U.S. patent application number 10/898277 was filed with the patent office on 2005-01-06 for method, chemistry, and apparatus for noble metal electroplating on a microelectronic workpiece.
This patent application is currently assigned to Semitool, Inc.. Invention is credited to Graham, Lyndon W., Jacobson, Curt W., Ritzdorf, Thomas L..
Application Number | 20050000818 10/898277 |
Document ID | / |
Family ID | 23703290 |
Filed Date | 2005-01-06 |
United States Patent
Application |
20050000818 |
Kind Code |
A1 |
Graham, Lyndon W. ; et
al. |
January 6, 2005 |
Method, chemistry, and apparatus for noble metal electroplating on
a microelectronic workpiece
Abstract
The present invention is directed to an improved electroplating
method, chemistry, and production worthy apparatus for depositing
noble metals (e.g., platinum) and their alloys onto the surface of
the workpiece, such as a semiconductor wafer, pursuant to
manufacturing a microelectronic device, circuit, and/or component.
The reliability of the noble metal material deposited using the
disclosed method, chemistry, and/or apparatus is significantly
better than the reliability of noble metal structures deposited
using the teachings of the prior art. This is largely attributable
to the low stress of films that are deposited using the teachings
disclosed herein. The metals, which can be deposited, include gold,
silver, platinum, palladium, ruthenium, iridium, rhodium, osmium
and alloys containing these metals.
Inventors: |
Graham, Lyndon W.;
(Kalispell, MT) ; Jacobson, Curt W.; (Kalispell,
MT) ; Ritzdorf, Thomas L.; (Bigfork, MT) |
Correspondence
Address: |
CHRISTENSEN, O'CONNOR, JOHNSON, KINDNESS, PLLC
1420 FIFTH AVENUE
SUITE 2800
SEATTLE
WA
98101-2347
US
|
Assignee: |
Semitool, Inc.
|
Family ID: |
23703290 |
Appl. No.: |
10/898277 |
Filed: |
July 23, 2004 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10898277 |
Jul 23, 2004 |
|
|
|
09429446 |
Oct 28, 1999 |
|
|
|
Current U.S.
Class: |
205/102 |
Current CPC
Class: |
C25D 17/001 20130101;
C25D 7/123 20130101 |
Class at
Publication: |
205/102 |
International
Class: |
C25D 005/18 |
Claims
1. In a method for electroplating a noble metal into submicron
features on a surface of a microelectronic workpiece, wherein the
surface of the workpiece to be plated is brought into contact with
an electroplating solution including ions and/or complexes of the
noble metal and an electroplating power is applied between the
surface of the workpiece and an anode spaced from the surface of
the workpiece and in contact with the electroplating solution, the
improvement comprising: reducing film stress within the
electroplated noble metal by applying electroplating power between
the surface of the workpiece and the anode using a low current for
a first predetermined period of time and applying higher current
electroplating power between the surface of the workpiece and the
anode for a second predetermined period of time, the noble metal
being deposited into the submicron features during the first and
second time periods.
2. The method of claim 1, wherein the low current is applied using
a pulsed waveform.
3. The method of claim 1, wherein the higher current electroplating
power has a current density between about 3 and 9 mA/cm.sup.2.
[ALKALINE PLATING CONDITIONS]
4. The method of claim 1, wherein the higher current electroplating
power has a current density between about 20 and 50 mA/cm.sup.2.
[ACIDIC PLATING CONDITIONS]
5. The method of claim 4, wherein the higher current electroplating
power is applied using a pulsed waveform.
6. The method of claim 5, wherein the pulsed waveform comprises an
on-time in a range of about 1-10 milliseconds and an off-time in a
range of about 1-10 milliseconds.
7. The method of claim 1, wherein the higher current electroplating
power is initiated immediately after the first predetermined period
of time.
Description
CROSS-REFERENCE TO RELATED APPLICATION
[0001] This application is a divisional of copending Application
Ser. No. 09/429,446. filed Oct. 28, 1999.
FIELD OF THE INVENTION
[0002] The present invention relates to a method of reducing stress
in electroplated noble metals.
BACKGROUND OF THE INVENTION
[0003] The present invention is directed to electroplating a
low-stress noble metal film onto the surface of a workpiece, such
as a semiconductor wafer, in the manufacture of microelectronic
devices and/or components. More particularly, the present invention
is directed to a method, chemistry and apparatus for electroplating
a noble metal, such as platinum, on a microelectronic
workpiece.
[0004] Production of semiconductor integrated circuits and other
microelectronic devices from workpieces, such as semiconductor
wafers, typically requires formation of one or more metal layers on
the workpiece. These metal layers are used, for example, to
electrically interconnect the various devices of the integrated
circuit. Further, the structures formed from the metal layers may
be elements of microelectronic devices such as read/write heads,
etc.
[0005] The microelectronic manufacturing industry has applied a
wide range of metals to form such structures. These metals include,
for example, nickel, tungsten, solder, and copper. Further, a wide
range of processing techniques have been used to deposit such
metals. These techniques include, for example, chemical vapor
deposition (CVD), physical vapor deposition (PVD), electroplating,
and electroless plating. Of these techniques, electroplating tends
to be the most economical and, as such, the most desirable.
Electroplating can be used in the deposition of blanket metal
layers as well as selectively deposited or patterned metal
layers.
[0006] One of the process sequences used in the microelectronic
manufacturing industry to form one or more metal structures on a
semiconductor wafer is referred to as "damascene" or "inlaid"
processing. In such processing, holes, commonly called "vias",
trenches and/or other microscopic-sized recesses are formed in a
workpiece surface and filled, either entirely or only partially,
with a metal. In the damascene process, the wafer is first provided
with a metallic seed layer which is used to conduct electrical
current during a subsequent metal electroplating step. When certain
metals that readily migrate into the surface of the wafer are used,
the seed layer is disposed over a barrier layer material, such as
Ti, TiN, Ta, TaN, etc.
[0007] The seed layer is a very thin layer of metal which can be
applied using one or more of several processes. For example, the
seed layer can be laid down using physical vapor deposition (PVD)
or chemical vapor deposition (CVD) processes to produce a layer on
the order of 100-1,000 angstroms thick. The seed layer can be
formed of copper, gold, nickel, palladium, platinum, or other
metals compatible with the subsequently applied metal. The seed
layer is formed over a surface which is convoluted by the presence
of the vias, trenches, or other recessed device features.
[0008] A metal layer may then be electroplated onto the seed layer.
The layer is plated to form an overlying layer, with the goal of
providing a metal layer that either entirely or partially fills the
trenches and vias.
[0009] After the blanket layer has been electroplated onto the
semiconductor wafer, excess metal material present outside of the
vias, trenches, or other recesses is removed. The excess plated
material can be removed, for example, using chemical mechanical
planarization, chemical etching, or plasma etching. Chemical
mechanical planarization is a processing step which uses the
combined action of a chemical removal agent and an abrasive which
grinds and polishes the exposed metal surface to remove undesired
parts of the metal layer applied in the electroplating step. The
metal is removed to provide a resulting pattern of metal layer in
the semiconductor integrated circuit being formed.
[0010] The electroplating of the semiconductor wafers takes place
in a reactor assembly. In such an assembly an anode electrode is
disposed in a plating bath, and the wafer with the seed layer
thereon is used as a cathode. Preferably, only a lower face of the
wafer contacts the surface of the plating bath. The wafer is held
by a support system that also conducts the requisite electroplating
power (e.g., cathode current) to the wafer. The support system may
comprise conductive fingers that secure the wafer in place and also
contact the wafer seed layer in order to conduct electrical current
for the plating operation. One embodiment of a reactor assembly is
disclosed in U.S. Ser. No. 08/988,333 filed Sept. 30, 1997 entitled
"Semiconductor Plating System Workpiece Support Having
Workpiece--Engaging Electrodes With Distal Contact Part and
Dielectric Cover".
[0011] An efficient process for electroplating of certain noble
metals is desirable in those microelectronic applications in which
nickel, copper, etc. are not the optimal metal. Such components
include, for example, sensors (electrochemical or
micro-mechanical), capacitor structures in memory cells, and some
interconnects for microelectronic devices. One application in which
electroplating of a noble metal onto a workpiece is particularly
useful is in the fabrication of platinum electrodes for capacitors
used, for example, in semiconductor memory devices. The work
function of platinum facilitates the formation of capacitor
electrodes that exhibit enhanced electrical characteristics,
including lower leakage currents and a higher breakdown voltage
when compared to electrodes of other metals. The low leakage
current minimizes the amount of charge lost between refresh cycles
of, for example, dynamic memory cells including such capacitors.
The higher breakdown voltage allows the capacitor to store a larger
charge without significant current leakage. Consequently capacitors
having smaller geometries are possible thereby allowing the
formation of a greater number of capacitors upon a workpiece of a
given size. Further benefits of platinum relate to the fact that it
has a low propensity to react with other materials or oxidize and,
as such, does not form an undesired oxide film at its surface when
it is exposed to the ambient environment. This can be important
where processing steps subsequent to the plating of the platinum
expose the workpiece to oxygen. Such exposure is possible if
subsequent processing steps, for example, try the workpiece and
expose it to oxygen, such as found in the ambient air.
[0012] Various platinum electroplating processes are known, though
efforts have mainly been directed at development of an appropriate
electroplating bath, and additives for the bath. For example, a
process of electroplating a platinum-rhodium alloy on a metal
substrate has been disclosed in U.S. Pat. No. 4,285,784; and a
procedure for electroplating platinum and platinum alloys involving
use of an organic polyamine as a platinum complexing agent has been
disclosed in U.S. Pat. No. 4,427,502. Moreover, there have been
some uses of platinum in semiconductor chip manufacture. For
example, selective deposition of platinum on a conductive or
semiconductive substrate was disclosed in U.S. Pat. No. 5,320,978,
and a method for depositing a coat of platinum on the surface of a
silicon substrate by dipping the substrate into an aqueous solution
of chloroplatinic acid and hydrofluoric acid was disclosed in U.S.
Pat. No. 3,963,523. Recently, noble metal plating on a pre-existing
seed layer for the fabrication of electrodes for use in DRAM and
FRAM was disclosed in U.S. Pat. No. 5,789,320.
[0013] Several technical problems must be overcome in designing
reactors used in the electroplating of semiconductor wafers with a
noble metal, such as platinum. For example, most noble metals tend
to be deposited in a state of high film stress. This film stress is
generally greatest at or near the point of contact where current is
applied to the seed layer during the electroplating process. This
stress can be detrimental to the function and reliability of the
microelectronic components produced using these materials.
