’ the memory material
will revert back to the high resistance state.
(f) Recovery time (tr)
This time pertains to threshold switching. A definite time is required by a sample
to come back to the high resistance state, after switching, by reducing the current below
holding current (Ih). This time is known as recovery time (tr) and is typically of the order
13
of 1µsec. If the material is switched again within this recovery time, it switches at a lesser
Vt [98-100].
14
(g) Set/ Reset time
The time required to ‘set’ the device from high resistance OFF state to the low
resistance ON state is known as the set time (ts).In case of a memory switching sample, it
is the sum of ‘tD’ and ‘tlo’. The reset time (tr) is relevant for memory switching only. It is
the time required to bring the device back to the high resistance OFF state by applying
suitable voltage/current pulse.
1.4.2.2 Optical Properties
The investigations on the optical properties, particularly the optical gap in

amorphous semiconductors is of particular interest because of their effect on the electrical
properties [103,104]. In general, amorphous films are found to be more transparent than
the crystalline ones [105,106]. It has been observed that the variations of compositions
[107-110] and annealing conditions [111-113] for glassy chalcogenide semiconductors
have a significant influence on the band gap. In general, Further, the studies on the optical
absorption in both the visible and infrared, and the determination of the optical parameters
have found interest in the context of a variety of applications including solar cells, infrared
optics, etc. [114].
Chalcogenide glasses have been found to exhibit the change in refractive index
under the influence of light, which makes it possible to use these materials to record not
only the magnitude but also the phase of illumination. The latter is especially important in
optical data storage applications; the optical memory effects in amorphous semiconductor
films have been investigated and utilized for various applications in recent years [115-
119]. The optical memories based on these materials have many distinct advantages, viz.,
large packing density, mass replication, fast data rate, high signal to noise ratio and high
immunity to defects [115-120]. Further, the use of amorphous chalcogenide films for
reversible optical recording by the amorphous to crystalline phase change has recently
been reported by a number of workers [121-123]. Recently, reversible phase change
optical recording materials such as GeSbTe [124,125] and AgInSbTe [126] have been
developed which have become the focus of attention in optical recording technology,
because of promising applications in digital video disc-random access memory (DVD-
RAM), power disc dual (PD), and compact disk-rewritable (CD-RW). In addition,
amorphous semiconductors are useful in the fabrication of various integrated optical
components and devices such as selective optical filters, mixers, couplers and modulators
[127-129].
15
1.4.2.3 Mechanical Properties
The suitability of glasses for an application often depends upon their mechanical
properties. Various studies have been done, both theoretically and experimentally, to
understand the mechanical behavior of covalent amorphous semiconductors on the basis of
their elastic constants, micro hardness, brittleness etc. [130-133]. Young’s Modulus and
hardness are two of the most important mechanical properties of a material. By definition,
Young's modulus is a measure of the stiffness of an isotropic elastic material. It is defined
as the ratio of the uniaxial stress over the uniaxial strain in the range of stress in which
Hooke's law holds; hardness is a measure of how resistant a solid matter is to various
kinds of permanent shape change when a force is applied. Macroscopic hardness is
generally characterized by strong intermolecular bonds; however the behavior of solid
materials under force is complex and therefore there are different measurements of
hardness such as scratch hardness, indentation hardness, rebound hardness, etc. The
indentation hardness measures the resistance of a sample to a permanent plastic
deformation due to a constant compression load from a sharp object; it is primarily used in
engineering and metallurgy fields.
Instrumented indentation, also known as depth sensing indentation or nano-
indentation, is increasingly being used to probe the mechanical response of different
materials including metals, ceramics, polymers and biological materials. An instrumented
indentation system allows the application of a specified force or displacement, such that
force, P, and the displacement, h, are controlled and/or measured simultaneously and
continuously over a complete loading cycle. Additionally, the extremely small force (≈1
μN) and displacement resolution (≈0.2 nm), are combined with very large ranges of
applied forces and displacements (tens of μN to hundreds of mN or larger in force and tens
of nm to tens of μm or larger in displacement) to allow a single instrument to be used to
characterize nearly all types of material systems [134].
During the instrumented indentation process, a record of the depth of penetration
is made, and then the area of the indent is determined using the known geometry of the
indentation tip. While indenting, various parameters such as load and depth of penetration
can be measured. A record of these values can be plotted on a graph to create a load-
displacement curve (such as the one shown in Figure 1.4). These curves can be used to
extract mechanical properties of the material.
16
Figure 1.4: Schematic of Load-Displacement curve
1.5 Application of Chalcogenides

Chalcogenide glasses are found applications in many fields like phase change
switches, hybrid amorphous/crystalline transistors, thermo switches, xerography,
holography, photo recording, acousto-optical devices, etc. Among them, the most
important application is the phase change switches (also known as Ovonic switches)
discovered by Ovshinsky [20]. Ternary Ovonic glasses with tellurium have been used in
preparing thin film switches such as Si35As30Te35, Ge18As42Te40, Si37As30Te33 and
Si12As24Te64 [135]. Hybrid amorphous/crystalline transistors are another important
application of chalcogenide amorphous semiconductors. This device consists of an
amorphous film of composition Te39As36Si17Ge7P1 as emitter deposited on crystalline Si or
Ge diode. As long as the amorphous emitter in the hybrid transistor is in the OFF state, it
has a very high resistance. Once the voltage between the base and emitter exceeds the
threshold voltage of the amorphous material the emitter switches into the ON state.
Thermo switches are devices which operate as thermistors on the basis of the exponential
dependence of the electric conductivity on temperature.
The chalcogenides have other applications also apart from the above. Lasers
operating at 10.6µm have relied on bulky articulated arms to deliver the power to a micro
manipulator for surgical purposes. These can be replaced by more compact fiber-optic
systems based on Ge-Se-Te/Ge-As-Se, core clad chalcogenide fibers [136]. Silver doped
Ge17Te83 device has found application as a versatile sensor. It can be operated as voltage
17
dependent resistor (VDR) or a temperature dependent resistor (TDR) at lower frequencies
[137]. Further, Ag doped Ge15As10Te75 has been found to be an excellent material for
varactors preferably at higher frequencies (> 10 kHz) [138].
Chalcogenide glasses have been widely exploited in the areas of Xerography and
computer memories [25, 89, 78, 92, 102, 139]. Xerography is the most widely used form
of electro-photography or electrostatic imaging. The heart of the process is a large area of
a thin film photo-conducting element. Typical materials employed as the photoconductor
are chalcogenide glasses such as Se, As2Se3, etc., deposited by vapor condensation on a
metal substrate. These glasses are semiconductors with band gaps of about 2 eV,
transparent in the infrared but highly absorbing the visible light.
Perhaps, the most important application of glassy chalcogenides is the Phase
Change Memory (PCM), which is considered to be the next generation Non-Volatile
Random Access Memory (NVRAM). These memories bridge the gap between non-
volatile Read Only memories (ROMs) and volatile Random Access Memories (RAMs)
and are referred as Read Mostly Memories (or non-volatile RAMs). An important aspect
of chalcogenide memories is that they can be operated by optical means as the ON and
OFF states are optically distinct with large difference in optical properties like
transparency and refractive index. Since light can be focused to a spot size of one micron
or less, it is possible to fabricate optical mass memories with based chalcogenide glasses
[140].
1.6 Phase Change Memory and Chalcogenides: A brief

Overview
The Phase change random access memory (PCRAM) based on rapid reversible
phase change effect in the chalcogenide film has become increasingly important and
attracted much attention as a new technology because of its low cost, low programming
voltage and excellent scalability to nanometer cell size [20,141-144]. In contrast, flash
memory technologies based on floating gate is the most extensively used nonvolatile
memory, which is also increasingly challenged to downscaling beyond the 65nm node.
The Phase change random access memory (PCRAM) (also known as Ovonic
unified memory) is a novel alternative, actively pursued due to its high endurance and
good compatibility with complementary metal-oxide semiconductor (CMOS) technology
18
[145-147]. Furthermore, embedded memory applications using PCRAMs can readily be
achieved due to the simple, planar structure.
The PCRAM is a nonvolatile device that uses a small volume of chalcogenide
material switched between low-resistance polycrystalline and high resistance amorphous
structural phases by resistive heating with programming current pulses. The read/write
performance of the PCRAM is comparable to the conventional dynamic RAM device
[148].
Phase change materials can be of two different types based on the crystallization
process; those which phase segregate and those which do not. The phase segregation
involves diffusion, causing the crystal growth rate to be comparatively lower. The
crystallization process can be speeded up by adding heterogeneous nucleation sites. While
tailoring the crystallization process, an important aspect to be kept in mind is that, the
materials which have very short crystallization time do not re-amorphize easily. However,
easy re-amorphization is needed for re-writing the information. Efforts are being made to
increase the quenching rate by adding a heat sinking layer to the device structure or by
designing a glass which has the same chemical composition in both amorphous and
crystalline state, as diffusion less crystal growth is preferred for direct over-write
capability.
The conventional chalcogenide film used in PCRAM is Ge2Sb2Te5 (GST). The
GST alloy is also used in optical re-writable CD/DVD RW disks [149], in which a laser
beam with different intensity is used to heat a small volume to switch the material between
crystalline and amorphous states. The memory state is determined by the reflectivity of the
memory layer, whereas in silicon based PCRAM, currents of different magnitudes are
passed through the chalcogenide material from a heater element and local joule heating is
used to change the programmable volume around the contact region.
A large writing current is a crucial issue for the commercial application of
PCRAM. In order to reduce the writing current, some approaches from two angles are
proposed, one is reducing the programming volume by improving the cell structure
[150,151] and the other is optimizing the properties of the phase change material, like
increasing the crystalline resistivity of the film etc. Attempts have been made by doping
nitrogen [151] or silicon [152] in GST film to reduce the writing current of phase change
memory [153]. Various other systems also like Si-Sb-Te [154], Ge-Se-Te [155] etc.,
have been studied in order to tailor the performance of a device for the phase change
memory application.
19
Materials that can be reversibly switched from an amorphous to crystalline phase
at an appropriate time scale are composed mainly of elements from groups IV to VI of the
Periodic Table. The Te-based chalcogenide materials have been widely studied because it
becomes vitreous by high speed quenching for the optical memory device material. In this
work also different Te-based binary and ternary chalcogenide glasses are studied in bulk
as well as thin film form for suitability in phase chance memory applications.
1.7 Objectives and Scope of the Present Work

The unique advantage of chalcogenide glasses is the flexibility to prepare them in
a wide range of compositions so that their properties can be tuned for any desired
application. In the present thesis work, efforts have been made to prepare different types
of tellurium based chalcogenide glasses in bulk and thin film form and to study their
electrical, thermal (bulk) and optical (film) properties for phase change memory
application.
Extensive studies on electrical switching and thermal properties have been
undertaken on bulk Ge-Te-Bi and Ge-Te-Sn glassy systems. The role of metallic dopants
(Bi and Sn) in the Ge-Te network and subsequent effect of the same on the electrical and
the thermal properties of the system have been investigated. Further, Al-Te, Ge-Te-
Se, Ge-Te-Si and Ge-Te-Sn systems are studied in amorphous thin film form to
understand their electrical switching behavior and optical properties.
20
References
[1] C. Schultz-Sellack, Ann. Phys. Chem., 139 (1870) 182.
[2] R.W. Wood, Phil Mag., 3 (1902) 607.
[3] W. Meier, Ann. Phys., 31 (1910) 1017.
[4] R. Frerichs, J. Opt. Soc. Amer., 43 (1953) 1153.
[5] W.A. Fraser, J. Opt. Soc. Amer., 43 (1953) 823.
[6] F.W Glaze, D.H. Blackburn, J.S. Osmalov, D. Hubbard and M.H. Black, J. Res.
N.B.S., 59 (1957) 83.
[7] B.T. Kolomiets and T.F. Nazarova, Sov. Phys. Solid State, 2 (1960) 369.
[8] N.A. Goryunova, V.P. Shylo and B.T. Kolomiets, Sov. Phys. Tech. Phys., 3 (1958)
912.
[9] S.S. Flaschen, A.D. Pearson and W.R. Northover, J. Amer. Ceram.Soc., 42 (1959)
450.
[10] T.F. Nazarova and B.T. Kolomiets, Fiz. Tverd. Tela., 2 (1959) 22.
[11] S.S. Flaschen, A.D. Pearson and W.R. Northover, J. Amer. Ceram.Soc., 43 (1960)
274.
[12] A.R. Hilton and M. Brau, Infrared Phys., 3 (1963) 69.
[13] A.R. Hilton, C.E. Jones and M. Brau, Infrared Phys., 4 (1964) 213.
[14] A.D. Pearson, Modern aspects of the Vitreous State, Vol 3( Edited by J.D.
Mackenzie, Butterworths), London, 1964.
[15] S.S. Flaschen, A.D. Pearson and W.R. Northover, J. Appl. Phys., 31 (1960) 219.
[16] A.D. Pearson, W.R. Northover, J.F. Dewald and W.F. Peck, Advances in Glass
Technology, Part I, Plenum, New York, 1962.
[17] F.C. Lin and S.M. Ho, Amer. Ceram. Soc., 46 (1963) 24.
[18] R. Meyerowitz, Amer. Mineral, 40 (1955) 398.
[19] A.G. Fischer and A.S. Mason, J. Opt. Soc. Amer., 52 (1962) 721.
[20] S.R. Ovshinsky, Phys. Rev. Lett., 21 (1968) 1450.
[21] C.A. Angell, Proc. Natl. Acad. Sci. USA, 92 (1995) 6675.
[22] A. Cooper, J. Non-cryst. Solids, 95 &96 (1987) 1-14.
[23] G.W. Scherer, Relaxation in Glass and Composites, Wiley, New York, 1986.
[24] S.A. Brawer, Relaxation in Viscous Liquids and Glasses, American Ceramic Society,
Columbus, OH, 1985.
[25] Richard Zallen, The Physics of Amorphous Solids, John Wiley and Sons, Inc.1983.
21
[26] W. Kauzmann, Chem. Rev., 43 (1948) 219.
[27] R. Parthasarathy, K.J. Rao and C.N.R Rao, Chem. Soc. Rev., 12 (1983) 361.
[28] D. Turnbull, Contemp. Phys., 10 (1969) 473.
[29] S. R. Elliott, Material Science and Technology: A comprehensive treatment, 9, VCH
Publishers Inc, New York, 1991.
[30] W.H. Zachariasen, J. Amer. Chem. Soc., 54 (1932) 3841.
[31] A.G. Smekal, J. Soc. Glass Tech., 35 (1951) 411.
[32] J.C. Phillips, J. Non-Cryst. Solids, 34 (1979) 153.
[33] A. Hruby, Czech. J. Phys. B, 32 (1972)1187.
[34] S.R. Elliot, The Physics of Amorphous Materials, Longman, London, 1984.
[35] A.C. Wright, Adv. Struct. Res. Diff. Meth. 5 (1974) 1.
[36] A.C. Wright and A.J. Leadbetter, Phys. Chem. Glass, 17 (1976) 122.
[37] S.R. Elliot, Phys. Rev. Lett., 67 (1991) 711.
[38] S.R. Elliot, Nature, 354 (1991) 445.
[39] A.B. Bhatia and D.E. Thornton, Phys. Rev. B, 2 (1970) 3004; A.C. Wright, R.N.
Sinclair and A.J. Leadbetter, J. Non-Cryst. Solids, 71 (1985) 295.
[40] S.R. Elliot, J. Non-Cryst Solids, 123 (1990) 149.
[41] J. Wong and C.A. Angell, Glass: Structure by Spectroscopy, Dekker, New York,
1976.
[42] S.S. Mitra, Physics of Structurally Disordered Solids, Plenum, New York, 1976.
[43] M. Tenhover, R.D. Boyer and R.S. Henderson, Solid State Commun. 65 (1988) 1517.
[44] H. Eckert, Encyclopedia of NMR, 2, (1995) 756.
[45] P.E. Stallworth, Glass Science and Technology (edited by D.R. Uhlmann and N.J.
Kreidil), Academic, New York, 1990.
[46] H. Eckert, Adv. Mater., 28 (1989) 1723.
[47] N.F. Mott, Phil. Mag., 19 (1969) 835.
[48] J.C. Maxwell, Phil. Mag., 27 (1864) 299.
[50] M.F. Thorpe, J. Non-Cryst. Solids, 57 (1983) 355.
[51] J.C. Phillips and M F Thorpe, Solid State Commun., 53 (1985) 699.
[52] J.C. Phillips, Phys. Rev. B, 31 (1985) 8157.
[53] A. Feltz, H. Aust and A. Blayer, J. Non-Cryst. Solids, 55 (1983) 179.
[54] R.A. Street, R.J. Nemanich and G A N Connell, Phys. Rev. B, 18 (1978) 6915 19.
[55] K. Tanaka, J. Non-Cryst. Solids, 103 (1988) 149.
22
[56] K. Tanaka, Phys. Rev. B, 39 (1989) 1270 9.
[57] C.N. Murthy, V. Ganesan and S. Asokan, Appl. Phys. A, 81 (2005) 939.
[58] S. Prakash, S. Asokan and D.B. Ghare, J. Phys. D: Appl. Phys., 29 (1998) 2004.
[59] S. Murugavel and S. Asokan, Phys. Rev. B, 55 (1998) 3022.
[60] R. Aravinda Narayanan, S. Asokan and A. Kumar, Phys. Rev. B, 54 (1996) 4413.
[61] P. Boolchand, D.G. Georgiev and B. Goodman, J. Optoelectron. Adv. Mater., 3
(2001) 703.
[62] P. Boolchand et al., C.R. Chemie, 5 (2002) 713.
[63] Y. Wang, M. Nakamura, O. Matsuda and K. Murase, J. Non-Cryst. Solids, 266-269
(2000) 872.
[64] Xingwei Feng, W.J. Bresser and P. Boolchand, Phys. Rev. Lett., 78 (1997) 4422.
[65] D. Selvanathan, W.J. Bresser and P. Boolchand, Phys. Rev. B, 61 (2000) 15061.
[66] D. Georgiev, P. Boolchand and M. Micoulaut, Phys. Rev. B, 62 (2000) R9228.
[67] M. Tatsumisago et al., Phys. Rev. Lett., 64 (1990) 1549.
[68] S. Chakravarty, D.G. Georgiev, P. Boolochand and M. Micoulaut, J. Phys.:Condens.
Matter, 17 (2005) 1.
[69] G. Lucovsky and T.M. Hayes, Amorphous Semiconductors, (edited by M.H.
Brodsky), Springer-Verlag, Berlin, 1979.
[70] N.F. Mott and E.A. Devis, Electronic Processes in Non-Crystalline Materials,
Clarendon, Oxford, 1979.
[71] R.M. White, J. Non-Cryst. Solids, 16 (1974) 387.
[72] R. Azoulay, H. Thibierge and A. Brenac, J. Non-cryst. Solids, 18 (1975) 33.
[73] R. Aravinda Narayanan, PhD thesis, Indian Institute of Science, India (2000)
[74] G. Lucovsky, F.L. Galeener, R.H. Giels and R.C. Keezer, Proc. Conf. on Structure
of Non-Cryst. Materials (edited by P.H. Gaskell and E.A. Davis), Taylor and Fransis,
London, 1977.
[75] G. Lucovsky and R.M. Martin, J. Non-Cryst. Solids, 8-10 (1972) 185.
[76] P.C. Taylor, S.G. Bishop and D.L. Mitchell, Solid State Commun., 16 (1975) 167.
[77] G. Lucovsky, R.J. Nemanich, S.A. Solin and R.C. Keezer, Solid State Commun., 17
(1975) 1567.
[78] R. Neale, J.A. Aseltine, IEEE Trans. Electron Devices, 20 (1973) 195.
[79] R.W. Pyros, Disordere and Order in Solid State: Concepts and Devices, Plenum
Press, New York, 1989.
[80] D. Adler, Amorphous Semiconductors, CRC, Butterworths, London, 1971.
23
[81] J. Stuke and W. Brenig (Editors), Amorphous and Liquid Semiconductors, Taylor
and Fransis, London, 1974.
[82] R.B. Laibowitz, S. R. Herd and P. Chaudhari, Phil. Mag., 28 (1973) 1155.
[83] T. Takamori, R. Roy and G.J. McCarthy, Mat. Res. Bull., 5 (1970) 529.
[84] S. Prakash, S. Asokan and D.B. Ghare, Semicond. Sci. Tech. 9 (1994) 1484.
[86] R. Chatterjee, S. Asokan and S.S.K. Titus, J. Phys. D: Appl. Phys., 27 (1994) 2624.
[87] J.T. Devaraju, B.H. Sharmila, K.V. Acharya and S. Asokan, Phil. Mag. B, 81 (2001)
583.
[88] N. Manikandan and S. Asokan, Phil. Mag., 87 (2007) 5109.
[89] H. Fritzsche, Switching and Memory in Amorphous Semiconductors, Plenum Press,
New York, 1974.
[90] M. Anbarasu, S. Asokan, Sudakshina Prusty and A. K. Sood, Appl. Phys. Lett., 91
(2007) 093520.
[91] P. Pattanayak, N. Manikandan, M. Paulraj and S Asokan, J. Phys.: Condens Matter,
19 (2007) 036224.
[92] D. Adler, Sci. Amer., 236 (1977) 36.
[93] H.K. Henisch, Sci. Amer., 221 (1969) 30.
[94] H.K. Henisch, E.A. Fagen and S.R. Ovshinsky, J. Non-Cryst. Solid., 2 (1970) 538.
[95] I. Lucas, J. Non-Cryst. Solid., 6 (1971) 136.
[96] A.C. Warren, Electron Lett., 5 (1969) 461.
[97] J.C. Male, Electron Lett., 6 (1970) 91.
[98] H. Fritzsche, Electronic and Structural Properties of Amorphous Semiconductors,
(edited by P.G. Lecomber and J. Mort), Academic Press, New York, 1973.
[99] A.E. Owen and J.M. Robertson, IEEE Trans. on Electron Devices, 20 (1973) 105.
[100] A.C. Warren, IEEE Trans. on Electron Devices, 20 (1973) 123.
[101] G.V. Bunton and R.M. Quillam, IEEE Trans. on Electron Devices, 20 (1973) 140.
[102] S.R. Ovshinsky and H. Fritzsche, IEEE Trans. on Electron Devices, 20 (1973) 91.
[103] S.K.M. Dehaldhar. and S.P. Sengupta, Ind. J. Pure, Appl. Phys. 17 (1979) 422.
[104] W. Beyer, H. Mell, and J. Stuke, Phys. Status Solidi B, 45 (1971) 153.
[105] L. Zdanowicz and B. Jahzabek, J. Non-crystalline Solids, 97/ 98 (1987) 1203.
[106] L. Bdanovic and T. Kwiecien, Vacuum, 27 (1977) 409.
[107] N. Tohge, T. Minami, Y. Yamamoto and M. Tanaka, J. Appl. Phys, 51 (1980) 1048.
[108] B.D. Muragi and J.K. Zope, J. Mat. Sci. Lett., 3, (1984) 633.
24
[109] N. Tohge, T. Minami and M. Tanaka, J. Non-cryst. Solids, 37 (1980) 23.
[110] M. Szukwei, Y. Hanmei and C. Zongcai., J. Non-cryst. Solids, 52, (1982) 131.
[111] S. Chaudhuri, S.K. Biswas and A .Choudhury, J. Mat. Sci., 23 (1988) 4470.
[112] M.Di. Giulio, D. Manno, R. Rella, P. Siciliano and A. Tepore, Solar Energy Mat, 15,
(1987) 209.
[113] M. M. Wakkad., J. Phys. Chem. Solids, 51, (1990) 1171.
[114] D.M. Petrovic, S.R. Lukic, M.J. Avramov, and V.V. Khiminets, J. Mater. Sci. Lett.,
5 (1986) 290.
[115] J. Feinleib, J. DeNeufville, S.C. Mose and S.R. Ovshinsky, Appl. Phys. Lett., 18
(1971) 254.
[116] K.A. Rubin and M. Chen, Thin Solid Films, 181 (1989) 129.
[117] S.Y. Suh, D.A. Snyder and D.L. Anderson, Appl. Opt., 24 (1985) 868.
[118] P.F. Carcia, F.D. Kalk, P.E. Bierstedt, A. Ferretti, G.A. Jones and D.G. SwarizFager,
J. Appl. Phys., 64 (1988) 1715.
[119] S. Zembutsu, Y. Toyoshima, T. Igo and H. Nagai, Appl. Opt., 14 (1975) 3073.
[120] Y. Maeda, H. Andoh, I. Ikuta and H. Minemura, Appl. Opt., 64 (1988) 1715.
[121] D.P. Gosain, T. Shimizu, M. Ohmura, M. Suzuki, T. Bando and S. Okano, J. Mater.
Sci., 26 (1991) 3271.
[122] N. Sharo and A.K. Chakravarti, Optical memories: Technologies and applications,
September,1991
[123] D.J. Gravesteijn, Appl. Opt., 27 (1988) 736.
[124] N. Yamada, E. Ohno, N. Akahira, K. Nishiuchi, K. Nagata and M. Takao, Proc.
Int.Symp. Optical Memories, Jpn. J. Appl. Phys., 26 (1987) 61.
[125] K. Nishimura, M. Suzuki, I. Morimooto and K. Mori, Jpn. J. Appl. Phys., 28 (1989)
135.
[126] H. Iwasaki, Y. Ide, M. Harigaya, Y. Kageyama and I. Fujimura, Jpn. J. Appl. Phys.,
31 (1992) 461.
[127] A.M. Andriesh, V.V. Ponomar, V.L. Smirnov and A.V. Mironos, Sov. J. Quamtum
Elect., 16 (1986) 721.
[128] I.V. Kityk, E. Golis, J. Filipecki, J. Wasylak and V.M. Zachark, J. Mater. Sci. Lett.,
14 (1995) 1292.
[129] J. Wasylak, J. Kucharski, I.V. Kityk and B. Sahraui, J. Appl. Phy., 85 (1999) 425.
[130] H. He and M.F. Thorpe, Phy. Rev. Lett., 54 (1985) 2107.
[131] K. Tanaka, Solid State Commun., 60 (1986) 295.
25
[132] A.K. Varshneya and D.J. Mauro, J. Non-Cryst. Solids, 353 (2007) 1291.
[133] A.N. Sreeram, A.K. Varhneya and D.R. Swiler, J. Non-Cryst. Solids, 130 (1991)
225.
[134] M. R. VanLandingham, J. Res. Nat. Inst. Stand. Technol., 108 (2003) 249.
[135] Mihai Popescu, Non-Crystalline Chalcogenides, Kluwer Academic Publisher, The
Netherlands, 2001.
[136] J. Nishii, S. Morimoto, I. Ingawa, R. Iizuka, T. Yamashita and T. Yamagishi, J. Non-
Cryst. Solids, 140 (1992) 199.
[137] A. Vohra and K.K. Srivastava, Sensors and Actuat. A, 24 (1990) 151.
[138] K.K. Srivastava and P. Khrunna, Proc. 17th Yugoslavia Conf.
Microelectronics’MIEL 89’, 153, 1989.
[139] C.L. Dargan, P. Burton and R.M. Redstall, Electronics, 38 (1975) 711.
[140] S.R. Ovshinsky, J. Non-Cryst. Solids, 141 (1992) 200.
[141] D. Adler, J. Vac. Sci. Technol., 10 (1973) 728.
[142] S. Lai, and T. Lowrey, IEEE IEDM Technol.Dig., 803, 2001.
[143] S. Lai, IEEE IEDM Technol.Dig., 225, 2003.
[144] Pirovano, A.L. Lacaita, A. Benvenuti, F. Pellizer, S. Hudgens and R. Bez, IEEE
IEDM Technol. Dig. 699, 2003.
[145] M. Wuttig, Nat. Mater., 4 (2005) 265.
[146] R. Bez and A. Pirovano, Mater. Sci. Semicond. Process, 7 (2004) 349.
[147] A.L. Lacaita, Solid-State Electron, 50 (2006) 24.
[148] H. Hrii et al., Tech Dig. VLSI Symp., 173, 2003.
[149] N. Yamada, E. Ohno, K. Nishiuchi, N. Akahira and M. Takao, J. Appl. Phys., 69
(1991) 2849.
[150] H.A Ha et al., Tech Dig. VLSI Symp., 175, 2003.
[151] Y.N. Hwang et al., Tech Dig. VLSI Symp., 173, 2003.
[152] B.W. Qiao, J. Feng, Y.F. Lai, Y. Ling, Y.Y. Lin, T.A. Tang, B.C. Cai and B. Chen,
Appl. Surf. Sci., 252 (2006) 8404.
[153] J. Feng, Y. Zhang, B.W. Qiao, Y.F. Lai, Y.Y. Lin, B.C. Cai, T.A. Tang and B. Chen,
Appl. Phys. A, 87 (2007) 57.
[154] B.W. Qiao, J. Feng, Y.F. Lai, Y. Cai, Y.Y. Lin, T.A. Tang, B.C. Cai and B. Chen,
Semicond. Sci. Technol., 21 (2006) 1073.
[155] H.B. Chung, K. Shin and J.M. Lee, J. Vac. Sci. Technol. A, 25 (2007) 48
26
Chapter 2
Experimental Techniques
2.1 Introduction
The chalcogenides generally have high vapor pressures around melting
temperature [1] and they have a tendency to react with oxygen especially at higher
temperature. Therefore, chalcogenide glasses are normally prepared by melting the
constituent high purity elements in quartz ampoules sealed under high vacuum conditions.
The conditions for preparing glasses depend on glass compositions, glass forming ability
etc. In this chapter, the methods adopted for synthesizing bulk and thin film chalcogenide
glassy samples and their characterization are discussed briefly. Also, the details of
experimental procedures used for this thesis work, namely electrical switching,
Alternating Differential Scanning Calorimetry, Optical Spectroscopy etc., are outlined.
2.2 Synthesis of Chalcogenide Glasses
2.2.1 Bulk Glass preparation

