Physics World Archive

Single-electron transistors
Michel H Devoret, Christian Glattli
From

Physics World
September 1998
IOP Publishing Ltd 2014
ISSN: 0953-8585

Institute of Physics Publishing

Bristol and Philadelphia
Downloaded on Wed Oct 15 16:32:56 GMT 2014 [131.193.117.248]

FEATURES
While the electronics industry wonders what will happen when transistors become
so small that quantum effects become important, researchers are building
new transistors that actively exploit the quantum properties of electrons

Single-electron transistors
Michel Devoret and Christian Glattli
THE INVENTION of the transistor by John Bardeen
and William Shockley in 1948 triggered a new era in
electronics. Originally designed simply to emulate
the vacuum tube, scientists soon found that this solidstate device could offer much more. The great potential of the transistor for speed, miniaturization and
reliability has been fully exploited since well controlled materials such as pure single-crystal silicon became available. According to the latest "road-map"
for the microelectronics industry, microchips containing one billion transistors and operating with a
clock cycle of a billionth of a second will be on the
market just a few years into the new millennium.
As transistors continue to shrink, a question naturally arises: will the quantum nature of electrons and
atoms become important in determining how the
devices are built? In other words, what will happen
when a transistor is reduced to the size of a few atoms
or a single molecule?
Researchers seeking to answer these questions have
devised the so-called single-electron tunnelling transistor - a device that exploits die quantum effect of
tunnelling to control and measure the movement of
single electrons. Experiments have shown that charge
does not flow continuously in these devices but in a
quantized way. Indeed, single-electron transistors are
so sensitive to charge that they can be used as
extremely precise electrometers.
At the beginning

1 Operation of a conventional transistor

channel

energy

distance
The metal-cxide-semiconductor field-effect transistor (MOSFET) is the basic switching
and amplification device of digital electronics. The current between the source and drain
electrodes is controlled by the gate voltage, (a) When the gate voltage is zero, no
conduction electrons are present in the channel, (b) When the gate is at a positive
voltage, electrons from the source and drain accumulate in the area of the channel
close to the gate, (c) As the gate voltage is increased further, the number of electrons in
the channel increases until saturation is reached. The potential seen by the electrons is
also shown along a line going from the gate to the channel. With no gate voltage,
electrons in the channel experience a potential that is higher than the bias potential,
shown by the dashed line (d). As the gate voltage increases, the potential in the channel
gradually lowers and electrons accumulate there (e-f).

The most common transistor in today's microchips is

the metal-oxide-semiconductor field-effect transistor (MOSFET). Its operation is surprisingly simple: not much
This field effect leads to an amplification mechanism in
quantum mechanics is required to understand how it works, which die gate voltage can control the current flowing
even though die size of a typical device is now just a few thou- between die source and drain when a bias voltage is applied
sand atoms placed side by side.
across these two electrodes (figure 2). The source-drain curTwo conducting electrodes, called the source and drain, are rent is determined by the conductance of the channel, which
connected together by a channel of material in which the in turn depends on two factors: die density of the conduction
density of free electrons can be varied in practice a semi- electrons and their mobility. The mobility of electrons
conductor (figure 1). A voltage is applied to the semiconduct- depends on how often the electrons collide with crystal
ing channel through the "gate", a third electrode that is imperfections, and is essentially independent of the gate voltseparated from the channel by a thin insulating layer. When age. In contrast, the density of electrons is controlled direcdy
the gate voltage is zero, the channel does not contain any by the gate voltage.
conduction electrons and is insulating. But as die voltage is
The transistor therefore works like a tap controlling the
increased, the electricfieldat the gate attracts electrons from flow of water between two tanks, where the opening of the
the source and drain, and the channel becomes conducting. tap is set by the pressure of the water in a third tank. The difPHYSICS