[0014] One factor affecting film stress is the occurrence of
varying current densities that occur during the plating process
while the workpiece is functioning as a cathode. In many reactors
used to electroplate metals onto the surface of a semiconductor
wafer, a small number of discrete electrical contacts (e.g., 6
contacts) are used to contact the seed layer about the perimeter of
the wafer. Such discrete contacts ordinarily produce higher current
densities near the contact points than at other portions of the
wafer. This non-uniform distribution of current across the wafer,
in turn, causes non-uniform deposition of the plated metallic
material and, further, produces a substantial film stress near the
contact locations. Such reactors are therefore not particularly
well-suited for plating noble metals, such as platinum.
[0015] Another problem with electroplating of noble metals onto
workpieces concerns efforts to prevent the electric contacts
themselves from being plated during the electroplating process. Any
material plated to the electrical contacts must be removed to
prevent changing contact performance. However, noble metals such as
platinum, unlike metals such as copper, cannot be reverse plated
from the electrical contacts. Rather, any electrical contact that
is plated with the noble metal must be replaced if the plating
process is to remain in a satisfactory working state.
[0016] The foregoing concern also applies to the use of current
thieving in the electroplating process. Current thieving, effected
by the provision of electrically-conductive elements other than
those which contact the seed layer, can be employed near the wafer
contacts to minimize non-uniformity of the deposited noble metal.
The electrically-conductive elements are generally exposed to the
electroplating solution and, as such, are plated with the noble
metal during the electroplating process. The elements must
therefore be replaced if the plating process is to remain in a
satisfactory operational state. As a result, current thieving,
while desirable to increase film uniformities, can be costly to
implement.
[0017] When electroplating a noble metal such as platinum, it is
desirable to prevent electroplating on any exposed barrier layer
near the edge of the semiconductor wafer. Electroplated material
may not adhere well to the exposed barrier layer material, and is
therefore prone to peeling off in subsequent wafer processing
steps. Further, metal that is electroplated onto the barrier layer
within the reactor may flake off during the electroplating process
thereby adding particulate contaminants to the electroplating bath.
Such contaminants can adversely affect the overall electroplating
process.
[0018] The specific metal used for the seed layer can also
complicate the electroplating process. For example, certain seed
layer metals have a relatively high electrical resistance. Still
further, some noble metals, such as platinum, have a high
electrical resistance. As a consequence, use of the typical
plurality of electrical wafer contacts (for example, six (6)
discrete contacts) may not provide adequate uniformity of the
plated metal layer on the wafer due to non-uniformities in the
plating current that result from the high electrical resistance of
the seed layer and/or noble metal layer (e.g., platinum).
[0019] Beyond the contact related problems discussed above, there
are also other problems associated with electroplating reactors. As
device sizes decrease, the need for tighter control over the
processing environment increases. This includes control over the
contaminants that affect the electroplating process. The moving
components of the reactor, which tend to generate such
contaminants, should therefore be subject to strict isolation
requirements. To control film stress, optimal process parameters
must be determined for parameters such as electrolyte temperature,
flow rate, cathode current density, current waveform and
electrolyte composition. Other factors that should be considered
include uniformity of deposition thickness, film resistivity,
surface roughness, micro-feature throwing power and particulate
contamination.
[0020] Still further, existing electroplating reactors are often
difficult to maintain. Such difficulties must be overcome if an
electroplating reactor design is to be accepted for large-scale
manufacturing.
[0021] The present inventors have recognized and addressed many of
the foregoing problems that exist in connection with the plating of
noble metals, particularly platinum. To this end, they have
developed an efficient method and production worthy apparatus for
electroplating noble metals onto the surface of a workpiece, such
as a semiconductor wafer. The disclosed method and apparatus
provide for a suitable deposition rate, excellent film
characteristics, and a reduction in the level of film stress that
could otherwise result in cracking, delamination, or poor device
reliability.
SUMMARY OF THE INVENTION
[0022] The present invention is directed to an improved
electroplating method, chemistry, and production worthy apparatus
for depositing noble metals onto the surface of the workpiece, such
as a semiconductor wafer, pursuant to manufacturing a
microelectronic device, circuit, and/or component. The reliability
of the noble metal material deposited using the disclosed method,
chemistry, and/or apparatus is significantly better than the
reliability of noble metal structures deposited using the teachings
of the prior art. This is largely attributable to the low stress of
films that are deposited using the teachings disclosed herein. The
metals, which can be deposited, include gold, silver, platinum,
palladium, ruthenium, iridium, rhodium, osmium and alloys
containing these metals.
[0023] In accordance with one aspect of the present invention, an
apparatus for plating a noble metal on a microelectronic workpiece
is disclosed that comprises a reactor chamber that contains an
electroplating solution containing ions or complexes of the noble
metal or noble metal alloy that is to be plated onto the workpiece.
The apparatus also includes a workpiece support having a contact
for providing electroplating power to a surface at a side of the
workpiece that is to be plated. The contact electrically contacts
the workpiece at a large plurality of discrete contact points and
each of the contact points is isolated from exposure to the
electroplating solution. To complete the electroplating cell, an
anode is provided and the electroplating solution and is spaced
from the workpiece support within the reaction chamber.
[0024] In accordance with a further aspect of the present
invention, a contact member for use in conducting electroplating
power to a surface of a microelectronic workpiece that is to be
electroplated with a noble metal is set forth. The contact member
comprises a conductive member and a removable conductive surface
material disposed about an exterior surface of the conductive
member. The removable conductive surface material may be in the
form of a removable conductive strip wound about the exterior
surface of the conductive member. In a preferred embodiment, the
conductive member and the removable conductive surface material
form a single, discreet contact.
[0025] In accordance with a still further aspect of the present
invention, an apparatus for plating a noble metal on a
microelectronic workpiece is disclosed that comprises a reactor
chamber that contains an electroplating solution having ions or
complexes of the noble metal or noble metal alloy that is to be
plated onto the workpiece. The apparatus also includes a workpiece
support including a contact assembly for providing electroplating
power to a surface at a side of the workpiece that is to be plated
and an anode spaced from the workpiece support within the reaction
chamber and contacting the electroplating solution. A chemical
delivery system is employed for supplying the electroplating
solution to the reactor chamber and recirculating electroplating
solution removed from the reactor chamber. To eliminate fouling of
the solution, as is quite prevalent when plating a noble metal, a
multi-stage filtration system is utilized. The filtration system is
disposed within the chemical delivery system for filtering
electroplating solution removed from the reactor chamber before it
is re-supplied to the reactor chamber. It includes at least a first
filter stage for filtering particles greater than or equal to a
first size and a second filter stage disposed downstream of the
first filter stage for filtering particles greater than or equal to
a second size, the first size being greater in magnitude than the
second size.
[0026] It may be desirable to use a current thief in any of the
foregoing electroplating apparatus. In accordance with a still
further aspect of the present invention, a disposable current thief
is set forth. The disposable current thief is disposed in the
electroplating solution between the anode and the contact assembly
and is formed from the conductive portions of a printed circuit
board. The disclosed current thief is manufactured from readily
available materials using simple manufacturing processes thereby
significantly reducing the costs of providing current thieving in
noble metal electroplating processes.
[0027] A method for electroplating a noble metal onto the surface
of a microelectronic workpiece is also set forth. Although the
method is generally apparatus independent, any of the foregoing
apparatus may be used to implement the method. Generally stated,
the method involves bringing the surface of the workpiece that is
to be plated into contact with an electroplating solution including
ions or complexes of a noble metal or noble metal alloy that is to
be plated on the surface of the workpiece. Electroplating power is
applied between the surface of the workpiece and an anode using a
low current for a first predetermined period of time. This is
subsequently followed at a later time by application of full-scale
electroplating power between the surface of the workpiece and the
anode for a second predetermined period of time. In many instances,
it is preferable, though not necessary, to provide the low current
as the initial electroplating power and to have the full-scale
electroplating power applied immediately thereafter. At a time
subsequent to the end of the second predetermined period of time,
electroplating power is removed and the surface of the workpiece is
disengaged from the electroplating solution. Suitable parameters
for electroplating the noble metal using an acidic electroplating
solution as well as an alkaline electroplating solution are also
set forth.
BRIEF DESCRIPTION OF THE DRAWINGS
[0028] The foregoing aspects and many of the attendant advantages
of this invention will become more readily appreciated as the same
become better understood by reference to the following detailed
description, when taken in conjunction with the accompanying
drawings, wherein:
[0029] FIG. 1 is a cross-sectional view through an electroplating
reactor that is constructed in accordance with various teachings of
the present invention.
[0030] FIG. 2 illustrates a specific construction of one embodiment
of a reactor bowl suitable for use in the assembly illustrated in
FIG. 1.
[0031] FIG. 3 illustrates one embodiment of a reactor head,
comprised of a stationary assembly and a rotor assembly that is
suitable for use in the assembly illustrated in FIG. 1.
[0032] FIGS. 4-10 illustrate one embodiment of a contact assembly
using flexure contacts that is suitable for use in the reactor
assembly illustrated in FIG. 1.
[0033] FIGURES 11A-12 illustrate two different embodiments of a
"Belleville ring" contact structure.
[0034] FIGS. 13-15 illustrate one embodiment of a contact assembly
using a "Belleville ring" contact structure, such as one of those
illustrated in FIGS. 11-12, that is suitable for use in the reactor
assembly illustrated in FIG. 1.
[0035] FIG. 16A is a schematic block diagram of a flow system for
supplying the plating solution to the reactor bowl.
[0036] FIGS. 16B-20D illustrate various aspects of one embodiment
of a quick-attach mechanism.
[0037] FIG. 21 is a cross-sectional view of the reactor head
illustrating the disposition of the reactor head in a condition in
which it may accept a workpiece.
[0038] FIG. 22 is a cross-sectional view of the reactor head
illustrating the disposition of the reactor head in a condition in
which it is ready to present the workpiece to the reactor bowl.
[0039] FIG. 23 illustrates an exploded view one embodiment of the
rotor assembly.
[0040] FIG. 24 illustrates one embodiment of a segmented current
thief suitable for noble metal plating.
[0041] FIG. 25 illustrates one embodiment of a finger contact that
may also function as a current thief in the plating of noble
metals.
[0042] FIGS. 26-28 are top plan views of integrated processing
tools that may incorporate electroless plating reactors and
electroplating reactors in combination.