The bulk chalcogenide glassy samples are prepared by vacuum sealed melt
quenching technique. Appropriate quantities of 99.999% pure constituent elements
weighed to 0.1 mg accuracy are taken in flattened quartz ampoules of 5 mm inner
diameter and 7 mm outer diameter. As the glass formation is easier for samples with lower
mass, the total weight of the constituent elements is kept at around 1gm. The ampoules
containing the constituent elements are sealed under a vacuum of 10-5 mbar and, slowly
heated in a horizontal rotary furnace (figure 2.1) up to 950 0C at the rate of 100 0C/hr. The
ampoules are maintained at this temperature for 36 hours under rotation (10 rpm) to ensure
homogeneity of the melt. The ampoules are subsequently quenched in ice-cold water
mixed with NaOH to obtain bulk glasses. The amorphous nature of the quenched samples
is confirmed by X-ray diffraction.
Ch.2: Experimental Techniques…
Figure 2.1: Schematic diagram of the rotary furnace used to prepare bulk glasses;
1 - Outer metal casing; 2 - asbestos powder; 3 - alumina tube; 4 - motor arrangement for
rotation; 5 - ampoule.
2.2.2 Thin film coating
The bulk ingots obtained from conventional melt quenching technique are then
milled to a powder. Two different types of deposition techniques namely, resistive
evaporation and flash evaporation are utilized for coating electrodes and chalcogenide
films respectively. These two methods of coating are briefly described below:
(i) Resistive evaporation technique:
The schematic diagram of a resistive evaporation system is shown in figure 2.2.

Typically, it consists of a thermally heated evaporation source to vaporize the desired
material. The substrates are located at an appropriate distance facing the evaporation
source. The substrate holder is heated using a power supply. Evaporation is carried out in
vacuum with pressure ranging from 10-5 to 10-8 mbar. In this pressure range, the mean free
path (MFP) is very large (500-105 cm) as compared to the source to substrate distance.
Hence, the evaporated atoms undergo collision less line of sight transport prior to
condensation on the substrate [2]. In this thesis work aluminum electrodes are coated
using this technique.
28
Figure 2.2: Schematic diagram of conventional resistive evaporation technique.
(i) Flash Evaporation Technique:
In the present work, flash evaporation is used to deposit different Te based

amorphous films. It is known that the vaporization of compounds is usually accompanied
by dissociation or association or both processes. The association processes help to
maintain the stoichiometry of the constituents, whereas, the dissociation processes lead to
non-stoichiometry especially if one of the constituents is less volatile than the other.
Consequently, the composition of the vapor condensed on the substrate is not the same as
that of the source material [3]. Hence, direct resistive thermal evaporation method is not
suitable and flash evaporation is preferred in systems like amorphous chalcogenides, in
which the constituents have different vapor pressures.
In the flash evaporation technique, the desired film composition is accomplished

by complete evaporation of a series of small micro-particulate quantities of the
constituents in the desired ratio. Only one filament is used to keep the boat at sufficiently
high temperature to evaporate the least volatile constituent among the desired film
material. Thus the net result of this simultaneous, discrete evaporation is a vapor stream
whose composition is uniform and identical to that of the source material. A necessary
condition to get an amorphous film is to keep the temperature of the substrate at liquid
nitrogen temperature. In the present thesis work, a custom built flash evaporation unit has
been used which is shown in figure 2.3. Thin film devices are coated by flash evaporation
technique in two different geometries, namely co-planar and sandwich, on a 25mm ×
75mm glass substrate. Few patterns are shown in figure 2.4.
29
Figure 2.3: A photograph of custom built flash evaporation Unit.

A view of inside chamber (top).
30
(a)
(b)
Figure 2.4: Different patterns of devices in (a) Coplanar geometry and (b) Sandwich
geometry.
2.3 Thickness Measurement

Veeco Dektak 150 surface profilometer has been used for determining thickness of
the films in this thesis work. Few important advantages of the profilometer are listed
below. It
• delivers programmable X-Y wafer positioning up to 150 mm
31
• accommodates samples up to 90 mm thick
• performs long scans of 55 mm
• has best standard Z performance
• has quick stylus exchange procedure eliminates risk of damage to sensor
• has 2X faster 3D maps with new enhanced software
The thickness of the films deposited has been measured by the profilometer to be in the
range of 100-350 nm.
2.4 Electrical Switching Studies

In this work, electrical switching studies of the glasses have been undertaken on
bulk Ge18Te82-xBix (1 ≤ x ≤ 4), Ge15Te85-xSnx (1 ≤ x ≤ 6) and Ge17Te83-xSnx (1 ≤ x ≤ 5)
glasses and also on GexSe35-xTe65 (17 ≤ x ≤ 23), Ge15Te85-xSix (1 ≤ x ≤ 6), Ge15Te85-xSnx (1
≤ x ≤ 5) and Ge17Te83-xSnx (1 ≤ x ≤ 4) amorphous thin films.
The I-V (current-voltage) characteristics can be generally studied either by

sourcing current and measuring voltage or vice versa. In this thesis, for studying the I-V
characteristics, a constant current is passed and the resulting voltage is measured [4]. For
Ohmic samples, any of the two options will produce the same result. However, in
amorphous semiconductors, current increases Ohmically up to a threshold voltage Vt. At
Vt, the resistance drops drastically or gradually indicating a distinct negative resistance
region. If a voltage source is used directly across the sample, the reduction in resistance
leads to a very high current. Therefore a series resistance is required to limit the current
flowing through the sample. On the other hand, no such precaution is needed if a constant
current source is used and also, the complete information about the curve is obtained.
The schematic of the sample cell used for electrical switching studies is shown in
figure 2.5. In the cell, samples polished to the required thickness are mounted in a brass
holder, in between a flat bottom electrode and a point contact top electrode using a spring
loading mechanism. Further, provisions are made in the cell, to study the temperature
dependence of switching voltages by placing a heater in contact with the bottom electrode;
a Chromel-Alumel (K-type) thermocouple is placed adjacent to the sample for measuring
the sample temperature. A Keithley Source-Measure Unit (Model 2410c), which can
deliver a maximum current of 20 mA at a compliance voltage of 1100V, is used for
32
measuring the I-V characteristics. The current sourcing is controlled by LabVIEW 7 (NI)
through a PC. Since switching voltage (Vt) is known to depend on thickness, samples
polished to the required thickness, in the range 0.25 mm to 0.3 mm are used. The thickness
of the samples is determined using a screw gauge of least count 0.01 mm.
Figure 2.5: Schematic diagram of the sample cell for electrical switching studies of bulk
glasses.
2.5 Scanning Calorimetry

Calorimetric techniques are widely used for thermal measurements. Generally,
calorimetric measurements provide information about the nature of the reaction
(exothermic or endothermic) and the heat capacity variations of any chemical or physical
reaction which involves transfer of heat and also a measure of difference in heat content,
in terms of heat capacity or heat of transition.
2.5.1 Differential Scanning Calorimetry
The Differential Scanning Calorimetry (DSC) has extensively been used in

industries and research for the determination of glass transition & crystallization
temperatures, phase transformations, heat capacity, melting point of crystals etc. [5]. This
is one of the calorimetric techniques in which the heat-flow rate to the sample is monitored
while the temperature of the sample is varied in a specified atmosphere at a given rate. The
thermal parameters of the sample enclosed in an aluminum pan (S), are measured in
comparison with that of a reference (R), both placed in the DSC cell. It can be operated
either in the heat flux mode or the power compensated mode, depending on the heating
arrangement [6]. In the heat flux DSC, the S and R are heated using a single heating
33
source whereas in a power compensated DSC, the sample and the reference are heated
separately. The schematic diagram of the measurement cell of a heat flux DSC instrument
is shown in figure 2.6.
The DSC suffers from few drawbacks like insufficient sensitivity, inadequate
resolution, requirement of multiple experiments to obtain heat capacity, lack of ability to
analyze complex transitions, etc.
Figure 2.6: The cross sectional diagram of the cell in a heat-flux DSC: 1 -dynamic
sample chamber, 2 - lid, 3 - purge gas inlet, 4 - constantan thermoelectric disc, 5-
chromel disc, 6 - heating block, 7 – chromel wire, 8 - alumel wire, and 9 - thermocouple
junction. S and R are the sample & reference pans respectively.
2.5.2 Alternating Differential Calorimetry
To overcome the problems of the conventional DSC, a modified version of DSC

called the Alternating differential Scanning Calorimetry (ADSC) or Temperature
Modulated Differential Scanning Calorimetry (TMDSC) has been employed [7,8]. Like
DSC, ADSC measures the difference in heat flow between the sample and reference.
However, the ADSC provides the possibility of deconvoluting the total heat flow rate into
a part that tracks the temperature modulation (known as the reversing heat flow rate),
leaving a part that does not track the temperature modulation (known as the non-reversing
heat flow rate). This is equivalent to two experiments being run simultaneously- one at
linear heating rate and other at sinusoidal heating rate. The heating profile of a ADSC
curve is shown in figure 2.7 [9].
The few advantages of ADSC over conventional DSC are, increased sensitivity to
detect weak transition, increased resolution without loss of sensitivity, an order of
magnitude reduction in scan rate (1-3 0C/min), direct measurement of heat capacity and
34
heat flow in a single experiment, separation of complex transition into a simpler ones, etc.
Recently, the composition dependence of thermal parameters of chalcogenide glasses
obtained by ADSC has revealed many interesting aspects of microscopic structure and the
local network connectivity such as the thermally reversing window. This aspect is deal
with in more detail in the subsequent chapters [10,11].
Figure 2.7: Typical modulated temperature profile of ADSC. The dashed line indicates
the underlying linear ramp.
2.6 X-Ray Diffraction Studies

In this work, the amorphous nature of the samples in both bulk glassy and thin film
form has been confirmed using a Philips powder diffractometer and a Bruker D8 Advance
diffractometer. The basic operating principle and analysis of XRD is well understood in
literature [12]. The crystalline nature of a material is studied from the diffraction peak
obtained from XRD. These peaks correspond to the Bragg reflection belonging to a
particular family of planes. The basic Bragg equation, which relates the lattice spacing and
the observed 2θ in a diffraction pattern, is given below:
2𝑑ℎ𝑘𝑙 sin 𝜃 = 𝑛𝜆
Where dhkl is the lattice spacing of a particular [h k l], θ is the glancing angle, λ is
the wavelength of X-ray source and n is the order of diffraction. In a typical
experiment, the diffraction pattern is recorded by scanning 2θ from 100 to 800 at a
scanning rate of 20/min with CuKα radiation (λ = 1.541 Å).
The diffraction pattern of thermally annealed samples has also been obtained after
annealing them for 1-2 hours at their respective crystallization temperature. The lattice
35
spacing and the unit cell parameters of the annealed samples are obtained from observed
diffraction pattern and matched with JCPDS data.
2.7 Scanning Electron Microscopy (SEM)

The Scanning Electron Microscopy is the one of the important tools for studying the micro
structural features of a material. The basic operating principle of a SEM is based on
scanning of a finely focused electron beam over the surface of a specimen [13]. Electrons
thermionically emitted from a cathode filament are drawn to an anode, focused by two
successive condenser lenses into a beam with a very fine spot size (~ 50 Ǻ). Pairs of
scanning coils located at the objective lens deflect the beam either linearly or in raster
fashion over a rectangular area of the specimen surface. Upon impinging on the specimen,
the primary electrons decelerate and in losing energy transfer it inelastically to other
atomic electrons and to the lattice. Through continuous random scattering events, the
primary beam effectively spreads with a multitude of electron excitations.The result is a
distribution of electrons that manage to leave the specimen with an energy spectrum. In
addition, target X-rays are emitted, and other signals such as light, heat and specimen
current are produced, and the sources of their origin can be imaged with appropriate
detectors. The beam current in a SEM is typically 10-12 to 10-7 A, with a beam energy of
5-50 keV.
In this thesis work, SEM images of the switched and un-switched region of the
bulk and thin films samples have been obtained from a Quanta LV/ ESEM scanning
microprobe machine. A high voltage of 15 kV has been applied for both the samples. The
working distance has been maintained about 11.0 mm during the SEM investigations.
2.8 Energy - Dispersive X-ray Analysis (EDAX)
In energy-dispersive X-ray analysis system, electron beam serves to excite

characteristic X-rays from the area of the specimen being probed. Most of the EDAX
systems are interfaced to SEMs. A liquid-nitrogen Dewar with its cooled Si(Li) detector is
attached to the SEM column, aimed to efficiently intercept emitted X-rays. The Si(Li)
detector is a reverse biased Si diode doped with Li to create a wide depletion region. An
incoming X-ray generates a photo electron that eventually dissipates its energy by
creating electron-hole pairs. The incident photon energy is linearly proportional to the
number of pairs produced or equivalently proportional to the amplitude of the voltage
36
pulse they generate when separated. The pulses are amplified and then sorted according to
the voltage amplitude by a multi channel analyzer, which also counts and stores the
number of pulses within given increments of the voltage (energy) range. Si(Li) detectors
typically have a resolution of about 150 eV, so overlap of peaks occurs when they are not
separated in energy by more than this amount.
In this thesis work EDAX analysis of the thin films are done for compositional analysis.
The data is obtained from same Quanta LV/ESEM scanning microprobe machine as
mentioned above.
2.9 Optical Spectroscopy

