Phase Change Materials

845 Pages·2009·24.978 MB·English

by  Dr. Chung H. Lam (auth.), Simone Raoux, Matthias Wuttig (eds.)

Phase Change Materials
Simone Raoux • Matthias Wuttig
Editors
Phase Change Materials
Science and Applications
123
Editors
Simone Raoux Matthias Wuttig
IBM Almaden Research Cente r 1. Physikalisches Institut (1A)
650 Harry Road RWTH Aachen University
San Jose, CA 95120 52056 Aachen
USA Germany
ISBN 978-0-387-84873-0 e-ISBN 978-0-387-84874-7
DOI 10.1007/978-0-387-84874-7

Library of Congress Control Number: 2008935619
© Springer Science+Business Media, LLC 2009
All rights reserved. This work may not be translated or copied in whole or in part without the written
permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York,
NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in
connection with any form of information storage and retrieval, electronic adaptation, computer
software, or by similar or dissimilar methodology now known or hereafter developed is forbidden.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they
are not identified as such, is not to be taken as an expression of opinion as to whether or not they are
subject to proprietary rights.

Printed on acid-free paper

springer.com
Foreword

It is a pleasure to write a foreword to this much needed book on “Phase Change
Materials: Science and Applications.” The book is a measure of the growing im-
portance of the field. Phase change materials cover almost every aspect of mate-
rial science from materials in the amorphous state to intermediate range order and
to nano-crystalline and micro-crystalline states. Nanotechnology is considered to
be a new science, yet we have utilized nano-materials in phase change memories
since 1960. The richness of material science involved cannot be overstated. We
are dealing with truly atomically engineered materials. Three examples are the
Ovonic threshold switch, the multi-state Ovonic Universal Memory (OUM), and
the Ovonic cognitive device which emulates the biological neurons with its plas-
ticity and synaptic activity.
The field of amorphous and disordered materials created not only a basic new
area of science, but also important new technologies. It should be kept in mind that
to do so we had to invent the materials, products and manufacturing technology.
Amorphicity and disorder opened up new degrees of freedom that by-pass the
constraints of crystalline periodicity. This freedom enabled me to synthesize ma-
terials with new physical, chemical, electronic and structural properties of unique
functionality.
Referring to the Ovonic electrical phase change memories, Intel’s chief tech-
nology officer in flash memory, Ed Doller, said at a meeting in 2007, “The phase
change memory gets pretty close to Nirvana.” Ironically, Bob Noyce and Gordon
Moore came out to visit Iris and me before they set up Intel because there existed
no solid state memory except our Ovonic Phase Change Memory. At that time, all
computer memories were wire wound magnetic core devices.
The background of my work goes back to 1955, where I was determined to lay
the basis of cognitive computing by creating an inorganic analog of biological
neurons and their synapses. Fortunately, my approach attracted great interest from
Dr. Ernest Gardner, head of the Department of Anatomy of Wayne State Medical
School of Detroit, a highly respected neurophysiologist, who invited me to join his
team to continue work on my ideas.
Such an analog neuronal device had to respond to various thresholds, exhibit
plasticity and non-volatility and, of course, reversibility between a conducting and
a non-conducting state. Since there was no prior literature to guide me, I had to
create new materials and initiate a field of atomically designed amorphous and
disordered materials.
vi Foreword

The phase change materials that I developed had to have plasticity and a re-
versible structural change. I applied polymer science to inorganic polymeric ma-
terials such as Te and Se. Te/ Se are lone pair, polymeric, divalent materials with
chain structures. They can be designed to bond with cross-linking elements of dif-
ferent bond strengths such as Ge and Sb. Such atomically designed, cross-linked
Te/Se alloys have a huge number of non-bonded lone-pair electrons which could
easily be excited by optical and electrical fields. When the amorphous state can-
not contain the excitation energy, a phase change to a crystalline state occurs.
It is important to understand that such polymeric lone-pair chain structures are
vibronic in nature wherein electronic transitions are made possible by vibrational
motion of the chains. These simultaneous vibrational and electronic transitions
are important to understanding the mechanism of both the optical and the electri-
cal Ovonic phase change memories.
Lone-pair electrons are crucial for both the non-volatile phase change memory
and the electronic Ovonic threshold switch which is volatile. The crucial differ-
ence between them is the amount of cross links and their bonding energies.
Strong and numerous cross-linking atoms such as Si, As and Ge provide an amor-
phous material that is structurally not crystallizable, but electronically switchable
when an electrical field beyond a certain threshold value is coupled to the large
number of non-bonding lone pairs in the material. This results in a highly con-
ducting filament, which expands and contracts in a servo manner to a load. These
conducting filaments have a plasma nature which can sustain about 50 times the
current density encountered in a typical transistor; an important implication for fu-
ture devices.
The Ovonic phase change memory is a non-volatile device, which is commonly
used for binary operations. But the phase change memory is capable of much
more. We have demonstrated that it can operate reproducibly in many resistance
states which correspond to different amorphous to crystalline fractions in the ma-
terial. The plasticity of phase change memory and its ability to have many other
states in both the amorphous and crystalline portions of the resistance vs. current
curve (the U curve) enabled me to develop the cognitive computing device that re-
sembles the biological neurons and synaptic activity.
Now, what is the future of our field? The Ovonic universal phase change mem-
ory is called universal because it can replace flash memory, DRAM and SRAM.
These are not only basic computer memory devices but also are becoming the
driving force for the ongoing revolutionary growth of cell phones and other mo-
bile devices, which are in desperate need of memory providing higher density,
faster speed and lower power consumption. As we write, the OUM has penetrated
those areas already and it will excel with its multi-state operations. It is also im-
portant to keep in mind the point that I’ve been making since the very beginning
that information is encoded energy.
What else will be coming along in the future? The phase change memories
hold great promise for the following reasons: our devices are scalable, the smaller
we make them, the better they are; they can be reduced in size to the quantum lim-
Foreword vii
it. The crystalline Si based transistors might be replaced by Ovonic three-terminal
devices. The various multi-states Ovonic devices will be able to replace the logic
and memory functions. In other words, logic and memory will become one. Huge
parallelism which is not available in conventional computing will become stan-
dard, more and more re-configurable and interactive circuits will be integrated
with the new parallelism. The Ovonic cognitive devices will make cognitive
computing possible. Recall that Bill Gates said, “If you invent a breakthrough in
artificial intelligence, so machines can learn, that is worth 10 Microsofts.”
The work described by the authors in this book illustrates the range of contribu-
tions that are being made throughout the world. The future of our field continues
to be expanded by ongoing creative work. The editors deserve much credit for as-
sembling the important work of noted authors in this book. The collegiality and
collaboration I have enjoyed is remarkable.