[0043] FIGS. 29-32 are various views of a further embodiment of a
reactor base for providing a flow of electroplating solution to the
surface of a workpiece in which the flow assists in increasing the
uniformity of the electroplated noble metal layer.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
[0044] With reference to FIGS. 1-3, there is shown a reactor
assembly 20 for electroplating a noble metal on the surface of a
microelectronic workpiece, such as a semiconductor wafer 25.
Generally stated, the reactor assembly 20 is comprised of a reactor
head 30 and a corresponding reactor bowl 35. This type of reactor
assembly is particularly suited for effecting electroplating of
semiconductor wafers or like workpieces, in which an electrically
conductive, thin-film seed layer of the wafer is electroplated with
a blanket or patterned noble metal layer, such as a layer of
platinum.
[0045] A specific construction of one embodiment of a reactor bowl
35 suitable for use in the reactor assembly 20 is illustrated in
FIG. 2. The electroplating reactor bowl 35 is that portion of the
reactor assembly 20 that contains electroplating solution, and that
directs the solution at a high flow rate against a generally
downwardly facing surface of an associated workpiece 25 to be
plated. To this end, electroplating solution is circulated through
the reactor bowl 35. Attendant to solution circulation, the
solution flows from the reactor bowl 35, over the weir-like
periphery of the bowl, into a lower overflow chamber 40 of the
reactor assembly 20. Solution is drawn from the overflow chamber
typically for re-circulation through the reactor.
[0046] The temperature of the electroplating solution is monitored
and maintained by a temperature sensor and heater, respectively.
The sensor and heater are disposed in the circulation path of the
electroplating solution. For electroplating noble metals and their
alloys, particularly platinum and platinum alloys, these components
maintain the temperature of the electroplating solution in a
temperature range between 40.degree. C. and 80.degree. C. Even more
preferably, these components maintain the temperature of the
electroplating solution at about 65.degree. C. 0.5. As will be
explained in connection with the preferred electroplating process,
the electroplating solution exhibits optimal deposition properties
within this latter temperature range.
[0047] The reactor bowl 35 includes a riser tube 45, within which
an inlet conduit 50 is positioned for introduction of
electroplating solution into the interior portion of the reactor
bowl 35. The inlet conduit 50 is preferably conductive and makes
electrical contact with and supports an electroplating anode 55.
Anode 55 is preferably an inert anode, and, in at least one of the
preferred methods, a platinized titanium inert anode is used. The
electrically-conductive surface of the workpiece functions as a
cathode.
[0048] Electroplating solution flows at a high flow rate,
preferably at a rate of 5 gal/min, from the inlet conduit 50
through openings at the upper portion thereof. From there, the
solution flows about the anode 55, and through an optional
diffusion plate 65 positioned in operative association with and
between the cathode (workpiece) and the anode.
[0049] The reactor head 30 of the electroplating reactor 20 is
preferably comprised of a stationary assembly 70 and a rotor
assembly 75, diagrammatically illustrated in FIG. 3. Rotor assembly
75 is configured to receive and carry an associated wafer 25 or
like workpiece, position the wafer in a process-side down
orientation within reactor bowl 35, and to rotate or spin the
workpiece while joining its electrically-conductive surface in the
plating circuit of the reactor assembly 20. The reactor head 30 is
typically mounted on a lift/rotate apparatus 80, which is
configured to rotate the reactor head 30 from an upwardly-facing
disposition, in which it receives the wafer to be plated, to a
downwardly facing disposition, in which the surface of the wafer to
be plated is positioned downwardly in reactor bowl 35, generally in
confronting relationship to diffusion plate 65. A robotic arm 418,
including an end effector, is typically employed for placing the
wafer 25 in position on the rotor assembly 75, and for removing the
plated wafer from the rotor assembly.
[0050] It will be recognized that other reactor assembly
configurations may be used with the inventive aspects of the
disclosed reactor head, the foregoing being merely illustrative.
Another reactor assembly suitable for use in the foregoing
configuration is illustrated in U.S. Ser. No. 60/143,769, entitled
"Workpiece Processor Having Improved Processing Chamber", filed
Jul. 12, 1999 (Attorney Docket No. SEM4492P0831 US) and further
reactor assembly illustrated in U.S. Ser. No. 60/120,955, filed
Apr. 13, 1999, both of which are incorporated herein by
reference.
Electroplating Solutions
[0051] The plating bath that is used in the reactor 35 depends upon
the particular noble metal or noble metal alloy that is to be
plated. Examples of suitable plating solutions for plating noble
metals include:1) for gold--cyanide-based or sulfite-based baths
(such as Enthone-OMI Neutronex 309);2) for ruthenium--a sulfamate,
nitrosyl sulfamate or nitroso-based bath (such as Technic's
Ruthenium U, Englehard's Ru-7 and Ru-8, and LeaRonal's Decronal
White 44 and Decronal Black 44);and 4) for platinum--a potassium
hydroxide-based based, ammonia-sulfamate-based, or sulfate-based
bath (such as Englehard's Platinum A bath or Technics Platinum S
bath).
[0052] The particular solution for each selected metal is partially
dependent upon the particular plating process being used. For
example, with respect to platinum, a potassium hydroxide-based
solution, such as Englehard's Platinum A, is well suited for use in
an alkaline plating process, while an ammonia-sulfamate-based
solution (such as Technic's Platinum S) is particularly well suited
for use in a photoresist template process. Careful control of
temperature and pH of the bath is often required for optimal
plating results. Such parameters are typically placed under the
control of a programmable control system.
Exemplary Process
[0053] An exemplary process sequence for plating a noble metal or
noble metal alloy, such as platinum or a platinum alloy, onto the
surface of a workpiece in a reactor assembly, such as the reactor
assembly illustrated in FIGS. 1-3, includes the following
processing steps:
[0054] (Optional) Pre-wet/Pre-clean the substrate material using
deionized water or acid and/or a surfactant to eliminate the dry
plating surface (about 30 seconds) (the pre-wet solution may be
heated to the same temperature at which electroplating will
occur);
[0055] Adjust and/or program (either manually or using the
programmable control system) the electroplating system for the
appropriate processing parameters, including electroplating
solution flow rate, pH, temperature, concentration of metal or
alloy to be deposited, current density and waveform of
electroplating power applied, and rotation rate of workpiece;
[0056] Bring the surface of the workpiece that is to be plated into
contact with the noble metal or noble metal alloy electroplating
solution;
[0057] (Optional) Apply an initial low electroplating current for a
first predetermined period of time to initiate electroplating of
the workpiece;
[0058] Apply full-scale electroplating current for the duration
necessary to achieve the desired depth of deposited material;
[0059] Halt electrolysis;
[0060] Disengage the workpiece from electroplating solution;
[0061] Spin the workpiece at a high spin rate (i.e., above about
200 rpm) to remove excess electroplating solution;
[0062] Rinse the workpiece in a spray of deionized water (about 2
min.) and spin dry at a high rotation rate;
[0063] (Optional) Subject the workpiece to a backside cleaning
process to remove any backside contamination, such as potassium
hydroxide contamination
[0064] For an alkaline platinum plating process, the preferred
processing parameters include a flow rate of about 5 gallons per
minute using a plating solution having a temperature of 65.degree.
C., a pH in the range of about 11-12, preferably about 11.5, and a
platinum concentration in the range of about 10-15 g/l, preferably
12.5 g/l. Electroplating power is applied having a current density
between about 3 and 9 mA/cm.sup.2, depending on the seed layer
type, with low current initiation using a pulsed waveform having a
pattern of 1 ms on, 1 ms off, or DC (again, depending on the seed
layer material).
[0065] Low current initiation is often desirable. If a seed layer
is to thin, lacks sufficient adhesion, or is too highly stressed,
cracking and peeling can occur at electroplating contact points
and/or structures (e.g., posts, trenches, etc.). This may be due to
high localized current densities. A low current initiation step
allows for a slow build-up of platinum, or other noble metal, in
these areas. When the thickness has increased beyond a
predetermined magnitude, the current (and plating rate) can be
increased without stress cracking. The predetermined magnitude can
be determined experimentally.
[0066] Deposition rates in excess of 320 angstroms/min are typical
using the above noted parameters. A process using the same
parameters, with the exception of applying a DC waveform, can
result in deposition rates that are in excess of 740 angstroms/min.
The current density that is used is principally limited by the
amount of hydrogen gas that evolves during the electroplating
process. Hydrogen gas may then be trapped in the plated film
thereby resulting in stress cracking. There is therefore a
trade-off that must be made between the deposition rate and the
risk of stress cracking.
[0067] The alkaline platinum plating process using the above
parameters has exhibited high throwing power with respect to
submicron features, making it well suited for the formation of 3D
plug capacitor electrodes. Such electrodes require a conformal
platinum layer that is defect free, and non-porous. A For 3D plug
capacitors, plating thickness in an ultra thin layer less than 500
.ANG. is typical.
[0068] The foregoing alkaline platinum plating process is generally
unsuitable for use in patterned plating in which photoresist is
used as the plating mask unless the photoresist has been
specifically chosen and treated (e.g., deep ultraviolet bake) for
use in an alkaline plating bath. Rather, an acidic plating bath is
preferred for such processes. The present inventors have likewise
developed an acidic platinum plating process that is suitable for
use in processes employing a photoresist mask. The preferred
processing parameters include a flow rate of 5 gallons per minute
for a plating solution having a temperature of 65.degree. C., a pH
in the range of about 2-4, preferably about 3.0, and a platinum
concentration in the range of about 2-16 g/l, preferably about 4.0
g/l. Electroplating power is applied having a current density in
the range of about 20-50 mA/cm.sup.2, preferably about 34.5
mA/cm.sup.2 using a pulsed waveform. The waveform may have an on
time of about 1-10 ms, preferably of 1 ms, and an off time of about
1-10 ms, preferably 1 ms off. Deposition rates in excess of 550
angstroms/min are typical using the above noted acidic platinum
plating parameters. The acidic platinum plating process is well
suited for the formation of patterned capacitors, where a plating
thickness of about 2500 angstroms is typical.
[0069] As noted above, the plating process is typically applied to
a workpiece having a seed layer on top of a barrier layer.
Typically the seed layer is applied to the barrier layer using
physical vapor deposition. The characteristics of the barrier layer
and the seed layer, including the type of material and thickness,
can have an impact on stress cracking and plating uniformity.
