PHOTOCOPY
`
`Jul 22, 2015
`
`RUSH service required?: RUSH (no extra fee)
`
`Type of request: Institutional request
`
`Contact name (required): Jonathan Germann
`
`E-mail address (required): jgermann@lrell.com
`
`Phone number (required): 310.203.7048
`
`Address:
`
`1800 Avenue of the Stars, Ste. 900
`Los Angeles, CA 90067
`
`Journal or book title: Design of high density plasma sources for materials
`processing
`Author & article title: Michael A. Lieberman and Richard A. Gottscho
`Call number: TK7800 .C32b no.93:1-4- NRLF - C3572407
`
`Notes: http://oskicat.berkeley.edu/search/o441 07283
`
`Delivery method: Send via e-mail/web delivery
`Destination e-mail jgermann@irell.com
`
`How did you find out about our service?: Ask a librarian
`Can you suggest ways to raise the visibility of our service?: Include
`information next to catalog results
`
`\ {o 0 () s:-s(cid:173)
`
`UCB reference number:
`Charge: 33-
`Pre-pay: ~ 't 'ES
`
`Copyright: The material requested may be protected by copyright law Title 17,
`U.S. Code
`
`LAM Exh 1002-pg 1
`
`

`
`DESIGN OF HIGH DENSITY PLASMA SOURCES
`FOR MATERIALS PROCESSING
`
`by
`
`Michael A. Lieberman and Richard A. Gottscho
`
`Memorandum No. UCB/ERL M93/3
`
`11 January 1993
`
`ELECTRONICS RESEARCH LABORATORY
`
`College of Engineering
`University of California, Berkeley
`94720
`
`LAM Exh 1002-pg 2
`
`

`
`DESIGN OF HIGH DENSITY PLASMA SOURCES FOR MATERIALS PROCESSING
`
`Michael A. Liebennan
`Deparunent of Electrical Engineering and Computer Sciences
`University of California
`Berkeley, CA 94720
`
`and
`
`Richard A. Gottscho
`AT&T Bell Laboratories
`Murray Hill, NJ 07974
`
`ABSTRACT
`
`In this review article, we focus on recent advances in plasma source technology for materials
`processing applications. The motivation behind new source development is discussed along with
`the limitations of conventional radio frequency diode systems. Then the fundamental principles
`underlying electron heating in electron cyclotron resonance, helicon wave, inductively coupled,
`helical resonator, and surface wave plasmas are discussed with some attention to design issues.
`The transpOrt of ions to device wafers and its influence on etching anisotrophy is discussed for all
`sources. Similarly, we examine the benefits of using high density sources for minimizing plasma
`process induced damage and discuss in particular, the effects of plasma unifonnity on charging
`damage.
`
`LAM Exh 1002-pg 3
`
`

`
`CONTENTS
`
`I. ~ODUCI'ION .
`.
`.
`.
`.
`.
`.
`.
`.
`.
`.
`.
`.
`•
`.
`.
`.
`I.A
`Capacitively Coupled Radio Frequency Discharge Sources • • •
`I.B
`Limitations of Capacitively Coupled Radio Frequency Discharges
`I.C
`Overview of High Efficiency Sources
`•
`•
`•
`•
`•
`• •
`• • •
`
`.