Bloomfield Hills, MI Stanford R. Ovshinsky
March 2008
Preface

Phase change materials are characterized by a unique property portfolio. They pos-
sess a pronounced difference of optical and electronic properties depending upon
their atomic arrangement, i.e. whether they are amorphous or crystalline. At the
same time they can be rapidly switched between the amorphous and crystalline state.
This combination of properties is attractive for applications and also provides a
unique opportunity to test our understanding of the relationship between bonding
and the atomic arrangement in solids and their correlation to solid state properties.
In rewriteable optical data storage employing phase change materials a short
pulse of a focused, high intensity laser beams locally heats the phase change mate-
rial above its melting temperature. Rapidly cooling the phase change alloy with
rates higher than 109 K/s quenches the liquid-like state into a disordered, amor-
phous phase. This amorphous state has different optical properties than the sur-
rounding crystalline state. Hence detecting regions with amorphous structure is
straightforward employing a laser beam of low intensity. To erase the stored in-
formation a laser pulse with intermediate power is utilized. The laser locally heats
the phase change film above the crystallization temperature. At elevated tempera-
tures above the glass transition temperature the atoms become increasingly mobile
and can revert to the energetically favourable crystalline state, erasing the re-
corded information. In electronic memory applications of phase change materials
a short and relatively high current pulse is used to locally melt the crystalline ma-
terial. The resulting amorphous state has a high resistance which exceeds the resis-
tance of the crystalline state by several orders of magnitude. An intermediate
power pulse is used to heat the material above its crystallization temperature to
switch it back to the crystalline, low resistance state, while a low power pulse is
used to determine the resistance of the phase change material.
While the simplicity of the storage concept has attracted the interest of industry
a number of questions will decide how far reaching the potential of phase change
materials is. For rewriteable optical data storage the most prominent questions are
linked to a strong optical contrast also at short laser wavelengths as required for
next generation optical storage media. Another important aspect is the quest for
very high speed materials that enable crystallization on the nanosecond time scale.
Finally mechanisms to bypass the diffraction limit are of paramount interest. For
electronic memories again the ultimate data transfer rate is of utmost importance.
Since crystallization of phase change materials is the time limiting step in storage
applications a detailed understanding of ultra-fast crystallization processes is a
prerequisite for the design of even faster materials and the identification of intrin-
sic material limits. It would be ideal if it was possible to correlate the atomic ar-
x Preface
rangement in the amorphous and crystalline state with the speed of the transforma-
tion process. Up to now this goal has not yet been reached. Nevertheless a detailed
understanding of the atomic arrangement in the amorphous and crystalline state
should facilitate an in-depth understanding of the bonding and the physical proper-
ties of phase change materials. To fully exploit the potential of phase change ma-
terials in electronic memory applications a detailed understanding of electronic
switching phenomena is a prerequisite.
Material scientists are faced with formidable challenges if phase change solid
state memory is considered for new applications. For example, in automotive ap-
plications rather high operation temperatures of about 150 ºC are required at
which none of the currently applied phase change materials is stable in the amor-
phous phase. In case of a possible DRAM replacement extremely high switching
speed and very high numbers of switching cycles (1016-1018) are necessary. The
quest for the ultimate scaling limits will decide if this is a viable technology that
can be developed for several next generations of lithographies following the
roadmap of the semiconductor industry.
This book covers our current understanding of the science and the status of the
applications of phase change materials. In the foreword, written by the founding
father of the field, Stanford Ovshinsky, the past, the present status and future per-
spectives of phase change materials are discussed. Chapter 1 provides a discussion
of the history of phase change materials. The theoretical understanding of phase
change materials and the nature of glasses are discussed in Chapters 2 and 3, while
the structure of amorphous Ge-Sb-Te alloys is presented in Chapter 4. Selection
strategies for phase change materials are provided in Chapters 5 and 10, with the
latter chapter focusing on the development of materials for third generation re-
writeable optical data storage. The crystallization kinetics which are important to
maximize data transfer rates are the subject of chapter 7. Chapters 8 and 9 discuss
the short and long-range order and the resulting optical and electrical properties
with an emphasis on the prototype material Ge Sb Te . Chapter 6 discusses the
2 2 5
scaling behavior of phase change materials, while novel deposition methods are
presented in Chapter 11. The following chapters focus on applications of phase
change materials. In Chapter 12 the first three generations of optical memories are
discussed, while 4th generation optical memories employing near-field effects are
presented in Chapter 13. The last five chapters focus on applications utilizing elec-
tronic switching phenomena for storage and logic. Chapter 14 provides an in-
depth discussion of phase change memory device modeling, while advanced proto-
type devices and their scaling are presented in Chapter 15. Phase change memory
cell concepts are introduced in Chapter 16 and integration aspects are provided in
Chapter 17. Finally applications in reconfigurable logic are the focus of Chapter 18.