Preferred barrier layer materials include TiN, Ta, TaN, Ti, and
TiO2.
[0070] Similar plating processes, or ones with only slight
modifications, using the particular plating bath required, can be
used to plate other noble metals, or noble metal alloys. As will be
set forth in further detail below, the foregoing processing steps
and sequence may be implemented in a single fabrication tool having
a plurality of similar processing stations and a programmable robot
that transfers the workpieces between such stations.
[0071] There are a number of enhancements that may be made to the
reactor assembly 20 described above that facilitate uniformity of
the noble metal deposits over the face of the workpiece. For
example, the reactor assembly 20 may use a contact assembly that
reduces non-uniformities in the deposits that occur proximate the
discrete contacts that are used to provide plating power to the
surface at the perimeter of the workpiece, including the
alternative use of a continuous or a semi-continuous ring contact.
Additionally, other enhancements to the reactor assembly 20 may be
added to facilitate routine service and/or configurability of the
system.
Improved Contact Assemblies
[0072] As noted above, the manner in which the electroplating power
is supplied to the wafer at the peripheral edge thereof is very
important to the overall film quality of the deposited metal. Some
of the more desirable characteristics of a contact assembly used to
provide such electroplating power include, for example, the
following:
[0073] uniform distribution of electroplating power about the
periphery of the wafer to maximize the uniformity of the deposited
film;
[0074] consistent contact characteristics to insure wafer-to-wafer
uniformity;
[0075] minimal intrusion of the contact assembly on the wafer
periphery to maximize the available area for device production;
and
[0076] minimal plating on the barrier layer about the wafer
periphery to inhibit peeling and/or flaking.
[0077] To meet one or more of the foregoing characteristics,
reactor 20 preferably employs a ring contact assembly 85 that
provides either a continuous electrical contact or a high number of
discrete electrical contacts with the wafer 25. By providing a more
continuous contact with the outer peripheral edges of the
semiconductor wafer 25, in this case around the outer circumference
of the semiconductor wafer, a more uniform current is supplied to
the semiconductor wafer 25 that promotes more uniform current
densities. The more uniform current densities enhance uniformity in
the depth of the deposited material.
[0078] Contact assembly 85, in accordance with a preferred
embodiment, includes contact members that provide minimal intrusion
about the wafer periphery while concurrently providing consistent
contact with the seed layer. Contact with the seed layer is
enhanced by using a contact member structure that provides a wiping
action against the seed layer as the wafer is brought into
engagement with the contact assembly. This wiping action assists in
removing any oxides at the seed layer surface thereby enhancing the
electrical contact between the contact structure and the seed
layer. As a result, uniformity of the current densities about the
wafer periphery are increased and the resulting film is more
uniform. Further, such consistency in the electrical contact
facilitates greater consistency in the electroplating process from
wafer-to-wafer thereby increasing wafer-to-wafer uniformity.
[0079] Contact assembly 85, as will be set forth in further detail
below, also preferably includes one or more structures that provide
a barrier, individually or in cooperation with other structures,
that separates the contact/contacts, the peripheral edge portions
and backside of the semiconductor wafer 25 from the plating
solution. This prevents the plating of metal onto the individual
contacts and, further, assists in preventing any exposed portions
of the barrier layer near the edge of the semiconductor wafer 25
from being exposed to the electroplating environment. As a result,
plating of the barrier layer and the appertaining potential for
contamination due to flaking of any loosely adhered electroplated
material is substantially limited.
Ring Contact Assemblies Using Flexure Contacts
[0080] One embodiment of a contact assembly suitable for use in the
assembly 20 is shown generally at 85 of FIGS. 4-10. The contact
assembly 85 forms part of the rotor assembly 75 and provides
electrical contact between the semiconductor wafer 25 and a source
of electroplating power. In the illustrated embodiment, electrical
contact between the semiconductor wafer 25 and the contact assembly
85 occurs at a large plurality of discrete flexure contacts 90 that
are effectively separated from the electroplating environment
interior of the reactor bowl 35 when the semiconductor wafer 25 is
held and supported by the rotor assembly 75.
[0081] The contact assembly 85 may be comprised of several discrete
components. With reference to FIG. 4, when the workpiece that is to
be electroplated is a circular semiconductor wafer, the discrete
components of the contact assembly 85 join together to form a
generally annular component having a bounded central open region
95. It is within this bounded central open region 95 that the
surface of the semiconductor wafer that is to be electroplated is
exposed. With particular reference to FIG. 6, contact assembly 85
includes an outer body member 100, an annular wedge 105, a
plurality of flexure contacts 90, a contact mount member 110, and
an interior wafer guide 115. Preferably, annular wedge 105, flexure
contacts 90, and contact mount member 110 are formed from
platinized titanium while wafer guide 115 and outer body member 100
are formed from a dielectric material that is compatible with the
electroplating environment. Annular wedge 105, flexure contacts 90,
mount member 110, and wafer guide 115 join together to form a
single assembly that is secured together by outer body member
100.
[0082] As shown in FIG. 6, contact mount member 110 includes a
first annular groove 120 disposed about a peripheral portion
thereof and a second annular groove 125 disposed radially inward of
the first annular groove 120. The second annular groove 125 opens
to a plurality of flexure channels 130 that are equal in number to
the number of flexure contacts 90. As can be seen from FIG. 4, a
total of 36 flexure contacts 90 are employed, each being spaced
from one another by an angle of about 10 degrees.
[0083] Referring again to FIG. 6, each flexure contact 90 is
comprised of an upstanding portion 135, a transverse portion 140, a
vertical transition portion 145, and a wafer contact portion 150.
Similarly, wedge 105 includes an upstanding portion 155 and a
transverse portion 160. Upstanding portion 155 of wedge 105 and
upstanding portion 135 of each flexure contact 90 are secured
within the first annular groove 120 of the contact mount member 110
at the site of each flexure channel 130. Self-adjustment of the
flexure contacts 90 to their proper position within the overall
contact assembly 85 is facilitated by first placing each of the
individual flexure contacts 90 in its respective flexure channel
130 so that the upstanding portion 135 is disposed within the first
annular groove 120 of the contact mount member 110 while the
transition portion 145 and contact portion 150 proceed through the
respective flexure channel 130. The upstanding portion 155 of wedge
member 105 is then urged into the first annular groove 120. To
assist in this engagement, the upper end of upstanding portion 155
is tapered. The combined width of upstanding portion 135 of the
flexure contact 90 and upstanding portion 155 of wedge 105 are such
that these components are firmly secured with contact mount member
110.
[0084] Transverse portion 160 of wedge 105 extends along a portion
of the length of transverse portion 140 of each flexure 90. In the
illustrated embodiment, transverse portion 160 of wedge portion 105
terminates at the edge of the second annular groove 125 of contact
mount member 110. As will be more clear from the description of the
flexure contact operation below, the length of transverse portion
160 of wedge 105 can be chosen to provide the desired degree of
stiffness of the flexure contacts 90.
[0085] Wafer guide 115 is in the form of an annular ring having a
plurality of slots 165 through which contact portions 150 of
flexures 90 extend. An annular extension 170 proceeds from the
exterior wall of wafer guide 115 and engages a corresponding
annular groove 175 disposed in the interior wall of contact mount
member 110 to thereby secure the wafer guide 115 with the contact
mount member 110. As illustrated, the wafer guide member 115 has an
interior diameter that decreases from the upper portion thereof to
the lower portion thereof proximate contact portions 150. A wafer
inserted into contact assembly 85 is thus guided into position with
contact portions 150 by a tapered guide wall formed at the interior
of wafer guide 115. Preferably, the portion 180 of wafer guide 115
that extends below annular extension 170 is formed as a thin,
compliant wall that resiliently deforms to accommodate wafers
having different diameters within the tolerance range of a given
wafer size. Further, such resilient deformation accommodates a
range of wafer insertion tolerances occurring in the components
used to bring the wafer into engagement with the contact portions
150 of the flexures 90.
[0086] Referring to FIG. 6, outer body member 100 includes an
upstanding portion 185, a transverse portion 190, a vertical
transition portion 195 and a further transverse portion 200 that
terminates in an upturned lip 205. Upstanding portion 185 includes
an annular extension 210 that extends radially inward to engage a
corresponding annular notch 215 disposed in an exterior wall of
contact mount member 110. A V-shaped notch 220 is formed at a lower
portion of the upstanding portion 185 and circumvents the outer
periphery thereof. The V-shaped notch 220 allows upstanding portion
185 to resiliently deform during assembly. To this end, upstanding
portion 185 resiliently deforms as annular extension 210 slides
about the exterior of contact mount member 110 to engage annular
notch 215. Once so engaged, contact mount member 110 is clamped
between annular extension 210 and the interior wall of transverse
portion 190 of outer body member 100.
[0087] Further transverse portion 200 extends beyond the length of
contact portions 150 of the flexure contacts 90 and is dimensioned
to resiliently deform as a wafer, such as at 25, is driven against
them. V-shaped notch 220 may be dimensioned and positioned to
assist in the resilient deformation of transverse portion 200. With
the wafer 25 in proper engagement with the contact portions 150,
upturned lip 205 engages wafer 25 and assists in providing a
barrier between the electroplating solution and the outer
peripheral edge and backside of wafer 25, including the flexure
contacts 90.
[0088] As illustrated in FIG. 6, flexure contacts 90 resiliently
deform as the wafer 25 is driven against them. Preferably, contact
portions 150 are initially angled upward in the illustrated manner.
Thus, as the wafer 25 is urged against contact portions 150,
flexures 90 resiliently deform so that contact portions 150
effectively wipe against surface 230 of wafer 25. In the
illustrated embodiment, contact portions 150 effectively wipe
against surface 230 of wafer 25 a horizontal distance designated at
235. This wiping action assists in removing and/or penetrating any
oxides from surface 230 of wafer 25 thereby providing more
effective electrical contact between flexure contacts 90 and the
seed layer at surface 230 of wafer 25.
[0089] With reference to FIGS. 7 and 8, contact mount member 110 is
provided with one or more ports 240 that may be connected to a
source of purging gas, such as a source of nitrogen. As shown in
FIG. 8, purge ports 240 open to second annular groove 125 which, in
turn, operates as a manifold to distribute the purging gas to all
of the flexure channels 130 as shown in FIG. 6. The purging gas
then proceeds through each of the flexure channels 130 and slots
165 to substantially surround the entire contact portions 150 of
flexures 90. In addition to purging the area surrounding contact
portions 150, the purge gas cooperates with the upturned lip 205 of
outer body member 100 to effect a barrier to the electroplating
solution. Further circulation of the purge gas is facilitated by an
annular channel 250 formed between a portion of the exterior wall
of wafer guide 115 and a portion of the interior wall of contact
mount member 110.
