`
`[19]
`United States Patent
`6,139,699
`[11] Patent Number:
`Chiang et al. *Oct. 31, 2000 [45] Date of Patent:
`
`
`
`
`
`
`
`OTHER PUBLICATIONS
`
`S.M. Rossnagel and J. Hopwood, “Metal ion deposition
`from ionized mangetron sputtering discharge” J. Vac. Sci.
`Technol. B, vol. 12, No. 1, Jan/Feb. 1994, pp. 449—453.
`S.M. Rossnagel, et al. “Thin, high atomic weight refractory
`film deposition for diffusion barrier, adhesion layer, and seed
`layer applications” J. Vac. Sci. Technol. B, 14(3), May/Jun.
`1996.
`
`Kyung—Hoon Min et al. “Comparative study of tantalum and
`tantalum nitrides (Ta2N and TaN) as a diffusion barrier for
`Cu metalization” J. Vac. Sci Technol. B 14(5), Sep./Oct.
`1996.
`
`Primary Examiner—Nam Nguyen
`Assistant Examiner—Gregg Cantelmo
`Attorney, Agent, or Firm—Shirley L. Church
`
`[57]
`
`ABSTRACT
`
`The present disclosure pertains to our discovery that residual
`stress residing in a tantalum film or tantalum nitride film can
`be controlled (tuned) by controlling particular process vari-
`ables during film deposition. By tuning individual film
`stresses within a film stack, it is possible to balance stresses
`within the stack. Process variables of particular interest
`include: power to the sputtering target; process chamber
`pressure (i.e., the concentration of various gases and ions
`present in the chamber); substrate DC offset bias voltage
`(typically an increase in the AC applied substrate bias
`power); power to an ionization source (typically a coil); and
`temperature of the substrate upon which the film is depos-
`ited. The process chamber pressure and the substrate offset
`bias most significantly affect the film tensile and compres-
`sive stress components, respectively. The most advanta-
`geous tuning of a sputtered film is achieved using high
`density plasma sputter deposition, which provides for par-
`ticular control over the ion bombardment of the depositing
`film surface. When the tantalum or tantalum nitride film is
`deposited using high density plasma sputtering, power to the
`ionization source can be varied for stress tuning of the film.
`We have been able to reduce the residual stress in tantalum
`or
`tantalum nitride films deposited using high density
`plasma sputtering to between about 6><10+9 dynes/cm2 and
`about —6><10+9 dynes/cm2 using techniques described
`herein. The tantalum and tantalum nitride films can also be
`tuned following deposition using ion bombardment of the
`film surface and annealing of the deposited film.
`
`14 Claims, 4 Drawing Sheets
`
`[54] SPUTTERING METHODS FOR DEPOSITING
`STRESS TUNABLE TANTALUM AND
`TANTALUM NITRIDE FILMS
`
`[75]
`
`Inventors: Tony Chiang, Mountain View; Peijun
`Ding, San Jose; Barry L. Chin,
`Saratoga, all of Calif.
`
`[73] Assignee: Applied Materials, Inc., Santa Clara,
`Calif.
`
`[*] Notice:
`
`This patent issued on a continued pros-
`ecution application filed under 37 CFR
`1.53(d), and is subject to the twenty year
`patent
`term provisions of 35 U.S.C.
`154(a)(2).
`
`[21] Appl. No.: 08/863,451
`
`[22]
`
`Filed:
`
`May 27, 1997
`
`Int. Cl.7 ................................................... .. C23C 14/34
`[51]
`[52] US. Cl.
`.............................. .. 204/192.15; 204/192.12;
`204/192.13; 204/192.22
`[58] Field of Search ....................... .. 204/192.12, 192.13,
`204/192.15, 298.12, 192.22, 298.13; 347/203;
`438/685
`
`[56]
`
`References Cited
`U.S. PATENT DOCUMENTS
`
`9/1971 Banks .................................... .. 117/215
`3,607,384
`10/1973 Cordes . . . . . . .
`. . . . .. 204/192
`3,763,026
`12/1976 Kumagai
`.... .. 204/192
`4,000,055
`7/1977 Feit et al.
`204/38 A
`4,036,708
`7/1987 Melton et al.
`. 204/192.15
`4,683,043
`2/1993 Tepman et al.
`29/25.01
`5,186,718
`8/1993 Nulman . . . . . . . . . .