`•
`
`ll. PRINCIPLES OF WW PRESSURE, lllGH EFFICIENCY SOURCE DESIGN • •
`ll.A Unified Analysis of Source Operation
`•
`•
`•
`•
`•
`• • • •
`•
`• •
`8
`ll.A.l
`Electron Temperature
`9
`ll.A.2
`Ion Bombarding Energy
`ll.A.3
`Plasma Density and Ion Current Density 10
`ll.B Discharge Heating •
`•
`•
`• •
`•
`•
`•
`•
`•
`
`ill. ELECIRON CYCLOTRON RESONANCE (ECR) DISCHARGES
`ill.A Source Configurations
`ill.B Electron Heating
`ill.C Resonant Wave Absorption
`
`IV. HELICON DISCHARGES
`IV .A Helicon Configurations
`IV .B Helicon Modes •
`•
`•
`IV .C Antenna Coupling •
`•
`IV.D Helicon Mode Absorption
`
`•
`•
`
`• •
`
`•
`
`V. INDUCTIVE DISCHARGES •
`•
`V.A
`Inductive Source Configurations
`V .B
`Power Absorption and Operating Regimes
`•
`V .C
`Source Operation and Coupling • •
`•
`V .D Low Density Operation and Source Efficiency
`
`VI. HELICAL RESONATOR DISCHARGES
`
`Vll. SURFACE WAVE DISCHARGES
`
`• • •
`
`•
`
`•
`
`1
`2
`4
`5
`6
`8
`
`11
`
`12
`12
`14
`15
`
`18
`18
`19
`21
`21
`
`23
`23
`24
`25
`26
`27
`
`30
`
`32
`33
`
`35
`36
`37
`
`• •
`
`•
`
`•
`
`•
`
`•
`
`•
`
`•
`
`39
`
`43
`43
`43
`43
`
`45
`46
`48
`51
`
`Vlll. PLASMA TRANSPORT
`• •
`•. • •
`•
`•
`•
`Vlll.A The Ion Energy Distribution Function •
`Vlli.A.l Ion Transport and Etching Anisotropy 34
`Vlll.B Methods for Measuring Ion Energy Distribution Functions •
`Vlll.C Methods for Measuring Plasma Potentials •
`• •
`•
`• • •
`Vlll.D Measurements of Energy Distributions and Potentials
`VIII.D.l Ion Acceleration Outside the Sheath 37
`Vlll.D.2 Transverse Ion Energy 38
`Vill.E Ion Energy Control
`•
`•
`• •
`• • •
`Vlli.E.l Plasma Anodization 42
`
`• •
`
`•
`
`• •
`
`• •
`•
`•
`•
`•
`IX. Device Damage •
`• •
`IX.A Atomic Displacement Damage
`IX.B Contamination •
`•
`•
`• • · • •
`DCC Charging
`•
`•
`•
`•
`• • •
`IX.C.l Plasma Uniformity 44
`IX.C.2 Biasing 44
`•
`•
`•
`IX.D Radiation
`• •
`• •
`X. SUMMARY AND REMAINING QUESTIONS
`
`XI. SYMBOL DEFINITIONS •
`
`XII. REFERENCES
`
`•
`
`•
`
`•
`
`•
`
`-i-
`
`LAM Exh 1002-pg 4
`
`

`
`-ii-
`
`and, in particular, etching. Heavy emphasis is placed on simple, analytical, unifying theories and
`Based on our interests and experience, we focus on applications of plasmas to microelectronics fabrication
`discussion meaningful. Finally, any review reflects the biases of the authors and this work is no exception.
`may adversely affect materials properties, but too little has been reported and analyzed to make a thorough
`sources is a matter of some concern and recent work illustrates that sudden mode changes and bistability
`has appeared in print and it is premature to review the field. Similarly, the stability of high efficiency
`simulation. While impressive results have been reported recently and we will draw on some of these, little
`areas that while important are not yet ready for review. Foremost amongst these is the field of numerical
`isms is the subject of its own review and clearly beyond the scope of this work. Third, there are pertinent
`induced electtical damage in microelectronic circuits but a comprehensive discussion of damage mechan(cid:173)
`use of high efficiency plasmas. For example, we discuss aspects of source design that affect plasma(cid:173)
`above. Second, we resttict our focus to those aspects of plasma processing that are uniquely affected by the
`reference every paper. Rather, we have opted for highlighting key results in line with our objectives stated
`diagnostics, and modeling of high efficiency sources is now so voluminous that we are not able to review or
`Before beginning such a review, several disclaimers must be made. Ftrst, the literature on applications,
`
`focus on problems of technological interesL
`choosing designs that can best meet their customers' expectations, and to inspire the research community to
`make wise choices in constructing or purchasing sources, to guide vendors of high efficiency sources in
`development is needed. In so doing, we hope to assist those who use plasma for materials processing to
`of source design that affect processing results; and, (3) highlight areas where additional research and
`framework from which all "high efficiency" sources may be viewed and compared; (2) outline key elements
`conventional counterparts operated at low pressure. The purpose of this review is to: (1) develop a unified
`differences between the novel sources, whose most distinctive characteristic is higher efficiency than their
`remain elusive and are the subject of much current research, one can clearly identify commonalities and
`cessing? And why are the conventional approaches inadequate? While the answers to these questions
`But which scheme is best? What are the key aspects to plasma source design that impact materials pro(cid:173)
`
`ing integrated circuit manufacturing.