[0090] As shown in FIGS. 4, 5 and 10, contact mount member 110 is
provided with one or more threaded apertures 255 that are
dimensioned to accommodate a corresponding connection plug 260.
With reference to FIGS. 5 and 10, connection plugs 260 provide
electroplating power to the contact assembly 85 and, preferably,
are each formed from platinized titanium. In a preferred form of
plugs 260, each plug 260 includes a body 265 having a centrally
disposed bore hole 270. A first flange 275 is disposed at an upper
portion of body 265 and a second flange 280 is disposed at a lower
portion of body 265. A threaded extension 285 proceeds downward
from a central portion of flange 280 and secures with threaded bore
hole 270. The lower surface of flange 280 directly abuts an upper
surface of contact mount member 110 to increase the integrity of
the electrical connection therebetween.
[0091] Although flexure contacts 90 are formed as discrete
components, they may be joined with one another as an integral
assembly. To this end, for example, the upstanding portions 135 of
the flexure contacts 90 may be joined to one another by a web of
material, such as platinized titanium, that is either formed as a
separate piece or is otherwise formed with the flexures from a
single piece of material. The web of material may be formed between
all of the flexure contacts or between select groups of flexure
contacts. For example, a first web of material may be used to join
half of the flexure contacts (e.g., 18 flexure contacts in the
illustrated embodiment) while a second web of material is used to
join a second half of the flexure contacts (e.g., the remaining 18
flexure contacts in the illustrated embodiment). Different
groupings are also possible.
Belleville Ring Contact Assemblies
[0092] Alternative contact assemblies are illustrated in FIGS.
11-15. In each of these contact assemblies, the contact members are
integrated with a corresponding common ring and, when mounted in
their corresponding assemblies, are biased upward in the direction
in which the wafer or other substrate is received upon the contact
members. A top view of one embodiment of such a structure is
illustrated in FIG. 11A while a perspective view thereof is
illustrated in FIG. 11B. As illustrated, a ring contact, shown
generally at 610, is comprised of a common ring portion 630 that
joins a plurality of contact members 655. The common ring portion
630 and the contact members 655, when mounted in the corresponding
assemblies, are similar in appearance to half of a conventional
Belleville spring. For this reason, the ring contact 610 will be
hereinafter referred to as a "Bellville ring contact" and the
overall contact assembly into which it is placed will be referred
to as a "Bellville ring contact assembly".
[0093] The embodiment of Belleville ring contact 610 illustrated in
FIG. 11A and 11B includes 72 contact members 655 and is preferably
in formed from platinized titanium. The contact members 655 may be
formed by cutting arcuate sections 657 into the interior diameter
of a platinized titanium ring. A predetermined number of the
contact members 658 have a greater length than the remaining
contact members 655 to, for example, accommodate certain flat-sided
wafers.
[0094] A further embodiment of a Belleville ring contact 610 is
illustrated in FIG. 12. As above, this embodiment is preferably
formed from platinized titanium. Unlike the embodiment of FIGURES
11A and 11B in which all of the contact members 655 extend radially
inward toward the center of the structure, this embodiment includes
contact members 659 that are disposed at an angle. This embodiment
constitutes a single-piece design that is easy to manufacture and
that provides a more compliant contact than does the embodiment of
FIGS. 11A and 11B with the same footprint. This contact embodiment
can be fixtured into the "Belleville" form in the contact assembly
and does not require permanent forming. If the Belleville ring
contact 610 of this embodiment is fixtured in place, a complete
circumferential structure is not required. Rather the contact may
be formed and installed in segments thereby enabling independent
control/sensing of the electrical properties of the segments.
[0095] A first embodiment of a Bellville ring contact assembly is
illustrated generally at 600 in FIGS. 13-15. As illustrated, the
contact assembly 600 comprises a conductive contact mount member
605, a Bellville ring contact 610, a dielectric wafer guide ring
615, and an outer body member 625. The outer, common portion 630 of
the Bellville ring contact 610 includes a first side that is
engaged within a notch 675 of the conductive base ring 605. In many
respects, the Belleville ring contact assembly of this embodiment
is similar in construction with the flexure contact assembly 85
described above. For that reason, the functionality of many of the
structures of the contact assembly 600 will be apparent and will
not be repeated here.
[0096] Preferably, the wafer guide ring 615 is formed from a
dielectric material while contact mount member 605 is formed from a
single, integral piece of conductive material or from a dielectric
or other material that is coated with a conductive material at its
exterior. Even more preferably, the conductive ring 605 and
Bellville ring contact 610 are formed from platinized titanium or
are otherwise coated with a layer of platinum.
[0097] The wafer guide ring 615 is dimensioned to fit within the
interior diameter of the contact mount member 605. Wafer guide ring
615 has substantially the same structure as wafer guides 115 and
115b described above in connection with contact assemblies 85 and
85b, respectively. Preferably, the wafer guide ring 615 includes an
annular extension 645 about its periphery that engages a
corresponding annular slot 650 of the conductive base ring 605 to
allow the wafer guide ring 615 and the contact mount member 605 to
snap together.
[0098] The outer body member 625 includes an upstanding portion
627, a transverse portion 629, a vertical transition portion 632
and a further transverse portion 725 that extends radially and
terminates at an upturned lip 730. Upturned lip 730 assists in
forming a barrier to the electroplating environment when it engages
the surface of the side of workpiece 25 that is being processed. In
the illustrated embodiment, the engagement between the lip 730 and
the surface of workpiece 25 is the only mechanical seal that is
formed to protect the Bellville ring contact 610.
[0099] The area proximate the contacts 655 of the Belleville ring
contact 610 is preferably purged with an inert fluid, such as
nitrogen gas, which cooperates with lip 730 to effect a barrier
between the Bellville ring contact 610, peripheral portions and the
backside of wafer 25, and the electroplating environment. As
particularly shown set forth in FIGS. 19 and 20, the outer body
member 625 and contact mount member 605 are spaced from one another
to form an annular cavity 765. The annular cavity 765 is provided
with an inert fluid, such as nitrogen, through one or more purge
ports 770 disposed through the contact mount member 605. The purged
ports 770 open to the annular cavity 765, which functions as a
manifold to distribute to the inert gas about the periphery of the
contact assembly. A given number of slots, such as at 780,
corresponding to the number of contact members 655 are provided and
form passages that route the inert fluid from the annular cavity
765 to the area proximate contact members 655.
[0100] FIGS. 14 and 15 also illustrate the flow of a purging fluid
in this embodiment of Bellville ring contact assembly. As
illustrated by arrows, the purge gas enters purge port 770 and is
distributed about the circumference of the assembly 600 within
annular cavity 765. The purged gas then flows through slots 780 and
below the lower end of contact mount member 605 to the area
proximate Bellville contact 610. At this point, the gas flows to
substantially surround the contact members 655 and, further, may
proceed above the periphery of the wafer to the backside thereof.
The purging gas may also proceed through an annular channel 712
defined by the contact mount member 605 and the interior of the
compliant wall formed at the lower portion of wafer guide ring 615.
Additionally, the gas flow about contact members 655 cooperates
with upturned lip 730 effect a barrier at lip 730 that prevents
electroplating solution from proceeding therethrough.
[0101] When a wafer or other workpiece 25 is urged into engagement
with the contact assembly 600, the workpiece 25 first makes contact
with the contact members 655. As the workpiece is urged further
into position, the contact members 655 deflect and effectively wipe
the surface of workpiece 25 until the workpiece 25 is pressed
against the upturned lip 730. This mechanical engagement, along
with the flow of purging gas, effectively isolates the outer
periphery and backside of the workpiece 25 as well as the Bellville
ring contact 610 from contact with the plating solution.
[0102] Other similar contact assembly designs that have a large
number of contacts and that isolate the contacts from the
electroplating environment are likewise suitable for use in the
disclosed reactor assembly. Such additional contact assembly
designs are set forth, for example, in PCT Application ______,
filed Jul. 9, 1999 (Attorney Docket No. SEM4492P0571PC), which is
hereby incorporated by reference.
Plating Bath Filtration System
[0103] While platinum will plate over the barrier layer the
platinum will not readily adhere to materials preferably used for
the barrier layer (i.e. Ti, Ta, TaN, TiO.sub.2, TiAlN). As a
result, the plated platinum tends to flake off of the barrier layer
and pollute the electroplating solution. By limiting the formation
of plated platinum on the barrier layer, a significant source of
platinum flakes is substantially reduced.
[0104] Prior to supplying the recirculated solution to the plating
module, the solution is filtered so as to limit pollutants in the
solution, like platinum flakes. The filtered particles will
eventually clog the filter and the filter will need to be replaced.
By limiting the flaking of plated material the operational life of
the filter is extended. So as to further extend the operational
life of the filter, and/or in instances where platinum is allowed
to form on the barrier layer, the use of a cascaded filter 201 has
been determined to be beneficial. As illustrated in FIG. 16A,
preferably a three-stage filter is used between the electroplating
reactor 20 and the plating solution source tank 22. In the
illustrated embodiment, the first stage 202 provides filtration of
particles of a first predetermined size or larger, the second stage
203 provides filtration of particles of a second predetermined size
or larger, and the third stage 204 provides filtration of particles
of a third predetermined size or larger. In the preferred
embodiment, the first stage 202 provides filtration of particles
4.5 .mu.m or larger, the second stage 203 provides filtration of
particles 1.0 .mu.m or larger, and the third stage 204 provides
filtration of particles 0.1 .mu.m or larger.
Rotor Contact Conection Assembly
[0105] In many instances, it may be desirable to have a given
reactor assembly 20 function to execute a wide range of noble metal
electroplating recipes. Execution of a wide range of electroplating
recipes may be difficult, however, if the process designer is
limited to using a single contact assembly construction. Further,
the plating contacts used in a given contact assembly construction
must be frequently inspected and, sometimes, replaced. This is
often difficult to do in existing electroplating reactor tools,
frequently involving numerous operations to remove and/or inspect
the contact assembly. This problem may be addressed by providing a
mechanism by which the contact assembly 85 is readily attached and
detached from the other components of the rotor assembly 75.
Further, a given contact assembly type can be replaced with the
same contact assembly type without re-calibration or readjustment
of the system.