`. . . . .. 437/190
`5,236,868
`11/1995 Mogab et al.
`.
`........ .. 430/5
`5,464,711
`2/1996 Suzuki et al.
`.... .. 347/203
`5,491,505
`..
`7/1996 Terakado et al.
`. 204/192.15
`5,540,820
`7/1997 Harada et al.
`. . . . .
`. . . . .. 437/210
`5,643,834
`11/1998 Cabral, Jr. et al.
`................... .. 438/685
`5,834,374
`FOREIGN PATENT DOCUMENTS
`
`
`
`........ .. G03F 1/00
`0346828 12/1989 European Pat. Off.
`........ .. G11B 5/60
`0644535
`3/1995 European Pat. Off.
`WO 9704143
`2/1997 WIPO .......................... .. C23C 16/30
`
`IMP-Ta Stress
`
`100
`
`
`
`Page 1 of 10
`
`TSMC V. IP Bridge
`
`IP Bridge Exhibit 2041
`
`|PR2016-01264
`
`
`
`Page 1 of 10
`
`
`
`US. Patent
`
`Oct. 31, 2000
`
`Sheet 1 0f4
`
`6,139,699
`
`IMP-Ta: Stress (dyne/cmz)
`
`IMP-Ta: Stress (dyne/cmz)
`
`200
`
`227
`W 70
`
`207
`70
`
`60
`
`50
`
`4o
`
`30
`
`20
`
`10
`
`_1
`
`'
`1.5
`
`.5
`
`1.0
`
`,:
`
`206
`
`2.0
`
`2.5
`
`3.0
`
`3.5
`
`Pressure
`
`(m7)
`
`208
`
`Pressure
`
`(M)
`
`DC(kW)
`FIG. 2A
`
`DC(kW)
`
`228
`
`Page 2 of 10
`
`FIG. 28
`
`
`
`
`
`60
`
`50
`
`40
`
`30
`
`20
`
`1o
`
`.5
`
`1.0
`
`1.5
`
`2.0
`
`2.5
`
`3.0
`
`3.5
`
`Page 2 of 10
`
`
`
`US. Patent
`
`Oct. 31, 2000
`
`Sheet 2 0f4
`
`6,139,699
`
`IMP-Ta Stress (dyne/cmz)
`Bias (W)
`
`/300
`
`;
`
`320
`
`0
`
`100
`
`200
`
`300
`
`400
`
`FIG. 3
`
`
`
`-1.4x1010
`
`Sfress
`(Dyne/cm 2)
`
`Page 3 of 10
`
`Page 3 of 10
`
`
`
`US. Patent
`
`Oct. 31, 2000
`
`Sheet 3 0f4
`
`6,139,699
`
`/400
`Gamma TaNX Composition
`70
`
`
`414 500
`450
`
` 413
`
`4O
`
`NHrogen
`30 Content
`20
`(A)
`‘0
`
`516
`
`400
`
`350
`
`300
`250
`200
`150
`1505)
`
`0
`
`p
`(MO-cm)
`
`
`
`O
`
`20
`
`
`
`10
`12
`14
`18
`N2 Flow (sccm)
`
`
`
`
`416
`
`FIG. 4
`
`IMP-TaNx Composition /
`
`500
`
`
`
`(ttfl’cm)
`
`N2 How @ccm)
`
`513
`
`70
`
`60
`
`50
`
`40
`
`FHW
`3O NHrogen
`Content
`no
`
`20
`
`10
`
`0
`
`Page 4 of 10
`
`Page 4 of 10
`
`
`
`US. Patent
`
`Oct. 31, 2000
`
`Sheet 4 0f4
`
`6,139,699
`
`Gamma TaNx Film Stress
`
`600
`
`Stress
`
`0.0
`
`502
`—1x10‘°
`
`—2x10‘0-
`
`—5x1010
`
`-6x10'0
`
`0
`6
`
`702
`
`Siress
`(dyne/cmz)
`
`(dyne/cmz) _3x I 010 -4x10'°
`
`
`Flow (sccm)
`
`IO
`
`20
`
`30
`N2 Flow (sccm)
`
`lMP-TaNx Film Stress
`
`40
`
`50
`
`50
`
`604
`
`700
`
`/ 708
`
`5
`
`10
`
`20
`
`FIGI 7
`
`Page 5 of 10
`
`Page 5 of 10
`
`
`
`
`
`6,139,699
`
`1
`SPUTTERING METHODS FOR DEPOSITING
`STRESS TUNABLE TANTALUM AND
`TANTALUM NITRIDE FILMS
`BACKGROUND OF THE INVENTION
`1. Field of the Invention
`
`The present invention pertains to tantalum and tantalum
`nitride films which can be stress tuned to be in tension or in
`compression or to have a particularly low stress, and to a
`method of producing such films. These stress tuned films are
`particularly useful in semiconductor interconnect structures
`where they can be used to balance the stress within a stack
`of layers which includes a combination of barrier layers,
`wetting layers, and conductive layers, for example. The low
`stress tantalum and tantalum nitride films are particularly
`suited for the lining of vias and trenches having a high aspect
`ratio.
