Ramanathan Sh. (Ed.) Thin Film Metal-Oxides: Fundamentals and Applications in Electronics and Energy

Подождите немного. Документ загружается.

Thin Film Metal-Oxides

Shriram Ramanathan

Editor

Thin Film Metal-Oxides

Fundamentals and Applications in Electronics

and Energy

123

Editor

Shriram Ramanathan

Harvard University

School of Engineering & Applied Sciences

29 Oxford St.

Cambridge MA 02138

Perice Hall

USA

shriram@seas.harvard.edu

ISBN 978-1-4419-0663-2 e-ISBN 978-1-4419-0664-9

DOI 10.1007/978-1-4419-0664-9

Springer New York Dordrecht Heidelberg London

Library of Congress Control Number: 2009941699

 Springer Science+Business Media, LLC 2010

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 is part of Springer Science+Business Media (www.springer.com)

Stoichiometry in Thin Film Oxides –

A Foreword

The present book edited by Shiram Ramanathan kills various but at least three birds

with one stone.

1. It reveals the importance of oxidic materials for a variety of modern applications

in solid state electronics and solid state ionics in view of energy research and

information technology.

2. It highlights the signiﬁcance of electronic and ionic defects and their coupling

through stoichiometry. Ionic point defects are relevant in two ways: At room

temperature they are typically frozen and act as dopants. At high temperatures

they are mobile, a property that is used in electrochemical devices. The latter

point is important for room-temperature applications, too, as the stoichiometry

can be tuned at high temperatures and then frozen. The same point, however,

does not only provide a new practical degree of freedom, but it also can, on a

long time scale, give rise to stoichiometric polarization and degradation.

3. This contribution devoted to thin ﬁlms addresses the crucial role of interfaces not

only for pathway reduction but also as far as variation of charge carrier concen-

trations and mobilities is concerned. The thickness of the layers, i.e., the spacing

of the interfaces is not just important for regulating the proportion of interfacial

effects, extreme thickness reduction can also lead to mesoscopic phenomena as

it is to the fore in the ﬁelds of nanoionics and nanoelectronics.

Let us, as an introduction to what is presented by the distinguished experts in this

book, brieﬂy consider, on a qualitative level, the inﬂuence of the most decisive con-

trol parameters on charge carrier concentration. (Note that the ionic and electronic

charge carriers are not just important for electrical or electrochemical properties,

they are also reﬂecting the internal redox and acid base chemistry and are thus cru-

cial for reactivity.)

For simplicity we refer to a binary oxide M

2

in which only the oxygen sub-

lattice exhibits disorder. As intrinsic disorder processes we have to face (1) electron

transfer from the valence to the conduction band forming excess electrons and holes

(for main group oxides this typical refers to a charge transfer from oxygen orbitals

to metal orbitals, i.e., from O

2

to M

/, as well as ion transfer from regular sites to

vi Stoichiometry in Thin Film Oxides – A Foreword

Fig. 1 Ionic disorder (top) and electronic disorder (bottom) in a thin ﬁlm displayed as (free)

energy-level diagram. The (free) energy levels are standard electrochemical potentials ( Q

/ of

excess or missing particles. Unlike the standard potentials the full electrochemical potential (Q/ of

ions or electrons contains conﬁgurational entropy. The distance to the upper and lower levels is (in-

verse) measure of the respective carrier concentrations. The difference between the electrochemical

potentials themselves (note the necessary factor of 2 as O corresponds to O

2

–2e



/ reﬂects the

chemical potential of neutral oxygen. For bulk properties the ﬂat middle part of the energy levels is

relevant. The level bendings at the l.h.s. and r.h.s. are due to interfacial space charge effects (here

symmetrical boundary conditions are assumed)

interstitial sites forming excess oxygen ions and oxygen ion vacancies (see Fig. 1).

In the following, let us consider the decisive parameters that inﬂuence the charge

carrier concentrations and the stoichiometry in a given oxide.

Inﬂuence of Temperature

As temperature increases, disorder is favored. Let us concentrate on the ionic dis-

order ﬁrst. A typical scenario is as follows: At low T the defect concentrations are

small and the defects at random (gaps in Fig. 1) are more easily crossed. Further

increase of temperature and hence increase of interstitial and vacancy concentra-

tion leads to attractive interactions that effectively lower the ionic gap in Fig. 1,now

increasing the defect concentrations even more and eventually leading to a phase

transformation into a superionic state. Normally, the crystal structure does not tol-

erate this state and melting occurs prior to this.

The electronic analog is creating electron and holes through increased tem-

perature. Similar to the ionic picture, excitonic interaction can lead to band gap

Stoichiometry in Thin Film Oxides – A Foreword vii

narrowing and perhaps eventually to a metallic state. Often neither the ionic picture

nor the electronic picture predominates, rather the case of a mixed conductor is met,

where one meets ionic and electronic carriers in comparable concentrations. Here

trapping of ionic and electronic defects can be substantial leading to only partially

ionized or even neutral defects at low temperatures.

Inﬂuence of Oxygen Partial Pressure

For our pure material MO, at constant total pressure and temperature there is, as

far as complete chemical equilibrium is concerned, one more degree of freedom the

disposition of which ﬁxes the stoichiometry. This degree of freedom is exhausted

if we deﬁne the oxygen partial pressure (implicitly contained in the chemical po-

tential of oxygen in Fig. 1). Even though affecting the total masses and energies

only marginally, the effect of P

variation which allows traversing the phase width

is of ﬁrst order on the electronic and ionic charge carriers. P

increase augments

oxygen interstitial concentration and hole concentration (note that O formally corre-

sponds to O

2

plus two holes) and decreases the concentrations of oxygen vacancies

and conduction electrons. These variations are typically orders of magnitude vari-

ations. Phase stability provided, the conductivity typically changes from n-type to

p-type potentially via a regime of mixed conductivity or even predominant ionic

conductivity.