[0106] To be viable for operation in a manufacturing environment,
such a mechanism must accomplish several functions including:
[0107] 1. Provide secure, fail-safe mechanical attachment of the
contact assembly to other portions of the rotor assembly;
[0108] 2. Provide electrical interconnection between the contacts
of the contact assembly and a source of electroplating power;
[0109] 3. Provide a seal at the electrical interconnect interface
to protect against the processing environment (e.g., wet chemical
environment);
[0110] 4. Provide a sealed path for the purge gas that is provided
to the contact assembly; and
[0111] 5. Minimize use of tools or fasteners which can be lost,
misplaced, or used in a manner that damages the electroplating
equipment.
[0112] FIGS. 16B and 17 illustrate one embodiment of a quick-attach
mechanism that meets the foregoing requirements. For simplicity,
only those portions of the rotor assembly 75 necessary to
understanding the various aspects of the quick-attach mechanism are
illustrated in these figures.
[0113] As illustrated, the rotor assembly 75 may be comprised of a
rotor base member 1205 and a removable contact assembly 1210.
Preferably, the removable contact assembly 1210 is constructed in
the manner set forth above in connection with contact assembly 85.
The illustrated embodiment, however, employs a continuous ring
contact. It will be recognized that both contact assembly
constructions are suitable for use with the quick-attachment
mechanism set forth herein.
[0114] The rotor base member 1205 is preferably annular in shape to
match the shape of the semiconductor wafer 25. A pair of latching
mechanisms 1215 are disposed at opposite sides of the rotor base
member 205. Each of the latching mechanisms 1215 includes an
aperture 1220 disposed through an upper portion thereof that is
dimensioned to receive a corresponding electrically conductive
shaft 1225 that extends downward from the removable contact
assembly 1210.
[0115] The removable contact assembly 1210 is shown in a detached
state in FIG. 16B. To secure the removable contact assembly 1210 to
the rotor base member 1205, an operator aligns the electrically
conductive shafts 1225 with the corresponding apertures 1220 of the
latching mechanisms 1215. With the shafts 1225 aligned in this
manner, the operator urges the removable contact assembly 1210
toward the rotor base member 1205 so that the shafts 1225 engage
the corresponding apertures 1220. Once the removable contact
assembly 1210 is placed on the rotor base member 1205, latch arms
1230 are pivoted about a latch arm axis 1235 so that latch arm
channels 1240 of the latch arms 1230 engage the shaft portions 1245
of the conductive shafts 1235 while concurrently applying a
downward pressure against flange portions 1247. This downward
pressure secures the removable contact assembly 1210 with the rotor
base assembly 1205. Additionally, as will be explained in further
detail below, this engagement results in the creation of an
electrically conductive path between electrically conductive
portions of the rotor base assembly 1205 and the electroplating
contacts of the contact assembly 1210. It is through this path that
the electroplating contacts of the contact assembly 1210 are
connected to receive power from a plating power supply.
[0116] FIGS. 18A and 18B illustrate further details of the latching
mechanisms 1215 and the electrically conductive shafts 1225. As
illustrated, each latching mechanism 1215 is comprised of a latch
body 1250 having aperture 1220, a latch arm 1230 disposed for
pivotal movement about a latch arm pivot post 1255, and a safety
latch 1260 secured for relatively minor pivotal movement about a
safety latch pivot post 1265. The latch body 1250 may also have a
purge port 270 disposed therein to conduct a flow of purging fluid
to corresponding apertures of the removable contact assembly 210.
An O-ring 275 is disposed at the bottom of the flange portions of
the conductive shafts 1225
[0117] FIGS. 19A-19C are cross-sectional views illustrating
operation of the latching mechanisms 1215. As illustrated, latch
arm channels 1240 are dimensioned to engage the shaft portions 1245
of the conductive shafts 1225. As the latch arm 1230 is rotated to
engage the shaft portions 1245, a nose portion 1280 of the latch
arm 1230 cams against the surface 1285 of safety latch 1260 until
it mates with channel 1290. With the nose portion 1280 and
corresponding channel 1290 in a mating relationship, latch arm 1230
is secured against inadvertent pivotal movement that would
otherwise release removable contact assembly 1210 from secure
engagement with the rotor base member 1205.
[0118] FIGS. 20A-20D are cross-sectional views of the rotor base
member 1205 and removable contact assembly 1210 in an engaged
state. As can be seen in these cross-sectional views, the
electrically conductive shafts 1225 include a centrally disposed
bore 1295 that receives a corresponding electrically conductive
quick-connect pin 1300. It is through this engagement that an
electrically conductive path is established between the rotor base
member 1205 and the removable contact assembly 1210.
[0119] As also apparent from these cross-sectional views, the
lower, interior portion of each latch arm 1230 includes a
corresponding channel 1305 that is shaped to engage the flange
portions 1247 of the shafts 1225. Edge portions of channel 1305 cam
against corresponding surfaces of the flange portions 1247 to drive
the shafts 1225 against surface 1310 which, in turn, effects a seal
with O-ring 1275.
Rotor Contact Drive
[0120] As illustrated in FIGS. 21, 22 and 23, the rotor assembly 75
includes an actuation arrangement whereby the wafer or other
workpiece 25 is received in the rotor assembly by movement in a
first direction, and is thereafter urged into electrical contact
with the contact assembly by movement of a backing member 310
toward the contact assembly, in a direction perpendicular to the
first direction.
[0121] As illustrated, the stationary assembly 70 of the reactor
head 30 includes a motor assembly 1315 that cooperates with shaft
1320 of rotor assembly 75. Rotor assembly 75 includes a generally
annular housing assembly, including rotor base member 1205 and an
inner housing 1320. As described above, the contact assembly is
secured to rotor base member 1205. By this arrangement, the housing
assembly and the contact assembly 1210 together define an opening
1325 through which the workpiece 25 is transversely movable, in a
first direction, for positioning the workpiece in the rotor
assembly 75. The rotor base member 1205 preferably defines a
clearance opening for the robotic arm as well as a plurality of
workpiece supports 3130 upon which the workpiece is positioned by
the robotic arm after the workpiece is moved transversely into the
rotor assembly by movement through opening 1325. The supports 1330
thus support the workpiece 25 between the contact assembly 1210 and
the backing member 1310 before the backing member engages the
workpiece and urges it against the contact ring.
[0122] Reciprocal movement of the backing member 1310 relative to
the contact assembly 1210 is effected by at least one spring which
biases the backing member toward the contact assembly, and at least
one actuator for moving the backing member in opposition to the
spring. In the illustrated embodiment, the actuation arrangement
includes an actuation ring 1335 which is operatively connected with
the backing member 1310, and which is biased by a plurality of
springs, and moved in opposition to the springs by a plurality of
actuators.
[0123] With particular reference to FIG. 21, actuation ring 1335 is
operatively connected to the backing member 1310 by a plurality
(three) of shafts 1340. The actuation ring, in turn, is biased
toward the housing assembly by three compression coil springs 1345
which are each held captive between the actuation ring and a
respective retainer cap 350. By this arrangement, the action of the
biasing springs 1345 urges the actuation ring 1335 in a direction
toward the housing, with the action of the biasing springs thus
acting through shafts 1340 to urge the backing member 1335 in a
direction toward the contact assembly 1210. The drive shaft 1360 is
operatively connected to inner housing 1320 for effecting rotation
of workpiece 25, as it is held between contact assembly 210 and
backing member 310, during plating processing. The drive shaft 360,
in turn, is driven by motor 315 that is disposed in the stationary
portion of the reactor head 30.
[0124] Rotor assembly 75 is preferably detachable from the
stationary portion of the reactor head 30 to facilitate maintenance
and the like. Thus, drive shaft 1360 is detachably coupled with the
motor 1315. In accordance with the preferred embodiment, the
arrangement for actuating the backing member 1310 also includes a
detachable coupling, whereby actuation ring 1335 can be coupled and
uncoupled from associated actuators which act in opposition to
biasing springs 1345.
[0125] Actuation ring 1335 includes an inner, interrupted coupling
flange 1365. Actuation of the actuation ring 1335 is effected by an
actuation coupling 1370 of the stationary assembly 70, which can be
selectively coupled and uncoupled from the actuation ring 1335. The
actuation coupling 1370 includes a pair of flange portions 1375
which can be interengaged with coupling flange 1365 of the
actuation ring 1335 by limited relative rotation therebetween. By
this arrangement, the actuation ring 1335 of the rotor assembly 75
can be coupled to, and uncoupled from, the actuation coupling 1370
of the stationary assembly 70 of the reactor head 30.
[0126] Actuation coupling 370 is movable in a direction in
opposition to the biasing springs 1345 by a plurality of pneumatic
actuators 1380 mounted on a frame of the stationary assembly 70.
Each actuator 1380 is operatively connected with the actuation
coupling 1370 by a respective drive member 1385, each of which
extends generally through the frame of the stationary assembly
70.
[0127] There is a need to isolate the foregoing mechanical
components from other portions of the reactor assembly 20. A
failure to do so will result in contamination of the processing
environment (here, a wet chemical electroplating environment).
Additionally, depending on the particular process implemented in
the reactor 20, the foregoing components can be adversely affected
by the processing environment.
[0128] To effect such isolation, a bellows assembly 1390 is
disposed to surround the foregoing components. The bellows assembly
1390 comprises a bellows member 1395, preferably made from Teflon,
having a first end thereof secured at 1400 and a second end thereof
secured at 1405. Such securement is preferably implemented using
the illustrated liquid-tight, tongue-and-groove sealing
arrangement. The convolutes 1410 of the bellows member 1395 flex
during actuation of the backing plate 1310.
Wafer Loading/Processing Operations
[0129] Operation of the reactor head 30 will be appreciated from
the above description. Loading of workpiece 25 into the rotor
assembly 75 is effected with the rotor assembly in a generally
upwardly facing orientation, such as illustrated in FIG. 3.
Workpiece 25 is moved transversely through the opening 325 defined
by the rotor assembly 75 to a position wherein the workpiece is
positioned in spaced relationship generally above supports 1330. A
robotic arm 418 is then lowered (with clearance opening 325
accommodating such movement), whereby the workpiece is positioned
upon the supports 1330. The robotic arm 418 can then be withdrawn
from within the rotor assembly 75.