`
`2. Brief Description of the Background Art
`A typical process for producing a multilevel structure
`having feature sizes in the range of 0.5 micron
`or less
`would include: blanket deposition of a dielectric material;
`patterning of the dielectric material to form openings; depo-
`sition of a diffusion barrier layer and, optionally, a wetting
`layer to line the openings; deposition of a conductive
`material onto the substrate in sufficient thickness to fill the
`openings; and removal of excessive conductive material
`from the substrate surface using a chemical, mechanical, or
`combined chemical-mechanical polishing techniques.
`Future technological requirements have placed a focus on
`the replacement of aluminum (and aluminum alloys) by
`copper as the conductive material. As a result, there is an
`increased interest in tantalum nitride barrier layers and in
`tantalum barrier/wetting layers which are preferred for use
`in combination with copper.
`Tantalum nitride barrier films, Ta2N and TaN, have been
`shown to function up to 700° C. and 750° C., respectively,
`without the diffusion of copper into an underlying silicon
`(Si) substrate. Tantalum barrier/wetting films have been
`shown to function at temperatures of approximately 500° C.
`It
`is advantageous in terms of processing simplicity to
`sputter the barrier and or wetting layers underlaying the
`copper. Tantalum nitride barrier layers are most commonly
`prepared using reactive physical sputtering, typically with
`magnetron cathodes, where the sputtering target is tantalum
`and nitrogen is introduced into the reaction chamber.
`S. M. Rossnagel and J. Hopwood describe a technique
`which enables control of the degree of directionality in the
`deposition of diffusion barriers in their paper titled “Thin,
`high atomic weight refractory film deposition for diffusion
`barrier, adhesion layer, and seed layer applications” J. Vac.
`Sci. Technol. B 14(3), May/June 1996. In particular,
`the
`paper describes a method of depositing tantalum (Ta) which
`permits the deposition of the tantalum atoms on steep
`sidewalls of interconnect vias and trenches. The method uses
`
`conventional, non-collimated magnetron sputtering at low
`pressures, with improved directionality of the depositing
`atoms. The improved directionality is achieved by increas-
`ing the distance between the cathode and the workpiece
`surface (the throw) and by reducing the argon pressure
`during sputtering. For a film deposited with commercial
`cathodes (Applied Materials Endura® class; circular planar
`cathode with a diameter of 30 cm) and rotating magnet
`defined erosion paths, a throw distance of 25 cm is said to
`be approximately equal to an interposed collimator of aspect
`ratio near 1.0. In the present disclosure, use of this “long
`throw” technique with traditional, non-collimated magne-
`tron sputtering at low pressures is referred to as “Gamma
`sputtering”.
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`2
`Gamma sputtering enables the deposition of thin, confor-
`mal coatings on sidewalls of a trench having an aspect ratio
`of 2.821 for 0.5 ym-wide trench features. However, we have
`determined that Gamma sputtered TaN films exhibit a rela-
`tively high film residual compressive stress, in the range of
`about —1.0><10+10 to about —5 .0><10+10 dynes/cm2. High film
`residual compressive stress, in the range described above
`can cause a Ta film or a tantalum nitride (e.g. Ta2N or TaN)
`film to peel off from the underlying substrate (typically
`silicon oxide dielectric). In the alternative, the film stress can
`cause feature distortion on the substrate (typically a silicon
`wafer) surface or even deformation of a thin wafer.