`coupled radio frequency discharge now used in many factories for etching and deposition of thin films dur(cid:173)
`areas. All these schemes purport to offer advantages over conventional approaches such as the capacitively
`generation schemes are now being offered to satisfy manufacturers' needs in these materials processing
`ing (46-59), surface passivation (60), and oxidation (53,61-63). Besides ECR, many other "novel" plasma
`beyond etching and deposition have also been reported and include ion implantation (41-45), surface clean(cid:173)
`silicon nittide, boron carbide, and Si02 to name just a few materials (34-40). Applications of ECR plasmas
`all been reported in recent years (7-33). Similarly, ECR plasmas have been used to deposit amorphous Si,
`Si, single crystalline Si, silicides, AI, Mo, W, Si02, polymers, and m-V compound semiconductors have
`tems based on electron cyclotron resonance (ECR) heating (3-6). ECR plasma etching of polycrystalline
`deposition of thin films. As a result, we have witnessed the introduction of a new generation of plasma sys(cid:173)
`optimization and control (1,2) which, in tum, have led to improved plasma reactors for the etching and
`The advent of sub-micron electronic device fabrication has brought unprecedented demands for process
`
`L INTRODUCTION
`
`j
`I
`
`i
`
`. Murray Hil~ NJ 07974
`AT&T Bell Laboratories
`
`Richard A. Gouscho
`
`and
`
`Berkeley, CA 94 720
`
`Department of Electtical Engineering and Computer Sciences
`
`University of California
`
`Michael A. Liebennan
`
`DESIGN OF illGH DENSITY PLASMA SOURCES FOR MATERIALS PROCESSING
`
`-1-
`
`. . . . . . . . 60
`
`XIII. FIGURE CAPTIONS . . . . . . . . . . . . . .
`
`LAM Exh 1002-pg 5
`
`

`
`1 Exceptions to this rule are also po~sible in low frequency ele~eganve an "th the plasma.(64). In the laner,the plasma potential
`d de discharges In the former, the build-up of negative
`
`~s lie between the two electrode potentials if sufficient current is drawn from the plasma (65).
`·
`
`can reduce the plasma potential below that of large surfaces m contact WI
`
`·
`
`. be
`
`° ·
`
`large ion bombardment energies are used to tailor film stress and stotc orne
`·
`for chip encapsulation is ordinarily performed at frequenctes ~e: try (7S)
`50 and 500 kHz where relauvely
`times lower than the commonly used stan
`(76,77). For deposition and isotropic etch ~h~~~·~r ~~pie, silicon nitride deposition .used
`range widely from less than 0.1% to near Y
`d to be higher and frequencies some(cid:173)
`
`pen
`
`•
`
`0
`
`.
`
`gies (sheath voltages) are high, >200 V, ~d :::o: •o;::u:: :as ~o~position and plasma conditions
`
`1 tron distributtons (e.g. two-tempera
`
`h 1
`
`we 1ail ~ ec
`
`tr
`
`Max
`
`•
`
`observed. with the bulk electron temperature some~es ~u~ ~ss . 1 w The' degree of dissociation can
`than 1 v (74 75). Ion acceleration ener-
`tures of order 3 V. However, non-
`tore) are also
`-1010 cm-3 and mean electron energtes are of orderS V, correspon. g
`quency is typically 13.6 MHz. and mul.uple wafer systems are
`to Maxwellian electron tempera(cid:173)
`clln
`tropic etching, pressures are 10 e rang~
`common Plasma densities are relatively low.
`0 1 1 W/cmz the driving fre-
`d tasma parameters are gtven. For antso-
`
`1(}-100 mTorr power denstttes are · -
`.
`
`ln the second column of Table 1, typtcal RF di e source an P . .
`
`•
`
`. th
`
`. od
`
`j
`
`.
`
`' .,
`J
`
`T
`1
`
`(RIE's)-a misnomer, since the etching is generally a chemical process enhanced by energetic ion
`electrode, and used to remove substrate material, such reactors are commonly called reactive ion etchers
`Fig. lb, are also in widespread use. When operated at low pressure, with the wafer mounted on the powered
`gases are removed by the vacuum pump. Coaxial discharge geometries, such as the "hexode" shown in
`strates are placed on one electrode, feedstock gases are admitted to flow through the discharge, and effluent
`chamber containing two electrodes separated by a spacing I and driven by an rf power source. The sub(cid:173)
`als processing. An idealized source in plane parallel geometry, shown in Fig. 1a, consists of a discharge
`Capacitively driven rf discharges-so-called rf diodes-are the most common sources used for materi(cid:173)
`
`I.A Capacitively Coupled Radio Frequency Discharge Sources
`
`tion, it does conform to common usage. We apologize to the purists.
