throbber

`
`Plasma Etching
`
`A n Introduction
`
`
`
`Edited by
`
`Dennis M. Manos
`Plasma Physics Laboratory
`Princeton University
`Princeton, New Jersey
`
`Daniel L. Flamm
`A T& T Bell Laboratories
`
`Murray Hill, New Jersey
`
`
`
`£9
`
`Academic Press
`San Diego New York Boston
`London Sydney Tokyo Toronto
`
`
`
`ILLETTE 11 12
`
`GILLETTE 1112
`
`

`

` This book is printed on acid-free paper.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Copyright © 1989 by Academic Press
`All rights reserved.
`No part of this publication may be reproduced or
`transmitted in any form or by any means, electronic
`or mechanical, including photocopy, recording, or
`any information storage and retrieval system, without
`permission in writing from the publisher.
`
`ACADEMIC PRESS
`
`A Division ofHarcourt Brace & Company
`525 B Street, Suite 1900
`San Diego. California 92101-4495
`
`United Kingdom Edition published by
`ACADEMIC PRESS INC. (LONDON) LTD.
`24-28 Oval Road, London NW1 7DX
`
`Library of Congress Cataloging—in-Publication Data
`
`Plasma etching.
`(Plasma: materials interactions)
`Bibliography: p.
`Includes index.
`
`I. Manos, Dennis M.
`1. Plasma etching.
`Daniel L.
`III. Series: Plasma.
`TA2020.P5
`1988
`621.044
`87~37419
`ISBN 0-12-469370-9
`
`II. Flamm,
`
`Alkaline paper
`
`PRINTED IN THE UNITED STATES OF AMERICA
`
`97 BB 9 8 7 6
`
`

`

`
`An Introduction to Plasma Physics
`for Materials Processing
`
`
`
`
`
`
`
`Samuel A. Cohen
`
`Plasma Physics Laboratoty
`Princeton University
`Princeton, New Jersey
`
`I. Introduction ................................. , ..............
`II. The Plasma State .............................................
`
`III. Single~Particle Motion .........................................
`A. E = constant, B = 0 ........................................
`B. E = 0, B = constant ........................................
`C. Isl-perpendicular to B ........................................
`D. Non—Uniform Fields and Other Forces ...........................
`E. Time-Varying Fields ........................................
`F. Adiabatic Invariants ........................................
`G. Summary of Particle Drifts ....................................
`IV. Plasma Parameters ............................................
`A. Temperature. Density, and Pressure ..............................
`B. Debye Length and Plasma Frequency ............................
`C. Skin Depth and Dielectric Constant .............................
`D. Collisions ...............................................
`E. Summary of Plasma Parameters in Practical Units ...................
`F.
`Instabilities i .............................................
`G. Plasma Waves ........................................... -.
`.
`
`V. Discharge Initiation ...........................................
`A. DC Glow ................................................
`B. Microwave Breakdown .......................................
`
`VI. An Application—«The Planar Magnetron ............................
`Acknowledgements ...............................................
`References .....................................................
`
`186
`187
`
`193
`194
`195
`197
`198
`201
`202
`204
`205
`205
`207
`213
`217
`229
`229
`238
`
`241
`241
`245
`
`248
`258
`258
`
`Plasma Etching:
`An Introduction
`
`185
`
`Copyright © 1989 by Academic Press. Inc.
`All rights of reproduction in any form reserved.
`ISBN 0-12~469370-9
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`

`

`techniques into a highly specialized art. Perhaps the basic foundation for
`this art is plasma physics,
`though chemistry, electrical engineering, and
`vacuum technology have defensible claims. Each must be understood and
`well-practiced for the material processing to succeed.
`
`’
`lengths, time scales, excitable modes (both stable and unstable), and atomic
`and surface processes important
`in the initiation and maintenance of
`plasma discharges. To discuss these topics in a practical way, we include
`information from a wide range of plasma configurations used in plasma
`processing, presenting material on dc- and rf-driven discharges with and
`without externally applied magnetic fields.
`
`Samuel A. Cohen
`
`developed in the later sections. Section III concerns single-particle motion;
`
`processing devices do not have even this simplicity, in large part due to the
`boundary between the plasma and the solid surfaces. It is at the boundary
`that our ultimate interests lie. However,
`the reader should find that the
`simplified situations described here will form a good understanding of the
`often counter—intuitive behavior of plasmas and will encourage improve—
`ments in existing equipment or processes.
`Plasmas are usually created in metal vacuum vessels, commonly used to
`attain the low pressures essential for particular plasma properties. Plasmas
`have a propensity to fill every crevice in these vacuum vessels. (The word
`“plasma” originates from a Greek root
`,
`'
`'
`efforts are made to constrain the plasma
`these are not completely successful. Device operation is considerably af-
`fected. To emphasize this, we shall use the label “containment” vessels to
`fully appreciate that some plasma reaches everywhere in them.
`We assume familiarity with college physics (especially Maxwell’s equa-
`tions) and introductory calculus. Most equations will be presented both in
`cgs and practical units to aid their easy application. Section II presents
`most of the basic ideas and definitions concerning plasmas. These are
`
`