WORLD

SEPTEMBER

1998

29

ference is that electrons in the 2 Amplification in a field-effect transistor simple system known as a singlechannel behave as a compressible
electron box (figure 3). If a voltage
fluid with a local density that
source charges a capacitor, Cg,
depends strongly on the electric
through an ordinary resistor, the
potential at that point. In other
charge on the capacitor is strictly
words, the electric field produced
proportional to the voltage and
by the gate does not generate a
shows no sign of charge quantiza"hard wall" for electrons inside the
tion. But if the resistance is rechannel, but a smoothly varying
placed by a tunnel junction, the
potential that is modified by the
metallic area between the capacitor
presence of electrons (figure 1 d-f).
plate and one side of the junction
forms a conducting "island" surNote that we have made no referrounded by insulating materials. In
ence to the wave-like properties of
this case the transfer of charge onto
electrons, nor to the fact that the
the island becomes quantized as the
channel is made from individual gate voltage
voltage
increases, leading to the soatoms. The only quantum property The current in a field-effect transistor varies with gate
called
Coulomb
staircase (figure 3c).
that comes into play is the Pauli voltage and with the bias voltage between the source and
This
effect
was
first observed by
drain.
For
a
fixed
bias
voltage,
the
current
is
turned
on
when
exclusion principle, which dictates
Philippe Lafarge and collaborators
that each possible state in the chan- the gate voltage is positive, and turned off when the gate
voltage is negative.
in our laboratory in 1991.
nel can be occupied by only one
electron. This means that only a
This Coulomb staircase is only
certain number of electrons can accumulate in the channel, seen under certain conditions. Firstly, the energy of the electrons due to thermalfluctuationsmust be significantly smaller
and this sets a limit on the current flow.
However, the quantum properties of electrons and atoms than the Coulomb energy, which is the energy needed to
will play a more important role as transistors are made smaller. transfer a single electron onto the island when the applied voltFor example, the wave-like nature of electrons will influence age is zero. This Coulomb energy is given by e2/2 C, where e is
the way in which they travel through the channel. When the the charge of an electron and Cis the total capacitance of the
width of the channel becomes comparable to the wavelength gate capacitor, Cg, and the tunneljunctions. Secondly, the tunof electrons (around 100 nm), electron propagation becomes nel effect itself should be weak enough to prevent the charge of
more sensitive to the atomic disorder in the device, which is the tunnelling electrons from becoming delocalized over the
created in the fabrication process. This will pose a major prob- two electrodes of the junction, as happens in chemical bonds.
lem if the reduction in size is not accompanied by an improve- According to recent work by theorists at the universities of
Freiburg and Karlsruhe in Germany, this means that the conment in the atomic structure of the fabricated devices.
should be much less than the
The technological constraints of moving towards the ductance of the tunnel junction
2
quantum
of
conductance,
2e
/h,
where h is Planck's constant.
atomic scale may force us to adopt a new physical principle
When
both
these
conditions
are
met, the steps observed in
for achieving the transistor's function. Alternatively, a new
principle might be found that can provide functions that are the charge are somewhat analogous to the quantization of
charge on oil droplets observed by Millikan in 1911. In a
not possible with current devices.
single-electron box, however, the charge on the island is not
random but is controlled by the applied voltage. As the temTowards single-electron devices
perature
or the conductance of the barrier is increased, the
In 1985 Dmitri Averin and Konstantin Iikharev, then working
steps
become
rounded and eventually merge into the straight
at the University of Moscow, proposed the idea of a new
line
typical
of
an ordinary resistor.
three-terminal device called a single-electron tunnelling (SET)
transistor. Two years later Theodore Fulton and Gerald Dolan
at Bell Labs in the US fabricated such a device and demon- A single-electron transistor
strated how it operates.
The SET transistor can be viewed as an electron box that has
Unlike field-effect transistors, single-electron devices are two separate junctions for the entrance and exit of single elecbased on an intrinsically quantum phenomenon: the tunnel trons (figure 4). It can also be viewed as afield-effecttransistor
effect. This is observed when two metallic electrodes are sep- in which the channel is replaced by two tunnel junctions
arated by an insulating barrier about 1 nm thick - in other forming a metallic island. The voltage applied to the gate
words, just 10 atoms in a row. Electrons at the Fermi energy can electrode affects the amount of energy needed to change the
"tunnel" through the insulator, even though in classical terms number of electrons on the island.
their energy would be too low to overcome the potential barrier. The SET transistor comes in two versions that have been
The electrical behaviour of the tunnel junction depends on nicknamed "metallic" and "semiconducting". These names
how effectively the barrier transmits electron waves, which are slightly misleading, however, since the principle of both
decreases exponentially with its thickness, and on the number devices is based on the use of insulating tunnel barriers to sepof electron-wave modes that impinge on the barrier, which is arate conducting electrodes.
In the original metallic version fabricated by Fulton and
given by the area of the tunnel junction divided by the square
of the electron wavelength. A single-electron transistor exploits Dolan, a metallic material such as a thin aluminium film is
the fact that the transfer of charge through the barrier becomes used to make all of the electrodes. The metal isfirstevaporated through a shadow mask to form the source, drain and
quantized when t^he junction is made sufficiently resistive.
This quantization process is shown particularly clearly in a gate electrodes. The tunnel junctions are then formed by
30