The absorption spectra of different compositions of the materials have been taken
with an Ocean optics high-resolution spectrophotometer (Model HR 4000,) in the range
of 200 nm to 1000 nm to calculate the band gap of the different compositions of the
material. The spectrometer has a 3648-element CCD-array detector that enables optical
resolution as precise as 0.02 nm (FWHM). The HR4000 is responsive from 200-1100 nm,
but the specific range and resolution depend on grating and entrance slit choices. This
novel combination of optics and electronics is ideal for applications such as characterizing
lasers, measuring gas absorbance, and determining atomic emission lines.
The optical band gap has been estimated from absorption coefficient data as a
function of wave length by using Tauc Relation [14].
αhυ = B( hυ - Eg )m
Where hυ is the energy of the incident photon, B is a constant which depends on

the material and nature of transition, Eg represents the band gap energy and m in an
exponent whose value is characteristic of the transition. m may take different values like
1/2, 3/2, 2 or more depending on whether the transition is direct or indirect and allowed or
forbidden. In the direct and allowed cases m = 1/2, whereas for the direct but forbidden
cases it is 3/2. In case of indirect allowed transition, m = 2 and for the forbidden case it
will be 3 or more.
2.10 Nano-indentation
Nano-indentation refers to depth-sensing indentation testing in the sub-micrometer
range and has been made possible by the development of
37
1) instruments that can make nanometric indentations while recording load
and displacement with very high accuracy and precision
2) analysis models by which the load displacement data can be interpreted to

obtain hardness, modulus, and other mechanical properties.
The idea of nano-indentation arose from the realization that an indentation test is
an excellent way to measure very small volumes of materials. In principle, if a very sharp
tip is used, the contact area between the sample and the tip, and thus the volume of
material that is tested, can be made arbitrarily small. The only problem is determining the
indentation area, for which a powerful microscope such as an Atomic Force Microscope is
required. To solve this problem, depth sensing indentation methods are developed. In this
method, the load and displacement of the indenter are recorded during the indentation
process and these data are analysed to obtain the contact area, which provide mechanical
properties, without having to view the indentations.
A nano-indenter (Triboindenter of Hysitron) was used to carry out the experiments

with an in-situ imaging capability. The machine continuously monitors and records the
load, P, and displacement, h, of the indenter with force and displacement resolutions of 1
nN and 0.2 nm, respectively. A Berkovich tip (three-sided pyramidal diamond tip) of
diameter 100 nm is used for nano-indentation. In order to identify flat regions for the
reliable and repeatable experiment, the pellet surfaces are imaged prior to indentation
using the same indenter tip. A peak load of 8 mN is employed with a loading and
unloading rate of 1 mN/s and a hold time of 10 s. A minimum of 10 indentations are
performed on each sample. The indentation impressions are captured immediately after
unloading, so as to avoid time-dependent elastic recovery of the residual impression.
38
References
[1] D. Lezal,et al., J. Non-cryst. Solids, 326 & 327 (2003) 47.
[2] J.L. Vossen and W. Kern, Thin Film Processes II, Academic Press, Inc. New York,
1991.
[3] L. Maissel and R. Glang, Hand Book of Thin Films McGraw-Hill, New York, 1983.
[4] S. Murugavel, Local Structure and Transport Properties of Chalcogenide Glasses,

PhD, Thesis, Indian Institute of Science, Bangalore, 1998.
[5] Hobart H. Willard, Lynne L. Merritt, Jr., John A. Dean and Frank A. Settle, Jr.,
Instrumental Methods of Analysis, CBS publishers and distributors, 1986.
[6] G.W.H. Hohne, W. Hemminger and H. Flammershem, Differential Scanning

Calorimetry, Springer-Verlag, Berlin,1996.
[7] P.S. Gill, S.R. Sauerbrunn and M. Reading, J. Therm. Anal., 40 (1993) 931.
[8] M. Reading, D. Elliot and V.L. Hill, J. Therm. Anal., 40 (1993) 949.
[9] S.L. Simon, Thermochim. Acta, 374 (2001) 55.
[10] P. Boolchand, D.G. Georgiev and B. Goodman, J. Optoelectron. Adv. Mater., 3

(2001) 703.
[11] U. Vempati and P. Boolchand, J. Phys.: Condens. Matter., 16 (2004) S5121.
[12] B.D. Cullity, Elements of X-ray diffraction, Addison-Wesley publishing company,

Inc.1978.
[13] M.H. Loretto, Electron Beam Analysis of Materials, Chapman and Hall, NewYork,
1984.
[14] J. Tauc, Amorphous and Liquid Semiconductor, Plenum Press, New York, 197
39
Chapter 3
Electrical Switching Behavior of Amorphous Al23Te77

Thin Film Sample
3.1 Introduction
In the group of amorphous solids, semiconducting chalcogenides are interesting
due to their applications in optoelectronics [1], infrared optical fibers [2], solar cells [3],
optical recording systems [4], Phase Change Memories (PCM), etc. The absence of long-
range order in these materials provides the convenience of changing the elemental ratio
and hence the properties over a wide range.
Several tellurium based binary amorphous semiconductors such as As–Te, Ge–Te,

Si–Te, In–Te and Ga–Te have been found to exhibit memory type of switching in their
bulk form, at electrical fields of the order of 105 V/cm [5–12]. Ge–Te–Cu, Ge–Te–Ag, and
Ge–As–Te are examples of ternary tellurium based glasses which exhibit memory
switching [13,14]. Apart from that, studies also have been done on tellurium based glasses
like As–Te [15], Ge–Te [16] to know the composition dependence of various properties
and the effect of metallic doping on these properties.
Earlier investigations indicate that bulk, glassy AlxTe100-x samples [17], exhibit
memory type electrical switching. This chapter deals with the deposition of amorphous
thin films of Al23Te77 by flash evaporation method and studies on the I–V characteristics
and electrical switching behavior of these films. In addition investigations on the surface
morphology and composition of the films have also been discussed. The reasons behind
choosing this particular stoichiometry are:
a) It can be prepared easily in bulk glass which can be used as a base material for
thermal evaporation.
b) It has shown a clean memory switching in the bulk glassy form [17].
c) The earlier investigations have indicated that the x = 23 composition in the

AlxTe100-x system corresponds to the rigidity percolation threshold (RPT) [18] and
interesting properties have been observed in chalcogenide glasses at RPT [19,20].
Ch.3: Electrical Switching Behavior of Amorphous…
3.2 Experimental techniques

The base-glassy material having the composition Al23Te77 has been prepared by
melt quenching technique. The bulk glassy ingot is then ground into a fine powder for use
as the evaporation source material. Glass slides of 25 mm×75 mm, cleaned with detergent
solution, acetone and ultrasound are used as substrates. Amorphous thin films are
deposited by flash evaporation technique with a mechanical mask used to obtain the
required structure of the device. An extra slide is also coated under identical conditions,
for structural and other characterization. The thickness of the films measured by the stylus
profilometer (form Taly-surf plus) has been found to be in the range of about 150 nm, for
about 5 min deposition time.
Aluminum electrodes have been deposited on the film in co-planar geometry, using
an appropriate mask. First, the chalcogenide film is coated in a rectangular shape on a
glass slide. Secondly, aluminum is coated as the electrode, with a thin mask wrapped on
the film. Subsequently, the electrode mask is removed. The active area of the device is the
film area between two aluminum electrodes.
In this work, the coplanar geometry has been deliberately chosen in order to
compare the electrical switching characteristics of bulk glassy and amorphous thin film
samples. Nevertheless, the sandwich geometry can be used to deposit the electrodes which
would bring down the switching voltages of the devices. During all the depositions, the
pressure inside the chamber is maintained at 10-5 mbar. The electrical contacts from the
aluminum electrodes have been made using silver paste and copper wire.
The switching behavior of the sample is studied by recording the current-voltage

(I-V) characteristics using a Keithley Source-Meter (Model 2410) with a basic accuracy of
0.012%, controlled by LabVIEW 6i (National Instruments). Scanning electron
microscopy, optical microscopy and Energy Dispersive X-ray analysis have also been
done on the sample for measuring the distance between the electrodes, surface
morphology and chemical composition of the film respectively.
3.3 Results and Discussion

Figure 3.1 shows the X-ray diffraction pattern of the as deposited Al23Te77 film
from which it is evident that the film is amorphous in nature. Table 3.1 gives the
composition of the different constituent elements in the as-deposited films, obtained by
42
Energy Dispersive X-ray Analysis (EDAX). It can be seen that the composition of the film
is close to the composition of starting material, Al23Te77. The resistance of the as-
deposited devices has been found to be in mega ohm range.
Figure 3.1: X-ray diffraction pattern of the as deposited Al23Te77 film
Element Atomic % in Film
Al 22.54
Te 77.46
Table 3.1: Atomic percentage of Al23Te77 film obtained by EDAX analysis
43
Figure 3.2: I-V characteristics of two representative Al23Te77 thin film devices, exhibiting
memory switching. Inset pictures show switching region of the graphs with error bar.
Figures 3.2 (a) and (b) show the typical I-V characteristics of two representatives
Al23Te77 thin film devices, which have been deposited in co-planar geometry with
aluminum as electrode.
It can be seen from these figures that the voltage across the devices initially
increases linearly with current, indicating an ohmic behavior. After a particular threshold
voltage, the voltage across the sample starts decreasing with current, which indicates a
negative resistance behavior. The negative resistance region leads to a low resistance state,
which latches the device in that state. The device remains in low resistance state even if
the current reduces to zero, showing that amorphous Al23Te77 films exhibit memory type
electrical switching.
44
As mentioned earlier, the application of a high electric field leads to the formation
of a conducting crystalline filament during memory switching, because of which sample
remains in the low resistance ON state even after the removal of the field. The memory
switching behavior of telluride glasses arises because of two general factors: the local
structure of these samples is characterized by long Te chains in which atomic
rearrangements and devitrification can occur easily [21]. Glassy tellurides also have a
higher electrical conductance in comparison with other chalcogenides like sulphides and
selenides. The larger Joule heating due to higher conductance and easy crystallizability,
lead to memory switching in telluride glasses such as Al23Te77.
Figures 3.3 (a) and (b) show the scanning electron micrographs of un-switched and
switched devices respectively. It is clearly seen that the un-switched film is featureless
with a smooth topography. On the other hand, the phase change in the electrode region is
revealed by an image contrast in the SEM picture; the image contrast after switching can
be due to any of the following aspects:
a) The phenomenon of electrical switching in chalcogenides, especially

memory, is associated with local melting and re-solidification into the
crystalline state of the active material. This process can lead to a surface
relief which can give rise to an image contrast.
b) The glass to crystal phase change is also associated with local structural
changes and densification in the electrode region as the crystalline phase is
denser than the corresponding amorphous phase. The density changes in the
switched region can contribute for the image contrast.
c) There is a large change in the electrical resistance of the active material in

the electrode region (about three orders of magnitude). The conductivity
changes can also lead to an image contrast in SEM.
Detailed investigations are necessary to comprehend which of the above mentioned

aspects is responsible for the observed changes in the SEM images of the switched sample.
However, it can be noted here that a similar SEM image has been obtained earlier during
the SET process of the device which has been attributed to the change of state of the
material from amorphous to crystalline [6].
45
Figure 3.3: Scanning electron micrograph of (a) un-switched and (b) switched Al23Te77
thin film samples respectively.
Earlier studies on bulk Al-Te glasses of different compositions reveal that these
samples exhibit memory type switching [17]. Figure 3.4 shows the I-V characteristic of
bulk Al23Te77 sample of thickness of 0.2 mm [17] and in ambient conditions, which
indicates that the switching behavior of bulk and thin film Al-Te samples is similar.
Table 3.2 shows the comparison of switching fields of the two representative thin
film devices and that of the bulk Al23Te77 glass [17]. It can be seen that the switching
fields of the two devices are comparable to each other. The difference in the switching
fields of the two films is 0.4 kV/cm; however, the switching field of the bulk glass is
much higher than that of the thin film samples. In bulk sample, the heat is likely to be
dissipated away faster from the electrode region to the larger thermal mass surrounding it.
Consequently the Joule heating and temperature rise in the electrode region for a given
applied voltage will be lower for the bulk, requiring a larger voltage for switching. Hence
the threshold field for the bulk sample is higher than that of the thin film samples.
Figure 3.4: I-V characteristics of bulk Al23Te77 glass showing memory switching [17].
Switching voltage Distance (d) between two Electric field

Device
(Vt) in volt electrodes in micron kV/cm
1 276 ± 2 977 ± 5 2.8
46
2 195 ± 2 789 ± 5 2.4
Bulk 135 ± 2 200 ± 5 6.75

Table 3.2: Comparison of switching fields of two representative amorphous thin film
samples and the bulk glass
3.4 Conclusions
Amorphous thin films of Al23Te77 have been deposited by flash evaporation
method. The amorphous nature and the composition of films are confirmed by X-ray
diffraction and EDAX respectively. The I-V characteristics and electrical switching
behavior of the films have been studied using a Keithley source-meter. It is found that thin
film amorphous Al23Te77 samples exhibit memory type electrical switching. The image
contrast in the Scanning Electron Micrograph of the switched region shows the phase
change (formation of a crystalline filament), which is responsible for switching of the
device from high resistance OFF state to low resistance ON state.
Further, the switching behavior of bulk and thin film Al23Te77 samples are found to
similar, with both types of samples exhibiting memory switching. The threshold fields of
bulk samples are much higher compared to those of thin film samples, which can be
understood based on the higher thermal conductance in bulk samples which reduces the
Joule heating and temperature rise in the electrode region.
47
References
[1] P. Sharma, M. Vashistha and I.P.Jain, Chalcogenide Letters, 2 (2005) 113.
[2] H.A. Abd El Ghani, M.M. Abd El Rahim, M.M.Wakkad, A. Abosehli and N.
Assraan, Physica B, 381 (2006) 156 .
[3] J. Fusong, and M. Okuda, Jpn. J. Appl. Phys., 30 (1991) 07.
[4] M.A. Majeed Khan, M. Zulfequar, S. Kumar and M. Husssain, J. Mod. Opt., 50
(2003) 251.
[5] S.R. Ovshinsky and H. Fritzche, IEEE Trans. Electron. Devices, 20 (1973) 91.
[6] H. Fritzsche, Amorphous and liquid semiconductors, Plenum Press, London, 1974.
[7] N.F. Mott and E.A. Davis, Electronic processes in Non-Crystalline Materials,
Clarendon Press, London, 1979.
[8] P. Dargan, P. Burton, S.V. Phillips, A.S. Bloor and P. Nesvadba, J. Mater. Science, 9
(1974) 1595.
[9] E. Bebenskas, S. Balyavichyus, A. Deksins, A. Poshkus and N. Shiktorov, J. Non-

Cryst. Solids, 90 (1987) 601.
[10] D.P. Gosain, M. Nakamura, T. Shimizu, M. Suzuki and S. Okano, Jpn. J. Appl.
Phys., 28 (1989) 1013.
[11] S. Balyavichyus, A. Deksins, A. Poshkus and N. Shiktorov, Sov. Phys. Semocond.,

18 (1984) 947.
[12] S.S.K. Titus, R. Chatterjee, S. Asokan and A. Kumar, Phys. Rev. B, 48 (1993) 1460.
[13] K. Ramesh, S. Asokan, K.S. Sangunni, et al., Appl. Phys. A, 69 (1999) 421.
[14] S. Prakash, S. Asokan and D.B. Ghare, J. Phys. D:Appl. Phys., 29 (1996) 2004.
[15] S.S.K. Titus, S. Asokan and E.S.R. Gopal, Solid State Commun., 83 (1992) 745.
[16] K. Ramesh, S. Asokan, K.S. Sangunni, et al., J. of Phys. Chem. Solids, 61 (2000) 95.
[17] S. Prakash, S. Asokan and D.B. Ghare, Semicond. Sci. Technol., 9 (1994) 1484.
[18] R. Ramakrishna, S. Asokan, G. Parthasarathy, S.S.K. Titus and E.S.R. Gopal, J.

Non-Cryst. Solids, 139 (1992) 120.
48
[19] S. Asokan, G. Parthasarathy and E.S.R. Gopal, Phil. Mag. B, 57 (1988) 49.
[20] S. Asokan, M.V.N. Prasad, G. Parthasarathy and E.S.R. Gopal, Phys. Rev. Lett., 62
(1989) 108.
[21] S.R. Ovshinsky and K. Sapree, Proc. 7th Int. Conf. on Amorphous and Liquid
Semiconductors, 1977, Institute of Physics, Bristol.
49
Chapter 4
Electrical Switching and Thermal Studies on Bulk Ge-Te-

Bi Glasses
4.1 Introduction
Chalcogenide glasses prepared by melt quenching technique are normally p-type
conductors. However, it has been found that n-type semiconducting glasses can be
obtained by adding Bi to certain chalcogenide glassy systems such as Ge-Se and Ge-Se-
Te, etc. [1,2]. The p-n transition has also been found in Ge-Te-Bi glasses [3], which has
prompted the present investigations on the electrical switching behavior on Ge18Te82-xBix
(1≤ x≤ 4) glasses, with a focus on the variation of threshold voltages with composition,
temperature and thickness.
In addition, the crystallization process in Ge18Te82-xBix (x = 1, 2, 3) alloys at
different heating rates has been studied by Differential Scanning Calorimetry (DSC), as it
is important for phase change applications; also, the composition dependence of
crystallization temperatures and the activation energies for crystallization (obtained using
the Kissinger’s method) have been investigated. Further, X-ray diffraction investigations
have been undertaken on crystallized Ge-Te-Bi samples to identify the phases formed
during devitrification. Finally, the scanning electron microscopic studies have been carried
out on the switched and un-switched regions of a representative Ge18Te81Bi1 sample.
4.2 Experimental procedures

As mentioned in chapter 2, chalcogenide glasses are reactive at high temperature,
particularly with O2, and also the constituent elements frequently differ widely in
volatility. This necessitates the preparation of these glasses in sealed containers. Further,
impurities can alter the electrical switching and thermal behavior of these materials,
though chalcogenide glasses are generally not as sensitive to impurities as intrinsic
semiconductors. Therefore, the elementary materials used for preparing bulk samples are
always of a high purity (99.999%).
Bulk Ge18Te82-xBix (1 ≤ x ≤ 4) glasses have been prepared by the conventional melt

quenching method. The amorphous nature of the quenched samples is confirmed by X-ray
diffraction (XRD). The electrical switching characteristics of these glasses have been
Ch.4: Electrical Switching and Thermal Studies on Bulk…
studied using a Keithley source-measure unit (Model 2410c). Samples polished to a

thickness of about 0.25mm are mounted between a flat bottom electrode and a point
contact top electrode. A constant current is passed through the sample and the voltage
developed across the sample is measured. The standard deviation method is used for the
analysis of errors in the measurement of switching voltages.
The thermal parameters such as crystallization temperatures (Tc) of Ge18Te82-xBix
glasses have been obtained using a Mettler Toledo Differential Scanning Calorimeter.
Samples have been heated at four different heating rates and the activation energy for
crystallization has been calculated using Kissinger’s equation, with the calorimetric scans
made from ambient temperature to 350 0C. The typical error in the measurement is found
to be ± 1 0C for Tc at different scanning rates.
Further, crystallization studies of two representative Ge18Te81Bi1 and Ge18Te79Bi3
samples have been undertaken by annealing the samples at the respective crystallization
temperatures for about two hours and X-ray diffraction studies have been carried out on
the annealed samples in a Philips powder diffractometer with CuK radiation. The
Scanning Electron Microscopic (SEM) images of the switched and un-switched regions of
a representative Ge18Te81Bi1 sample has been taken by ESEM Quanta machine.
4.3.1. Electrical switching behavior of Ge-Te-Bi glasses
Figure 4.1 shows the I-V characteristics of Ge18Te82-xBix series of glasses, from
which it is clear that these samples exhibit memory type electrical switching. As
mentioned earlier, the memory switching in chalcogenide glasses has a thermal origin
involving the formation of a conducting crystalline channel between the electrodes, with
the energy required for crystallization provided by the electric field. The Te-rich
chalcogenide glasses are normally found to exhibit memory switching because of their
comparatively larger electrical conductance which results in a higher power dissipation
and therefore the conducting channel could be formed easily. It is also known that weaker
bonds, poor structural cross-linking and more lone-pair interactions [4] favor memory
switching in telluride glasses. The present studies reveal that Ge18Te82-xBix glasses also
exhibit memory type electrical switching like many other binary and ternary telluride
glasses such as Ge-Te, Si-Te, As-Te, Al-Te, As-Se-Te, Ge-As-Te, Ge-Te-Cu [5-12].
52
Figure 4.1: I-V characteristics of (a) Ge18Te81Bi1 (b) Ge18Te80 Bi2 (c) Ge18Te79Bi3
(d) Ge18Te78Bi4 glass representing the Ge18Te82-xBix series.
In this context, it is interesting to note that the thermal diffusivity (α) of

representative Ge-Te-Bi samples, namely Ge18Te81Bi1, Ge18Te80Bi2 and Ge18Te79Bi3, has
been found to be 0.015, 0.0108 and 0.00926 respectively, which is comparatively lower
than other chalcogenide glasses [13,14]. Usually, chalcogenide glasses which have higher
thermal diffusivity exhibit threshold type electrical switching and those with a lower α
show memory behavior [15]. The memory behavior seen in Ge-Te-Bi glasses is therefore
consistent with thermal diffusivity values of these samples.
Figure 4.2 shows the variation of threshold voltage of Ge18Te82-xBix glasses with Bi
content, which indicates that switching voltage decreases with increase in bismuth content.
The composition dependence of threshold/switching voltage in chalcogenide glasses is
determined by three main factors, namely the resistivity of the additive elements (known
as the metallicity factor) [16], network connectivity, rigidity [17] and chemical ordering
[18].
The addition of more metallic elements usually brings down the switching voltages
of chalcogenide glasses; this is also the case in the present Ge18Te82- xBix system in which
53
Bi is more metallic compared to Te (Bi = 1.29 × 10-6 Ωm and Te = 1× 10-4 Ωm) and the
replacement of Te with Bi leads to the reduction in the threshold voltages (figure 4.2).
The rigidity percolation deals with dimensionality and rigidity of a glassy network.
For a covalent network glass, the network connectivity and rigidity increases with the
increase in average coordination number and at a mean co-ordination number rc = 2.4
(known as the Stiffness Threshold or Rigidity Percolation Threshold) (chapter 1) [19-21],
the glassy system exhibits a transition from a floppy polymeric glass to a rigid amorphous
solid. In general, the memory switching voltages of chalcogenide glasses are found to
increase with an increase in network connectivity and rigidity as the structural
rearrangements become more difficult with increasing network rigidity [17]. In the
Ge18Te82-xBix system, the addition of Bi with a higher coordination number of 3 [22], at
the expense of Te with the coordination number of 2 [23], results in a progressive increase
in network connectivity and rigidity. Hence, based on the network connectivity rigidity,
one could expect an increase in switching voltages with the addition of Bi. However, the
present results show that the switching voltages of Ge18Te82-xBix glasses show a decrease
with the increase in Bi concentration, which indicates that in this system the effect of
resistivity of the dopant is more pronounced than the network connectivity & rigidity.
Further, in many chalcogenide systems, a sharp signature of the rigidity percolation

is seen in the variation of switching voltages at rc. However, in the present Ge18Te82-xBix
system, it has not been possible to probe the effect of stiffness transition on switching
voltages as the composition range of bulk glass formation is limited only to lower rc values
(rc ≤ 2.4).
54
Figure 4.2: The composition dependence of threshold voltages of Ge18Te82-xBix glasses