`

`ll. The Plasma State
`
`Plasmas are a state of matter that consists of a large group of electrons and
`ions with nearly equal numbers of opposite charges, each particle moving at
`a high rate of speed relative to the others. It is the precise electric field of
`the individual charged particles that gives the plasma its unique properties.
`The electric field of each particle influences the motion of distant particles,
`whether they have like or opposite charge. This action-at-a-distance causes
`a wide variety of waves and instabilities to be possible in a plasma. And
`because each particle is influenced by the electric and magnetic fields of
`many particles, the term used to describe the kinematicslS collective motion.
`The electric field of a single isolated electron1s proportional to r 2.The
`volume of a spherical shell a distance r
`from that electron increases
`proportional to r2. Thus, the product of the electric field times the volume,
`a measure of the effectiveness of the field at a distance, is constant (Fig. la).
`It
`is the same near the electron as it
`is far away, showing how the
`action-at—a-distance arises.
`
`An Introduction to Plasma Physics for Materials Processing
`
`187
`
`
`
`Section IV gives details of plasma parameters; Section V is devoted to
`plasma formation; and Section VI applies the previous four sections to the
`magnetron device.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`What differentiates a plasma from a group of neutral atoms that also has
`equal and large numbers of electrons and ions? It is the distance at which
`the electric field is felt strongly. Neutral atoms (and molecules) have an
`electric field no stronger than a dipole. This falls-off proportional to the
`distance cubed or faster. Hence at large distances, it is weak compared to
`the Coulomb electric field of the bare electrons found in a plasma. Because
`of the very short range of their electric and magnetic fields, molecules
`interact with each other only by “hard” collisions, meaning close encoun-
`ters, typically at separations of about 1 A. Free electrons and ions in a
`plasma interact over much greater distances, typically 1000 A or more, as
`well as less, of course!
`Numerous distant interactions will change a charged particle’s trajec-
`tory more than the infrequent hard collisions (Fig. lb). For this reason close
`encounters may be unimportant to the charged particles in a plasma. (This
`is related to another reason why a group of neutral atoms does not behave
`like a plasma. The quantal nature of the electronic energy levels in an atom
`precludes the small changes in energy required by distant encounters so
`important to plasma behavior.) Hence plasmas are frequently termed colli-
`
`
`
`

`

`Samuel A. Cohen
`
`\
`SPHEMCALSHELL
`VOLUME=47r2m
`
`'
`
`'
`'
`
`.
`'
`
`TRAJECTORY
`BENTBY“CLUMPS”
`OFPARTmLES
`ATiANDfi
`
`'
`
`trajectory of the test particle more than the infrequent hard collisions with nearby particles
`
`a) The electric field of an isolated electron falls OH proportional to r”. The
`FIGURE 1.
`volume of a spherical shell around that electron increases as r2. This shows the importance of
`distant particles to the motion of that single electron. b) When a charged test particle moves
`through a cloud of charged particles,
`the electric field of the distant particles alters the
`
`TEST
`
`PARHCLEB
`
`.
`
`‘/’/_fl_______jMPROBABLE
`”HARD" COLLISION
`
`