PHYSICS

WORLD

SEPTEMBER

1998

critical voltage needed to transfer an electron onto the island,

equal to e/C, is called the Coulomb gap voltage.
capacitor
resistor
Now imagine diat the bias voltage is kept below die
/
Coulomb gap voltage. If the gate voltage is increased, the
\
energy of the initial system (with no electrons on the island)
gradually increases, while the energy of the system with one
excess electron on die island gradually decreases. At die gate
3 ft
voltage corresponding to die point of maximum slope on the
Coulomb staircase, bom of these configurations equally qual/J
E 0
ify as me lowest energy states of the system. This lifts the
Coulomb blockade, allowing electrons to tunnel into and out
tunnel
capacitor
,'j
/ junction
1
of
die island.
/
ra o
The Coulomb blockade is lifted when die gate capacitance
is charged widi exacdy minus half an electron, which is not as
0
1

3e
5e
surprising as it may seem. The island is surrounded by insulae
20,
2CO
tors, which means that the charge on it must be quandzed in
voltage
units of e, but the gate is a metallic electrode connected to a
plentiful supply of electrons. The charge on die gate capacitor merely represents a displacement of electrons relative to a
(a) When a capacitor is charged through a resistor, the charge on the
background of positive ions.
capacitor is proportional to the applied voltage and shows no sign of
If we furmer increase the gate voltage so that die gate capaquantization, (b) When a tunnel junction replaces the resistor, a
citor
becomes charged widi -e, die island again has only one
conducting island is formed between the junction and the capacitor plate.
stable configuration separated from die next-lowest-energy
In this case the average charge on the island increases in steps as the
voltage is increased (c). The steps are sharper for more resistive barriers
states by the Coulomb energy. The Coulomb blockade is set
and at lower temperatures.
up again, but the island now contains a single excess electron.
The conductance of die SET transistor dierefore oscillates
introducing oxygen into the chamber so that the metal between minima for gate charges mat are integer multiples of
becomes coated by a thin layer of its natural oxide. Finally, a e, and maxima for half-integer multiples of e (figure 5).
second layer of the metal - shifted from the first by rotating
Accurate measures of charge
the sample - is evaporated to form the island.
In the semiconducting versions, the source, drain and Such a rapid variation in conductance makes die single-elecisland are usually obtained by "cutting" regions in a two- tron transistor an ideal device for high-precision electrometry.
dimensional electron gas formed at the interface between two In this type of application the SET has two gate electrodes,
layers of semiconductors such as gallium aluminium arsenide and the bias voltage is kept close to the Coulomb blockade
and gallium arsenide. In this case the conducting regions are voltage to enhance the sensitivity of the current to changes in
defined by metallic electrodes patterned on the top semicon- die gate voltage.
ducting layer. Negative voltages applied to these electrodes
The voltage of the first gate is initially tuned to a point
deplete the electron gas just beneath them, and the depleted where the variation in current reaches a maximum. By
regions can be made sufficiently narrow to allow tunnelling adjusting the gate voltage around this point, the device can
between the source, island and drain. Moreover, the electrode measure die charge of a capacitor-like system connected to
that shapes the island can be used as the gate electrode.
die second gate electrode. A fraction of this measured charge
In this semiconducting version of the SET, the island is is shared by the second gate capacitor, and a variation in
often referred to as a quantum dot, since the electrons in the charge of Vie is enough to change die current by about half
dot are confined in all three directions. In the last few years the maximum current diat canflowthrough the transistor at
researchers at the Delft University of Technology in the the Coulomb blockade voltage. The variation in current can
Netherlands and at NTT in Japan have shown that quantum be as large as 10 billion electrons per second, which means
dots can behave like artificial atoms. Indeed, it has been poss- that tiiese devices can achieve a charge sensitivity that outible to construct a new periodic table that describes dots con- performs other instruments by several orders of magnitude.
taining different numbers of electrons (see "Quantum dots" Indeed, a collaboration between researchers at Yale Uniby Leo Kouwenhoven and Charles Marcus Physics Worldjuneversity in the US and Chalmers University in Gothenburg,
pp35-39).
Sweden, recendy showed diat charge variations smaller tiian
10~5 e can be detected in a measurement period of just one
second and widi a bandwiddi of several hundred megahertz.
Operation of a SET transistor
So how does a SET transistor work? The key point is that
SET transistors have already been used in mesoscopic
charge passes through the island in quantized units. For an physics experiments that have required extreme charge senelectron to hop onto the island, its energy must equal the sitivity. For example, earlier this year Robert Westervelt and
Coulomb energy e2/2C. When both the gate and bias voltages co-workers at Harvard University in the US used this type of
are zero, electrons do not have enough energy to enter the device to measure the rounding of steps observed in a
island and current does not flow. As the bias voltage between Coulomb staircase.
the source and drain is increased, an electron can pass through Electrometers based on SET transistors could also be used
the island when the energy in the system reaches the Coulomb to measure die delicate quantum superpositions of charge
energy. This effect is known as the Coulomb blockade, and the states in a superconducting island connected by a tunnel junc3 An electron in a box