Santa Clara, CA Simone Raoux
July 2008
Aachen, Germany Matthias Wuttig
July 2008
Contents

1. History of Phase Change Memories.............................................................1
Chung H.Lam
1.1 The Discovery of Phase Change Materials.............................................1
1.2 Early Electronic Computers and Memory Systems................................2
1.3 Pioneers in Phase Change Memory........................................................4
1.4 Early Attempts with Phase Change Memory..........................................9
1.5 Rebirth of Phase Change Memory........................................................10
References.....................................................................................................14
Part I: Material Science: Theory and Experiment
2. Density Functional Theory Calculations
for Phase Change Materials........................................................................17
Wojciech Wełnic
2.1 Introduction..........................................................................................17
2.2 The Theorem of Hohenberg and Kohn.................................................18
2.3 The Kohn-Sham Equation....................................................................20
2.4 The Local Density Approximation.......................................................22
2.5 Beyond Density Functional Theory......................................................23
2.6 Application of DFT in the Field of Phase Change Materials................24
2.6.1 Structure Determination............................................................25
2.6.2 Electronic Properties.................................................................29
References.....................................................................................................36

3. Nature of Glasses.........................................................................................39
Punit Boolchand, Matthieu Micoulaut, and Ping Chen
3.1 Introduction..........................................................................................39
3.2 Thermodynamics of the Glass Transition.............................................41
3.3 Glass Transition from Dynamics..........................................................43
3.4 Glass Forming Tendency......................................................................44
3.4.1 Compositional Trends of the Glass Transition Temperature....46
3.5 Calorimetric Measurement of the Glass Transition
Temperature and Related Thermal Properties......................................48
3.6 Three Generic Classifications of Glasses and Glass
Transitions ............................................................................................51
xii Contents
3.7 Elastic Phases in Ionic and Super-ionic Glasses...................................54
3.8 Ideal Glasses and Self-organization of Networks.................................54
3.9 Does the View Below the Glass Transition Temperature
Correlate with the View above the Glass Transition Temperature?.....56
3.10 Glass Formation in Hydrogen Bonded Networks.................................57
3.11 Epilogue...............................................................................................59
References.....................................................................................................59

4. Structure of Amorphous Ge-Sb-Te Solids.................................................63
Stephen R.Elliott
4.1 Introduction..........................................................................................63
4.2 Structural Order in Amorphous Materials............................................64
4.2.1 Short-range Order.....................................................................64
4.2.2 Medium-range Order................................................................65
4.2.3 Long-range Structure................................................................66
4.3 Experimental Structural Probes............................................................67
4.4 Structural Modeling..............................................................................68
4.5 The Structure of Amorphous Phase-change Materials.........................69
4.5.1 Experimental Studies................................................................69
4.5.2 Simulational Studies.................................................................72
4.6 Summary..............................................................................................78
References.....................................................................................................79

5. Experimental Methods for Material Selection
in Phase-change Recording.........................................................................81
Liesbeth van Pieterson
5.1 Introduction..........................................................................................81
5.2 Reversible Switching............................................................................82
5.3 Phase-change Materials........................................................................84
5.3.1 Crystallization by Nucleation and Growth...............................86
5.3.2 Crystallization Dominated by Crystal Growth.........................88
5.4 Archival Life Stability..........................................................................89
5.5 Crystallization Rate..............................................................................91
5.6 Material Optimization..........................................................................93
5.7 Outlook.................................................................................................97
References.....................................................................................................98

6. Scaling Properties of Phase Change Materials.........................................99
Simone Raoux
6.1 Introduction..........................................................................................99
6.2 Thin Films of Phase Change Materials...............................................100
6.2.1 Crystallization Temperature as a Function of Film Thickness...101
6.2.2 Crystallization Rate as a Function of Film Thickness............105