`A method of reducing the residual stress in a Ta barrier/
`wetting film or a Ta2N or TaN barrier film would be
`beneficial in enabling the execution of subsequent process
`steps without delamination of such films from trench and via
`sidewalls or other interconnect features. This reduces the
`
`number of particles generated, increasing device yield dur-
`ing production. In addition, a film having a near zero stress
`condition improves the reliability of the device itself.
`SUMMARY OF THE INVENTION
`
`We have discovered that the residual stress residing in a
`tantalum (Ta)
`film or a tantalum nitride (TaNx, where
`0<><§1.5) film can be controlled (tuned) by controlling
`particular process variables during deposition of the film.
`Process variables of particular interest for sputter applied Ta
`and TaNx films include the following. An increase in the
`power to the sputtering target (typically DC) increases the
`compressive stress component in the film. An increase in the
`process chamber pressure (i.e. the concentration of various
`gases and ions present in the chamber) increases the tensile
`stress component in the film. An increase in the substrate DC
`offset bias voltage (typically an increase in the applied AC
`as substrate bias power) increases the compressive stress
`component in the film. The substrate temperature during
`deposition of the film also affects the film residual stress. Of
`these variables, an increase in the process chamber pressure
`and an increase in the substrate offset bias most significantly
`affect
`the tensile and compressive stress components,
`respectively. The most advantageous tuning of a sputtered
`film is achieved using Ion Metal Plasma (IMP) as the film
`deposition method. This sputtering method provides for
`particular control over the ion bombardment of the depos-
`iting film surface. When it is desired to produce a film
`having minimal residual stress, particular care must be taken
`to control the amount of ion bombardment of the depositing
`film surface, as an excess of such ion bombardment can
`result
`in an increase in the residual compressive stress
`component in the deposited film.
`Tantalum (Ta) films deposited using the IMP method
`typically exhibit a residual stress ranging from about +1><10+
`10 dynes/cm2 (tensile stress) to about —2><10+ dynes/cm2
`(compressive stress), depending on the process variables
`described above. Tantalum nitride (TaNx) films deposited
`using the IMP method typically can be tuned to exhibit a
`residual stress within the same range as that specified above
`with reference to Ta films. We have been able to reduce the
`
`residual stress in either the Ta or TaNx films to low values
`ranging from about +1><10+9 dynes/cm2 to about —2><10+9
`dynes/cm2 using tuning techniques described herein. These
`film residual stress values are significantly less than those
`observed for traditionally sputtered films and for Gamma-
`sputtered films. This reduction in film residual compressive
`stress is particularly attributed to bombardment of the film
`surface by IMP-generated ions during the film deposition
`process. Heavy bombardment of the film surface by IMP-
`
`Page 6 of 10
`
`
`
`Page 6 of 10
`
`
`
`
`
`6,139,699
`
`3
`generated ions can increase the film residual compressive
`stress, so when it is desired to minimize the film compres-
`sive stress, the ion bombardment should be optimized for
`this purpose.
`Other process variables which may be used in tuning the
`film stress include the spacing between the sputter target and
`the substrate surface to be sputter deposited; ion bombard-
`ment subsequent to film deposition; and annealing of the
`film during or after deposition.
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`FIG. 1 is a graph showing the residual stress in an IMP
`deposited Ta film as a function of DC power to the Ta target,
`RF power to the IMP ionization coil, and the pressure in the
`process chamber.
`FIG. 2A is a contour plot showing the IMP deposited Ta
`film residual stress in dynes/cm2 as a function of the DC
`power to the Ta target and the process chamber pressure,
`when the RF power to the ionization coil is 1 kW.
`FIG. 2B is a contour plot showing the residual stress in an
`IMP deposited Ta film as a function of the same variables
`illustrated in FIG. 2A, when the RF power to the ionization
`coil is 3 kW.
`
`FIG. 3 is a graph showing the residual stress in an IMP
`deposited Ta film as a function of the substrate offset bias,
`and in particular as a function of the AC bias power
`(typically the AC power is coupled to the substrate through
`the substrate heater which is in electrical contact with the
`
`substrate).
`FIG. 4 is a graph showing the chemical composition of a
`Gamma-sputtered tantalum nitride film, as a function of the
`nitrogen gas flow rate to the sputtering process chamber. In
`addition, FIG. 4 shows the resistivity and the structure of the
`tantalum nitride compound, which is in conformance with
`the nitrogen content of the compound.