`citly written. Pressures are given in Torr or mTorr. While this does not conform to international conven(cid:173)
`electron charge in coulombs. Energies are usually given in units of volts, not e V, so the value of e is expli(cid:173)
`sistent set of units. Generally, magnetic field is expressed in gauss, distances in m, em, or mm, and the
`Throughout this paper we strive to be consistent with dimensional analysis despite not using a con(cid:173)
`
`investigation is needed.
`charge-induced damage. In the final section, we highlight remaining issues and the areas where further
`by source design, to processing results such as etching anisotropy, atomic displacement damage, and
`available from other systems. In sections Vlli and IX, we relate ion energy and plasma uniformity, dictated
`in section vm on measurements of ion energy distributions, mostly in ECR systems since little data are
`source design parameters such as the magnetic field and power absorption profiles. We focus our attention
`Obtaining such control is largely independent of the electron heating mechanism but depends critically on
`In section vm, we tum to the issue of plasma transport and independent control of ion energy and flux.
`tron heating and power absorption, since these are the primary differences between one source and another.
`helicon, inductive, helical resonator, and surface wave sources, respectively. Emphasis is placed on elec(cid:173)
`with which one source can be compared to another. In sections m-vn we discuss in greater detail ECR,
`absorbed power, and source dimensions. In this way, we provide an approximate but common framework
`perature, ion bombardment energy, and plasma density in terms of the gas phase cross sections, gas density,
`compare one source with another, use a simple analysis in Sec. n that allows estimation of electron tem(cid:173)
`tions, we first discuss the fundamental principles underlying high efficiency plasma source design and, to
`quency plasmas and elaborate further on the advantages of high efficiency sources. In the following sec(cid:173)
`In the remainder of this section, we review briefly the properties of capacitively coupled radio fre(cid:173)
`
`sources and processing systems.
`trol of ion energy and decoupling it from ion flux control is fmther motivation for developing new plasma
`placement damage while meeting linewidth and throughput specifications. Therefore, gaining superior con(cid:173)
`topography. Ion energy control is used in plasma etching to optimize selectivity and minimize atomic dis(cid:173)
`in plasma deposition to tailor film properties such as stress, composition, refractive index, crystallinity, and
`In conventional rf systems, ion energy and flux are inexorably linked. But, ion energy control is needed
`
`operate at lower pressure and higher efficiency.
`tivity in etching and produce electrical damage that reduces device yield. Thus, new sources are needed to
`power input to a capacitively coupled system leads to high ion bombarding energies that can degrade selec(cid:173)
`zation and dissociation necessary for high throughput, low-cost manufacturing. Unfortunately, excessive
`rf sources are inefficient at low pressure so that high powers must be used to achieve the high rates of ioni(cid:173)
`tional capacitively coupled rf reactor. This means operating plasmas at lower pressures. But, conventional
`trol, we need to increase the anisotropy of ion ttanspon to the device wafer from what it is in the conven(cid:173)
`plasma etching of gate electrodes, contact windows, and metallic interconnections. To achieve such con(cid:173)
`only 0.25 J.UD, variations in CD must be less than 0.025 J.UD. This requires unprecedented anisotropy in the
`linewidth variation of 0.1 J.UD can be tolerated but by the end of the 20th century when the CD should be
`width of the ttansistor gate is specified to better than 10%. For yesterday's CD of 1 J.UD, this means a
`placed unprecedented demands on process control. Consider critical dimension (CD) control where the
`Why new sources? In plasma etching, the shrinking dimensions of micro-electronic devices have
`
`quantitative diagnostic measurements.
`
`tron temperature in units of volts.
`.
`c(Bondih uo) n lSI 1? req~ (e T IM)ltZ where e and M are the charge and mass of the ion and T e is the elec-
`modifications (66-73). The plasma an~ ~eath ~odels thmustlasmJOmheath edge be equal to the ion-sound
`ous high voltage sheath models (for RF diodes), such as co~~~ at their interface. The usual joining
`by various ion space charge sheath laws, ~eluding low voltag~ .sh
`s ~d collisional Child laws and their
`low pressures. In the positive space charg~ sh~. high fields eXlS~~~~ high=ity sources) and vari(cid:173)
`!
`bulk plasma dynamics are described by ambtpolar . uston
`ics that are described
`discharges. The bulk regton ts quast-neutral, ~ bo~~~ gh pressures and by free-fall ion loss at
`The separation ofthe.~harge ~to bulk and sh~ regtons 18 =:time-averaged fields are low. The
`· portant paradigm that applies to all
`ence of positive space charge sheaths near all surfaces.