(1 ≤ x ≤ 4).
4.3.2. Temperature dependence of threshold voltages
The threshold voltage (Vt) of Ge18Te81Bi1 glass is found to decrease with increase
in temperature as shown in figure 4.3, which is a common feature exhibited by many
memory switching chalcogenides glasses [24-26]. In case of memory switching glasses,
the decrease in switching voltages with temperature has been explained on the basis of a
configurational free-energy diagram [8], according to which the decrease in Vt is due to
the decrease in the energy barrier required for crystallization of a sample with the increase
in temperature. In addition, at higher temperature, the charged defect centers are filled up
by thermally excited charge carriers, which are in addition to field–injected carriers,
resulting in decrease in switching voltage [27].
4.3.3. Thickness dependence of switching voltages
Figure 4.4 shows thickness dependence of threshold voltage, indicating an almost

linear variation. It is known that in memory switching samples, the threshold voltage (Vt)
is found to be proportional to “t” or “t1/2”, where “t” is the thickness of the sample. A
linear variation of Vt with t, has been seen in samples like Ge-Te-Al [28], As-Se [29], Ge-
Se-Tl [30]. However, a few samples like Ge-As-Te [31], Ge-Te [5] have been found to
exhibit a t1/2 variation with Vt . In the present sample, the variation of threshold voltage Vt
with thickness has been observed to be linear (figure 4.4).
55
Figure 4.3: Temperature dependence of threshold voltage of a Ge18Te81Bi1 sample.
Figure 4.4: Thickness dependence of threshold voltages of a representative Ge18Te81Bi1

glass
.
4.3.4. Activation energy for crystallization
Table 4.1 shows the crystallization temperature (Tc) of three different compositions
of Ge-Te-Bi glassy system, obtained at four different heating rates (β). The Kissinger’s
plot [32] of the variation of ln (β/Tc2) vs. (1000/Tc) for a representative Ge18Te81Bi1 has
56
been shown in figure 4.5. The activation energy for crystallization (Ec) has been
calculated using the Kissinger’s equation [33], ln (β/Tc2) = -Ec / RTc + D.
The activation energy for devitrification of the Ge-Te-Bi samples, estimated from
the Kissinger’s plots, has been found to decrease with the addition of Bi (Table 4.1). The
observed decrease in Ec with Bi content is consistent with the decrease in memory
switching voltages with Bi content.
x (at %
4 K/min 6 K/min 8 K/min 10 K/min Ec (kJ/mol)
of Bi)
1 512.68 515.44 519.1 520.87 27.5
2 473.28 484.14 480.48 483.13 12.9
3 517.85 514.43 517.48 526.5 11.8
Table 4.1: The crystallization temperature (Tc) of three different compositions of

Ge-Te-Bi glassy system, at four different heating rates (β).
Figure 4.5: The Kissinger’s plot showing the variation of ln (β/Tc2) vs. (1000/Tc) for
Ge18Te81Bi1 glass.
4.3.5. Crystallization behavior of Ge18Te82-xBix glasses
Figure 4.6 shows the X-ray diffraction patterns of two representative Ge18Te81Bi1
and Ge18Te79Bi3 samples respectively, annealed for two hours under vacuum at their
respective crystallization temperatures. It is interesting to note the presence of hexagonal
57
Te and hexagonal GeTe, in annealed Ge-Te-Bi samples. In addition, hexagonal BiTe and
hexagonal Bi2Te3 have also been found in the crystallized Ge-Te-Bi samples.
Figure 4.6: The X-ray diffraction patterns of (a) Ge18Te81Bi1 and (b) Ge18Te79Bi3 glasses.
In this context, it is interesting to compare the earlier studies on the crystallization

behavior of a similar metal doped Ge-Te glass, namely Ge-Te-Ag [34]; X-ray
investigations on crystallized Ge-Te-Ag samples show that silver forms a ternary phase
along with possible binary phases in this system [34]. However, in the case of Ge-Te-Bi
samples, no ternary phase has been found and in fact, more number of binary phases is
58
found. The possible reason for not finding any ternary phases of bismuth or more number
of binary phases in Ge18Te79Bi3 could be that bismuth is not taking part in network
formation to a greater extent. This aspect is also reflected in the variation of switching
voltages which shows no increase with the addition of Bi.
4.3.6. Electron Microscopic analysis of Ge-Te-Bi

Figure 4.7 shows the scanning electron micrograph of switched and un-switched
regions of Ge18Te81Bi1 sample. It is clear from the SEM images of switched and un-
switched regions of Ge18Te81Bi1 sample that there are morphological changes in the
switched region. The morphological changes are generally due to the formation of the
crystalline channel between two electrodes during switching.
Figure 4.7: Scanning electron microscopy of (a) un-switched and (b) switched regions of
Ge18Te81Bi1 glass.
4.4 Conclusions
Melt quenched Ge18Te82-xBix glasses have been found to exhibit memory switching
which is supported by lower thermal diffusivity values of these samples. A linear increase
has been found in switching voltages with increase in sample thickness which is also
consistent with the memory switching exhibited by Ge-Te-Bi samples. Also, the switching
voltages have been found to decrease with increase in temperature due to decrease in the
activation energy for crystallization at higher temperatures. Further, the switching voltages
of the Ge-Te-Bi samples have been found to decrease with increase in Bi content, which
indicates that resistivity of the additive has a stronger role to play in comparison with the
network connectivity and rigidity factors.
The composition dependence of activation energy for crystallization of Ge18Te82-

xBix glasses, exhibits a decrease with increase in Bi content, which is consistent with the
decrease in switching voltages with Bi content. Based on the X-ray diffraction studies of
thermally crystallized samples, it is proposed that bismuth is not taking part in network
formation in a greater extent, which is reflected in the variation of switching voltages with
the addition of Bi. SEM studies on switched and un-switched regions of Ge-Te-Bi sample
indicate that there are morphological changes in the switched region, which can be
attributed to the formation of the crystalline channel formation between two electrodes
during switching.
59
60
References
[1] N. Tohge, T. Minami and M. Tanaka, Jpn. J. Appl. Phys., 16 (1977) 977.
[2] N. Tohge, T. Minami, Y. Yamamoto and M. Tanaka, J. Appl. Phys., 51 (1980) 1048.
[3] K.L. Bhatia, G. Parthasarathy, Ashwani Sharma and E.S.R. Gopal, Phy. Rev. B, 38
(1988) 6342.
[4] S.R. Ovshinsky, Phys. Rev. Lett., 36 (1976) 1469.
[5] E. Babenskas, S. Balevicius, A. Cesnys, A. Poskus and N. Siktorov, J. Non- Cryst.
Solids, 90 (1987) 601.
[6] S. Balyavichyus, A. Deksins, A. Poshkus and N. Shiktorov, Fiz. Tekh. Poluprovodn.,
18 (1984) 1513; Sov. Phys. Semicond. 18 (1984) 947.
[7] S.S.K. Titus, R. Chatterjee, S. Asokan and A. Kumar, Phys. Rev. B, 48 (1993) 14650.
[8] S. Prakash, S. Asokan and D. B. Ghare, Semicond. Sci. Technol., 9 (1994) 1484.
[9] R. Chatterjee, S. Asokan and S. S. K. Titus, J. Phys. D: Appl. Phys., 27 (1994) 2624.
[10] R. Uttecht, H. Stevenson, C.H. Sie, J.D. Griener and K.S. Raghavan, J. Non-Cryst.
Solids, 2 (1970) 358.
[11] K. Tanaka, Y. Okada, M. Sugi, S. Iizima and M. Kikuchi, J. Non-Cryst. Solids, 12
(1973) 100.
[12] K. Ramesh, S. Asokan, K.S. Sangunni and E.S.R Gopal, Appl. Phys. A, 69 (1999)
421.
[13] K.N. Madhusoodanan, Jacob Philip, S. Asokan and E.S.R. Gopal, J. Mat. Sci. Lett., 7
(1988) 1333
[14] Pulok Pattanayak, N. Manikandan, M. Paulraj and S. Asokan, J. Phys.: Condens.
Matter, 19 (2007) 036224.
[15] S. Manohar, S. Murugavel and S. Asokan, Solid State Commun., 135 (2005) 323.
[16] S.B. Bhanu prashanth and S. Asokan, J. Non-Cryst. Solids, 355 (2009) 164.
[19] J.C. Phillips and M.F. Thorpe, Solid State Commun., 53 (1985) 699.
[20] P. Boolchand and M.F. Thorpe, Phys. Rev. B, 50 (1994) 10366.
[21] J.C. Phillips, Rigidity Theory and Applications, Kluwer Academic/Plenum, New
York, 1999.
[22] Ambika and P.B. Barman, J. Ovonic. Res., 3 (2007) 21.
61
[23] K. Tsunetomo, R. Shimizu, T. Imura and Y. Osaka, J. Non-Cryst. Solids, 116 (1990)
262.
[24] M. Fadel, K. Sedeek and N.A. Hegab, Vacuum, 57 (2000) 307.
[25] K. Tanaka, S. Iizima, M. Sugi and M. Kikuchi, Solid State Commun., 8 (1970) 75.
[26] A. Alegria, A. Arruabarrena and F. Sanz, J. Non-Cryst. Solids, 58 (1983) 17.
[27] S.R. Ovshinky and H. Fritzsche, IEEE Trans. Elect. Dev., 20 (1973) 91.
[28] J.R. Bosnell and C.B. Thomas, Solid-State Electron, 15 (1972) 1261.
[29] A. Giridhar and K.J. Rao, J. Non-Cryst. Solids, 33 (1979) 177.
[30] M.F. Kotkata, M.A. Afifi, H.H. Labib, N.A. Hegab and M.M. Abdel-Aziz, Thin Solid
Films, 240 (1994) 143.
[31] H.J. Stocker, C.A. Barlow and D.F. Weirauch, J. Non-Cryst. Solids, 4 (1970) 523.
[32] H.E. Kissinger, J. Res. NBS, 57 (1956) 217.
[33] T. Ozawa, Polymer, 12 (1971) 150.
[34] K. Ramesh, S. Asokan, K.S. Sangunni and E.S.R Gopal, J. Phys. Chem. Solids, 61
(2000) 95
62
Chapter 5
Electrical Switching and Optical Studies on Amorphous

GexSe35-xTe65 (17≤ x ≤ 23) Thin Films
5.1 Introduction
Phase Change Memories (PCM) based on electrical switching in chalcogenide
glasses [1], have proven to be replacement for conventional Random Access Memories,
due to their advantages such as better read/write performance, low programming energy,
high data retention, low cost, compatibility with complementary metal oxide
semiconductor (CMOS) technology, etc. [2,3].
Good glass forming ability and at the same time, easy crystallizability are required
for a good PCM material. The materials which are suitable for the phase change memory
applications are usually chalcogenide glasses exhibiting memory switching, which consist
of long Te chains in which atomic rearrangement and devitrification can occur more easily
[4]. In particular, Te-based chalcogenide glasses such as Ge-Te, Si-Te, As-Te, Al-Te,
As-Se-Te, Ge-As-Te etc., which exhibit memory type switching, have been widely studied
[5-11], in order to find their suitability for PCM applications. Among these glasses,
germanium telluride based chalcogenide glasses have attracted much attention due to their
interesting electrical, optical and thermal properties [1,12].
Earlier studies on bulk GexSe35-xTe65 glasses (17 ≤ x ≤ 25) indicate that these
samples exhibit memory switching behavior [13]. These studies have also explored the
dependence of the electrical switching behavior of bulk GexSe35-xTe65 glasses on other
material properties such as network rigidity, thermal parameters, etc. In this chapter, the
electrical switching behavior of amorphous GexSe35-xTe65 thin films has been studied
systematically in the composition range of 17 ≤ x ≤ 23. The composition dependence of
the switching voltages of amorphous GexSe35-xTe65 thin film samples has been
investigated. In addition, the optical band gap of these samples calculated from their
absorption spectra and its variation with composition is compared with the composition
dependence of switching voltages.
Ch.5: Electrical Switching and Optical Studies on…
The intention behind these studies is to compare the switching behavior of
amorphous thin film and bulk glassy samples and also find the suitability of amorphous
thin film samples for Phase Change Memory applications.
5.2 Experimental Techniques

Bulk GexSe35-xTe65 glassy samples (17 ≤ x ≤ 23) has been prepared by vacuum
sealed convectional melt–quenching technique as described in chapter 2. Thin film devices
are coated by flash evaporation technique in sandwich geometry on a 25mm × 75mm
glass substrate. Aluminum is used for both top and bottom electrodes. The thickness of the
films has been found to be in the range of 150-300 nm. The amorphous nature of the film
has been confirmed by X-ray diffraction, using Bruker AXS D8 Advance X-ray
diffractometer.
The electrical switching studies of the samples have been carried out using a
Keithley source-measure unit in the current range 0-20 mA. The optical absorption spectra
of the five representative compositions of the GexSe35-xTe65 series (19 ≤ x ≤ 23) have been
obtained using a spectrophotometer (Model HR 4000, Ocean optics) in the range of 200
nm to 1000 nm; the band gaps of the samples have been estimated from the absorption
spectra.
EDAX analysis of the samples has been done using Quanta LV/ ESEM microscope.
The EDAX values of different elements of a representative sample Ge20Se15Te65 are
tabulated in table 5.1. A minor variation in composition has been found in films from
their bulk counterparts.
Ge20Se15Te65 Ge20Se15Re65
Element
(bulk) (film)
Ge 20 22
Se 15 15
Te 65 63
Table 5.1: Energy dispersive X-ray analysis of one representative bulk and film
Ge20Se15Te65 sample.
64
5.3. Results and Discussion
5.3.1. Electrical Switching
Figure 5.1 shows the X-ray diffraction patterns of two representative GexSe35-
xTe65 thin film samples, Ge20Se15Te65 and Ge21Se14Te65, confirming their amorphous
nature. The electrical switching behavior of Ge20Se15Te65 and Ge19Se16Te65 amorphous
thin film samples, representing the GexSe35-xTe65 series (17 ≤ x ≤ 23) is shown in figure 5.2.
Figure 5.1: XRD patterns of (a) Ge20Se15Te65 and (b) Ge21Se14Te65 thin film samples
showing their amorphous nature.
65
It is seen from figure 5.2 that after an initial Ohmic response and above a threshold
voltage, amorphous thin film GexSe35-xTe65 samples exhibit memory type electrical
switching. In this context, it is interesting to compare the electrical switching behavior of
bulk glassy and amorphous thin film GexSe35-xTe65 samples. The earlier studies [13]
indicate that bulk GexSe35-xTe65 glasses exhibit a composition dependent switching
behavior; GexSe35-xTe65 glasses with x < 20 are found to exhibit memory type electrical
switching, whereas samples with x ≥ 20 show threshold type switching behavior.
Figure 5.2: I-V characteristics of (a) Ge19Se16Te65 and (b) Ge20Se15Te65 thin film samples
showing their memory type electrical switching.
66
In bulk GexSe35-xTe65 glasses, the observed crossover in switching behavior
(from memory to threshold), has been understood structurally in terms of a change in
steric hindrances and energy barriers to matrix rearrangement across the rigidity
percolation or the stiffness threshold [14].
The present results, however, indicate that all the thin film amorphous samples of
the GexSe35-xTe65 system, exhibit memory type of electrical switching behavior. Similar
memory type of switching behavior has been observed in many binary and ternary
chalcogenide glasses such as Ge-Te, Si-Te, Ge-Te-Cu, Ge-Te-Ag, Ge-As-Te etc [15-17] in
bulk form and in several amorphous thin films including Al-Te, Ge-Se-Tl, etc [18,19].
The variation of switching fields of GexSe35-xTe65 amorphous thin film samples

with composition is shown in figure 5.3. Figure 5.4 shows the composition dependence of
switching fields of bulk GexSe35-xTe65 glasses [13] for comparison.
Figure 5.3: Variation of threshold field with composition in GexSe35-xTe65 ( 17 ≤ x ≤ 23 )

thin film samples.
67
Figure 5.4: Variation of threshold field with composition in GexSe35-xTe65 ( 17 ≤ x

≤ 25 ) bulk samples [13].
It can be noted that the threshold fields of bulk GexSe35-xTe65 glasses increase
continuously with composition, with a sudden upward jump seen around the composition
x = 20. Further, the thermal studies indicate that the glass transition temperature of bulk
GexSe35-xTe65 glasses increases with the addition of Ge-atoms [20], confirming an increase
in network connectivity; with the addition of 4-fold coordinated Ge atoms, there is a
continuous change from a 1-dimensional distorted chain structure to a 3-dimensional
tetrahedron network [21]. The composition x = 20 in GexSe35-xTe65 bulk glasses
corresponds to the stiffness threshold of the system [13], at which there is a sudden change
in the steric hindrances and energy barriers for structural rearrangement required for the
glass to crystal phase change [14]. For compositions below the stiffness threshold (x < 20),
the low network connectivity permits an easy creation of nucleation sites for the structural
phase change required for the memory type electrical switching [13].
Though the switching fields of amorphous GexSe35-xTe65 thin films remain a

constant in the composition range 17 ≤ x ≤ 21 (pre stiffness transition regime), a sharp
increase is seen in the composition dependence of the switching fields of amorphous
GexSe35-xTe65 films above x = 21, which can be associated with the stiffness transition. The
similar trend, namely a sharp increase in switching fields across the stiffness transition,
indicate that similar factors are responsible for the composition dependence of switching
fields of amorphous thin film and bulk glassy GexSe35-xTe65 samples. It is worthy of
68
pointing out here that this is the first observation of the manifestation of the stiffness
transition in the composition dependence of switching fields of amorphous chalcogenide
films though there are several reports on the effect of stiffness transition on the switching
voltages/fields of bulk chalcogenide glasses [22,23].
Table 5.2 consolidates the value of electric fields, thicknesses of the films and
threshold voltages for different compositions of amorphous GexSe35-xTe65 thin films. It can
be seen that the threshold voltages of thin film samples are in the range of 0.8V to 10V
(for approximately 150nm-300nm thick samples, whereas those of bulk samples are in the
range of 350V to 550V (for 0.3mm thick sample); there is a remarkable reduction in
switching voltages, of thin film Ge-Se-Te samples, as compared to the switching voltages
of bulk samples of same composition. This indicates that Ge-Se-Te thin film samples are
interesting from the viewpoint of phase change memory applications, as the low switching
voltage is an important factor for phase change memory devices.
Threshold
Thickness of the Electric Field
Composition voltage
film (nm) (kV/cm)
(V)
Ge17Se18Te65 300 0.8248 27.5
Ge18Se17Te65 300 0.9233 30.78
Ge19Se16Te65 162 0.9133 56.38
Ge20Se15Te65 147 0.715 48.64
Ge21Se14Te65 198 0.9314 47.04
Ge22Se13Te65 100 1.011 101.07
Ge23Se12Te65 259 9.7 374.52
Table 5.2: The variation of threshold voltages and fields of amorphous thin film
GexSe35-xTe65 samples with composition.
5.3.2. Optical band gap
The optical band gap has a prime effect on the electrical properties of
semiconducting materials [24,25]. It has been observed that the composition has a
significant influence on the band gap of glassy chalcogenide semiconductors [26-28].
Studies on the composition dependence of optical band gap of bulk Ge-Se-Te glasses have
been reported previously [28]. In the present work, optical gap of Ge-Se-Te amorphous
thin films has been estimated from absorption coefficient data as a function of wave length
by using Tauc Relation [29].
69
αhυ = B(hυ - Eg )m
Where hυ is the energy of the incident photon, B is a constant which depends on

the material and nature of transition, Eg represents the band gap energy and m in an
exponent whose value is characteristic of the transition. In direct case m = 1/2 and for
indirect, m = 2.
Figure 5.5 shows the plot of (αhυ)2 vs hυ of a representative as-deposited