`FIG. 5 is a graph showing the film composition of a
`reactive IMP-deposited tantalum nitride film, as a function
`of the nitrogen gas flow rate to the process chamber. Again,
`the resistivity of the film is indicative of the various film
`structures created as the nitrogen content of the film is
`increased.
`
`FIG. 6 is a graph showing the residual film stress for
`Gamma-sputtered tantalum nitride film as a function of the
`nitrogen gas flow rate to the sputtering process chamber and
`as a function of the temperature at which the film is
`deposited.
`FIG. 7 is a graph showing the residual film stress for
`reactive IMP sputtered tantalum nitride film as a function of
`the nitrogen gas flow rate to the sputtering process chamber.
`DETAILED DESCRIPTION OF THE
`PREFERRED EMBODIMENTS
`
`The present invention pertains to stress tunable tantalum
`and tantalum nitride films and to a method of producing such
`films. In particular, applicants have discovered that residual
`film stress can be tuned by controlling particular process
`variables such as process chamber pressure, DC offset bias
`voltage, power to the sputtering target and substrate tem-
`perature during film deposition. When IMP sputtering is
`used, a variation in the power to the ionization coil can be
`used for tuning. Ion bombardment of the depositing film
`surface is particularly useful in controlling residual film
`stress.
`
`I. Definitions
`
`As a preface to the detailed description, it should be noted
`that, as used in this specification and the appended claims,
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`a :a
`a , “an”, and “the” include plural
`the singular forms
`referents, unless the context clearly dictates otherwise. Thus,
`for example, the term “a semiconductor” includes a variety
`of different materials which are known to have the behav-
`ioral characteristics of a semiconductor, reference to a
`“plasma” includes a gas or gas reactants activated by an RF
`glow discharge, and reference to “copper” includes alloys
`thereof.
`
`Film stress values were measured using a Tencor® Flexus
`FLX 3200 machine available from Tencor Corporation,
`Mountain View, Calif.
`Specific terminology of particular importance to the
`description of the present invention is defined below.
`The term “aspect ratio” refers to the ratio of the height
`dimension to the width dimension of particular openings
`into which an electrical contact is to be placed. For example,
`a via opening which typically extends in a tubular form
`through multiple layers has a height and a diameter, and the
`aspect ratio would be the height of the tubular divided by the
`diameter. The aspect ratio of a trench would be the height of
`the trench divided by the minimal travel width of the trench
`at its base.
`
`The term “completely filled” refers to the characteristic af
`a feature such as a trench or via which is filled with a
`
`conductive material, wherein there is essentially no void
`space present within the portion of the feature filled with
`conductive material.
`
`The term “copper” refers to copper and alloys thereof,
`wherein the copper content of the alloy is at least 80 atomic
`% copper. The alloy may comprise more than two elemental
`components.
`The term “feature” refers to contacts, vias, trenches, and
`other structures which make up the topography of the
`substrate surface.
`
`The term “Gamma or (y) sputtered copper” refers to the
`“long throw” sputtering technique described in the paper by
`S. M. Rossnagel and J. Hopwood, which was discussed
`previously herein. Typically the distance between the sub-
`strate and the target is about the diameter of the substrate or
`greater; and, preferably, the process gas pressure is suffi-
`ciently low that the mean free path for collision within the
`process gas is greater than the distance between the target
`and the substrate.
`
`The term “ion metal plasma” or “IMP” refers to sputter
`deposition, preferably magnetron sputter deposition (where
`a magnet array is placed behind the target). A high density,
`inductively coupled RF plasma is positioned between the
`sputtering cathode and the substrate support electrode,
`whereby at least a portion of the sputtered emission is in the
`form of ions at the time it reaches the substrate surface.
`
`The term “IMP sputtered tantalum” refers to tantalum
`which was sputtered using the IMP sputter deposition
`method.
`
`The term “IMP sputtered tantalum nitride” refers to
`tantalum nitride which was sputtered using the IMP sputter
`deposition method.
`The term “reactive IMP sputtered tantalum nitride” refers
`to ion-deposition sputtering wherein nitrogen gas is supplied
`during the sputtering of tantalum, to react with the ionized
`tantalum, producing an ion-deposition sputtered tantalum
`nitride-comprising compound.