PHVSICS

WORLD

S L P I I M B E R

1998

31

tion to a superconductor. Such an 4 Principle of the SET transistor

University have confirmed that
island can accommodate not only
quantum fluctuations in the charge
several charge states corresponding
of the tunnelling electrons reduce
to different numbers of Cooper
the Coulomb energy.
pairs, but also coherent quantum
superpositions of these states. SuperTowards room temperature
conducting islands could therefore
Until recently single-electron tranprovide a means for implementing
sistors had to be kept at temperatunnel
the quantum bits needed for a quantures of a few hundred millikelvin
y junctions
tum computer (see "Fundamentals
to maintain the thermal energy of
of quantum information" by Anton
the electrons below the Coulomb
Zeilinger Physics M^rWMarchpp35energy of the device. Most early
40). The feasibility of the idea has
devices had Coulomb energies of a
been shown by experiments at Delft
few hundred microelectronvolts
University, the State University of
because diey were fabricated using
New York at Stony Brook, the PTB
conventional electron-beam lithoLaboratory in Germany, the NEC
graphy, and the size and capacitance
Laboratory in Japan and in our lab. Like a MOSFET. the single-electron tunnelling (SET)
of the island were relatively large.
transistor consists of a gate electrode that electrostatically
For a SET transistor to work at room
As we have seen, the charge sen- influences electrons travelling between the source and
temperature, the capacitance of the
sitivity of the SET is ultimately drain electrodes. However, the electrons in the SET
island must be less than 10~17 F and
linked to the fact that electrons tra- transistor need to cross two tunnel junctions that form an
therefore its size must be smaller
verse the island one at a time. In isolated conducting electrode called the island. Electrons
through the island charge and discharge it, and the
than 10 nm.
1990 Bart Geerligs, Valerie Ande- passing
relative energies of systems containing 0 or 1 extra
regg, Peter Holweg and Hans Mooij electrons depends on the gate voltage. At a low
This year two experiments have
at Delft, together with Hugues source-drain voltage, a current will only flow through the
demonstrated that SET transistors
Pothier, Daniel Esteve, Cristian Ur- SET transistor if these two charge configurations have the
can work at room temperature.
same energy.
bina and one of us (MD) at CEA
Lei Zhuang and Lingjie Guo of
Saclay showed that electrons can 5 Counting electrons with a singlethe University of Minnesota, and
be counted one by one by creating
Stephen Chou of Princeton Unidevices that combine several SET electron transistor
versity in the US fabricated a SET
transistors. And in 1996 John Martransistor in a similar way to a fieldcurrent
i
tinis and colleagues at the National
effect transistor with a channel just
bias
Institute for Standards and Technol16 nm wide. The fabrication process
~ = voltage
ogy in Boulder, Colorado, showed
generated variations in the channel
that a device called the electron