Ge22Se13Te65 thin film sample. From figure 5, it can be concluded that these samples are
direct band gap semiconductors as the plot of (αhυ)2 vs hυ is linear; The band gap of the
material is calculated from the x-intercept of the linear portion of the curve.
Figure 5.5: (αhυ)2 Vs hυ plot for as deposited Ge22Se13Te65 thin film sample.
Table 5.3 consolidates the band gap of amorphous GexSe35-xTe65 thin film samples
obtained in the present study; It is seen that the band gap of the films varies from 1.56 to
1.82 eV. Optical band gaps obtained from another Ge-Se-Te based series varies from 1eV
to 1.5 eV and is comparable to the values obtained in this series of sample [28]. So it can
be said that optical band gap values are consistent with the data found from literature.
70
Thin film Band Gap
composition (eV)
Ge19Se16Te65 1.56
Ge20Se15Te65 1.6
Ge21Se14Te65 1.64
Ge22Se13Te65 1.75
Ge23Se12Te65 1.82
Table 5.3: The variation of optical band gap of amorphous thin film GexSe35-xTe65
samples with composition.
Figure 5.6 shows the variation of optical band gap of amorphous GexSe35-xTe65 thin
films, from which it is clear that the band gap increases with the increase in Ge content.
The earlier study on the bulk Ge-Se-Te glasses indicates that, the glass transition
temperature and optical band gap increases with the progressive replacement of Te by Se
[28]. A similar increase in the band gap Eg has also been found in as-deposited and
annealed Ge-Se-Te films upon progressively replacing Se with Te [30-34 ].
Figure 5.6: Variation of band gap with composition of GexSe35-xTe65 ( 19 ≤ x ≤ 23 ) thin

film samples.
In the present composition tie-line, Se is replaced by Ge. The bond strength of Ge-
Se is 484.7±1.7 kJ/mol and that of Ge-Te is 396.7±3.3 kJ/mol; Assuming that Ge atoms
combine with Se as Ge is incorporated into the Ge-Se-Te system, the average bond
strength of the system increases continuously with Ge addition. This could be a reason for
the observed increase in optical band gap of GexSe35-xTe65 samples, with the increase in
71
Ge concentration. It is interesting to note here that other Se-Te based samples such as Se1-
xTex (0.2 ≤ x ≤ 0.8) [35], Se96-xTe4Agx (0 ≤ x ≤ 12) [36], Se80Te20-xPbx (0 ≤ x ≤ 10) [37],
etc., have also been seen to exhibit an increasing trend in optical band gap with
composition.
The chalcogenide glassy semiconductors are intrinsic like and the conductivity
activation energy is half the optical band gap [38]; therefore an increase in optical band
gap indicates an increase in conductivity activation energy and consequently the resistivity
of the semiconducting material. It is known that the switching field/voltages of
amorphous/glassy chalcogenides are directly related to the resistivity of the sample. Thus,
the composition dependence of optical band gap and the variation of switching voltages
with Ge concentration of amorphous GexSe35-xTe65 thin films are consistent.
5.4. Conclusions
Amorphous GexSe35-xTe65 thin film samples GexSe35-xTe65 (17 ≤ x ≤ 23), deposited
with sandwich geometry of electrodes, are found to exhibit memory switching behavior,
which is similar to that of bulk Ge-Se-Te glasses. The switching voltages of GexSe35-xTe65
thin film samples are found to be lower compared to those of bulk samples. In both thin
film amorphous and bulk glassy samples, the switching voltages are found to increase with
the increase in Ge concentration, which is consistent with the increase in network
connectivity with the addition of higher coordinated Ge atoms. A sharp increase is seen in
the composition dependence of the switching fields of amorphous GexSe35-xTe65 films
above x = 21, which can be associated with the stiffness transition. This is the first
observation of the manifestation of the stiffness transition in the composition dependence
of switching fields of amorphous chalcogenide films. Further, the optical band gap of a-
GexSe35-xTe65 thin film samples, calculated from the absorption spectra, is found to show
an increasing trend with the increase in Ge concentration, which is consistent with the
variation of switching voltages with composition.
72
References
[1] S.R. Ovshinsky, Phy. Rev. Lett., 21 (1968) 1450.
[2] S. Hudgens and B. Johnson, MRS Bull., 29 (2004) 1.
[3] D.-H. Kang, B.-k. Cheong, J.-h. Jeong, T. S. Lee, I. H. Kim, W. M. Kim and J-Y.
Huh, Appl. Phys. Lett., 87 (2005) 253504.
[4] S.R. Ovshinsky and K. Sapree, Proc. 7th Int. Conf. on Amorphous and Liquid
Semiconductors, 1977, Institute of Physics, Bristol.
[5] E. Babenskas, S. Balevičius, A. Česnys, A. Poškus and N. Šiktorov, J. Non-Cryst.
Solids, 90 (1987) 601.
[6] S. Balyavichyus, A. Deksins, A. Poshkus and N. Shiktorov, Fiz. Tekh. Poluprovodn.,
1513 (1984) 18; Sov. Phys. Semicond., 18 (1984) 947.
[7] S.S.K. Titus, R. Chatterjee, S. Asokan and A. Kumar, Phys. Rev. B, 48 (1993)
14650.
[9] R. Chatterjee, S. Asokan, and S.S.K. Titus, J. Phys. D: Appl. Phys., 27 (1994) 2624.
[10] R. Uttecht, H. Stevenson, C.H. Sie, J.D. Griener and K.S. Raghavan, J. Non-Cryst.
Solids, 2 (1970) 358.
[11] K. Tanaka, Y. Okada, M. Sugi, S. Iizima and M. Kikuchi, J. Non-Cryst. Solids, 12
(1973) 100.
[12] D. Adler, J. Vac. Sci. Technol., 10 (1973) 728.
[13] S.B. Bhanu Prashanth and S. Asokan, Solid State Commun., 147 (2008) 452.
[14] J. Bicerano and S.R. Ovshinsky, J. Non-cryst. Solids, 74 (1985) 75.
[16] K. Ramesh, S. Asokan, K.S. Sangunni and E.S.R. Gopal, Appl. Phys. A, 69 (1999)
421.
[17] S. Prakash, S. Asokan and D.B. Ghare, J. Phys. D: Appl. Phys., 29 (1996) 2004.
[18] Chandasree Das, R. Lokesh, G. Mohan Rao and S. Asokan, J. Non-cryst. Solids, 356
(2010) 2203.
[19] M.A. Afifi, M.M. Abdel-Aziz, H.H. Labib, M. Fadel and E.G. El-Metwally,
Vacuum, 61 (2001) 45.
[20] S.B. Bhanu Prashanth and S. Asokan, J. Non-Cryst. Solids, 355 (2009) 1227.
[21] S. Asokan, PhD Thesis, Department of Physics, Indian Institute of Science, (1986).
[22] M. Anbarasu and S. Asokan, J. Phys. D: Appl. Phys., 40 (2007) 7515.
73
[24] S.K.M. Dehaldhar and S.P. Sengupta, Ind. J. Pure Appl. Phys,. 17 (1979) 422.
[25] W. Beyer, H. Mell and J. Stuke, Phys. Status Solidi B, 45 (1971) 153.
[26] N. Tohge, T. Minami, Y. Yamamoto and M. Tanaka, J. Appl. Phys., 51 (1980)
1048.
[27] N. Tohge, T. Minami and M. Tanaka, J. Non-Cryst. Solids, 37 (1980) 23.
[28] M. Szukwei, Y. Hanmei and C. Zongcai, J. Non-Cryst. Solids, 52 (1982) 131.
[29] J. Tauc, Amorphous and Liquid Semiconductor, Plenum Press, New York, 1974.
[30] B.D. Muragi and J.K. Zope, J. Mat. Sci. Lett., 3 (1984) 633.
[31] S. Prassad and A. Bhaskera Rao, Ind. J. Pure Appl. Phys., 10 (1972) 36.
[32] J.C. Grag, S.K. Jain and B.C. Jain, Ind. J. Pure Appl. Phys., 19 (1981) 319.
[33] R.M. Mehra, P.C. Mathur, A.K. Kathuria and R. Sham, Phys. Rev. B, 18 (1978)
5620.
[34] E. Kh. Shokr, J. Phys. Chem. Solids, 53 (1992) 1215.
[35] H. El-zahed, M.A. Khaled, A. El-Korashy, S.M. Youssef, M. El Ocker, Solid State
Commun., 89 (1994) 1013.
[36] A.A. Al-Ghamdi, Vacuum, 80 (2006) 400.
[37] Shamshad A. Khan, M. Zulfequar, M. Husain, Current Appl. Phys., 5 (2005) 583.
[38] N.F. Mott and E.A. Davis, Electronic Processes in Non-crystalline Materials,
Clarendon Press, Oxford, 1979.
74
Chapter 6
Electrical Switching and Optical Band gap studies on

Ge15Te85-xSix (1 ≤ x ≤ 6) Amorphous Thin Film Samples
6.1 Introduction
Easy devitrification, good glass forming ability, low thermal diffusivity, low
melting temperature, etc. [1-3] are some of the important attributes of a good PCRAM
material. Commercially, the Ge2Sb2Te5 (GST) is one of the resistance variable
chalcogenide samples, used widely in rewritable compact disks (CD) and digital versatile
disks (DVD) [4-7]. The high RESET current requirement due to comparatively high
melting temperature and the low crystalline resistivity of the GST film leads to a higher
power consumption, which is not desirable in high density PRAMs, due to the current
limitation of drive transistors [8]. A few other challenges in the implementation of a phase
change memories with the Ge2Sb2Te5 alloy are high programming current requirement,
variation in switching voltages and ON/OFF resistance ratios, thermal stresses on the
materials, etc.
Considerable effort has been already made to improve the performance of
commercially used Ge2Sb2Te5 sample [9-13]; for example, doping of Si in Ge2Sb2Te5
films have been found to reduce the RESET current of the device so that the programming
volume can be reduced [14,15]. It has also been reported that Si doping has a considerable
effect on the phase change properties, such as an increase in the crystalline resistivity and
the decrease the melting temperature of the GST film [16]. The earlier studies on the
switching behavior of bulk Ge15Te85-xSix (0 ≤ x ≤ 12) glasses indicate that these samples
could be suitable candidates for phase change memory applications, as they have been
found to withstand several SET-RESET cycles without any degradation [17-19].
In this investigation, efforts have been made to replace Sb with Si in the Ge-Sb-Te
system and study the electrical switching properties of these samples in amorphous thin
film form. Electrical switching studies have been carried out on six representative
amorphous thin film compositions of the Ge15Te85-xSix series, concentrating mainly on the
SET operation of the samples in thin film form.
Attempt has also been made in this work, to compare the electrical switching
behavior of thin film Ge-Te-Si samples with the electrical switching and thermal behavior
Ch.6: Electrical Switching and Optical Band gap…
of bulk Ge15Te85-xSix glasses [17,18]. Particularly, a comparative study has been made
on the composition dependence of switching fields of bulk and thin film samples.
In addition, the measurement of the optical band gap of Ge15Te85-xSix amorphous
thin film samples has been carried out for different compositions and the variation with
composition of optical band gap has been compared with that of switching fields. Further,
Scanning Electron Microscopy (SEM) has been done on a representative sample to
understand the local structural changes that occur in the switched region.
6.2 Experimental Details

The source glassy materials in bulk form for the subsequent thin film preparation
are prepared by melt quenching technique. The bulk ingots obtained from melt quenching
technique are powdered and used as a source material for thin film coating. Thin films of
six compositions of Ge15Te85-xSix series (1 ≤ x ≤ 6) are prepared by the flash evaporation
technique at 2 × 10-6 mbar on glass substrates (75 × 25 mm) cooled to liquid nitrogen
temperatures. The amorphous nature of the as-deposited films is confirmed by X-ray
diffraction using Bruker (AXS D8 Advance) X-ray diffractometer. Figure 6.1 shows the
X-ray diffraction pattern of a representative Ge15Te81Si4 thin film sample indicating its
amorphous nature.
Figure 6.1: XRD pattern of Ge15Te81Si4 thin film sample confirming amorphous nature.
The optical absorption of the films is obtained in the wavelength range 200 - 1000
nm with an Ocean Optics HR 4000 spectrometer. Switching devices are made in sandwich
76
geometry with aluminum used as the top and bottom electrodes. The device geometry
employed in switching studies is shown in figure 6.2.
Figure 6.2: Schematic diagram of a memory device in sandwich mode.

The I-V characteristics and the electrical switching behavior of the devices have
been studied using a Keithley (2410c) source-measure unit using Lab VIEW 7 (National
Instruments). The sourcing capacity of the source meter is in the range of 0-20 mA at a
compliance voltage of 1100V (max). A constant current is passed through the sample and
the voltage developed across the sample is measured. The thickness of the films is
measured by Dektak 150 and found to be in the range of 150 - 450nm. The scanning
electron microscopic studies are done using Quanta LV/ ESEM microscope.
6.3.1. Electrical switching behavior of amorphous Ge15Te85-xSix thin films

Figures 6.3 (a and b) show the I-V characteristics of two representative Ge15Te85-
xSix amorphous thin film samples. The behavior of other samples in the series, is similar. It
can be seen from figure 6.3 that the initial amorphous state of the film has a high
resistance, which represents the OFF state; only very small current flows through the
device in the OFF state showing the Ohmic behavior. Upon increasing the current and at
the switching or threshold voltage (Vt), the device switches to the low resistance ON
state, where current increases drastically showing a non-Ohmic behavior. The device
remains in the ON state even after the electric field is removed, which is a typical
characteristic of memory switching materials. The behavior of devices of other
compositions are similar and the observed electrical switching voltages of a-Ge15Te85-xSix
thin film samples are comparable with other samples used in PCM applications such as
Ge-Sb-Te, Si-Sb-Te, Si-Ge-Sb-Te, etc. [16].
77
Figure 6.3: The I-V characteristics of representative (a) Ge15Te81Si4 and (b) Ge15Te79Si6
amorphous thin film samples.
As the thickness of the films are different for different samples of the series, the
threshold electric field of the samples was calculated and plotted against different
compositions of the series (figure 6.4). It can be seen that the switching fields of
amorphous Ge15Te85-xSix samples increase with the increase in silicon content.
78
Figure 6.4: The composition dependence of threshold fields of Ge15Te85-xSix (1≤

x≤ 6) amorphous thin film samples.
The inset in figure 6.4 shows the threshold electric field versus composition for
bulk glassy samples reported earlier [17]. The threshold voltages of bulk Ge15Te85-xSix
glasses are reported to lie in the range of 100-600 V for 0.3 mm thick samples. Further,
with increase in Si content, the switching voltages/fields of the bulk Ge15Te85-xSix glasses
have been found to increase [17]. It is evident from figure 4 and the inset that the variation
of the switching fields with composition of bulk glassy and amorphous thin film Ge15Te85-
xSix samples are similar.
As mentioned earlier, the key factors which dictate the composition dependence of
switching fields/voltages of multi component phase change glasses, are the resistivity and
coordination of the dopants, network connectivity and rigidity, thermal diffusivity and
thermal stability, etc. [20-28]. Particularly, the network connectivity and rigidity are found
to influence switching fields/voltages in two ways: the first by increasing the resistivity of
the material and the second by making the structural network less flexible for any
reorganization required for phase change. Consequently, the switching fields/voltages of
memory glasses are found to increase with increasing network rigidity [29]. Such an
increase in switching voltage with increase in network connectivity and rigidity, has been
observed in a variety of phase change glasses, including As-Te-Si [30] , As-Te [23], Ge-
Te [24], Si-Te [24], Al-Te [25], Al-As-Te [26] and Ge-Te-As [27], etc.
79
In bulk Ge15Te85-xSix glasses, the observed increase in switching voltages, has been
understood on the basis of metallicity of the dopant and also the rigidity percolation [17].
In this context, it is interesting to consider the results of the ADSC studies on Ge15Te85-
xSix glasses which show that the glass transition temperature (Tg) and crystallization
temperature Tc increase with the incorporation of Si in Ge-Te network [18]. There is also a
report in literature that the addition of Si to the eutectic glass (Ge15Te85) increases both Tg
and Tc [31]; the observed increase in Tg has been explained on the basis of the increase in
network connectivity and rigidity of Ge15Te85-xSix glasses with Si addition. Based on
these, it can be concluded that the network rigidity and, metallicity factors are responsible
for the increase in the switching fields of amorphous Ge15Te85-xSix thin films with Si
content.
Another earlier study on Si doped Ge2Sb2Te5 films has also shown an increase in
threshold voltage with increase in Si-doping concentration [16]. The observed behavior
has been explained on the basis that Si doping results in the increase of the mobility gap
and the resistivity of the amorphous state. Since the switching occurs at an almost constant
current density, the higher OFF state resistance can result in a higher threshold voltage
[32,33]. This could be an alternate explanation for the observed composition dependence
of switching voltages/field of amorphous Ge-Te-Si thin films.
In the context of composition dependence of electrical switching behavior, it is

interesting to note that bulk Ge15Te85-xSix glasses have been found to exhibit two kinds of
switching behavior [17]; a normal memory switching is seen in samples with x ≤ 5. On the
other hand, an instability is seen in glasses with x > 5, during the transition to the ON state
above the threshold voltage Vt. Further, in bulk Ge15Te85-xSix glasses, the splashing of the
material in the electrode region is seen in samples with composition x > 5. However, the
present study indicates that amorphous Ge15Te85-xSix thin film samples exhibit a clean
memory type of electrical switching behavior without any fluctuations in the I-V
characteristic (figure 6.3). The possible reason behind this could be the drastic reduction
in the switching voltage due to scaling down of the thickness of the material in thin film
form. The fluctuations in the I-V characteristic and splashing of the material in the active
region are usually connected with over-heating in the electrode region.
Though the switching time measurements have not been undertaken in the present
study, it is observed that the switching is faster in thin film samples, compared to the bulk
samples; no data points can be obtained in the amorphous thin film samples during the
80
transition to the ON state which implies that fast phase change takes place during SET
process which is an important aspect for PCM applications.
6.3.2. Optical Band gap
The optical properties of glassy tellurides have been studied extensively, in the
context of exploiting these materials in optical recording media [34-38]. The spectral
variations in absorption coefficient α can be described by the following equation
αhυ = B ( hυ - Eg)m ….. (1)
Where hυ is the photon energy, B is a parameter that depends on the transition

probability and Eg is the energy gap. Figure 6.5 shows the plot of (αhυ)2 Vs hυ for a
representative Ge15Te81Si4 sample. It can be seen that the present a-Ge15Te85-xSix thin film
samples obey the rule of direct transition (m < 1). Other systems such as GeSe0.5Te3.5 [39],
Se100-xInx (0 ≤ x ≤ 20) [40], etc., have also been found to exhibit a similar kind of
transition.
The value of optical band gap (Eg) has been calculated by taking the intercept on
the x-axis. Figure 6.6 shows the composition dependence of optical band gap of
amorphous Ge15Te85-xSix thin film samples. Table 6.1 compares the calculated optical
band gap values and the observed electrical switching fields of amorphous Ge15Te85-xSix
thin film samples. It can be seen that there is a direct correlation between the switching
fields and optical band gap of Ge15Te85-xSix thin films; both these parameters increase with
the increase in Si concentration. The increase in network connectivity and rigidity could
be the main reason for increasing optical band gap as Si is added into the Ge-Te system. A
similar trend has been found in many other chalcogenide glasses such as Ga-Te [41], Se-
Te-Ag [42], Se-Te-Pb [43], etc.
81
Figure 6.5: The Plot of (αhυ)2 vs photon energy (hυ) of a representative a-Ge15Te81Si4
thin film.
Figure 6.6: The variation of optical band gap with composition of Ge15Te85-xSix (1≤ x≤ 6)
thin films.
The variation in optical band gap with composition is usually explained on the
basis of structure [40,44], electron affinity [45,46], electro negativity [42,43] etc. As
proposed by Mott and Davis [47], the increase in the optical band gap can be understood
on the basis of density of state model. Chalcogenide thin films always contain a high
concentration of unsaturated bonds or defects. These defects are responsible for the
presence of localized states in the amorphous band gap. Attempts have been made in
82
literature to correlate directly the variation in optical band gap with composition with the
density of defect states [43,48]. However, these effects are not fully understood.
Switching Switching Switching Optical

Thickness of
Compositions voltages for field for film field for bulk band gap
the film (nm)
film (V) (kV/cm) (kV/cm) (eV)
Ge15Te84Si1 132 2.89 219.24 2.81 1.18
Ge15Te83Si2 451 13.44 297.65 3.64 1.20
Ge15Te82Si3 348 10.54 302.87 3.77 1.23
Ge15Te81Si4 249 7.79 313.05 4 1.26
Ge15Te80Si5 151 5.35 353.66 4.76 1.27
Ge15Te79Si6 220 7.99 362.64 7.67 1.29
Table 6.1: The variation of threshold electric field and optical band gap with composition
of amorphous Ge15Te85-xSix thin films.
6.3.3. Electron Microscopic Studies