Inﬂuence of Doping Content

Consideration of frozen-in defects allows for further degrees of freedom; this is

addressed in this and the following paragraphs.

Implementing immobile impurities, by (homogeneous) doping, is a most efﬁcient

and established way to modify electronic and ionic properties. For not too complex

a defect chemistry of a given system, it is straightforward to predict the effect of

a given dopant for dilute concentrations: If the effective charge of the dopant is

positive (negative) then the concentrations of all the positively (negatively) charged

defects are depressed and that of all negatively (positively) charged defect increased.

The less trivial aspect here is that this holds individually for any carrier.

Frozen-in Native Defects

Also native defects can be considered as dopants if immobile. As already mentioned

in the beginning, the connection between the high temperature defect chemistry

where these defects are mobile and the frozen-in situation is essential. A simple but

viii Stoichiometry in Thin Film Oxides – A Foreword

important point refers to the establishment of a proﬁle-free situation. Already this

requires knowledge of the kinetics. Let us suppose the kinetics to be diffusion con-

trolled and the effective (chemical) diffusion coefﬁcient to be known as a function of

temperature. Then one has to select an annealing temperature at which equilibration

takes as long as one can afford to wait (e.g., 1 week). If then the sample is quenched

(e.g., within 1 min), one can neglect the proﬁles. Were the temperature higher, dif-

fusion would still occur during quenching; were the temperature lower there would

not have been complete equilibrium to freeze in.

Interfacial Effects

A fascinating area concerns the spatial redistribution of defects, i.e., not just elec-

trons and holes but also ionic defects, such as interstitials and vacancies in the

vicinity of higher-dimensional defects. Here enormous concentration variations can

occur that – given a high enough density – in thin ﬁlms do and in composites (non-

trivial percolation) may result in huge effects on overall transport properties. Ion

conductivities can be greatly enhanced by admixing even insulating particles (“het-

erogeneous doping”), in this way also the type of ion conductivity (from interstitial

to vacancy or even from anionic to cationic) can be varied. Even more strikingly:

the overall conductivity can be varied from ionic to electronic by particle size re-

duction. As to the concentration changes of all the individual carriers, one again just

needs to know the charge of the higher-dimensional “dopant,” here the charge of the

interface. If the interfacial excess charge is positive then all the positively charged

carriers such as oxygen vacancies and holes are depleted, while the concentrations

of the conduction electrons and of the oxygen interstitial are increased. In Fig. 1,

a thin ﬁlm with symmetrical boundaries is assumed. (The sign of the bending cor-

responds to a concentration enhancement of oxygen vacancies and electron holes.)

While this is straightforward, clarifying the reason for the excess charge density or

even controlling it is challenging. Also synergistic storage phenomena can be met

at interfaces that are based on charge separation. Charge carrier concentrations may

also be inﬂuenced by elastic effects. Curvature effects do not play a role if we con-

sider thin ﬁlms (note, however, the signiﬁcance for the crystallites in the case of

nanocrystalline ﬁlms), but strain effects do.

Mesoscopic Effects

While mesoscopic effects are well known for electronics characterizing the well

established ﬁeld of nanoelectronics, true size effects also occur as regards ion con-

ductivity. Concentrating on the latter (“nanoionics”) means dealing with overlap of

accumulation or depletion space charge layers (ﬂat part in Fig. 1 disappears) (lead-

ing in the extreme to artiﬁcial crystals) as well as with mesoscopic heterogeneous

Stoichiometry in Thin Film Oxides – A Foreword ix

storage forming the bridge between electrostatic capacitor and battery storage.

Conﬁnement of ionic carriers leading to variation of local formation free energies

(then the ionic and electronic bendings can be very different) or statistical ﬂuctua-

tions are two elements of a much larger list.

Focusing on concentration effects should not lead to the assumption that effects

on mobilities are marginal. Yet they are quite speciﬁc to the individual situation.

In addition, all these situations described also lead to equally fascinating kinetic

effects. Yet the just given compilation may already sufﬁce to arouse interest in what

is to be described in the following chapters.

Stuttgart, Germany Joachim Maier

Preface

Metal oxides are an important class of materials: from both scientiﬁc and techno-

logical perspectives they present interesting opportunities for research. Thin ﬁlms

are particularly attractive owing to their relevance in devices and also for the abil-

ity to pursue structure–property relations studies using controlled microstructures.

The inherent compositional complexity (due to the presence of ionic species) leads

to rich set of properties, while in several cases, coupling of structural complex-

ity with dynamic electronic properties leads to unexpected interfacial phenomena.

Oxide semiconductors are gaining interest as new materials that may challenge the

supremacy of silicon. A further recent area of research is in understanding interfaces

in oxides and how they inﬂuence carrier transport. Thin ﬁlm oxides are extensively

used to probe strong electronic correlations.

While the book does not attempt to cover every single aspect of oxides research,

it does aim to present discussions on selected topics that are both representative and

possibly of technological interest. Ranging from synthesis, in-situ characterization

to properties such as electronic and ionic conduction, catalysis is discussed. Theo-

retical treatments of select topics as well as relevance to emerging electronic devices

and energy conversion are highlighted.

We expect the book to be of interest to scientists and technologists working

broadly in the ﬁeld of metal oxides. I would like to acknowledge the authors for their

timely contributions and the Springer editorial team for their patience and valuable

suggestions.

Cambridge, MA Shriram Ramanathan