that act as tunnel junctions defining
pump can count electrons with an
several different islands, and the
accuracy of 15 parts in a billion. The
behaviour of the device is dominsame group is now attempting to
ated by the smallest island.
measure the charge of the electron
As expected from theory, the
2Cg
2Cg
2Cg
with an accuracy better than 1 part voltage 2Cg
conductance of the device shows a
per 10 million by combining an elec- The current flowing in a single-electron transistor increases series of peaks as a function of gate
tron pump with a specially calibra- with the bias voltage between the source and drain, and
voltage. The Coulomb energy of
varies periodically with the gate voltage. For low bias
ted capacitor. Other metrology labs voltages, current flows when the charge on the gate
the device is around 100 meV,
are aiming to use arrays of single- capacitor is a half-integer multiple of e, but is suppressed
which is large enough to reveal
electron transistors to establish a for integer multiples of e. Each time an electron is added to single-electron effects at room temthe gate, an electron tunnels into the island, which sets the
standard for electric current.
perature. However, the Coulomb
field in the gate capacitor back to its initial value. Peaks in
The precision with which elec- the conductance are observed for half-integer multiples of
energy is too small to provide a large
trons can be counted is ultimately e, and minima are seen at integer multiples of e. For bias
Coulomb gap, so the transistor is
voltages largerthan e/C, conduction occurs independently
limited by the quantum delocal- of the gate voltage.
not sensitive enough to be used as
ization of charge that occurs when
an electrometer. Interestingly, the
the tunnel-junction conductance bewavelength of electrons is comparcomes comparable with the conductance quantum, 2e2/h. able with the size of the dot, which means that their conHowever, the current through a SET transistor increases with finement energy makes a significant contribution to the
the conductance of the junctions, so it is important to under- Coulomb energy.
stand how the single-electron effects and Coulomb blockade
Meanwhile, Jun-ichi Shirakashi and colleagues at die
disappear when the tunnel conductance is increased beyond Electrotechnical Laboratory and the Tokyo Institute of
2e2/h. In 1991 Konstantin Matveev, now at Duke University in Technology fabricated a metallic SET from a very thin layer
North Carolina, drew a parallel between the suppression of of niobium. Tunnel junctions were created by oxidizing areas
the Coulomb blockade and the Kondo effect, in which mag- of the niobium with die tip of a scanning electron micronetic impurities in metals are screened by conduction elec- scope. The tip had almost atomic resolution, allowing the
trons. Experiments by Philippe Joyez and co-workers in our researchers to make very narrow oxide barriers between the
lab, and by Leonid Kuzmin and collaborators at Chalmers island and the source and drain, and a wider barrier between
32

PHYSICS WORLD

SEPTEMIEI

1998

the island and gate.