Figure 6.7 (a) and (b) show the Scanning Electron Microscopic image of the switched
and un-switched Ge15Te80Si5 thin film sample respectively from which it is clear that there
is an image contrast in the electrode region indicating morphological changes at the place
of switching.
(a) (b)
Figure 6.7: The Scanning electron micrograph of the switched region (a) and unswitched
region (b) of a representative Ge15Te80Si5 thin film sample showing some morphological
changes.
83
6.4 Conclusions
Amorphous Ge15Te85-xSix thin films have been deposited in sandwich geometry on
glass substrates with aluminum electrodes. These devices have been found to exhibit
memory type electrical switching, as exhibited by bulk Ge15Te85-xSix glasses. Unlike the
bulk Ge15Te85-xSix glasses of higher Si content which show fluctuations in I-V
characteristics and splashing of the material in the electrode region, a-Ge15Te85-xSix films
exhibit a smooth electrical switching behavior for all compositions. Also, the observed
electrical switching voltages of a-Ge15Te85-xSix thin film samples are comparable with the
samples used in PCM applications such as Ge-Sb-Te, Si-Sb-Te, Si-Ge-Sb-Te, etc.
It is observed that the switching is relatively faster in thin film samples, compared
to the bulk sample which implies that fast phase change takes place during the SET
process in amorphous films which is an important aspect for PCM applications.
Further, the switching fields of a-Ge15Te85-xSix films have been found to increase
with increasing Si concentration, which is similar to the composition dependence of
switching fields of bulk Ge15Te85-xSix glasses. The variation with composition of the
optical band gap of a-Ge15Te85-xSix films has been found to be similar to that of electrical
switching fields. The observed composition dependence of switching fields and optical
band gap of both bulk glassy & amorphous thin film Ge15Te85-xSix samples can be
understood on the basis of increase in network connectivity and rigidity with Si content.
Scanning Electron Microscopic studies on a representative switched Ge15Te80Si5

sample shows an image contrast in the electrode region indicating morphological changes
at the place of switching.
84
References
[1] K. Ramesh, S. Asokan, K.S. Sangunni and E.S.R. Gopal, J. Phys. Chem. Solids, 61
(2000) 95.
[2] S. Manohar, S. Murugavel and S. Asokan, Solid State Commun., 135 (2005) 323.
[3] P. Pattanayak and S. Asokan, J. Non-Cryst. Solids, 354 (2008) 3824.
[4] N. Yamada, E. Ohno, K. Nishiuchi, N. Akahira and M. Takao, J. Appl. Phys., 69
(1991) 2849.
[5] I. Friedrich, V. Weidenhof, W. Njoroge, P. Franz and M. Wuttig, J. Appl. Phys., 87
(2000) 4130.
[6] A. Kolobov, P. Fons, A. Frenkel, A. Ankudinov, J. Tominaga and T. Uruga, Nat.
Mater., 3 (2004) 703.
[7] H.Y. Cheng, C.A. Jong, R.J. Chung, T.S. Chin and R.T. Huang, Semicond. Sci.
Technol., 20 (2005) 1111.
[8] R. Bez and A. Pirovano, Mater. Sci. Semicond. Process., 7 (2004) 349.
[9] K.S. Min, M.J. Shin, D.J. Choi, K.N. Lee, S.K. Hong and Y.J. Park, Thin Solid
Films, 470 (2004) 322.
[10] S. Privitera, E. Rimini and R. Zonca, Appl. Phys. Lett., 85 (2004) 3044.
[11] Y.F. Lai, B.W. Qiao, J. Feng, Y. Ling, Y.Y. Lin, T.A. Tang, B.C. Cai and B. Chen,
J. Electron. Matter., 34 (2005) 176.
[12] M.J. Shin, S.M. Kim, D.H. Choi, K.N. Lee, S.K. Hong and Y.J. Park, J. Mater. Sci.,
40 (2005) 1543.
Chin. Phys. Lett., 23 (2006) 172.
[14] J. Feng, Y. Zhang, B.W. Qiao, Y.F. Lai, Y.Y. Lin, B.C. Cai, T.A. Tang, and B.
Chen, Appl. Phys. A, 87 (2007) 57.
[15] B.W. Qiao, J. Feng, Y.F. Lai, Y. Cai, Y.Y. Lin, T.A. Tang, B.C. Cai and B. Chen,
Semicond. Sci. Technol., 21 (2006) 1073.
Appl. Surf. Sci., 252 (2006) 8404.
[18] M. Anbarasu, K.K. Singh and S. Asokan, Phil. Mag., 88 (2008) 599.
[19] M. Anbarasu, S. Prusty, S. Asokan and A.K. Sood, Appl. Phys. Lett., 91 (2007)
093520.
85
[20] J.C. Philips, J. Non-Cryst. Solids, 34 (1979) 153.
[22] R. Aravinda Narayanan, PhD Thesis, Indian Institute of Science, Bangalore (2001).
[23] S.S.K. Titus, R. Chatterjee, S. Asokan and A. Kumar, Phys. Rev. B, 48 (1993)
14650.
[25] S. Murugavel and S. Asokan, J. Non-Cryst. Solids, 249 (1999) 145.
[26] S. Murugavel and S. Asokan, J. Mater. Res., 13 (1998) 2982.
[28] B.J. Madhu, H.S. Jayanna and S. Asokan, J. Non-Cryst. Solids, 355 (2009) 459.
[30] M. Anbarasu and S. Asokan, Appl. Phys. A, 80 (2005) 249.
[31] C.L. Dargan, P. Burton, S.V. Phillips, A.S. Bloor and P. Nesvadba, J. Mater. Sci., 9
(1974) 1595.
[32] A. Pirovano, A.L. Lacaita, F. Pellizzer, S.A. Kostylev, A. Benvenuti and R. Bez,
IEEE Trans. Electr. Dev., 51 (2004) 714.
[33] A. Pirovano, A.L. Lacaita, D. Merlani, A. Benvenuti, F. Pellizzer and R. Bez, IEDM
Tech. Dig., 923 (2002).
[34] T. Matsuhita, A. Suzuki, M. Okuda, J.C. Rhee and H. Naito, J. Appl. Phys., 24
(1985) L504.
[35] M. Chen, K.A. Rubin and R.W. Barton, Appl. Phys. Lett., 49 (1986) 502.
[36] N. Yadama, M. Takao and M. Takenage, Proc. SPIE, 695 (1986) 79.
[37] M. Okuda, H. Naito and T.J. Matsushita, J. Appl. Phys., 31 (1992) 466.
[38] T. Ide, M. Suzuki and M. Okada, J. Appl. Phys., 34 (1995) L529.
[39] E. Kh. Shokr, J. Phys. Chem. Solids, 53 (1992) 1215.
[40] Kedar Singh, N.S. Saxena, O.N. Srivastava, D. Patidar and T.P. Sharma,
Chacogenide Letters, 3 (2006) 33.
[41] M. Ilyas, M. Zulfequar, Zishan H Khan and M. Husain, Opt. Mater., 11 (1998) 67.
[42] A.A. Al-Ghamdi, Vacuum, 80 (2006) 400.
[43] Shamshad A Khan, M. Zulfequar and M. Husain, Current. Appl. Phys., 5 (2005)
583.
[44] H. El-Zahed, M.A. Khaled, A. El-Korashy, S.M. Youssef and M. ElOcker, Solid
State Commun., 89 (1994) 1013.
[45] S.H. Nguyen and F.P. Koffyberg, Solid State Commun., 76 (1990) 1243.
86
[46] F.P. Koffyberg, J. Phys. Chem. Solids, 53 (1992) 1285.
[47] N.F. Mott and E.A. Davis, Electronics Processes in Non-Crystalline Materials,
Clarendon Press, Oxford, 1979.
[48] A.S. Maan, D.R. Goyal, S.K. Sharma and T.P. Sharma, J. Non-Cryst. Solids, 183
(1995) 186.
87
Chapter 7
The Effect of Sn Doping on the Electrical and Thermal

Properties of Bulk Ge-Te Glasses
7.1 Introduction
The phenomenon of electrical switching has been discovered nearly four decades
ago [1]; however, the realization of Phase Change Memories (PCMs), based on this
phenomenon became a reality only few years back. PCMs are considered to be more
advantageous compared to conventional DRAMs or flash memories, due to the scaling
limits of DRAM and low cycle life time, high power requirements & radiation sensitivity
of flash memories.
Studies on the electrical switching behavior of different chalcogenide glasses help

us in identifying newer glasses which could be used for phase change memory
applications [2]. In particular, studies on composition dependence of thermal properties
and electrical switching parameters are necessary to facilitate the selection of proper
compositions in a given glassy system having easy crystallizability, good glass forming
ability, good thermal diffusivity, lower switching voltages, etc., required for a good
memory material.
In this work, glasses belonging to two composition tie-lines of the Ge-Te-Sn

system, namely Ge15Te85-xSnx (1 ≤ x ≤ 6) and Ge17Te83-xSnx (1 ≤ x ≤ 5), have been taken
up for thermal and electrical switching investigations. Earlier investigations on bi-layers of
Ge and Sn-chalcogenides have revealed the possibility of obtaining the multi-state
behavior with enhanced ability to withstand thermal cycling and the possibility of
achieving desired phase change response at lower voltages [3]. The motivation behind the
present study is to systematically investigate the effect of Sn-doping on electrical
switching behavior of Ge-Te glassy system and also to understand the variation of thermal
parameters and its influence on the composition dependence of switching voltages.
Though the studies on the switching behavior of thin film samples are more relevant to
phase change memory applications, investigations on bulk samples help us to understand
the basic electrical and thermal properties of phase change glasses better and also correlate
these properties with the switching behavior.
Ch.7: The Effect of Sn Doping on the Electrical and Thermal…
7.2 Experimental Procedure

Bulk glasses of Ge15Te85-x Snx (1 ≤ x ≤ 6) and Ge17Te83-xSnx (1 ≤ x ≤ 5) have been
synthesized by the conventional melt quenching technique as discussed in chapter 2. The
amorphous nature of the sample prepared was confirmed by X-ray diffraction.
The bulk Ge15Te85-xSnx and Ge17Te83-xSnx glassy samples are polished to a

thickness of 0.3 mm and mounted in the sample cell described earlier. The current-voltage
characteristic and electrical switching behavior have been undertaken using a
programmable DC-source-meter (Keithley 2410c). The measurement on the I-V
characteristics has been repeated at least 3 times for each sample and the switching
voltages have been found to be repeatable within ± 2%.
The thermal parameter like crystallization temperature of the as quenched glassy

samples is obtained by ADSC experiments employing Mettler Toledo’s alternating
differential scanning calorimeter (model DSC 822c). The glass transition temperature (Tg)
and crystallization temperature (Tc) are measured from reversing and non-reversing heat
flow curve respectively. Samples in the weight range of 10 - 15 mg are encapsulated in Al
pans and heated in argon atmosphere (flow rate of 75 ml/min). Temperature ranges of 50-
3400C with a heating rate of 30C/min and 10C modulation rate have been used. The
temperature calibration is done using high purity indium. Typical error in the measurement
of Tg and Tc is within ± 2 0C.
To identify the devitrified phases, which appear during crystallization, two

representative samples from Ge15Te85-xSnx and Ge17Te83-xSnx series are sealed under a
vacuum of more than 10-4 mbar in a quartz ampoule and annealed at their respective
crystallization temperature for about 1 hour. These annealed samples are then analyzed by
X-ray diffraction using a Bruker AXS D8 Advance diffractometer.
7.3.1. Memory switching behavior of Ge-Te-Sn glasses
The I-V characteristics of representative samples of Ge15Te85-xSnx (1 ≤ x ≤ 6) and

Ge17Te83-xSnx (1 ≤ x ≤ 5) glasses are shown in figures 7.1 (a) and (b) respectively. It is
evident from these figures, that Ge15Te85-xSnx and Ge17Te83-xSnx samples exhibit an Ohmic
behavior initially (OFF state). At the threshold or switching voltage (Vt), a deviation from
90
the Ohmic behavior and a negative resistance region is observed and the samples switch to
a high conducting ON state.
Upon reducing the current after switching, Ge15Te85-xSnx and Ge17Te83-xSnx glasses
latch-on to the high conducting state, and do not revert back to their original high-
resistance OFF state. It is clear from the above that Ge15Te85-xSnx and Ge17Te83-xSnx
glasses exhibit a current controlled negative resistance and memory switching. The
electrical switching in both Ge15Te85-xSnx and Ge17Te83-xSnx samples has been found to be
smooth without any fluctuations in the negative resistance zone (figure 7.1).
(a)
(b)
Figure 7.1: I-V characteristics of (a) Ge15Te84 Sn1 and Ge15Te83 Sn2 glasses representing
the Ge15Te85-xSnx series and (b) Ge17Te82Sn1 and Ge17Te80 Sn3 glasses representing
Ge17Te83-xSnx series.
91
It is well known that poor structural cross linking, weaker bonds and more lone
pair interactions [4] favor memory switching in chalcogenide glasses. It is also understood
that memory switching phenomenon is more probable in chalcogenide glass compositions
in which energy barrier is low for local structural changes [5].
The earlier investigations indicate that chalcogenide glasses such as the present
Ge15Te85-xSnx and Ge17Te83-xSnx samples, with an abundance of divalent elements and
having a low network connectivity/rigidity, are more susceptible for local structural
changes due to poor bond strength and flexibility [5]. These floppy glasses [6-9], which
also have lesser thermal stability [10], lower activation energy for crystallization, etc. can
permit easy creation of nucleation sites for an extended structural phase transition of the
type required for memory switching [5].
7.3.2 Composition dependence of switching voltages
Figures 7.2 (a) and (b) show the variation of switching voltages of Ge15Te85-xSnx
and Ge17Te83-xSnx glasses with Sn content, which indicates that Vt decreases with the
increase in Sn content. A similar decrease in threshold voltage with metallic additives has
also been observed in other Ge-Te based chalcogenide systems such as Ge-Te-Cu, Ge-Te-
Ag [11], etc.
The inset in figures 7.2 (a) and (b) show the variation of initial electrical resistance
(resistance in the OFF state) with composition, indicating that the resistance in the OFF
state of Ge15Te85-xSnx and Ge17Te83-xSnx glasses decrease with Sn content. It is well known
that there is a direct empirical relationship between the switching voltages and resistance
of the chalcogenide glasses [12] and the present Ge15Te85-xSnx and Ge17Te83-xSnx glasses
also follow this rule.
As revealed by earlier studies [6,8,13-18] and as described in earlier chapters, the

composition dependence of switching voltages of chalcogenide glasses is determined by
factors such as the resistivity of the additive element, the network connectivity, rigidity
percolation etc. In Ge15Te85-xSnx and Ge17Te83-xSnx glasses, the addition of more metallic
Sn at the expense of Te, leads to the decrease in switching voltages as the resistivity of tin
is much lower than that of Te (ρSn= 11.5 µΩcm and ρTe=104 µΩcm). Such a behavior is
seen in many other systems, including Ge-Se-Tl, Ge-Se-Sb, As-Te-Tl, As-Te-Cu, etc. [19-
22].
92
Figure 7.2: Composition dependence of switching voltage (Vt) of (a) Ge15Te85-x Snx and
(b) Ge17Te83-xSnx glasses. The insets show the variation of starting electrical resistance
with composition.
Apart from the resistivity of the additive element, network connectivity and
rigidity play an important role in the composition dependence of switching voltages of
chalcogenide glasses. The addition of higher coordinated elements increases the network
connectivity and rigidity, leading to an increase in switching voltages.
In general, metallic elements such as Al, In, etc., are considered to enter the
chalcogen glassy network samples, with a four-fold or higher coordination, increasing the
network connectivity [23-25]. The earlier work suggests that Sn atoms are either four-fold
or six-fold coordinated in a similar Ge-Se(S)-Sn glassy system [26]. Based on the above, it
93
can be assumed that Sn is four-fold or higher coordinated in Ge-Te-Sn system. Therefore,
the addition of higher coordinated Sn (at the expense of Te) is expected to increase the
network connectivity and consequently the switching voltages in the Ge-Te-Sn glassy
system. However, the present results show a decrease in Vt with Sn content. It can be
concluded that the metallicity of the dopant and the resistivity factor contribute more to
the composition dependence of Vt of Ge-Te-Sn samples in comparison to the network
connectivity.
Figure 7.3: Variation of crystallization temperature Tc of (a) Ge15Te85-x Snx and (b)
Ge17Te83-xSnx glasses with Sn content
The variation of crystallization temperatures (Tc) obtained from ADSC
measurements, with composition, of Ge15Te85-xSnx (1 ≤ x ≤ 5) and Ge17Te83-xSnx (1 ≤ x ≤
94
4) glasses is shown in figures 7.3 (a) and (b) respectively. From these figures, it is clear
that the crystallization temperatures decrease with the addition of Sn atom into the
network. Figures 7.4 (a) and (b) show the variation of (Tc-Tg), which reflects the thermal
stability of glasses with composition. It is found that (Tc-Tg) of Ge15Te85-xSnx and
Ge17Te83-xSnx glasses decrease with increase in Sn concentration, indicating that the Ge-Te
glasses become more easily devitrifiable with the addition of Sn. The decrease in thermal
stability also contributes for the observed decrease in the switching voltages of Ge15Te85-
xSnx and Ge17Te83-xSnx glasses with Sn content.
Figure 7.4: Variation of (Tc - Tg) of (a) Ge15Te85-x Snx and (b) Ge17Te83-xSnx glasses with
Sn content.
In order to know about the crystalline phases formed after devitrification,
representative Ge15Te85-xSnx and Ge17Te83-xSnx samples are annealed at their respective
crystallization temperatures for 1 hour and X-ray diffraction studies were performed on
95
these samples. Figure 7.5 (a) shows the X-ray diffraction patterns of a representative as-
quenched Ge17Te80Sn3 sample, confirming its amorphous nature. Figure 7.5 (b) and (c)
show the diffraction patterns of Ge15Te82Sn3 and Ge17Te80Sn3 samples respectively, after
annealing. The diffraction patterns of crystallized Ge-Te-Sn samples can be fully indexed
with hexagonal Te and hexagonal Ge-Te phases. No elemental Sn or Sn compounds with
Te or Ge have been found in Ge15Te82 Sn3 and Ge17Te80Sn3 glasses.
Figure 7.5: X-Ray Diffraction pattern of (a) as-quenched Ge17Te80Sn3 glass (b)
Ge15Te82Sn3 glass and (c) Ge17Te80Sn3 glass after annealing upto 1 hour at their respective
crystallization temperature.
When a dopant atom is added to a binary chalcogenide glass, it can form its own
structural units in the system. The glass formation in the system will be enhanced, if the
newly formed structural units interact with the parent glass matrix and increase the
network connectivity [27]. However, if the added impurity does not interact with the host
matrix, it forms a phase separated network of its own. This can remain in the parent glass
matrix as a micro-inclusion. The glass formation in Ge15Te85-xSnx and Ge17Te83-xSnx
composition tie-lines is limited to a maximum of 6 at % and 5 at % of Sn respectively.
Based on the X-ray diffraction results, it can be suggested that in Ge15Te85-xSnx and
Ge17Te83-xSnx glasses, the interaction of Sn atoms with the Ge-Te network is not
significant and they do not take part actively in enhancing the network connectivity.
Another possibility is that the Sn atoms randomly replace Ge atoms in the Ge-Te
network, as tin takes four fold co-ordination using the sp3 hybrid orbital [28] and Ge also
bonds with four fold co-ordination in the Ge-Te network [29]. Further, the electro-
negativity of Ge and Sn are 2.01 and 1.96 respectively, which are fairly closer. It is
96
possible that the similar co-ordination and electro-negativity of Ge and Sn in Ge15Te85-
xSnx and Ge17Te83-xSnx samples lead to random replacement of Ge by Sn, and hence there
is no net increase in the overall network connectivity with the addition of Sn. The random
replacement of Ge by Sn also limits the glass formation in Ge-Te-Sn system to less than 6
at % of Sn.
In the absence of the influence of network connectivity, the switching voltages of