`~:~~~!::ec::::.~::=~~d~uk;:~~uL ~tron confinement is ensured by the pres-
`escape to
`t to any large electrode and wall sur-
`the instanbalantaneous thsh~th egn:!l~::! :::· Ex:t for such brief moments, the instantane(cid:173)
`are lost to the electrode only wh~n the oscillating cloud clos: a~7. :Cfic~ent num~r of electrons ~
`th elec~e During that time
`bard th 1 trode over an rf cycle In contrast, electrons
`
`a-s
`
`that the mean ton velocity at e p
`
`•
`
`e
`
`.
`
`to d
`
`·
`
`·
`
`·
`
`o m ve OClty Us -
`·
`
`· ·
`
`·
`
`.th
`
`. .
`
`ce e ton c
`
`h
`
`e e ec
`
`We note that positive ions continuously born
`
`between the two electrodes might typically vary between 100 V and 1 kV.
`powered electrode for asymmetnc systems
`bombarding energies E; can be~ high as V(Fr~/21fo)r s::~c :~m:ol~:~ amplitude (peak rf ~oltage)
`energetic-ion bombardment, which can enhance, mhibtt, oro ~se m ;. 2) and as high as V f at the
`discharge can be acce era
`odif surface reactions. Typtcalton
`ted b th sheath .fields to high energies as they flow to the substrate, l~ng. to
`.
`directed fro
`appearing within the sheaths. ThlS excess producesfla s~ng ef the. bulk plasma near the center of the
`oscillation period; i.e., the post~ve charge exceeds the ne:n v-avera:ed electric field within each sheath
`cloud creates sheath regtons n~ eac
`ti e char e in the system, with the excess
`13.6 MHz. the masstve ~ons respon h eiectrode that contain net positive charge when averaged over an
`(13.6 MHz) drivmg ~ol~ge, osc ~ o~ to the time-averaged electric fields. Oscillation of the electron
`frequencies, th~ ~obile plasma e:tro;s· ;::~~: :th~ •:e positive space charge cloud of the ions. At
`Flg. 2 for a symmetncally
`tantaneous electric fields produced by the rf
`·
`The physical operation of capacitively driven discharges is ~nabebtwly welthl eunt·odnersantoadd ei~~op:O:
`
`.
`
`-
`
`een
`
`r/•
`
`t w
`
`tg.
`
`·
`
`th
`
`. .
`
`.
`
`Y e
`
`Iasma to the electrode Ions owmg out o
`
`th
`
`m e P 1
`
`di harg operated at frequenctes
`
`.
`
`e
`
`sc
`
`th .
`
`ven
`
`dri
`
`·
`
`~lasma etchers. In terms of the physical properties of these systems, this disuncuon 1S somewhat arbitrary:
`bombardment of the substrate, rather than a removal process due to reactive ions. When Tra~ ~hi!h:
`
`essure with the wafer mounted on the grounded electrode, such n:ac~~ ~ co~mo Y re err .
`
`-3-
`
`-2-
`
`•
`
`.
`
`LAM Exh 1002-pg 6
`
`

`
`(coupling) across dielectric windows over a wide operating range of plasma parameters. Degradation of and
`sport in such geometries is inherently radially nonuniform. Another critical issue is efficient power transfer
`cylindrical systems with length-to-diameter ratios of order or exceeding unity. Plasma formation and tran(cid:173)
`geometry of typical RF diodes (two closely spaced parallel electrodes), high density sources are often
`required process uniformity over 200-300 mm wafer diameters. In contrast to the nearly one dimensional
`sity, source development. there are many issues that need to be resolved. A critical issue is achieving the
`Although the need for low pressures, high fluxes and controllable ion energies has motivated high den-
`
`-
`
`(Sec. IX).
`
`and improving process uniformity. But, the wafer is exposed to higher levels of damaging radiation as well
`configurations are useful for obtaining increased ion and radical fluxes, reducing the spread in ion energy.
`process chambers are sometimes combined, or the wafer is placed very near to the source exiL Such
`not actually remote. For reasons that are discussed further in Sees. ll.A.2, VIII.D, and IX.C. the source and
`cal production occurs within the process chamber near the wafer (See Sec. VIII.D). Hence such sources are
`often called ''remote' • sources. another misnomer since at low pressures considerable plasma and free radi(cid:173)
`faces, while providing a magnetic field-free plasma environment at the wafer. Such configurations are
`rounded by de multidipole magnetic fields to enhance plasma confinement near the process chamber sur(cid:173)
`process chamber in which the wafer is located. As shown in the figure, the process chamber can be sur(cid:173)
`
`Figure 3 also illustrates the use of high density sources to feed plasma into a relatively distinct. separate
`
`helical resonator or inductive sources.