The team measured the current through the device as a
function of both bias and gate voltages, and the results closely
match theoretical predictions. The Coulomb energy of the
device is 250 meV, which means that single-electron effects
can readily be seen at room temperature. However, tunnel
barriers fabricated in this way are highly resistive, which
means that the current is about 100 times smaller than in
devices operating at low temperatures. This problem also limits die usefulness of the device for electrometry

Delft and Richard SmaJley at Rice University in the US made

a crucial step in this direction by observing Coulomb blockade in an island consisting of a single carbon nanotube.
It is not yet clear whether electronics based on individual
molecules and single-electron effects will replace conventional circuits based on scaled-down versions of field-effect
transistors. Only one thing is certain: if the pace of miniaturization continues unabated, the quantum properties of
electrons will become crucial in determining the design of
electronic devices before the end of the next decade.

Perspectives on the future

Further reading

Researchers have long considered whether SET transistors M H Devoret, D Esteve and C Urbina 1992 Single-electron transfer in metallic
could be used for digital electronics. Although the current nanostructures Nature 360 547
varies periodically with gate voltage in contrast to the thresh- D C Glattli, M Sanquer and J Tran Thanh Van (ed) 1994 Coulomb and
old behaviour of thefield-effecttransistor - a SET could still Interference Effects in Small Electronic Structures (Moriond series, Editions
form a compact and efficient memory device. However, even Frontieres, Gif-sur-Yvette)
the latest SET transistors suffer from "offset charges", which H Grabert and M H Devoret (ed) 1992 Single Charge Tunnelling{Nat.o series,
means that the gate voltage needed to achieve maximum cur- Plenum, New York)
rent varies randomly from device to device. Such fluctuations J Shirakashi et a/. 1998 Single-electron charging effects in Nb/Nb oxide-based
make it impossible to build complex circuits.
single-electron transistors at room temperature Appl. Phys. Lett. 721893
One way to overcome this problem might be to combine L LSohn, L P Kouwenhoven and G Schoen 1997 Mesoscopic Electron
the island, two tunnel junctions and the gate capacitor that Transport (Nato series, Kluwer, Dordrecht)
comprise a single-electron transistor in a single molecule - MTinkham 1996 Introduction to Superconductivity (McGraw-Hill, New York)
after all, the intrinsically quantum behaviour of a SET tran- L Zhuang, L Guo and S Y Chou 1998 Silicon single-electron quantum-dot
sistor should not be affected at the molecular scale. In prin- transistor switch operating at room temperature Appl. Phys. Lett. 721205
ciple, the reproducibility of such futuristic transistors would
be determined by chemistry, and not by the accuracy of the MichelHtk
and
are at the Service de Physique del'Etat
fabrication process. Last year a team led by Cees Dekker at Condense, CEA-Saclay, F-91191 Gif-sur-Yvette, France

Multimag
Multipurpose Magnet Systems

A permanent solution
to your needs :
*

Programmable variable fields

Fixed fields
Rapidly rotating fields
Fields up to 2.0 Tesla (20 kG)
Bores up to 355 mm ( 1 4 " )

Homogenities better than 10 ~

Frequencies up to 3,000 RPM
Multimag magnet systems provide
programmable magnetic fields in a highly
versatile, compact and energy efficient
package. They need no water cooling, no
high current power supply and a minimum
of system integration. Multimag systems
replace electromagnets in many application
areas; saving money and time in process
development.
Since 1994 Magnetic Solutions have been
supplying advanced magnet technology for
semiconductor characterisation, laboratory
analysis and high energy physics. Call us
and see how we can contribute to your
magnet system needs.