`The term “stress tuned” refers to a TaNx or Ta film which
`has been treated during processing to adjust the residual
`stress within the deposited film to fall within a particular
`desired range. For example, at times it is desired to use the
`
`Page 7 of 10
`
`
`
`Page 7 of 10
`
`
`
`
`
`6,139,699
`
`5
`TaNx or Ta film to balance the overall stress within a stack
`of layers, so the film may be tuned to be in compression or
`tension. At other times it may be desired to reduce the stress
`in the film to be as near to zero as possible.
`The term “traditional sputtering” refers to a method of
`forming a film layer on a substrate wherein a target is
`sputtered and the material sputtered from the target passes
`between the target and the substrate to form a film layer on
`the substrate, and no means is provided to ionize a substan-
`tial portion of the target material sputtered from the target
`before it reaches the substrate. One apparatus configured to
`provide traditional sputtering is disclosed in US. Pat. No.
`5,320,728, the disclosure of which is incorporated herein by
`reference. In such a traditional sputtering configuration, the
`percentage of target material which is ionized is less than
`10%, more typically less than 1%, of that sputtered from the
`target.
`
`II. An Apparatus For Practicing the Invention
`
`A process system in which the method of the present
`invention may be carried out is the Applied Materials, Inc.
`(Santa Clara, Calif.) Endura® Integrated Processing System.
`The system is shown and described in US. Pat. No. 5,186,
`718,
`the disclosure of which is hereby incorporated by
`reference.
`
`The traditional sputtering process is well known in the art.
`The Gamma sputtering method is described in detail by S.
`M. Rossnagel and J. Hopwood in their paper titled “Thin,
`high atomic weight refractory film deposition for diffusion
`barrier, adhesion layer, and seed layer applications”, as
`referenced above. The IMP sputtering method is also
`described by S. M. Rossnagel and J. Hopwood in their paper
`“Metal ion deposition from ionized magnetron sputtering
`discharge, J. Vac. Sci. Technol. B, Vol. 12, No. 1 (January/
`February 1994).
`III. The Structure of the Tantalum and Tantalum
`Nitride Films
`
`We have been able to create a copper filled trench or via,
`which is completely filled, at a feature size of about 0.4 M and
`an aspect ratio of greater than 1:1 (up to about 3:1 presently).
`To facilitate the use of a copper fill,
`the trench or via
`(constructed in a silicon oxide surface layer) was lined with
`a reactive IMP sputtered TaNx barrier layer, followed by a Ta
`barrier/wetting layer,
`to create a bilayer over the oxide
`surface layer. The copper fill layer was applied using a
`sputtering technique in the manner described in applicants’
`co-pending US. application Ser. No. 08/855,059, filed May
`13, 1997, pending.
`To ensure the overall dimensional stability of the
`structure, we investigated various factors which affect the
`residual film stress in a TaNx barrier layer and in a Ta layer
`(which can serve as a barrier layer, a wetting layer, or both,
`depending on the application).
`One skilled in the art can envision a combination of a
`
`number of different layers underlaying the copper fill mate-
`rial. Whatever the combination of layers, they provide a
`stack of layers; and tuning the stress of individual layers
`within the stack can provide a more stress balanced and
`dimensionally stable stack. Although the preferred embodi-
`ment described above is for the lining of trenches and vias,
`one skilled in the art will appreciate that the stress tuned
`TaNx and Ta films described herein have general application
`in semiconductor interconnect structures. The method of
`
`controlling and reducing the residual film stress in tantalum
`nitride and tantalum films can be used to advantage in any
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`6
`structure in which a layer of such a film is present. The
`concept of tuning the residual stress in a sputter-deposited
`film comprising at least one metal element has broad appli-
`cability.