[0130] The workpiece 25 is now moved perpendicularly to the first
direction in which it was moved into the rotor assembly. Such
movement is effected by movement of backing member 1310 generally
toward contact assembly 1210. It is presently preferred that
pneumatic actuators 1380 act in opposition to biasing springs 1345
which are operatively connected by actuation ring 1335 and shafts
1340 to the backing member 1310. Thus, actuators 1380 are operated
to permit springs 1345 to bias and urge actuation ring 1335 and,
thus, backing member 1310, toward contact 210. FIG. 22 illustrates
the disposition of the reactor head 30 in a condition in which it
may accept a workpiece, while FIG. 21 illustrates the disposition
of the reactor head in a condition in which it is ready to present
the workpiece to the reactor bowl 35.
[0131] In the preferred form, the connection between actuation ring
1335 and backing member 1310 by shafts 1340 permits some "float".
That is, the actuation ring and backing member are not rigidly
joined to each other. This preferred arrangement accommodates the
common tendency of the pneumatic actuators 1380 to move at slightly
different speeds, thus assuring that the workpiece is urged into
substantial uniform contact with the electroplating contacts of the
contact assembly 1210 while avoiding excessive stressing of the
workpiece, or binding of the actuation mechanism.
[0132] With the workpiece 25 firmly held between the backing member
1310 and the contact assembly 1210, lift and rotate apparatus 80
rotates the reactor head 30 and lowers the reactor head into a
cooperative relationship with reactor bowl 35 so that the surface
of the workpiece is placed in contact with the surface of the
plating solution (i.e., the meniscus of the plating solution)
within the reactor vessel. FIG. 1 illustrates the apparatus in this
condition. If a contact assembly such as contact assembly 85 is
used in the reactor 20, the contact assembly 85 seals the entire
peripheral region of the workpiece. Depending on the particular
electroplating process implemented, it may be useful to insure that
any gas which accumulates on the surface of the workpiece is
permitted to vent and escape. Accordingly, the surface of the
workpiece may be disposed at an acute angle, such as on the order
of two degrees from horizontal, with respect to the surface of the
solution in the reactor vessel. This facilitates venting of gas
from the surface of the workpiece during the plating process as the
workpiece, and associated backing and contact members, are rotated
during processing. Circulation of plating solution within the
reactor bowl 35, as electrical current is passed through the
workpiece and the plating solution, effects the desired
electroplating of the noble metal or noble metal alloy on the
surface of the workpiece.
[0133] A number of features of the present reactor facilitate
efficient and cost-effective electroplating of a noble metal or
noble metal alloy on workpieces such as semiconductor wafers. By
use of a contact assembly having substantially continuous contact
in the form of a large number of sealed, compliant discrete contact
regions, a high number of plating contacts are provided while
minimizing the required number of components. The actuation of the
backing member 1310 is desirably effected by a simple linear
motion, thus facilitating precise positioning of the workpiece, and
uniformity of contact with the contact ring. The isolation of the
moving components using a bellows seal arrangement further
increases the integrity of the electroplating process.
[0134] Maintenance and configuration changes are easily facilitated
through the use of the detachable contact assembly 1210. Further,
maintenance is also facilitated by the detachable configuration of
the rotor assembly 75 from the stationary assembly 70 of the
reactor head. The contact assembly provides excellent distribution
of electroplating power to the surface of the workpiece, while the
preferred provision of the peripheral isolation region protects the
contacts from the plating environment (e.g., contact with the
plating solution), thereby desirably preventing build-up of noble
metal onto the electrical contacts. The perimeter seal also
desirably prevents plating onto the peripheral portion of the
workpiece.
Current Thieving In Noble Metal Plating Reactors
[0135] FIG. 24 illustrates an embodiment of a current thief that
may be used in the plating of noble metals, such as platinum, to
enhance the uniformity of the plated film. The embodiment of the
current thief illustrated here may be exposed to the electroplating
solution and may be used in conjunction with one or more of the
contact assemblies described above or with a plurality of discrete
figure contacts such as those described below. Current thieving can
be particularly useful where a large number of discrete electrical
contacts is not practical. Beneficial features of current thieving
are discussed in connection with U.S. Patent Application Ser. No.
08/933,450, similarly assigned to Semitool, Inc., the disclosure of
which is incorporated herein by reference.
[0136] However, in an electroplating environment providing for the
deposition of noble metals, certain difficulties associated with
the use of current thieves are experienced. One such difficulty is
that certain noble metals, like platinum, once plated cannot be
readily deplated. The present inventors have addressed this problem
and have developed a segmented current thief 415, illustrated in
FIG. 24, that is suitable for use in the plating of noble metals,
such as platinum. The segmented current thief 415 provides for
multiple pads 420 located about the periphery of the semiconductor
wafer 25. Each of the pads 420 can be individually provided with a
controlled amount of electroplating power to promote uniform
current densities and/or uniform deposition of plated material.
[0137] In operation, current thief 415 is a contact with the noble
metal plating solution. Plating material will therefore plate the
pads 420. As a result, the current thief 415 has a limited useful
life before the plating material accumulates to a degree in which
it begins to interfere with the optimal plating process parameters.
Accordingly, the current thief 415 is designed to be readily
manufactured from inexpensive materials and, as such, is
disposable. To this end, current thief 415 is comprised of a
printed circuit board with the individual pads separately formed on
the printed circuit substrate. Such a current thief 415 could be
produced relatively inexpensively, and changed as necessary as part
of the regular maintenance. The lower costs of the current thief
would help mitigate the expense of more frequent replacement.
[0138] In instances where discrete finger contacts are used,
current thieving may be provided by a portion of the discrete
finger contact. FIG. 25 shows an example of a discrete finger
contact 425. Portions 430 of the finger 425 may have exposed metal
for performing a current thieving function.
[0139] Because the finger contact 425 often may not be readily
replaced with an inexpensive alternative, the finger contacts 425
includes multiple separate conductive wrap layers 435, only one of
which will be exposed at any given time. After sufficient build up
of deposited material has accumulated one of the conductive wrap
layers 435 may be individually removed, exposing a fresh wrap layer
underneath. As the individual wrap layers 435 are removed, the
accumulation of deposited material is removed with it. In this way
the useful life of the finger contact 425 is recycled or
renewed.
Integrated Noble Metal Plating Tool
[0140] FIGS. 26 through 28 are top plan views of integrated
processing tools, shown generally at 1450, 1455, and 1500 that may
be used to deposit a noble metal on the surface of a
microelectronic workpiece, such as a semiconductor wafer.
Processing tools 1450 and 1455 are each based on tool platforms
developed by Semitool, Inc., of Kalispell, Mont. The processing
tool platform of the tool 1450 is sold under the trademark
LT-210.TM., the processing tool platform of the tool 1455 is sold
under the trademark LT-210C.TM., and the processing tool 1500 is
sold under the trademark EQUINOX.TM.. The principal difference
between the tools 1450, 1455 is in the footprints required for
each. The platform on which tool 1455 is based has a smaller
footprint than the platform on which tool 1455 is based.
Additionally, the platform on which tool 1450 is based is
modularized and may be readily expanded. Each of the processing
tools 1450, 1455, and 1500 are computer programmable to implement
user entered processing recipes.
[0141] Each of the processing tools 1450, 1455, and 1500 include an
input/output section 1460, a processing section 1465, and one or
more robots 1470. The robots 1470 for the tools 1450, 1455 move
along a linear track. The robot 1470 for the tool 1500 is centrally
mounted and rotates to access the input/output section 1460 and the
processing section 1465. Each input/output section 1460 is adapted
to hold a plurality of workpieces, such as semiconductor wafers, in
one or more workpiece cassettes. Processing section 1465 includes a
plurality of processing stations 1475 that are used to perform one
or more fabrication processes on the semiconductor wafers. The
robots 1470 are used to transfer individual wafers from the
workpiece cassettes at the input/output section 1460 to the
processing stations 1475, as well as between the processing
stations 1475.
[0142] One or more of the processing stations 1475 are configured
as electroplating assemblies, such as the electroplating assembly
described above, for electroplating a noble metal, such as
platinum, onto the semiconductor wafers. For example, each of the
processing tools 1450 and 1455 may include eight noble metal
plating reactors and a single pre-wet/rinse station. The
pre-wet/rinse station is preferably one of the type available from
Semitool, Inc. Preferably, one of the stations may be configured to
execute a pre-wet/rinse process, and one of the stations may be
configured as a spin rinser/dryer (SRD). Further, one or more of
the processing chambers can be configured as an annealing station
that can be used to anneal the noble metal layer. It will now be
recognized that a wide variation of processing station
configurations may be used in each of the individual processing
tools 1450, 1455 and 1500 to execute pre-noble metal electroplating
and post-noble metal electroplating processes. As such, the
foregoing configurations are merely illustrative of the variations
that may be used.
Alternative Processing Container
[0143] FIG. 29 illustrates the basic construction of an alternative
processing container 35 and the corresponding flow velocity contour
pattern resulting from the processing container construction. As
illustrated, the processing container 35 generally comprises a main
fluid flow chamber 2505, an antechamber 2510, a fluid inlet 2515, a
plenum 2520, a flow guide 2525 separating the plenum 2520 from the
antechamber 2510, and a nozzle/slot assembly 2530 separating the
plenum 2520 from the main chamber 2505. These components cooperate
to provide a flow (here, of the electroplating solution) at the
wafer 25 with a substantially radially independent normal
component. In the illustrated embodiment, the impinging flow is
centered about central axis 2535 and possesses a nearly uniform
component normal to the surface of the wafer 25. This results in a
substantially uniform mass flux to the wafer surface that, in turn,
enables substantially uniform processing thereof.
[0144] Electroplating solution is provided through inlet 2515
disposed at the bottom of the container 35. The fluid from the
inlet 2515 is directed therefrom at a relatively high velocity
through antechamber 2510. In the illustrated embodiment,
antechamber 2510 includes an accelerated region 2540 through which
the electroplating solution flows radially from the fluid inlet
2515 toward fluid flow region 2545 of antechamber 2510. Fluid flow
region 2545 has a generally inverted U-shaped cross-section that is
substantially wider at its outlet region proximate flow guide 2525
than at its inlet region proximate region 2540. This variation in
the cross-section assists in removing any gas bubbles from the
electroplating solution before the electroplating solution is
allowed to enter the main chamber 2505. Gas bubbles that would
otherwise enter the main chamber 2505 are allowed to exit the
processing container 35 through a gas outlet (not illustrated in
FIG. 29, but illustrated in the embodiment shown in FIGS. 30-32)
disposed at an upper portion of the antechamber 2510.