`

` An Introduction to Plasma Physics for Materials Processing
`
`TEST PARTICLE
`
`
`
`
`
`__ _— _
`_
`—
`__+
`+.++ + + +_—+++ + —+ +
`+~+*+— +71 ++\
`+ __~
`_
`+
`_
`+ +++
`++-—_ ‘+
`—_+__+_\+ +
`+
`_+ L, _+,
`++
`+- -~ -
`—
`—~— +-—+* +‘“++ +
`++_+ :+‘ +-+_ ‘+‘+
`__+_+_+:+_ +_+ *+_:;_+
`i:_++— -+ +_+__ _.+
`+
`_+‘ +~_.+ + + +
`
`DEBYE LENGTH
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`FIGURE 2. When a negative test charge is placed in a plasma,
`a volume around it, of radius
`equal
`to the Debye length,
`is partially depleted of negative
`charges. This results in the
`shielding of that test particle’s electric field at distances larger th
`an the Debye length. Outside
`that volume the negative and positive charges are more nearly e
`qual in abundance
`
`sionless, meaning that individual hard collisions are unimportant compared
`to the numerous distant soft ones. Exact criteria, which hinge on the
`characteristic lengths in the problem, must be examined before a particular
`plasma can be correctly labeled collisionless. For example, is the size of the
`containment vessel larger than the mean—free~path between hard collisions?
`When many charged particles are present,
`they alter their positions,
`like-charged particles being repelled and oppositely charged ones being
`attracted,
`to reduce the distance over which an applied electric field is
`effective. The source of this field could be external metal plates attached to
`a battery, or a single electron placed in the plasma as a test particle. The
`shielding (Fig. 2) occurs in a distance called the Debye length, whose size
`determines many properties of the plasma relevant for material processing.
`The plasmas typically used in materials processing have a Debye length in
`the range of .01 to 1.0 mm. Within a sphere of this radius there are still
`many (typically more than a million) charged particles to influence and to
`be influenced by the test charge.
`Shielding does not prevent the penetration of all fields into a plasma.
`Certain electrostatic and electromagnetic waves, for example, can penetrate
`into and propagate through plasmas. This is essential to many schemes for
`plasma heating.
`By extrapolation it is clear that, at too high a density, plasma particles
`may be too close together and thus “appear” like dipoles to the distant
`particles. Also, at room temperature electrons and ions will rapidly recom-
`bine to form neutral atoms and molecules. Hence, to sustain a plasma, its
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`

`

`
`
`
`
`
`190
`Samuel A. Cohen
`
`
`temperature must be kept above some minimum, about 10,000 K (about
`
`1 eV), which depends on the density. The higher the temperature, the higher
`
`is the allowed density. From this, one can estimate that most laboratory
`
`plasmas have densities in the range of 108 to 1012 cm‘3.
`
`In astrophysical situations, plasmas exist at much lower densities
`
`(10‘3 cm”) as in the interstellar medium, and at much higher densities
`
`(above 1020 cm”) as in certain stars (Fig. 3). Other systems, such as
`
`electrons in metals or ions in liquids, also have certain properties like those
`
`of our gaseous plasmas.
`
`The approximate equality between oppositely charged particles is termed
`
`quasineutrality. It is one of the most basic tenets of plasma physics. A 1%
`
`excess of either charged species in a plasma with parameters like a mag-
`
`netron planar etcher would cause an electric field in excess of about 1000
`
`volts/cm. This large field would cause the electrons to rearrange their
`
`positions to restore a more balanced distribution of charges. Small electrical
`
`imbalances do occur. The resulting restoring force causes the plasma
`
`electrons to oscillate internally at a frequency (naturally enough) called the
`
`plasma frequency. These occur, again for the magnetron etcher, at about
`
`1010 Hz.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Three phases of mattermgas, liquid, and solid—are commonly experi—
`enced first hand, i.e., they can be readily touched if not too hot or cold.
`Plasmas cannot. They are generally too hot, too tenuous, and too fragile,
`almost like a soap bubble. A further similarity to soap bubbles is that the
`plasma possesses at its boundaries a skin, called the plasma or Debye
`sheath. The sheath is about 5 Debye lengths in thickness. Trying to touch a
`plasma by penetrating the sheath can destroy the plasma or, if the plasma
`has enough stored energy, the finger! Hence one of our best ways to test
`material properties is unavailable. But visual inspection with the naked eye
`can reveal the presence and dimensions of the sheath, properties of which it
`has taken scientists decades to confirm and quantify accurately with other
`diagnostic equipment.
`As might be inferred from their densities alone, plasmas behave more like
`gases than solids. But their modes of internal motion are complex because
`of collective motion. One similarity is that both support
`longitudinal
`compressional (sound) waves. Another is that both expand to fill
`their
`containers. On encountering the walls of the containment vessel, the charged
`particles of a plasma are neutralized by attachment to charged particles
`from the solid, often metallic, surfaces. Thus, the rate at which the plasma
`expands will determine how long, and if, it can be sustained. The charged
`particles lost by contact with the wall must be replaced if the plasma is to
`continue in its existence (Fig. 4). The steady state thus achieved will depend
`
`