`
`IV. The Method of Tuning Residual Stress in
`Tantalum and Tantalum Nitride Films
`
`The preferred embodiments described herein were pro-
`duced in an Endura® Integrated Processing System avail-
`able from Applied Materials of Santa Clara, Calif. The
`physical vapor deposition (sputtering in this case) process
`chamber is capable of processing an 8 inch (200 mm)
`diameter silicon wafer. The substrate was a silicon wafer
`
`having a silicon oxide surface coating with trenches in the
`surface of the silicon oxide. Sputtering was carried out using
`a tantalum target cathode having approximately a 35.3 cm
`(14 in.) diameter, and DC power was applied to this cathode
`over a range from about 1 kW to about 18 kW. The substrate
`was placed at a distance of about 25 cm (9.8 in.) from the
`tantalum target cathode in the case of gamma sputtering, and
`at a distance of about 14 cm (5.5 in.) from the cathode in the
`case of IMP sputtering. During IMP sputtering, an AC bias
`power ranging from about 0 W to about 400 W was applied
`to the substrate, to produce a substrate offset bias ranging
`from about 0 V to about —100 V. The substrate offset bias
`
`attracts ions from the plasma to the substrate.
`
`Example One
`
`When Gamma-sputtered tantalum film was produced, the
`film was sputtered using conventional (traditional) magne-
`tron sputtering, with rotating magnet-defined erosion paths
`(for better uniformity and cathode utilization). Two hundred
`(200) mm sample surfaces were sputter-deposited at a
`sample surface temperature of about 25° C., in argon, at
`pressures of about 1.5 mT or less. The cathode to sample or
`“throw” distance was typically about 25 cm. The DC power
`to the tantalum target was approximately 4 kW. No substrate
`offset bias was used. Under these conditions, the residual
`film stress of the tantalum film was about —1.5><10+10
`dynes/cm2.
`
`Example Two
`
`When IMP-sputtered tantalum film was produced, a high
`density, inductively coupled RF plasma was generated in the
`region between the target cathode and the substrate by
`applying RF power to a coil (having from 1 to 3 turns) over
`a range from about 400 kHz to about 13.56 MHz (preferably
`about 2 MHz). Two hundred (200) mm sample surfaces were
`IMP sputter-deposited at a sample surface temperature of
`about 25° C., in argon, at pressures ranging from about 10
`mT to about 60 mT. The distance from the cathode to the
`
`sample was typically about 14 cm. The DC power to the
`tantalum target was adjusted over a range from about 1 kW
`to about 8 kW (preferably about 1 kW to about 3 kW). The
`wattage to the RF power coil was adjusted over a range from
`about 1.0 kW to about 5 kW (preferably about 1.0 kW to
`about 3 kW). An AC bias power ranging from about 0 W to
`about 500 W was used. FIG. 1 shows a graph 100 of the
`residual film stress 101 of the tantalum film in dynes/cm2, as
`a function of the RF power 108 to the ionization coil, as
`illustrated by the curve numbered 102; the pressure 110 in
`the sputtering chamber, as illustrated by the curve numbered
`104; and the DC power 112 to the sputtering target
`(cathode), as illustrated by the curve numbered 106.
`As indicated in graph 100,
`the residual stress in the
`deposited Ta film can be tuned over a wide range, for
`
`Page 8 of 10
`
`
`
`Page 8 of 10
`
`
`
`
`
`6,139,699
`
`7
`
`example (but not by way of limitation), from about 1.0><10+
`10 dynes/cm2 to about —2><10+10 dynes/cm2, and can be set
`at a low stress nominal value, for example, between about
`6><10+9 dynes/cm2 and about —6><10+9 dynes/cm2, a range
`over which the residual stress can approach zero. At a
`residual stress of about —6><10+9 dynes/cm2, by way of
`example, the IMP sputtered film residual compressive stress
`is a factor of three lower than the residual compressive stress
`of a typical gamma-sputtered Ta film. The process variables
`which affect film residual stress can be optimized to produce
`the desired residual film stress in Ta films.
`FIGS. 2A and 2B show the effect of an increase in the RF
`
`power to the IMP ionization coil, which is directly related to
`the amount of ion bombardment at the tantalum film surface.
`
`FIG. 2A, graph 200, shows the Ta residual film stress in
`curves 201 through 206, when the power to the ionization
`coil is 1 kW, as a function of process chamber argon pressure
`207 and the DC power to the tantalum target 208. FIG. 2B,
`graph 220, shows the Ta residual film stress interior of
`ellipses 221 and 222, when the power to the ionization coil
`is 3 kW, as a function of process chamber argon pressure 227
`and the DC power to the tantalum target 228.