AV;i|iii;tit Siilnf ions

PHYSICS WORLD

SEPTEMBER

1998

Unit 13, IDA Centre, Pearse St. Dublin 2, Ireland Tel: 353 1 670 4046 Fax: 353 1 670 4047
Email: contact@magnetic-solutions.com Website: http://www.magnetic-solutions.com

33

Books from Springer

F. Pobell

Matter and Methods at

Low Temperatures
2nd ed. 1996. XIII, 371 pp. 197 figs., 28 tabs.,
77 problems.
Softcover 34.00
ISBN 3-540-58572-9

lectron
Microscopy
Physics of
Image Formation
and
Microanalysis

This textbook contains a wealth of information essential for successful experiments

at low temperatures. The major part is
devoted to refrigeration techniques and the
physics on which they rely, the definition of
temperature, thermometry, and a variety of
design and construction techniques.
L. Reimer

Scanning Electron Microscopy

Bo-run Jiang

Physics of Image Formation and

Microanalysis

The
M.S. Longair

Galaxy Formation

Least-Squares
Finite Element
lod

2nd completely rev. and updated ed.

1998. Approx. 510 pp. 260 figs.
(Springer Series in Optical Sciences, Vol. 45)
Hardcover 57.50
ISBN 3-540-63976-4

1998. XVIII, 510 pp. 142 figs.

(Astronomy and Astrophysics library)
Hardcover 37.50
ISBN .1-540-63785-0
7"^,

Written by a well-known astrophysicist

who is also a superbly talented writer, this
work deals with the matter and radiation
content of the universe and the formation
of galaxies, providing a comprehensive
introduction into relativistic astrophysics
as needed for the clarification of cosmological ideas.

R.J. leVeque, D. Mihalas, E.A. Dorfi,

E.Muller

Computational Methods for

Astrophysical Fluid Flow
Saas-Fee Advances Course 17.
Lecture Notes 1997 Swiss Society for
Astrophysics and Astronomy
1998. XIV, 508 pp. 124 figs., 6 in color.
Hardcover 46.00
ISBN 3-540-64448-2

This book leads directly to the most modern numerical techniques for compressible
fluid flow, with special consideration given
to astrophysical applications. Emphasis is
put on high-resolution shock-capturing
finite-volume schemes based on Riemann
solvers. The applications of such schemes,
in particular the PPM method, are given
and include large-scale simulations of
supernova explosions by core collapse and
thermonuclear burning and astrophysical
jets. Parts two and three treat radiation
hydrodynamics. The power of adaptive
(moving) grids is demonstrated with a
number of stellar-physical simulations
showing very crispy shock-front structures.

34

B.- Jiang

The Least-Squares Finite

Element Method
Theory and Applications in
Computational Fluid Dynamics and
Electromagnetics
1998. XVI, 418 pp. 130 figs.
(Scientific Computation)
Hardcover 157.00
ISBN 3-540-63934-9

M-

Here is a comprehensive introduction to

the least-squares finite element method
(LSFEM) for numerical solution of PDEs.
It covers the theory for first order systems,
particularly the div-curl and the div-curlgrad system. Then LSFEM is applied systematically to permissible boundary conditions
for the incompressible Navier-Stokes
equations, to show that the divergence
equations in the Maxwell equations are
not redundant, and to derive equivalent
second-order versions of the Navier-Stokes
equations and the Maxwell equations.

Please order from

Springer-Verlag London Ltd.
Fax: + 4 4 / 1 4 8 3 / 4 1 5 1 5 1
e-mail: alex@svl.co.uk
or through your bookseller

A description of the physics of electronprobe formation and of electron-specimen

interactions. The different imaging and
analytical modes using secondary and
backscattered electrons, electron-beaminduced currents, X-ray and Auger electrons, electron channelling effects, and
cathodoluminescence are discussed to
evaluate specific contrasts and to obtain
quantitative information.

K.Yosida

Theory of Magnetism
1st ed. 1996. Corr. 2nd printing 1998.
% 320 pp. 47 figs.
(Springer Series in Solid-State Sciences, Vol. 122)
Hardcover 37.50
ISBN 3-540-60651-3

An important subfield of the theory of

solids that has guided a long history of
research into the phenomenon of magnetism. This book provides the foundation
for further development in this field.

p
Prices subject to change without notice.
In EU countries the local VAT is effective

Springer

PHYSICS WORLD

SEPTcatc*

1998