Ge15Te85-xSnx and Ge17Te83-xSnx glasses decrease with Sn content, as determined by the
resistivity of the additive element and the decrease in thermal stability.
7.3.3 Variation of threshold voltage with thickness
The variation of switching voltages (Vt) with thickness is shown in figure 7.6 (a)
and (b) for two representative samples, Ge15Te82Sn3 and Ge17Te80Sn3 respectively. The
switching voltages will vary as t, t1/2 or t2, depending on whether the mechanism
responsible for switching is electronic, thermal or based on carrier injection [30].
Generally Vt is found to be proportional to t or t1/2 in samples exhibiting memory
switching with thermal origin [31].
Memory switching samples like Ge-Te-Al [32], As-Se [33], Ge-Se-Tl [34] etc. are
found to show a linear variation of switching voltages with thickness whereas in memory
samples like Ge-As-Te [35], Ge-Te [36] threshold voltage varies as t1/2. In the present
Ge15Te85-xSnx and Ge17Te83-xSnx glasses also, it is observed that the switching voltages
vary linearly with thickness which are in good agreement with memory type of switching
behavior shown by both the series of glasses.
97
Figure 7.6: Thickness dependence of Vt in (a) Ge15Te82Sn3 and (b) Ge17Te80Sn3 glasses
representing the Ge15Te85-x Snx and Ge17Te83-xSnx series respectively.
7.4 Conclusions
Melt quenched Ge15Te85-x Snx (1 ≤ x ≤ 6) and Ge17Te83-xSnx (1 ≤ x ≤ 5) glasses are
found to exhibit memory type of electrical switching with threshold voltage in the range of
25-100V (for 0.3 mm thick sample). The composition dependence and thickness
dependence of threshold voltage has been studied for both the series of samples. The
observed variation of Vt with composition is understood on the basis of network
98
flexibility, metallicity factor and lower resistivity of additive metallic element. Further, the
variation with composition of switching voltages is in agreement with the composition
dependence of crystallization temperature (Tc) for both the series of samples. X-ray
diffraction studies on thermally devitrified samples indicate that the interaction of Sn with
the Ge-Te matrix is not significant and they do not take part actively in enhancing the
network connectivity. This is also a reason for the observed composition dependence of
switching voltages.
The thickness dependence of switching voltages is found to be linear in these two

series of glasses, which is in good agreement with memory type of switching behavior
shown by both the series of glasses. The lower threshold switching voltages and memory
type of switching behavior of these glasses can make them a suitable candidate for phase
change memory application.
99
References
[1] S.R. Ovshinsky, Phys. Rev.Lett., 21 (1968) 1450.
[2] M. Wuttig, Nat. Mater., 4 (2005) 265.
[3] K.A. Campbell and C.M. Anderson, J. Microelectron., 38 (2007) 52.
[4] S.R. Ovshinsky, Phys. Rev. Lett., 36 (1976)1469.
[5] J. Bicerano and S.R. Ovshinsky, J. Non-Cryst. Solids, 74 (1985) 75.
[9] J.C. Phillips, Phys. Rev. B, 54 (1996) R6807.
[10] D. Adler, Sci. Am., 236 (1977) 36.
[11] K. Ramesh, S. Asokan, K.S. Sangunni and E.S.R. Gopal, Appl. Phys A, 69 (1999)
421.
[12] A. Alegria, A. Arruabarrena and F. Sanz, J. Non-Cryst. Solids, 58 (1983) 17.
[13] R. Aravinda Narayanan, PhD Thesis Indian Institute of Science, Bangalore (2000).
[14] S.S.K.Titus, R. Chatterjee, S. Asokan and A. Kumar, Phys. Rev. B, 48 (1993) 14650.
[21] B.H. Sharmila and S. Asokan, Appl. Phys. A, 82 (2006) 345.
[22] A.S. Soltan and A.H. Moharram, Physica B, 349 (2004) 92.
[23] B.H. Sharmila, J.T. Devaraju and S. Asokan, J. Non-Cryst. Solids, 326–327 (2003)
154.
[24] J.T. Devaraju, B.H. Sharmila, S. Asokan and K.V. Acharya, Phil. Mag. B, 81 (2001)
583.
[25] S. Murugavel, Ph. D Thesis, Indian Institute of Science, Bangalore (1998).
[26] Mark Steevns, J. Grothaus and P. Boolchand, Solid State Commun., 47 (1983) 199.
[27] Z.U. Borisova, Glassy Semiconductors, Plenum Press, New York, 1985.
[28] T. Fukunaga, Y. Tanaka and K. Murase, Solid State Commun., 42 (1982) 513.
100
[29] I. Kaban, Th. Halm, W. Hoyer, P. Jovari and J. Neuefeind, J. Non-Cryst. Solids,
326&327 (2003) 120.
[30] G. Jones and R.A. Collins, Phys. Stat. Sol. (a), 53 (1979) 339.
[31] J.J. Odwyer, The Theory of Electrical Conduction and Breakdown in Solid
Dielectrics, Clarendon Press, Oxford, 1973.
[32] J.R. Bosnell and C.B. Thomas, Solid-State Electron., 15 (1972) 1261.
[33] A. Giridhar and K.J. Rao, J. Non-Cryst. Solids, 33 (1979) 177.
[34] M.F. Kotkata, M.A. Afifi, H.H. Labib, N.A. Hegab and M.M. Abdel-Aziz, Thin Solid
Films, 240 (1979) 143
[35] H.J. Stocker, C.A. Barlow and D.F. Weirauch, J. Non-Cryst. Solids, 4 (1970) 523.
[36] E. Babenskas, S. Balevicius, A. Cesnys, A. Poskus and N. Siktorovl., J. Non-Cryst.
Solids, 90 (1987) 601.
101
Chapter 8
Studies on Electrical Switching Behavior and Optical

Band gap of Amorphous Ge-Te-Sn Thin Films
8.1 Introduction
It has been suggested that the addition of metallic elements such as Cu, Ag, Tl, In,
Bi, etc. in the Ge-Te network changes the electrical switching properties of parent glass to
a large extent [1-4]. In particular, the Ge-Te-Sn system is found to be interesting from the
point of view of optical information storage and phase change memories, even though the
number of optically induced write/erase cycles that could be achieved is quite low in these
samples (< 500) [5]. It is also reported that Sn doped Ge2Sb2Te5 samples have lower
transition temperatures and ultra-fast crystallization (< 10 ns) [6], which are beneficial for
memory applications. PCM devices employing bi-layers of Ge-Chalcogenide and Sn-
Chalcogenide have been studied earlier, to explore the possibility of a multistate behavior,
enhanced ability to withstand thermal cycling and use of lower voltages for achieving
desired phase change response [7]. In addition to Ge-Te-Sn, there are other Sn based
alloys studied for PCM applications which include SnSe, SnSe2, Sn doped Ge2Sb2Te5 etc.
[8,9].
In this work, efforts have been made to investigate systematically, the electrical
switching and optical properties of amorphous Ge15Te85-xSnx (1 ≤ x ≤ 5) and Ge17Te83-xSnx
(1 ≤ x ≤ 4) thin films to find their applicability for phase change memory applications.
8.2 Experimental Details

Bulk Ge15Te85-xSnx (1 ≤ x ≤ 5) and Ge17Te83-xSnx (1 ≤ x ≤ 4) glasses are prepared
using melt quenching technique. These bulk glassy ingots are powdered and used as a
source material for depositing amorphous thin film Ge15Te85-xSnx and Ge17Te83-xSnx
switching devices in sandwich geometry using flash evaporation technique. Aluminum is
used as the material for both top and bottom electrodes. During the thin film deposition,
the vacuum inside the chamber is maintained at 2×10-6 mbar and the substrate is
maintained at liquid nitrogen temperature. First, aluminum is coated on a 25×75mm glass
slide. On top of that, the chalcogenide film is coated in a rectangular pattern. Finally, the
Ch.8: Studies on Electrical Switching Behavior and Optical Band gap of Amorphous…
top electrodes (dots of 5 mm diameter) have been deposited using a mechanical mask,
resulting in at least ten active devices on a single substrate.
The I-V characteristics and electrical switching behavior of the devices have been
undertaken using a Keithley (2410c) source-measure unit using LabVIEW 7 (National
Instruments). The thickness of the films (measured by Veeco Dektak 150 profilometer)
has been found to be in the range of 100-300 nm. The optical absorption of the films has
been studied in the wavelength range 200-1000 nm with an Ocean optics HR4000
spectrometer.
8.3.1. Electrical Switching behavior of amorphous Ge15Te85-xSnx and

Ge17Te83-xSnx Thin Films
Figures 8.1 (a) and (b) show the XRD patterns of representative Ge15Te82Sn3 and
Ge17Te81Sn2 thin films, confirming their amorphous nature.
Figure 8.1: XRD figures of as deposited (a) Ge15Te82Sn3 and (b) Ge17Te81Sn2 thin films.
The electrical switching behavior of Ge15Te82Sn3 and Ge17Te80Sn3 switching
devices, representing Ge15Te85-xSnx and Ge17Te83-xSnx series respectively are shown in
figure 8.2 (a) and (b). From the I-V characteristics, it is clear that the amorphous thin film
devices, belonging to both the series exhibit memory type electrical switching behavior
with switching voltages in the range 0 – 8 V. The comparatively lower switching voltages
of Ge15Te82Sn3 and Ge17Te80Sn3 switching devices indicate that these samples may be
suitable for phase change memory applications.
104
Figure 8.2: I-V characteristics of (a) Ge15Te82Sn3 and (b) Ge17Te80Sn3 glasses
representing
Ge15Te85-xSnx and Ge17Te83-xSnx series respectively.
Figures 8.3 (a) and (b) show the composition dependence of switching fields of
amorphous Ge15Te85-xSnx and Ge17Te83-xSnx thin film switching devices respectively. It
can be seen from these figures that the switching fields of Ge15Te85-xSnx and Ge17Te83-xSnx
samples decrease with increase in Sn content.
As discussed in earlier chapters and reported in many papers earlier [10-17] the
composition dependence of switching voltages/fields of bulk chalcogenide glasses are
determined mainly by factors such as resistivity & coordination number of the additive
element, the network connectivity & rigidity, thermal stability, etc. With the addition of
more metallic dopants, the switching voltages/fields of bulk chalcogenide glasses are
found to decrease [1,4]. The present results clearly indicate that this trend (known as
metallicity effect) is valid in amorphous thin films also.
Apart from the metallicity effect, the factors which influence the composition
dependence of switching voltages/fields of bulk chalcogenide glasses is the connectivity
and rigidity of the glassy network. With increase in network connectivity and rigidity, the
switching voltages/fields of chalcogenide glasses (particularly memory switching samples)
are found to increase. Further, interesting variations, such as sharp slope change, maxima,
etc., have been seen in the composition dependence of switching voltages/ fields of bulk
chalcogenide glasses at the rigidity percolation threshold [3,18,19]. Also, in the present
thesis work, the first observation of the manifestation of the stiffness transition in the
composition dependence of switching fields of amorphous chalcogenide films has been
found in the Ge-Se-Te system [chapter 5,20].
105
Figure 8.3: Variation of threshold voltage with composition for (a) Ge15Te85-xSnx and
(b) Ge17Te83-xSnx amorphous thin films.
It is well known that Ge and Te atoms bond with 4-fold and 2-fold coordination
respectively in bulk glassy and amorphous thin film chalcogenide systems. Further,
metallic elements are considered to enter the chalcogenide glassy networks, with a four-
fold or higher coordination, increasing the network connectivity [21-23]. An earlier work
also suggests that Sn atoms are either four-fold or six-fold coordinated in a similar Ge-
Se(S)-Sn glassy system [24]. The addition of higher coordinated Sn (at the expense of Te)
is therefore expected to increase the network connectivity and consequently the switching
voltages in the Ge-Te-Sn glassy system. However, the present results indicate that no such
106
increase in switching fields and also no manifestation of rigidity percolation is seen in the
composition dependence of switching fields of Ge15Te85-xSnx and Ge17Te83-xSnx thin film
series of samples.
The composition dependence of switching fields of amorphous Ge15Te85-xSnx (1 ≤

x ≤ 5) and Ge17Te83-xSnx (1 ≤ x ≤ 4) thin films can be understood on the basis of the
chemically ordered network (CON) model [25], according to which heteropolar bonds
formation is always favored over the homopolar bond formation. The bond strengths of
various possible bonds in the Ge-Te-Sn system, namely Ge-Te, Te-Te, Sn-Sn, Sn-Te, Ge-
Ge and Ge-Sn, are 396.7 ± 3.3 kCal / mol, 257.6 ± 4.1 kCal/mol, 187.1±0.3 kCal/mol,
338.1±6.3 kCal/mol, 264.4±6.8 kCal/mol and 230.1±13 kCal/mol respectively [26]. As the
Ge-Te and Sn-Te heteropolar bonds have higher bond strength, compared to Te-Te, they
are more probable. Therefore, as the Sn content increases, more and more Te-Te bonds are
broken and the probability of ionic Sn-Te bond formation increases.
Table 8.1 shows the difference between the Ge-Te and Sn-Te systems from
chemical bonding point of view. The RσAB and RπAB correspond to the ionicity and
metallicity of the A-B bond [27], respectively. The metallicity of Ge-Te bonds differ little
from that of Sn-Te bond but the ionicity of the Sn-Te is much larger than that of the
former. A similar effect has been seen in (Ge,Sn)Se system also [28].
RσAB (a.u) RπAB (a.u)
Ge-Te 0.02 0.48
Sn-Te 0.25 0.51
Table 8.1: Classification by the chemical bonding; 𝑅𝜎𝐴𝐵 = |(𝑟𝑠𝐴 + 𝑟𝑝𝐴 ) − (𝑟𝑠𝐵 − 𝑟𝑝𝐵 )| and
𝑅𝜋𝐴𝐵 = {|𝑟𝑝𝐴 − 𝑟𝑠𝐴 | + |𝑟𝑝𝐵 − 𝑟𝑠𝐵 |}. The orbital dependent radii rs and rp determined by the
Pauli force model potential [27].
It is also known earlier that, the increase in the number of heteropolar bonds leads
to the decrease in threshold switching voltage, whereas, an increase in the number of
homopolar bonds leads to the growth of chemical disorder resulting in an increase of
switching voltages [17]. In the present Ge-Te-Sn system, as the Sn content increases, the
number of heteropolar Sn-Te bond formed increases as compared to the Te-Te homopolar
bonds. Due to the increase in the heteropolar bonds, the chemical disorder is reduced and
therefore charge carriers are less localized resulting in an enhanced conductivity [17]. This
resultant increase in the conductivity is responsible for decreasing switching fields with
107
increase in Sn content. A similar decrease in switching voltages due to increase in the
chemical order and due to the metallicity of the additive element has been observed in
other chalcogenide systems also [29-32].
Thus, the increase in chemical order due to heteropolar bonding and metallicity
factor of the additive, contribute together to the observed significant decrease in switching
fields of Ge15Te85-xSnx and Ge17Te83-xSnx samples with Sn addition.
8.3.2. Optical band gap of amorphous Ge15Te85-xSnx and Ge17Te83-xSnx

Thin Films
The optical band gap of amorphous Ge15Te85-xSnx and Ge17Te83-xSnx thin films has
been calculated from the Tauc plot [33]. The spectral variations in absorption coefficient α
are described by the following equation
αhυ = B ( hυ - Eg)m ….. (1)
Where hυ is the photon energy, B is a parameter that depends on the transition

probability and Eg is the energy gap. Figures 8.4 (a) and (b) show the plots of (αhυ)2 Vs hυ
for two representative samples Ge15Te80Sn5 and Ge17Te81Sn2 from Ge15Te85-xSnx and
Ge17Te83-xSnx series respectively. The value of optical band gap (Eg) has been calculated
by taking the intercept on the x-axis. The band gap values calculated from the plots for all
the samples for both the series are tabulated in Table 8.2 along with their switching fields.
The variation of optical band gap of amorphous Ge15Te85-xSnx and Ge17Te83-xSnx

thin films with composition are shown in figure 8.5 (a) and (b), which indicate that optical
band gap decreases with increase in atomic percent of Sn for both the series of thin films.
The reason behind this could be that the Sn-atoms bind their valence electrons more
loosely than the Ge-atoms. As the number of Sn atoms increases, the difference in energy
between the bonding and anti-bonding states at the band edge decreases.
108
Figure 8.4: Plots of (αhυ)2 Vs hυ for (a) Ge15Te83Sn2 and (b) Ge17Te81Sn2 samples
representing the Ge15Te85-xSnx and Ge17Te83-xSnx series respectively.
The decrease in optical band gap with composition is consistent with the decrease
in the switching fields of amorphous Ge15Te85-xSnx and Ge17Te83-xSnx thin films. The
conductivity activation energy of amorphous and glassy chalcogenides is half of the
optical gap and the decrease in optical gap corresponds directly to a decrease in activation
energy for electrical conduction and better conductivity. The switching fields are known to
decrease with improved conductivity of the chalcogenide samples. A similar behavior has
been seen in Ge1-xSnxSe2 system also [34].
109
Figure 8.5: Variation of the optical band gap of (a) Ge15Te85-xSnx and (b) Ge17Te83-xSnx
thin films with composition.
110
Optical Band gap
Threshold Filed (𝐸𝑡 )
Series Composition (𝐸𝑔 )
(kV/cm)
(eV)
Ge15Te84Sn1 161.40 1.52
Ge15Te83Sn2 133.02 1.47
Ge15Te85-xSnx Ge15Te82Sn3 127.08 1.36
Ge15Te81Sn4 117.98 1.28
Ge15Te80Sn5 113.33 1.2
Ge17Te82Sn1 425.62 1.27
Ge17Te81Sn2 377.77 1.26
Ge17Te83-xSnx
Ge17Te80Sn3 258 1.17
Ge17Te79Sn4 57 1.12
Table 8.2: The Switching fields and optical band gaps of amorphous Ge15Te85-xSnx and
Ge17Te83-xSnx thin films showing a decreasing trend with increasing Sn concentration.
8.4. Conclusions
Electrical switching studies on amorphous Ge15Te85-xSnx and Ge17Te83-xSnx thin
films, indicate that these samples exhibit memory type electrical switching. The switching
fields for both the series have been found to decrease with increase in Sn concentration,
which confirms that the “metallicity effect”, namely the decrease in switching
voltages/fields with the addition or metallic dopants, is valid in amorphous chalcogenide
thin films also. In addition, no manifestation of rigidity percolation is seen in the
composition dependence of switching fields of Ge15Te85-xSnx and Ge17Te83-xSnx
amorphous thin film samples. Efforts have been made to understand the observed
composition dependence of switching fields of amorphous Ge15Te85-xSnx and Ge17Te83-
xSnx thin films on the basis of Chemically Ordered Network model.
Further, the optical band gap for these samples, calculated from absorption
spectra, has been found to exhibit a decreasing trend with increasing Sn concentration
supporting the composition dependence of switching fields. The observed variation of
optical gap with Sn content has been explained on the basis of loosely bound valence
electrons of Sn atoms which decrease the energy difference between bonding and anti-
bonding states at the band edge. Finally, it can be concluded that the comparatively lower
switching voltages and lower optical band gap of the amorphous Ge-Te-Sn thin film
samples, make them suitable for phase change memory applications.
111
References
[1] K. Ramesh, S. Asokan, K.S. Sangunni and E.S.R. Gopal, Appl. Phys. A, 69 (1999)
421.
[2] K. Ramesh, S. Asokan, K.S. Sangunni and E.S.R. Gopal, J. Phys. Chem. Solids, 61
(2000) 95.
[4] Chandasree Das, G.Mohan Rao and S. Asokan, J. Non-Cryst. Solids, 357 (2011) 165.
[5] M. Chen and K.A. Rubin, Progress of erasable phase-change materials, SPIE Vol.
1078; Optical Data Storage Topical Meeting, 1989.
[6] K. Wang, C. Steimer, D. Wamwangi, S. Ziegler, M. Wuttig, J. Tomforde and W.
Bensch, Mycrosyst. Technol., 13 (2007) 203.
[7] Kristy A. Campbell and Christopher M. Anderson, Microelectronic Journal, 38
(2007) 52.
[8] Kyung-Min Chung, Daniel Wamwangi, Michael Woda, Matthias Wuttig and
Wolfgang Bensch, J. Appl. Phys., 103 (2008) 083523.
[9] K. Wang, D. Wamwangi, S. Ziegler, C. Steimer, M. J. Kang, S. Y. Choi and M.
Wuttig, Phys. Stat. sol. (a), 201 (2004) 3045.
[11] M.F. Thorpe, J.Non-Cryst.Solids, 57 (1983) 355.
[12] R. Aravinda Narayanan, PhD Thesis Indian Institute of Science,Bangalore, 2001.
[13] S.S.K. Titus, R. Chatterjee, S. Asokan and A. Kumar, Phys.Rev. B, 48 (1993) 14650.
[20] Chandasree Das, M.G. Mahesha, G. Mohan Rao and S. Asokan, Thin Solid Films
(Communicated).
[21] B.H. Sharmila, J.T. Devaraju and S. Asokan, J. Non-Cryst. Solids, 326–327 (2003)
154.
[22] J.T. Devaraju, B.H. Sharmila, S. Asokan and K.V. Acharya, Phil. Mag. B, 81 (2001)
583.
112
[23] S. Murugavel, Ph. D Thesis, Indian Institute of Science, Bangalore (1998).
[24] Mark Steevns, J. Grothaus and P. Boolchand, Solid State Commun., 47 (1983) 199.
[25] G. Lucovsky, F.L. Galeener, R.H. Geils and R.C. Keezer, (edited by P.H. Gaskell),
The Structure of Non-Crystalline Materials, Taylor and Francis, London, 1977.
[26] David R. Lide, CRC Handbook of Chemistry and Physics 90 Internet Version,
CRC Press/Taylor and Francis, Boca Raton, FL, 2010.
[27] J.R. Cehlikowsky and J.C. Phillips, Phys. Rev. B, 17 (1978) 2453.
[28] T. Fukunaga. Y. Tanaka and K. Murase, Solid State Commun., 42 (1982) 513.
[31] M.A. Afifi, M.M. Abdel-Aziz, H.H. Labib, M. Fadel and E.G. El-Metwally, Vacuum,
61 (2001) 45.
[32] A.S. Soltan and A.H. Moharram, Physica B, 349 (2004) 92.
[33] J. Tauc, Amorphous and Liquid Semiconductor, Plenum Press, New York,1974.
[34] J.M. Mikrut, L.E. McNeil, J. Non-Cryst. Solids, 114 (1989) 127
113
Chapter 9
Nano-Indentation Studies on Bulk Ge-Te-Si Chalcogenide

Glasses
9.1 Introduction
Bulk chalcogenide glasses are prepared by melting together Group 16 elements (S,
Se and Te) with other elements, generally from Group 14 (e.g. Ge, Si) and Group 15 (e.g.
Sb, As) of the periodic table. When mixed in an oxygen and water deficient atmosphere,
glasses will form with unique optical and semiconducting properties.
Extensive studies have been done on optical and electrical characterization of

chalcogenide glasses. The suitability of the glasses for an application often depends upon
their mechanical properties. Various studies have been done theoretically and
experientially to understand the mechanical behavior of covalent glasses on the basis of
their elastic constants, micro hardness, brittleness etc. [1-4].
Indentation tests are extremely useful to understand the elasto-plastic behavior of

materials, for gaining insights into the micro-mechanisms of plastic deformation as well as
developing an understanding of the material’s response to multi-axial loading. Because of
these advantages, the indentation response of amorphous solids has been extensively
studied. This chapter reports systematic study on the variation of elastic modulus and
hardness of Ge15Te85-xSix glasses (0 ≤ x ≤ 9) with composition.
9.2 Experimental techniques

Bulk, glassy Ge15Te85-xSix samples (0 ≤ x ≤ 9) are prepared by melt quenching
technique. The amorphous nature of the quenched glasses is confirmed by x-ray
diffraction. A comparatively big piece of sample from each composition of the series is
embedded in polymer mould, and the surface is polished using a diamond suspension
starting from 9 µm to 100 nm particle size, followed by colloidal silica of 25 nm particle
size. A Triboindenter (Hysitron, Minneapolis, USA), with an in-situ imaging capability, is
used to carry out nano indentation studies.
The Triboindenter continuously monitors and records the load, P, and

displacement, h, of the indenter with force and displacement resolutions of 1 nN and 0.2 ,
Ch.9: Nano-Indentation Studies on Bulk…
respectively. A Berkovich tip (three-sided pyramidal diamond tip) of diameter 100 nm has
been used for nano-indentation. In order to identify flat regions for the reliable and
repeatable experiment, the pellet surfaces are imaged prior to indentation using the same
indenter tip. A peak load of 8 mN is employed with a loading and unloading rate of 1
mN/s and a hold time of 10 s. A minimum of 10 indentations have been performed and the
standard deviations of the values obtained are calculated. The indentation impressions are
also captured immediately after unloading, so as to avoid time-dependent elastic recovery
of the residual impression.
The density of the samples in the series is measured using a density measurement
kit (model AG64, METTLER TOLEDO) with C2H5OH as a reference liquid. The error in
the density is estimated to be less than 2%.