`traSt to the ECR and helicon sources, a de magnetic field is not required for efficient power coupling in the
`discharge chamber; rf power is inductively coupled to the plasma by transformer action (Sec. V). In con(cid:173)
`single-tum, lossy conductor that is coupled to a multitum non-resonant rf coil across the dielectric
`(Sec. VI). An inductive (or transformer) coupled source is shown in Fig. 3d. Here the plasma acts as a
`1ighL Efficient coupling of the RF power to the plasma is achieved by excitation of a resonant axial mode
`wave structure, i.e. supporting an electromagnetic wave with phase velocity much less than the velocity of
`3c, the external helix and conducting cylinder surrounding the dielectric discharge chamber form a slow
`transfer the wave energy to the plasma (82-86) (Sec. IV.D). For the helical resonator source shown in Fig.
`con wave within the source plasma. Resonant wave-particle interaction (Landau damping) is believed to
`driven anteMa placed near the dielectric cylinder that forms the source chamber allows excitation of a heli(cid:173)
`A helicon source is shown in Fig. 3b. A weak (50 -200 G) de axial magnetic field along with an rf(cid:173)
`
`which the wafer is located.
`the resonant magnetic field is B = 875 G. The plasma streams out of the source into the process chamber in
`the electron gyration frequency at resonance. For the typical microwave frequency f = 2450 MHz used.
`at ro = roce where the wave is absorbed. Here ro = 27t/is the applied radian frequency and roce = eBim is
`excites a right hand circularly polarized (RHP) wave that propagates to a resonance zone, for cold electrons.
`field. Microwave power is injected axially through a dielectric window into the source plasma, where it
`electromagnet coils surrounding the cylindrical source chamber generate an axially varying de magnetic
`coupled to the plasma. For the electron cyclotron resonance (ECR) source shown in Fig. 3a, one or more
`electrode are illustrated in Fig. 3. However, sources differ significantly in the means by which power is
`The common features of power transfer across dielectric windows and separate bias supply at the wafer
`
`greater length in Sec. Vlll.
`the ion bombarding energy (through the wafer electrode power) is possible. This subject is discussed at
`tively coupled rf solii'Ce. Hence independent control of the ion/radical fluxes (through the source power) and
`control the ion energy, the electrode on which the wafer is placed can be independently driven by a capaci(cid:173)
`surfaces. DC voltages, and hence ion acceleration energies. are then typically 20 -30 Vat all surfaces. To
`capacitive power transfer is key to achieving low voltages across all plasma sheaths at electrode and wall
`tric window, rather than by direct coMection to an electrode in the plasma, as for an rf diode. This non(cid:173)
`low pressure, a common feature is that the rf or microwave power is coupled to the plasma across a dielec(cid:173)
`in Fig. 3, and typical source and plasma parameters are given in Table 1. In addition to high density and
`new generation of low pressure, high efficiency plasma sources. A few examples are shown schematically
`The limitations of rf diodes and their magnetically enhanced variants have led to the development of a
`
`I.C Overview or High Efficiency Sources
`
`)
`
`I
`
`J
`
`1
`
`1
`
`·
`
`B
`
`'
`
`n~u
`
`.
`
`.
`
`•
`
`fte
`
`11 acuum tu
`
`· ed .
`
`be
`
`10-4 -10-1
`
`20-500
`
`2-7
`
`1010 -1012
`
`0-1
`
`300-500
`
`2-50
`
`0-2450
`100-5000
`
`0.5-50
`
`High Density Source
`
`10-6 -10-3
`200-1000
`
`109 -1011
`
`1-5
`
`300-2000
`
`1-10
`
`0.05-13.6
`50-2000
`10-1000
`RF Diode
`
`0
`
`v
`v
`cm-3
`kG
`cm2
`I
`
`MHz
`w
`mTorr
`Units
`
`Fractional Ionization X· II
`Ion Acceleration Energy E;
`Electron Temperature T.