`

`An Introduction to Plasma Physics for Materials Processing
`
`191
`
`
`
`Magnetic
`Fusion
`Reactors
`
`Process
`Plasmas
`
`(cm
`
`26
`IO
`
`22
`IO
`
`I8
`
`IO
`
`I4
`,0
`
`IO
`
`IO
`
`)
`
`
`
`ELECTRONDENSITY
`
`Yohen
`
`bout
`
`gher
`Ltory
`
`;ities
`;ities
`h as
`hose
`
`'med
`l 1%
`
`nag-
`1000
`their
`rical
`isma
`
`1 the
`bout
`
`peri—
`cold.
`
`tgile,
`t the
`
`ebye
`lch a
`isma
`test
`
`1 eye
`ch it
`)ther
`
`: like
`:ause
`
`:linal
`their
`
`trged
`:icles
`isma
`
`tI'ng
`is to
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`10‘2
`
`too
`
`I02
`
`IO
`
`ELECTRON TEMPERATURE (eV)
`
`FIGURE 3. Types of plasmas, categorized by their temperatures and densities. The corre-
`sponding Debye lengths are the diagonal lines.
`
`pend
`
`
`
`

`

`
`
`
`
`
`
`
`
`+aW
`
`all
`N“.
`/
`Containment Vessel
`
`strongly on the properties of the containment vessel, e.
`electrons and the right ions
`when and where needed.
`
`g., will it resupply
`
`Samuel A. Cohen
`
`9‘
`
`asma Source
`
`Plasma Loss To Wall
`
`

`

`
`
`nuel A. Cohen
`
`An Introduction to Plasma Physics for Materials Processing
`
`
`
`193
`
`Electron Impact Ionization
`
`IUPCE
`
`e‘ + Atom :> 2e” + Ion
`
`FIGURE 5. The impact of a plasma electron on a neutral atom may result in ionization of
`that atom.
`
`temperatures and densities of these difl‘erent species. Though the interaction
`between the charged particle pairs is much stronger than between a charged
`particle and a neutral, if the neutral density is very high, then electrons will
`collide more often with them, and, for example, the electrical resistivity will
`be aflected. The neutral particles in materials processing plasmas typically
`outnumber the charged ones by more than 104 to 1. At the extreme, if the
`neutral to charged particle ratio exceeds about 10“, the neutral collision
`frequency will exceed the plasma frequency, collective aspects of the motion
`will be lost, and the plasma state destroyed.
`
`III. Single-Particle Motion
`
`Bulk plasma motion is dominated by collective effects. Yet the motion of
`each charged particle under the influence of the local electric and magnetic
`fields is the correct description of what is happening at a microscopic level.
`And so the Lorentz force law and Maxwell’s equations provide the proper
`way to predict a single test particle’s motion. What this approach lacks is
`the back-effect of the test particle’s own motion on the local fields as caused
`by its fields acting on the distant particles. To improve the single-particle
`picture, one could start with a description of the plasma as a fluid or as an
`ensemble of particles. Then fluid equations or the kinetic equations would
`be used to describe the plasma’s evolution. These approaches are usually
`reserved for the more advanced students. Several references are listed at the
`end of this chapter that should satisfy the more ambitious. Many results
`achieved by a kinetic or fluid analysis can be reproduced in a single-particle
`description by choosing the proper initial conditions based on the known
`answer. This is often done for pedagogical reasons because the single
`particle picture is so easy to visualize and hence to remember. We use this
`
`: vessel. There
`ial source.
`
`it resupply
`
`me devices
`on into the
`be used to
`
`lensity gas.
`picking up
`usual way
`1e contain’
`act of the
`3 ions thus
`3k and the
`
`0 types of
`1e plasma
`I a dozen
`s electrons
`
`, processes
`
`)n plasma
`lC relative
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`194
`
`Samuel A. Cohen
`
`approach here.
`the motion of an
`In the single-particle approach to plasma physics,
`individual charged particle under the influence of externally applied electric
`and magnetic fields is examined. The fields are allowed to vary in space and
`time but do not change to reflect the subsequent motion of the charged
`particle. The motion of the single particle is readily obtained from the
`Lorentz force law, which (in cgs units) is
`
`mdv
`v X B
`
`where c is the speed of light (3 X 1010 cm/s), m is the mass of the charged
`particle (in gm), q its charge (—4.8 X 10‘10 statcoul for a single electron),
`0 its velocity (cm/s), and E (in statv/cm) and B (in gauss) are the applied
`electric and magnetic fields, respectively. We now discuss several simple
`cases of this equation.
`
`A. E = CONSTANT, B = 0
`
`The application of a constant and homogeneous electric field, but no
`magnetic field, results in the constant acceleration of a charged particle. The
`particle gains energy from the field at an ever increasing rate. If allowed to
`continue, the particle would eventually reach relativistic speeds. Then the
`simple Newtonian description fails and the particle gains mass rather than
`speed, and also energy and momentum. This relativistic limit is beyond the
`scope of processing plasma physics. Integrating Eqn. (1) for E parallel to
`the x-direction gives
`
`2
`
`(2)
`,
`x(t) = x0 + 0th + (12E;
`where x0 and 0x0 are the initial x-position and x-velocity of the particle. If
`the initial velocity U0 is zero or is in the direction parallel to E, this motion
`is in a straight line. Otherwise it is parabolic (Fig. 6). Though in most
`
`Applied Electric Field
`————————-———>
`
`@—-—-———-> Trajectory if Vyo=o
`
`
`
`
`Trajectory if vyo¥ 0
`
`FIGURE 6. Trajectories of charged particles in a constant electric field.
`
`