Figure 9.1 (a) and (b) shows the X-ray diffraction patterns of two representative
Ge15Te83Si2 and Ge15Te80Si5 samples, confirming their amorphous nature. The
representative load, P–displacement, h curves of the three representative samples, namely
Ge15Te85, Ge15Te80Si5 and Ge15Te77Si8 are shown in figure 9.2.
While the maximum penetration depth is highest (~0.550 µm) for GeTe system, it
is lowest (~0.45 µm) when GeTe is doped with 8 at % of silicon. The loading portion of
the curves reveals that the plastic flow is homogeneous and there is no signature of
displacement bursts. The elastic recoveries after unloading are 23 % and 29 % for GeTe
and Ge15Te77Si8, respectively. It is clear that the elastic recovery upon unloading is
increased with increasing silicon content in the GeTe glass.
Figure 9.3 shows the residual impression on the specimen surface obtained using
the inbuilt Triboscanner imaging facility, which shows that there are no cracks produced
in the sample during indentation. As expected, no pile-up around the sides or corners of
the indenter is observed, because glasses contain large amount of free volume. The P-h
curves have been analyzed using the Oliver–Pharr method [5] to extract the hardness, H,
and the elastic modulus, E.
116
Figure 9.1: X-ray diffraction pattern of two representative sample Ge15Te83Si2 and
Ge15Te80Si5 samples confirming their amorphous nature.
117
Figure 9.2: Load-displacement curve of three representative samples Ge15Te85,

Ge15Te80Si5, Ge15Te77Si8.
Figure 9.3: Residual impression of nano-indentation on the specimen surface of

Ge15Te80Si5 sample.
Figure 9.4 shows the variation of Young’s Modulus (E) of Ge15Te85-xSix glasses,
with composition. It is clear from figure 9.4 that E increases with silicon content
(x)/average coordination number (rc), with a plateau exhibited in the composition
dependence of E in the range x = 2 to x = 6,
118
Figure 9.4: Variation of Young’s Modulus of Ge15Te85-xSix glasses with

composition.
Earlier studies on the composition dependence of enthalpy change during glass
transition obtained from non-reversing heat flow (ΔHNR) have shown a broad through in
ΔHNR in the composition range x = 2 (rc = 2.34) to x = 6 (rc = 2.42) [6]. Further, ΔCp, the
specific heat of the system which can be defined as function of the flexibility of the glassy
network has also been found to show and minimum around x = 5 [6].
As discussed in more detail in chapter 1, in certain glassy systems, the rigidity

percolation span over a range of compositions [7,8]. In these systems, two distinct
transitions are found, namely from floppy to isostatically rigid state and from isostatically
rigid to stressed rigid phase. The composition range between the isostatically rigid to
stressed rigid phase constitutes the intermediate phase, which is also known as thermally
reversing window. The configurational entropy difference between a glass and its liquid
state is supposed to be minimal in the thermally reversing window and the glass
compositions in the thermally reversing window are stress free. The width of the thermally
reversing window varies from one glassy system to another. Generally ternary glassy
systems have a much wider window, whereas thermally reversing windows in
chalcohalide are extremely narrow [9,10]. Various theoretical calculations based on
techniques like graph theory, constraints counting, cluster approximations etc. [11],
validate the existence of thermally reversing window, which has also been confirmed by
experiments like alternating differential scanning calorimetry (ADSC), Raman scattering
etc [12-15]. From the thermal studies of many different glassy systems, the non-reversing
119
heat flow (ΔHNR) has been found to be a minimum for composition in the thermally
reversing window [8].
The plateau exhibited in the composition dependence of E in the range x = 2 to x =

6, is associated with the thermally reversing window identified on the basis of variation of
ΔHNR with composition. Since elastic modulus is also a measure of stiffness of a material,
the variation of modulus in the range of x = 2 to x = 6 is consistent with thermal results.
Figure 9.5 shows the variation of hardness with compositions, which also shows a
trend similar to elastic modulus. The hardness (H) of Ge15Te85-xSix glasses increases Si
concentration; however, a decrease in slope in the variation of H with x is seen at the onset
of extended stiffness transition. Further, a pronounced cusp is seen in H at the mean-field
stiffness threshold at x = 5. It is interesting to note here that a similar cusp is seen in the
Ge-Se binary system, in the hardness variation with composition plot [3].
Figure 9.5: Variation of hardness of Ge15Te85-xSix glasses with composition.
Figure 9.6 shows the variation of mass density with composition. Since density of
the material and elastic modulus are correlated, a decreasing trend in density is observed in
the composition range corresponding to the thermally reversing window (x = 2 to x = 6).
Thus, the variation of mass density with composition of Ge15Te85-xSix glasses supports the
Young’s modulus and hardness variations.
120
Figure 9.6: The variation of density with composition of Ge15Te85-xSix glasses.
It is to be noted here that while the variation in certain properties across the
stiffness threshold and across the thermally reversing window are universal, the variation
in elastic constants with rc is not generally unique; it depends on the actual compositions
being considered [16,17]. In literature, attempts have been made to understand the
composition dependence of the elastic modulus and hardness on the basis of mechanical
threshold, molar volume, etc., in many glassy systems [18,19]. Mahadevan and Giridhar
[20] have noted that, when the sizes of the atoms did not differ very much, the elastic
constants in the mean coordination range 2 < rc < 2.4 remains relatively unchanged. Since
atomic radii of Ge, Te and Si are 1.22, 1.36, 1.11 Ǻ respectively, there is not much
variation in E in the range of rc 2.32 and 2.4 in Ge-Te-Si series. The sharp increase in E
above rc = 2.4 is due to the rigidity effect.
9.4 Conclusions
In summary, composition dependence of elastic modulus and hardness of Ge15Te85-
xSix glasses has been studied systematically, in the composition range of 0 ≤ x ≤ 9. The
variation of Young’s modulus has been found to exhibit a plateau in the composition range
x = 2 (rc = 2.34) to x = 6 (rc = 2.42), which has been identified earlier as the thermally
reversing window on the basis of the composition dependence of enthalpy change during
glass transition obtained from the non-reversing heat flow (ΔHNR). Further, the hardness
(H) of Ge15Te85-xSix glasses increases Si concentration with a decrease in slope in the
variation of H with x at the onset of extended stiffness transition. Further, a pronounced
121
cusp is seen in H at the mean-field stiffness threshold at x = 5. Also, the variation of mass
density with composition of Ge15Te85-xSix glasses supports the Young’s modulus and
hardness variations.
122
References
[1] H. He and M.F. Thorpe, Phy. Rev. Lett., 54 (1985) 2107.
[2] K. Tanaka, Solid State Commun., 60 (1986) 295.
[3] A.K. Varshneya and D.J. Mauro, J. Non-Cryst. Solids, 353 (2007) 1291.
[4] A.N. Sreeram, A.K. Varhneya and D.R. Swiler, J. Non-Cryst. Solids, 130 (1991)
225.
[5] W.C. Oliver and G.M. Pharr, J. Mater. Res., 7 (1992) 1564.
[6] M. Anbarasu, K.K. Singh and S. Asokan, Phil. Mag., 88 (2008) 599.
[7] P. Boolchand, D. Georgiev and B. Goodman, J. Optoelectron. Adv. Mater., 3 (2001)
703.
[9] P. Pattanayak, N. Manikandan, M. Paulraj, et al., J. Phys. Condens. Matter, 19
(2006) 036224.
[10] M. Micoulaut and J.C. Phillips, J. Non-Cryst. Solids, 353 (2007) 1732.
[11] Y. Wang, P. Boolchand and M. Micoulaut, Europhys. Lett., 52 (2000) 633.
[12] Y. Wang, J. Wells, D. Georgiev, et al., Phys. Rev. Lett., 87 (2001) 185503.
[13] D. Selvanathan, W.J. Bresser and P. Boolchand, Phys. Rev. B, 61 (2000) 15061.
[15] P. Boolchand, M. Jin, D.I. Novita, et al., J. Raman Spectrosc., 38 (2007) 660.
[16] Sudha Mahadevan, A. Giridhar and A.K. Singh, J. Non- Cryst. Solids, 57 (1983)
423.
[17] U. Tille, G.H. Frischat and K.J. Leers, Proc. Fourth Int. Conf. on Phys. of Non-
Cryst. Solid, (edited by G.H. Frischat), 1976, Clausthal-Zellerfeld.
[18] A.N. Sreeram, A.K. Varshneya and D.R. Swiler, J. Non-Cryst. Solids, 128 (1991)
294.
[19] A.N. Sreeram, A.K. Varshneya and D.R. Swiler, J. Non-Cryst. Solids, 130 (1991)
225.
[20] S. Mahadevan and A. Giridhar, J. Non-Cryst. Solids, 110 (1989) 118.
123
Chapter 10
Summary and Future Scope

Chalcogenide glasses are one of the most widely investigated materials due to their
applications in Xerogarphy, IR optics, fiber optics, solar cells, CD ROMs and more
recently in electronic resistance variable memories known as Phase Change Memories
(PCM). The most important advantage of chalcogenide glasses is that their properties can
be tuned over a wide range by varying the dopants, their concentration, etc. Among the
different chalcogenide glassy systems, glassy tellurides, in particular Ge-Te based glassy
systems, have been widely investigated for its suitability for various applications.
The present thesis work mainly deals with electrical switching, thermal and optical
studies on certain tellurium based bulk ternary chalcogenide glasses (Ge-Te-Bi and Ge-Te-
Sn) and amorphous chalcogenide thin films (Ge-Te-Se, Ge-Te-Si and Ge-Te-Sn), with an
emphasis on PCM applications. The experiments undertaken in this thesis work, are
electrical switching studies on samples in bulk glassy and amorphous thin film forms,
DSC and ADSC investigations and also optical band gap measurements. The focus of the
investigations has been the composition dependence of switching voltage, crystallization
temperature, glass forming ability, optical band gap, Young’s Modulus, hardness etc. The
summary of the salient results obtained are given below:
1) Al23Te77 amorphous thin film samples deposited in coplanar geometry have

been found to exhibit memory type electrical switching, which is similar to
that of bulk samples. The switching fields of thin film samples is lower
compared to the bulk, which has been understood on the basis of higher
thermal conductance in bulk, which reduces the Joule heating and temperature
rise in the electrode region.
2) Bulk Ge-Te-Bi glasses have been found to exhibit memory type electrical
switching behavior, which is in agreement with the lower thermal diffusivity
values of these samples. Further, the switching voltages of Ge18Te82-xBix
glasses have been found to decrease with the increase in Bi content, indicating
that in the Ge-Te-Bi system, the resistivity of the additive has a stronger role
to play in the composition dependence of Vt, in comparison with the network
connectivity and rigidity factors.
Ch.10: Summary and Future…
3) Studies on the electrical switching and glass forming ability of Ge15Te85-x Snx
and Ge17Te83-xSnx glasses indicate that these glasses exhibit memory type of
electrical switching behavior. Further, the crystallization tendency of these
samples has been found to increase with Sn concentration. XRD studies reveal
the fact that Sn atoms do not take part actively to enhance the network
connectivity and rigidity. The composition dependence of crystallization
temperature, metallicity factor and XRD results are used to explain the
variation with composition of threshold voltages for both the series of samples.
4) GexSe35-xTe65 (17 ≤ x ≤ 23) amorphous thin film samples, coated in sandwich
geometry of electrodes are found to switch with very low voltages as
compared to bulk samples. The switching voltages and optical band gap are
found to increase with the addition of Ge at the expense of Se. High structural
cross linking with progressive addition of 4-fold coordinated Ge atoms could
be the one of the reasons of increasing switching voltage and stronger Ge-Se
bond strength could be the reason of increasing band gap for these
chalcogenide glasses.
5) Electrical switching studies on amorphous Ge15Te85-xSix (1 ≤ x ≤ 6) thin film
samples indicate a similar trend in the switching voltages of both bulk and film
samples when the threshold field is varied with composition. Optical band gap
has been measured as a function of composition for these films, which also
shows a behavior similar to that of switching voltages. The increasing trend in
the variation with composition of electrical switching voltages and optical
band gap are due to the increase in network connectivity and rigidity as Si
atoms are incorporated into the Ge-Te system.
6) Ge15Te85-xSnx (1 ≤ x ≤ 5) and Ge17Te83-xSnx (1 ≤ x ≤ 4) amorphous thin films
have been found to exhibit memory type of electrical switching behavior. The
composition dependence of threshold voltage shows a decreasing trend, which
has been explained on the basis of the Chemically Ordered Network (CON)
model, bond strength and the metallicity factor. The optical band gap variation
of both the series also exhibits a similar decreasing trend with composition.
7) Nano-indentation studies on Ge15Te85-xSix (0 ≤ x ≤ 9) bulk glasses have been
undertaken to study systematically the composition dependence of young’s
modulus and hardness. The density of the samples of different compositions
has also been measured, which strongly supports the variation of Young’s
126
Ch.10: Summary and Future…
Modulus and hardness with composition. The composition dependence of
mechanical properties of Ge-Te-Si samples has been understood on the basis
of the presence of an intermediate phase and a thermally reversing window in
this glassy system.
Scope for Future Work
• In the present work, the set process in bulk glasses and amorphous films belonging
to the Ge-Te-Se, Ge-Te-Si and Ge-Te-Sn systems have been studied. In future, the
reset process for these samples can be studied, particularly in thin film form, which
can lead to the development of a PCM cell.
• Time measurements during the set-reset processes need to be studied which is also
very important for PCM application.
• Detailed investigations on structural transformations that happen during switching

in bulk and thin film samples, which can throw light on the phase change
mechanism can be carried out.
127
List of Publications
1. Chandasree Das, G. Mohan Rao and S. Asokan, ‘A Comparative Study of

Electrical Switching Behavior of Certain Tellurium based Chalcogenide Thin
Films for Phase Change Memory (PCM) Applications’, Advanced Materials
Research, 123-125 (2010) 1207.
2. Chandasree Das, R. Lokesh, G. Mohan Rao and S. Asokan, ‘Electrical switching
behavior of amorphous Al23Te77 thin film sample’, Journal of Non-Crystalline
Solids (Letter to Editor), 356 (2010) 2203.
3. Chandasree Das, G. Mohan Rao and S. Asokan, ‘Electrical Switching and
Thermal Studies on Bulk Ge-Te-Bi glasses’, Journal of Non-Crystalline Solids,
357 (2011) 165.
4. Chandasree Das, M.G. Mahesha, G. Mohan Rao and S. Asokan, ‘Electrical
Switching Studies On Amorphous Ge-Te-Sn Thin Films’, 55th DAE Solid State
Physics Symposium, 2010, AIP Conference Proceedings, 1349 (2011) 633.
5. Chandasree Das, M.G. Mahesha, G. Mohan Rao and S. Asokan, ‘Electrical
Switching and optical studies on Amorphous GexSe35-xTe65 Thin films’, Thin Solid
Films, (Communicated).
6. Chandasree Das, G. Mohan Rao and S. Asokan, ‘Electrical Switching and
Optical Studies on Amorphous Ge15Te85-xSix thin films’, Semiconductor Science
and Technology, (Communicated).
7. Chandasree Das, G. Mohan Rao and S. Asokan, ‘The Electrical Switching and
Thermal Behavior of Bulk Ge15Te85-xSnx and Ge17Te83-xSnx glasses’, Journal of
Non-Crystalline Solids, (Communicated).
8. Chandasree Das, M.G. Mahesha, G. Mohan Rao and S. Asokan, ‘Studies on
Electrical Switching Behavior and Optical Band gap of Amorphous Ge-Te-Sn Thin
films’, Applied Physics A: Material Science & Processing, (Communicated).
9. Chandasree Das, Kiran Mangalampalli, U. Ramamurty, G. Mohan Rao and S.
Asokan, ‘Nano-indentation Studies on Bulk Ge-Te-Si Chalcogenide Glasses’, (to
be communicated).
127

Mam Thesis PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mam Thesis PDF

Uploaded by

Copyright:

Available Formats

Investigations on Certain Tellurium based Bulk

Chalcogenide Glasses and Amorphous Chalcogenide

Department of Instrumentation and Applied Physics

my father mr. mihir ranjan das,

my mother mrs. ilora das

September 2011 (Chandasree Das)

My sincere thanks goes to Prof. K. Rajanna, Chairman, Department of

I sincerely thank Dr. K. Narasimha Rao for allowing me to use optical

I wish to express my warm and sincere thanks to Professor U.

I am very thankful to Prof. Vikram Jayaram, Department of Materials

I am heartily thankful to Kesab Barai, Somnath, Amit of Nano Centre for

It’s my immense pleasure to acknowledge the support I received from Office

I amply acknowledge the invaluable help and ever memorable company

I am very much thankful to my lab mates specially Srinivas, Tamil and

I always enjoyed the company of my present lab mates Srividya, Lakshmi,

I especially thank Dr. M.G. Mahesha, Department of Physics, Manipal

I will be failing in my duties if I don’t express my gratitude to this premier

It gives me a great deal of pleasure to acknowledge all my teachers for

It would not be complete if I don’t express my gratitude to my parents. This

Finally and most importantly, I heartily acknowledge my husband Deep for

In chapter 6, electrical switching studies on amorphous Ge15Te85-xSix (1 ≤ x ≤ 6)

A summary of the significant results obtained in the present thesis work is

1.2.2 Glass Forming Ability .............................................................................................. 3

1.3 Synthesis of Glassy Alloys ............................................................................................. 4

1.4.1.1 The Continuous Random Network (CRN) Model ............................................. 5

1.4.1.3 Topological Thresholds ..................................................................................... 6

1.4.2 Electrical, Optical and Mechanical Properties ......................................................... 8

1.4.2.1 Electrical Switching ........................................................................................... 9

1.4.2.1.1 Different models proposed for electrical switching .................................. 11

1.4.2.1.2 Different time constants and Switching parameters for chalcogenide

1.4.2.2 Optical Properties ............................................................................................ 15

1.4.2.3 Mechanical Properties...................................................................................... 16

1.5 Application of Chalcogenides ....................................................................................... 17

2.2.2 Thin film coating .................................................................................................... 28

2.3 Thickness Measurement................................................................................................ 31

2.5.2 Alternating Differential Calorimetry ...................................................................... 34

2.6 X-Ray Diffraction Studies ............................................................................................ 35

4.3.2. Temperature dependence of threshold voltages .................................................... 55

4.3.3. Thickness dependence of switching voltages........................................................ 55

4.3.4. Activation energy for crystallization ..................................................................... 56

4.3.5. Crystallization behavior of Ge18Te82-xBix glasses ................................................. 57

4.3.6. Electron Microscopic analysis of Ge-Te-Bi .......................................................... 59

4.4 Conclusions ................................................................................................................... 59

5.3.2. Optical band gap .................................................................................................... 69

5.4. Conclusions .................................................................................................................. 72

6.3.2. Optical Band gap ................................................................................................... 81

6.3.3. Electron Microscopic Studies ................................................................................ 83

6.4 Conclusions ................................................................................................................... 84

7.3.2 Composition dependence of switching voltages .................................................... 92

7.3.3 Variation of threshold voltage with thickness ........................................................ 97

7.4 Conclusions ................................................................................................................... 98

8.4. Conclusions ................................................................................................................ 111

Glassy Chalcogenides and Phase Change Memories - A

The earliest report by Schulz-Sellack on an oxygen-free glass dates back to 1870

The research on chalcogenide glasses took a new direction when the

1.2 Glass Transition and Glass Forming Ability

1.2.1 Glass Transition Phenomenon

Even though several terms including amorphous, vitreous, non-crystalline, etc.,

As mentioned earlier, the reverse transformation from a glass to a super cooled

1.2.2 Glass Forming Ability

A macroscopic model of glass formation proposed by Zachariasen predicts that

1.3 Synthesis of Glassy Alloys

Figure 1.2: Different methods of preparation of amorphous/glassy solids: (a) Slow