`Plasma Density n
`Magnetic Field B
`Cross Sectional Area A
`VolumeV
`Frequency/
`PowerP
`Pressurep
`Parameter
`
`AND CONVENTIONAL RF PLASMA SOURCES
`
`TABLE 1: TYPICAL PARAMETERS FOR ffiGH EFFICIENCY
`
`-5-
`
`-4-
`
`significant problem fo~ processes h g mthtegmfi:..-~ CJI'Cwts (79). ~inally' low fractional ionization poses a
`significant cost savin s in fabricatin
`to have the potential for
`chambe ~ enVIrorednment. m which a smgle wafer is moved by a robot through a series of process
`~es are generally required for single wafer processing in a
`clustered
`reprodu:ibility :;er ~o~~~tes m:;afer or batch processing, with consequent loss of wafer-to-wafer
`many proc~s applications. The low process rates resulting from the limited
`ion flux ·n rf diod
`relatively narrow window f~ Urthennore, the ~mb~natlon of low Ion flux and high ion energy leads to a
`or loss of :: ~~~or Fwafers placed on the ":riv~n electrode: this can result in undesirable damage,
`. • as we as a reasonable diSSOCiation of the feedstock gas, sheath voltages at the
`driven elec
`or sc:mlc~~uctor pn Junctions. Hence, for a reasonable (but
`relatively low) ion flux
`tage and curr:n~::l in ~ode :depend~. The ~tuation is ~ogo_us to the lack of independent vol(cid:173)
`e ton acce era-
`bon energy E
`. A crucial limiting feature of RF diodes is that the ion bombarding flux r. = nu and th
`I.B Limitations or Capacitively Coupled Radio Frequency Discharges
`
`. an are sat
`·d
`
`d
`
`tools are used to control interface quality
`
`.
`
`.
`
`·
`
`uste
`
`rs.
`
`s are m use. processmg rates are still low at low pressures and sputtering contamina-
`WI a second RF solii'Ce. Although these so-
`Y putting the wafer on the
`Sg energy • an Its ~nergy distribution, should be controllable independently of the ion
`' se ec:'?ty ~d ~ag~ c~iltrol demands for next-generation fabrication, the
`
`·
`
`ac tev
`
`"th
`
`uxes. orne control oveuon bombarding energy can be h · ed b
`
`e ec
`
`•
`
`•
`
`stem
`
`lion is an issue. sy
`called rf triode
`undriven electrode and independently biasing this 1 trode
`and neutral fl
`mean ion bombardin WI
`
`· ·
`w ere e ~tock costs and disposal of effiuents are issues.
`
`I
`
`To meet the line · dth
`
`1he ~ ~eld is applied (80,81). However, the plasma ~ ':
`
`can damag~ thin gate o::!o~: ~:~~~~~rna over the wafer can give rise to a lateral de current that
`fabrication I!deed,u~formny whtch ~ay hmn therr applicability to next-generation, sub-micron device
`do not have ood
`• While thiS IS an Improvement. MERlE systems
`is rota:' in the ~~~s:::::r !oa ~r:: ~~ :formi~ (at_Ic:ast ~muthally), the magnetic field
`ly due to E x B drifts, where E and B are the local electric
`and m
`
`y an :wmu
`
`.
`
`etic fi
`
`. •
`
`. CJ 0
`
`·
`
`·
`
`·
`
`and
`
`~ enhances plasma confinemenL This results in a reduced sheath volta
`
`an
`
`increased
`
`rform
`
`ve
`
`pro_callpe
`
`Various magnetically. enhanced rf diodes and triodes have also been developed to im
`
`trode on
`netron. In the
`reacbve ton etcher (MERlE)
`of ~ rf. reactor. These mclude, for example, the Applied Materials' AMT-5000
`
`to th~ p~ th: :er sits. The magnetic field increases the efficiency of power transfer from the so~-
`h" hAMf MERlE, a DC magnetic field of 50-300 GIS applied parallel to the powered elec
`~lectromcs Center of ~orth Carolina's split cathode RF mag-
`magneti Y enhanced
`ance
`
`.
`
`e
`
`d th Mi
`
`an
`
`strongly.!,~=% =1
`
`LAM Exh 1002-pg 7
`
`

`
`diffusive) pressures, e magm u
`vana~on along ran tha co
`"t de of the ion transport velocity u; much exceeds the ion thermal velocity
`eve:, ~e usua
`I ~ton sine variation along z, does not describe the profile well, because at these low ~ut
`At higher pre~s:r~s ~r ~~ution for a constant diffusion coefficient (97), consisting of a J o Bessel funcuon
`er =tors such that A; ~< R, L, transport is diffusive and ambipolar. H~w­
`n,ln_o = 0.50 for R >>
`bulk plasma. This profi~ ~lrelau;::metry) and n !no = 0.40 for L >> R (infinite cylinder geometry).