`

`
`
`An Introduction to Plasma Physics for Materials Processing
`
`195
`
`an
`
`trio
`and
`
`ged
`the
`
`ged
`3n),
`lied
`
`iple
`
`no
`
`The
`i to
`the
`ban
`the
`1 to
`
`(2)
`
`:. If
`tion
`lost
`
`situations it is adequate to ignore a particle’s past history, i.e., x0 and 00,
`this is not true in the initiation of a discharge or in most microwave-driven
`discharges.
`
`B. E = 0, B = CONSTANT
`
`As seen from the Lorentz law, if there is no electric field and no initial
`velocity, the particle remains at rest. More precisely, the magnetic force acts
`only if the particle’s motion is perpendicular to B. The result is that a
`particle moving initially along B is unaffected by B while a particle moving
`perpendicular to B has its trajectory turned into a circle. For a particle
`moving at an angle to B the net result is a corkscrew or helical motion.
`There is no gain or loss of total particle kinetic energy from a static or
`slowly varying magnetic field (Fig. 7). Integrating Eqn. (1) for this choice of
`parameters gives
`
`where
`
`2(t) = 20 + vzot
`
`x(1)= x0 + rL sin(th/mc)
`
`y(t) = y0 + rLcos(th/mc)
`
`mu C
`
`a: q;
`
`,
`
`(3)
`
`(4)
`
`v i is the component of the velocity perpendicular to B, and we assumed B
`parallel to z, B = Bz.
`Notice that the circular motion has a characteristic radius, rL, called the
`Larmor radius, and a characteristic frequency,
`036 = 27rf = qB/mc, called
`
`'1‘
`
`B
`
`-—«>
`Vu
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`FIGURE 7. Trajectory of a (negatively) charged particle in a constant magnetic field.
`
`
`
`

`

`
`
`
`
`Electron-Genera ted
`Magnetic Field
`
`
`
`
`
`
`
`Samuel A. Cohen
`
`
`
`Externally Applied
`Magnetic Fieid
`
`
`
` Expelled
`
`
`
`
`
`Field Net
`Magnetic
`
`inside the orbit.
`
`the Larmor-, cyclotron-, or gyro-frequency, which are related by
`v
`rL= i'
`
`(5)
`
`
`
`
`
`
`
`
`
`
`
`The circular motion of the charged particle produces a dipole magnetic
`field in such a way as to reduce the strength of the field inside the circular
`orbit and to increase it outside (Fig. 8). Then one can View the circular
`motion as being caused by a higher magnetic pressure on the outside of the
`orbit. This property of a charged particle or a plasma to partially expel
`magnetic fields from their vicinity (interior) is called diamagnetism. The
`dipole field has a strength called the magnetic moment equal to
`2
`
`__ quirL _ mUJ.
`4‘“
`2c
`‘ 213'
`
`(6)
`
`Another convenient picture of the motion is to separately consider the
`circular (cyclotron) and linear motions. By ignoring the cyclotron motion
`one has the guiding center motion remaining. This nearly linear motion is a
`worthwhile concept if the cyclotron radius is small compared to the other
`
`
`
`