`These curves show that, with the other process values held
`constant, an increase in RF power to the ionization coil from
`1 kW to 3 kW results in an increase in the film residual
`
`compressive stress. Even so, under all of the process con-
`ditions shown, the residual film stress for the IMP-sputtered
`tantalum is less than that of a Gamma-sputtered tantalum
`film. We have concluded, then, that there is an optimum
`amount of ion bombardment of a tantalum film surface to
`
`produce a Ta film having only minor residual stress (whether
`compressive or in tension). Process pressure appears to have
`the greatest effect of the variables tested. It is believed that
`an increase in the process pressure leads to an increase in
`ionization within the process chamber, which leads to
`increased ion bombardment of the depositing film surface.
`
`Example Three
`The effect of the increase in ion bombardment of a
`
`depositing film surface, which can be achieved by increasing
`the DC offset bias voltage of the substrate onto which the
`film is deposited, is illustrated in FIG. 3. Graph 300 shows
`the residual stress 311 in dynes/cm2 310 as a function of the
`AC bias power 320 in Watts. The corresponding substrate
`DC offset bias voltage ranges from about 0 V to about —150
`V.
`
`Example Four
`
`When tantalum nitride films are produced, the structure of
`the tantalum nitride depends on the amount of nitrogen in the
`tantalum nitride compound
`FIGS. 4 and 5 show the
`chemical composition and resistivity of tantalum nitride
`films produced using Gamma sputtering and IMP sputtering
`techniques, respectively. The chemical composition (atomic
`nitrogen content) of the film is shown as a function of the
`nitrogen gas flow rate to the process chamber in which the
`TaNx film is produced.
`FIG. 4, graph 400, shows the nitrogen content 410 of the
`Gamma-sputtered tantalum nitride film in atomic % 413, as
`a function of the nitrogen flow rate 416 in sccm to the
`process vessel. A two hundred (200) mm diameter sample
`surface was Gamma sputter-deposited at a sample surface
`temperature of about 25° C.,
`in an argon/nitrogen
`atmosphere, at a pressure of about 1.5 mT, where the Argon
`gas feed was about 15 sccm and the nitrogen flow rate 416
`was as shown on graph 400. The “throw” distance between
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`8
`the tantalum target and the sample surface was approxi-
`mately 250 mm. The DC power to the tantalum target was
`about 4 kW.
`
`In addition, graph 400 shows the resistivity 412 in M Q-cm
`414 of the tantalum nitride film as the nitrogen content 413
`increases. The resistivity corresponds with the change in the
`tantalum nitride structure, as indicated on Graph 400, where
`402 represents B-Ta; 404 represents bcc—Ta(N); 406 rep-
`resents amorphous TaNx; and 408 represents nanocrystalline
`fcc—TaNx (xz1).
`FIG. 4 shows that when the atomic nitrogen content
`exceeds about 45% to about 50%, the resistivity of the TaNx
`film increases drastically (to above 1,000 M Q-cm).
`FIG. 6, graph 600, shows the residual film stress in
`dynes/cm2 602 of a Gamma sputtered TaNx film, as a
`function of the nitrogen flow rate to the process chamber in
`sccm 604, and as a function of the substrate temperature at
`the time of film deposition, when the other process variables
`are held at the values described with reference to FIG. 4.
`
`Curve 610 represents the TaNx film Gamma sputtered at
`a substrate temperature of about 25° C.; Curve 612 repre-
`sents the TaNx film Gamma sputtered at a substrate tem-
`perature of about 250° C., and Curve 614 represents the
`TaNx film Gamma sputtered at a substrate temperature of
`about 450° C.
`
`Line 606 constructed at a nitrogen flow rate 604 of about
`16 sccm, represents the atomic nitrogen content in excess of
`which the resistivity of the TaNx film increases drastically
`(as illustrated in FIG. 4 for a nitrogen flow rate of 16 sccm).
`Thus,
`the gamma-sputtered TaNx films having reduced
`residual compressive stress (in the direction of arrow 608)
`occur at nitrogen contents at which the resistivity of the film
`is unacceptably high (greater than about 1,000 M Q-cm).
`Looking at the residual film stress of TaNx films having a
`resistivity lower than about 1,000 M Q-cm, it is evident that
`residual film stress can be reduced by increasing the sub-
`strate temperature at the time of film deposition. This is in
`contrast with TaNx films having a resistivity higher than
`about 1,000 M Q-cm, where the residual film stress increases
`when the substrate temperature is higher during film depo-
`sition. Considering this unexpected