`A;·>> R, L, e ~on trans~ I
`. 1 flat near the plasma center and dips near the sheath edge, with
`pressure and :e ~alues or rt~s coilisionless and well described by an ion free-fall profile (96) within ~e
`
`plex, because the ambt~larR ;rn Assuming uniform ionization at very low pressures or small reactors,
`
`of ions and electrons spans the regime A; -R • L, depending on e
`between the density n.s at the sheath edge and the density n0 at the plasma cen~r IS co:-
`
`tran
`
`.
`
`e re a on
`
`1 ti
`
`Th
`
`density source, without substantially increasm~ the ton m ar ng
`sources.
`I
`netic field IS applied to mhib!t radial parti
`typ"cal RF diode to 2 kW or more for a high
`ing_radius ~and ~ength_L ~be ~ffecb~~!~ los:c This reduction in A•ff can be important for certain ECR
`V, this results in a five-fold mcrease m ~·· S=d,ed ~= ~2 + 21tRL to 27tR2 if a strong axial mag-
`example, E; -
`for a cylindrical unmagneuzed source hav(cid:173)
`mer
`-40
`easing p
`·
`mass and for the typical limited ~ge of Tused.:s. HF~~: ~· r:l:~ by ~cing sheathL~oltages fro~. for
`the effecuve area or partie e oss.
`reducing A If• or
`where p ~IS the pofi wer a~~ by Th: ~ohm ~io~ loss) velocity us is relatively constant for a given ion
`lasm n is the ion density at the plasma-sheath edge, and A •If is
`
`can be mcreased from, say •
`
`tr • abs
`
`Th. d p
`
`~-360 V (fi RF diodes) to= 40 V (for high density sources). For Ar, with ~c + 2T • -
`
`All three strategies are
`
`or .
`
`·
`
`be understood from the overall source power balance, written in terms of EL as
`
`The principle of high density source operatiOn, sue as
`
`h" h ·
`
`·
`
`vm). Summing the three contnbutions yields the total energy lost per ton lost from the system.
`energy lost per ion lost IS E;. w IC
`•
`·
`For Maxwellian electrons, the mean kinetic energy lost per electron. lost IS E~th= 2hT •• th (Seecmr:.A.2 and
`kinetic
`kinetic energy to the walls (Sec. Vlii).
`
`used. The elastic cross section is from the data m Hayashi (95).
`(93), with an excitation energy of 11.97 V. For_toruzauon_, e cross sec 0
`alone, is shown for Arm Ftg. 5. For e excttab~n. .
`'th
`ti n from PeterSOn and Allen (94) is
`. used from Eggarter
`to be made. The quantity Ec• which wtthm the framework o
`many excited states and mult-step .~o~?n pathways must f ~~n~um tions above, is a function of T •
`Sllllple, rapt means or comparm .
`· dered if quantitative comparisons are
`• a e
`·
`tation channels into one effective level charactenzed by energy Euceral·
`di etai"ledts energy balance including
`:::!~:r :e :a::Or ~:on the left hand side ~f (2.1). Note that for =-rn:~ =~= ~~t a:rf:1~
`·
`.
`pair is subsequently lost In a monatomic gas such as Ar, the dbeombalanmantcedlossby the rate of formation which
`peratures, it is important for T • < 2 V, where it is dominanL The _energy Ec :~::e~ ~:: ;::o~
`lion and elastic collisions, respectively. While the last term may appear neg ~
`w ere u• QC
`li "ble for large electron tern-
`h
`and E = (3m/ M) T are the energies lost by an electron as a result of ionization, excita-
`
`·
`
`·
`
`· gone source to another. In gen
`
`fi
`
`"d
`
`d this loss rate in steady-state must
`
`all
`
`th
`
`K;:Ec = K;:E;: + KucEuc + K.,(2m!M)(3T.I2),
`
`energy Ec lost per electron-ion pair created in the system. For single-step, electron-Impact mmzauon,
`
`For the sake of comparing the efficiency of one source Wl ano er, I
`
`•'
`
`. E
`
`E
`
`In addition to collisional energy losses, electrons an
`
`.
`
`)
`
`1
`
`I
`
`mean free path is
`(CJ; = w-IB m2) and relatively constant for the ion energies of interest The corresponding ion-neutral
`scattering. The combined ionic momentum transfer cross section CJ; for these two processes is large
`In argon, the cross section for resonant charge transfer of Ar+ on Ar somewhat exceeds that for elastic
`
`where K (m3/s) is the rate constant and N (m -3) is the neutral Ar concentration.
`
`v = KN
`
`tic scatt