`

` An Introduction to Plasma Physics for Materials Processing
`
`197
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`characteristic dimensions. And so it is of most use for electrons whose
`Larmor radii are small because of their small mass.
`
`C. E-PERPENDICULAR TO B
`
`If E and B are parallel to each other, there is .little change in a particle’s
`motion from the above description. That is, a particle will accelerate along
`E and B and have an unchanging circular motion around them.
`If E and B are perpendicular, however, a new effect takes place that is
`contrary to most everyday experiences, except that of the common toy top.
`What happens is that the charged particle starts to accelerate parallel to E
`until its velocity is large enough for the magnetic force to bend it. The bend
`becomes so large that the particle ends up moving in a direction that is
`perpendicular to both E and B (Fig. 9). Similarly, a spinning toy top, if
`pushed by a finger or by gravity, responds by precessing perpendicular to
`both the applied force and its axis of rotation, and not by falling over.
`So the motion of the charged particle is the sum of three distinct
`motions: cyclotron motion around B; linear motion along B; and a drift
`
` Y
`
`FIGURE 9. Drift of a charged particle in the presence of crossed E and B fields. Both
`electrons and ions drift in the same direction and at the same speed. Note that the Larmor
`radius changes as the particle gains and loses energy from the electric field. It is this difference
`in Larmor radius that causes the drift.
`
`Net
`Magnetic
`Field
`\
`
`iirected to
`18 reduced
`
`(5)
`
`iagnetic
`circular
`circular
`e of the
`
`.y expel
`'m. The
`
`(6)
`
`der the
`motion
`ion is a
`e other
`
`
`
`
`
`
`

`

`
`
`198
`
`Samuel A. Cohen
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`perpendicular to both B and E. The B-parallel and B-perpendicular (drift)
`motions now comprise the guiding-center motion.
`Integrating the equation of motion for this case is more complicated than
`in the previous paragraphs. Instead, one can take the vector cross—product
`of B with the Lorentz law, Eqn. (1),
`ignoring the dv/dt
`term, which
`describes the cyclotron motion. Then one obtains for the transverse compo-
`nent of the velocity (the drift velocity)
`
`
`
`(7)
`
`This drift velocity is independent of the particle’s energy, charge, and
`mass because the electric and magnetic forces are both proportional to a
`particle’s charge and independent of a particle’s mass. Though a heavy
`particle has a larger gyroradius, its cyclotron frequency is slower by the
`same amount so the gain and loss of energy from the electric field during
`each cycle are balanced.
`
`D. NON-UNIFORM FIELDS AND OTHER FORCES
`
`The same derivation could have been carried out for a different force F
`acting on a charged particle in a B field. The result can simply be obtained
`by replacing E by F/q, the equivalent electric field. The classic example
`most quoted is that of a charged particle in crossed magnetic and gravita-
`tional fields (Fig. 10a). The result is a drift velocity of magnitude cmg/qB,
`where g is the gravitational acceleration. This drift does depend on charge
`and mass because the gravitational force does not depend on charge but
`does depend on mass. Charges separate and the heavier particles drift
`faster. Hence, a net current flows, in contrast to the E X B case.
`Another drift would occur if the particle experiences a force due to a
`pressure gradient, vp. This would arise in a plasma with temperature or
`density gradients. The resulting drift is called the diamagnetic drift and has
`a velocity equal to
`
`UD—
`
`ch X B
`7?
`
`<8)
`
`
`
`Similarly, the drift resulting from gradients or curvature in the B field
`itself can be written down immediately after identifying the forces they
`cause. A charged particle moving along a curved field line (Fig. 10b) feels a
`centrifugal force of magnitude mvfi/Rc , where RC is the radius of curvature
`of the field. And magnetic field gradients (Fig. 100) cause a force equal to
`
`

`

`rift)
`
`:han
`duct
`hich
`
`1pc—
`
`(7)
`
`and
`to a
`
`:avy
`the
`
`ned
`
`1ple
`'ita-
`
`irge
`but
`[rift
`
`o a
`z or
`
`(8)
`
`An Introduction to Plasma Physics for Materials Processing
`
`(a)
`
`Z
`
`X
`
`
`ring
`has
`
`
`
`
`. 1d
`[e
`hey
`
`
`
`FIGURE 10. Drift motion of positively charged particles under the combined influences of a
`magnetic field and a) gravity, b) curved magnetic field, and c) magnetic field with a gradient
`(continued on next page).
`
`
`
`
`
`
`

`

`
`
`Samuel A. Cohen
`
`
`
`,4
`
`FIGURE 10c. Continued from previous page.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`,uV B. Hence the drifts are:
`
`and
`
`U r B X VB
`00: i4?
`
`mvfiRCXB
`0-w-
`
`(9)
`
`(10)
`
`In contrast to the E X B drift, these depend on a particle’s mass, charge,
`and energy. The more massive a particle, the faster it drifts. This is another
`counter-intuitive property of plasmas. Generally one expects the electrons,
`the lighter particle, to carry all the electrical currents. Not so. The ions can
`play a dominant role in current flows.
`A non-uniform electric field will alter the drift velocity by giving weight
`to the time the particle spends in the regions of different field strength and
`direction. The result for a sinusoidally varying electric field with a periodic—
`ity distance d is that the drift is proportional to the usual E X B drift times
`(1 —- rLZ/4d2). This is called a finite Larmor radius correction because of
`the presence of rL in the equation. This is different for each species and
`may cause charge separation and thus plasma waves.
`
`
`
`

`

`An Introduction to Plasma Physics for Materials Processing
`
`E. TIME-VARYING FIELDS
`
`
`
`
`
`When the electric field varies in time the drift is modified. One can again
`understand the motion from a microscopic picture. Each time the field
`“turns-on” the particle slowly accelerates parallel to E and then the E X B
`drift develops. When the electric field reverses the processes again occur,
`but with the drifts in the opposite directions. This can be seen quantita—
`tively from the following derivation (Fig. 11a and b). The magnetic field is
`along the z—axis;
`the electric field is along the y-axis. The equations of
`motion for the x and y directions are:
`x = we)?
`
`(11a)
`
`and
`
`y‘ = ‘17? — we.
`
`(11b)
`
`
`
`
`
`
`
`
`
`
`
`
`la-‘l Cyclotron Orbit
`—'l
`TIME
`
`FIGURE 11. Polarization drifts for positively charged particles when E increases monotoni-
`cally with time. If E increases slowly, the Larmor radius is unchanged Two drifts develop: the
`E X B drift and the polarization drift (parallel to E). When E increases at a fast rate, relative
`to the period of the cyclotron motion, the Larmor radius grows as the particle gains energy
`from the field, Again two drifts develop. Note that when d E/d! = 0 the polarization drift also
`is 0.
`
`
`
`
`
`

`

`
`
`202
`
`Samuel A. Cohen
`
`the
`
`By differentiating (11a) and substituting in (11b) and vice—versa,
`equations can be made separable yielding,
`
`55 + wfx = 41;”
`
`(12a)
`
`and
`
`which have solutions
`
`and
`
`y + way = 1,5,
`
`)3 = —ui sin(wct) + 2:?
`
`
`
`(121a)
`
`(13a)
`
`(13b)
`)6 = +1)i cos(wct) + £5.
`It is now easy to identify the E X B drift motion in the x—direction and
`the polarization drift in the y-direction, mczE/qBZ. The orbit shapes will
`vary depending on the rate of change of the electric field. Figure 11 shows
`two cases, both for monotonically increasing fields, not oscillating ones. In
`Fig. 11b the rapidly changing intense field increases the Larmor radius
`because it adds an appreciable amount of energy to the particle in a time
`faster than the cyclotron period.
`The polarization drift is in opposite directions for oppositely charged
`particles. Hence charge separation occurs. Again the more massive particle
`carries the current.
`
`F. ADIABATIC INVARIANTS
`
`opposite change in parallel energy. So if a particle moves into a region of
`
`A similar derivation carried out for slowly varying magnetic fields shows
`that no additional drift arises. Instead the Larmor radius grows or shrinks,
`depending on whether B decreases or increases. This is associated with a
`decrease or increase of the particle’s transverse energy (see Eqn. (4)). The
`change is such that the magnetic moment is unchanged. So vi
`increases
`proportional to B.
`The magnetic moment is called the first adiabatic invariant. It is constant
`as long as changes in E or B occur slowly compared to a cyclotron orbit. Its
`constancy reflects the symmetry and periodicity of the cyclotron orbit.
`Sitting in the particle’s frame-of-reference, a change of B can come from
`a change in its position as well as a change in the local field strength. Recall
`that static magnetic fields add no energy to a particle. It is then clear that
`the change in perpendicular energy must be accompanied by an equal but
`
`

`

` An Introduction to Plasma Physics for Materials Processing
`
`
`
`'sa,
`
`the
`
`(12a)
`
`(12b)
`
`(13a)
`
`(13b)
`
`n and
`:s will
`shows
`tes. In
`radius
`l time
`
`arged
`trticle
`
`hows
`
`inks,
`ith a
`The
`eases
`
`stant
`t. Its
`
`:‘rom
`ecall
`that
`but
`it of
`
`FIGURE 12. Magnetic mirror formed by two coaxial coils with co-directed currents, I. As a
`charged particle approaches a region of higher B,
`its perpendicular energy grows at
`the
`expense of its parallel energy. Reflection will occur if B reaches a high enough value. Note that
`the Larmor radius shrinks as the particle moves into a region of higher B.
`
`increasing B, it will continue to gain perpendicular energy at the expense of
`its parallel energy. At some point it may have lost all its parallel energy.
`WithV B still causing a force on the particle it is reflected. This property of
`spatially varying magnetic fields is called the mirror eflect and forms the
`basis for many plasma confinement configurations (Fig. 12). For the mirror
`to “wor ,” the particle must start with sufficient perpendicular energy
`because the mirror force only acts on the magnetic moment (F = av B).
`From conservation of energy and p. we can show that a particle starting
`from a region of magnetic field strength BO, will only be reflected in the
`region of higher field if its pitch angle, 0 = 1)” /vi , is less than
`B
`_
`7° = sin20= (RM)
`M
`
`(14)
`
`1
`
`where
`
`R M = the mirror ratio, and
`
`BM = the maximum field strength.
`
`The underlying principles for these statements are the laws of conserva-
`tion of energy and angular momentum. These are very powerful tools when
`a physical situation has symmetry. For example, calculations of cosmic ray
`trajectories in the earth’s vicinity are easy because of the size of the earth
`coupled with the symmetry of its magnetic field. The symmetry is such that
`other invariants of the motion, the so-called second and third adiabatic
`invariants, help to solve the problem. They are based on the symmetry and
`periodicity of the large-scale motion around the earth (bouncing back and
`forth between the magnetic mirrors at the poles and circulating around the
`equator), not just of the cyclot

This document is available on Docket Alarm but you must sign up to view it.


Or .

Accessing this document will incur an additional charge of $.

After purchase, you can access this document again without charge.

Accept $ Charge
throbber

Still Working On It

This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.

Give it another minute or two to complete, and then try the refresh button.

throbber

A few More Minutes ... Still Working

It can take up to 5 minutes for us to download a document if the court servers are running slowly.

Thank you for your continued patience.

This document could not be displayed.

We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.

Your account does not support viewing this document.

You need a Paid Account to view this document. Click here to change your account type.

Your account does not support viewing this document.

Set your membership status to view this document.

With a Docket Alarm membership, you'll get a whole lot more, including:

  • Up-to-date information for this case.
  • Email alerts whenever there is an update.
  • Full text search for other cases.
  • Get email alerts whenever a new case matches your search.

Become a Member

One Moment Please

The filing “” is large (MB) and is being downloaded.

Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.

Your document is on its way!

If you do not receive the document in five minutes, contact support at support@docketalarm.com.

Sealed Document

We are unable to display this document, it may be under a court ordered seal.

If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.


Access Government Site

We are redirecting you
to a mobile optimized page.





Document Unreadable or Corrupt

Refresh this Document
Go to the Docket

We are unable to display this document.

Refresh this Document
Go to the Docket