[0145] Electroplating solution within antechamber 2510 is
ultimately supplied to main chamber 2505. To this end, the
electroplating solution is first directed to flow from a relatively
high-pressure region 2550 of the antechamber 2510 to the
comparatively lower-pressure plenum 2520 through flow guide 2525.
Nozzle assembly 2530 includes a plurality of nozzles or slots 2555
that are disposed at a slight angle with respect to horizontal.
Electroplating solution exits plenum 2520 through nozzles 2555 with
fluid velocity components in the horizontal, vertical and radial
directions.
[0146] Main chamber 2505 is defined at its upper region by a
contoured sidewall 2560 and a slanted sidewall 2565. The contoured
sidewall 2560 assists in preventing fluid flow separation as the
electroplating solution exits nozzles 2555 (particularly the
uppermost nozzle(s)) and turns upward toward the surface of wafer
25. Beyond breakpoint 2570, fluid flow separation will not
substantially affect the uniformity of the normal flow. As such,
sidewall 2565 can generally have any shape, including a
continuation of the shape of contoured sidewall 2560. In the
specific embodiment disclosed here, sidewall 2565 is slanted and,
as will be explained in further detail below, is used to support
one or more anodes.
[0147] Electroplating solution exits from main chamber 2505 through
a generally annular outlet 2570. Fluid exiting outlet 2570 may be
provided to a further exterior chamber for disposal or may be
replenished for re-circulation through the electroplating solution
supply system.
[0148] In those instances in which the processing container 35
forms part of an electroplating reactor, the processing container
35 is provided with one or more anodes. In the illustrated
embodiment, a principal anode 2580 is disposed in the lower portion
of the main chamber 2505. If the peripheral edges of the surface of
the wafer 25 extend radially beyond the extent of contoured
sidewall 2560, then the peripheral edges are electrically shielded
from principal anode 2580 and reduced plating will take place in
those regions. However, if plating is desired in the peripheral
regions, one or more further anodes may be employed proximate the
peripheral regions. Here, a plurality of annular anodes 2585 are
disposed in a generally concentric manner on slanted sidewall 2565
to provide a flow of electroplating current to the peripheral
regions. An alternative embodiment would include a single anode or
multiple anodes with no shielding from the contoured walls to the
edge of the wafer.
[0149] The anodes 2580, 2585 may be provided with electroplating
power in a variety of manners. For example, the same or different
levels of electroplating power may be multiplexed to the anodes
2580, 2585. Alternatively, all of the anodes 2580, 2585 may be
connected to receive the same level of electroplating power from
the same power source. Still further, each of the anodes 2580, 2585
may be connected to receive different levels of electroplating
power to compensate for the variations in the resistance of the
plated film. An advantage of the close proximity of the anodes 2585
to the wafer 25 is that it provides a high degree of control of the
radial film growth resulting from each anode.
[0150] Anodes 2580, 2585 may be consumable, but are preferably
inert and formed from platinized titanium or some other inert
conductive material. However, as noted above, inert anodes tend to
evolve gases that can impair the uniformity of the plated film. To
reduce this problem, as well as to reduce the likelihood of the
entry of bubbles into the main processing chamber 2505, processing
container 35 includes several unique features. With respect to
anode 2580, a small fluid flow path 2590 is provided between the
underside of anode 2580 and antechamber 2510. This results in a
Venturi effect that causes the electroplating solution proximate
the surfaces of anode 2580 to be drawn into antechamber 2510 and,
further, provides a suction flow that affects the uniformity of the
impinging flow at the central portion of the surface of the wafer.
Gas bubbles forming at the surfaces of anode 2580 are thus swept
into antechamber 2510 and are prevented from entering main chamber
2505. Rather than entering main chamber 2505 where they would
disturb the boundary layer conditions at the surface of wafer 25,
the gas bubbles enter antechamber 2510 and exit the gas outlet at
the upper region of antechamber 2510. The Venturi flow path 2590
may be shielded to prevent any large bubbles originating from
outside the chamber from rising through region 2590. Instead, such
bubbles enter the bubble-trapping region of the antechamber 2510.
Similarly, electroplating solution sweeps across the surfaces of
anodes 2585 in a radial direction toward fluid outlet 2570 to
remove gas bubbles forming at their surfaces. Further, the radial
components of the fluid flow at the surface of the wafer assists
and sweeping gas bubbles therefrom.
[0151] The foregoing reactor design effectively de-couples the
fluid flow from adjustments to the electric field. This occurs due
to the absence of a diffuser disposed between the anode and the
cathode (workpiece). Further, the use of multiple anodes
contributes to this result as well. An advantage of this approach
is that a chamber with nearly ideal flow for electroplating and
other processes (i.e., a design which provide substantial uniform
diffusion layer across the wafer) may be designed that will not be
degraded when electroplating or other process applications require
significant changes to the electric field.
[0152] There are numerous processing advantages with respect to the
illustrated flow through the reactor chamber. As illustrated, the
flow through the various system components is directed away from
the wafer surface and, as such, there are no jets of fluid created
to disturb the uniformity of the diffusion layer. Although the
diffusion layer may not be perfectly uniform, any non-uniformity
will be relatively gradual as a result.
[0153] As is also evident from the foregoing reactor design, the
flow that is normal to the wafer has greater a magnitude near the
center of the wafer and creates a dome-shaped meniscus. The
dome-shaped meniscus assists in minimizing bubble entrapment as the
wafer or other workpiece is lowered into the processing solution
(here, the electroplating solution). The flow pattern resulting in
the dome-shaped meniscus is influenced by the Venturi flow at the
bottom of the chamber 2505. This flow at the bottom of the main
chamber 2505 influences the flow at the centerline thereof. The
centerline flow velocity is otherwise difficult to implement and
control. However, the strength of the Venturi flow provides a
non-intrusive design variable that may be used to affect this
aspect of the flow.
[0154] A still further advantage of the foregoing reactor design is
that it assists in preventing bubbles that find their way into the
main chamber from reaching the wafer. To this end, the flow pattern
is such that the solution travels downward just before entering the
main chamber. As such, bubbles remain in the antechamber and escape
through holes at the top thereof. Further, bubbles are prevented
from entering the main chamber through the Venturi flow path
through the use of the shield that covers the Venturi flow path
(see description of the embodiment of the reactor illustrated in
FIGS. 30-32). Still further, the upward sloping inlet path (see
FIG. 32 and appertaining description) to the antechamber prevents
bubbles from entering the main chamber through the Venturi flow
path.
[0155] There are also advantages associated with the electric field
in the foregoing reactor design. Multiple concentric anodes are
used so that a uniform film can be plated by making adjustments to
the current passing through each anode. Generally, the more
resistive the plated film, the more the magnitude of the current at
the central anodes should be increased to yield a uniform film.
Some further reasons for adjusting the electric field include
changes to the following:
[0156] seed layer thickness;
[0157] open area of plating surface (pattern wafers, edge
exclusion);
[0158] final plated thickness;
[0159] bath conductivity, metal concentration; and
[0160] plating rate.
[0161] The particular reactor embodiment disclosed herein is
readily adapted to compensate for the foregoing changes.
[0162] FIGS. 30-32 illustrate a specific construction of a complete
processing chamber assembly 2610. As illustrated, assembly 2610 is
comprised of the processing container 35 shown in FIG. 29 along
with a corresponding exterior cup 2605. Processing container 35 is
disposed within exterior cup 2605 to allow exterior cup 2605 to
receive spent electroplating solution that overflows from the
processing container 35. A flange 2615 extends about the assembly
2610 for securement with, for example, the frame of the
corresponding tool.
[0163] With particular reference to FIGS. 31 and 32, the flange of
the exterior cup assembly 2605 is formed to engage or otherwise
accept rotor portion 75 of head assembly 25 and allow contact
between the wafer 25 and the processing solution, such as
electroplating solution, in the main chamber 2505. The exterior cup
assembly 2605 also includes a main cylindrical housing 2625 into
which a drain cup member 2627 is disposed. The drain cup member
2627 includes an outer surface having channels 2629 that, together
with the interior wall of housing 2625, form one or more helical
flow chambers 2640 that serve as an outlet for the processing
solution. Electroplating solution overflowing a weir member 2739 at
the top of processing cup 35 drains through the helical flow
chambers 2640 and exits an outlet (not illustrated) where it is
either disposed of or replenished and re-circulated. This
configuration is particularly suitable for systems that include
fluid re-circulation since it assists in reducing the mixing of
gases with the processing solution thereby further reducing the
likelihood that gas bubbles will interfere with the uniformity of
the diffusion layer at the workpiece surface.
[0164] In the illustrated embodiment, antechamber 2550 is defined
by the walls of a plurality of separate components. More
particularly, antechamber 2550 is defined by the interior walls of
drain cup member 2627, an anode support member 2697, the interior
and exterior walls of a mid-chamber member 2690, and the exterior
walls of flow guide 2550.
[0165] FIG. 31 illustrates the manner in which the foregoing
components are brought together to form the reactor. To this end,
the mid-chamber member 2690 is disposed interior of the drain cup
member 2627 and includes a plurality of leg supports 2692 that sit
upon a bottom wall thereof. The anode support member 2697 includes
an outer wall that engages a flange 630 that is disposed about the
interior of drain cup member 2627. The anode support member 2697
also includes a channel 2705 that sits upon and engages an upper
portion of flow guide 2550, and a further channel 2710 that sits
upon and engages an upper rim of nozzle assembly 2530. Mid-chamber
member 2690 also includes a centrally disposed annular receptacle
2715 that is dimensioned to accept the lower portion of nozzle
assembly 2530. Likewise, an annular channel 2725 is disposed
radially exterior of the annular receptacle 2715 to engage a lower
portion of flow guide 550.
[0166] In the illustrated embodiment, the flow guide 2550 is formed
as a single piece and includes a plurality of vertically oriented
slots 2670. Similarly, the nozzle assembly 2530 is formed as a
single piece and includes a plurality of horizontally oriented
slots that constitute the nozzles 2555.
[0167] The anode support assembly 2697 includes a plurality of
annular grooves that are dimensioned to accept corresponding
annular anode assemblies 2785. Each anode assembly 2785 includes an
anode 2585 (preferably formed from platinized titanium or in other
inert metal) and a conduit 2730 extending from a central portion of
the anode 2585 through which a metal conductor may be disposed to
electrically connect the anode 2585 of the assembly 2785 to an
external source of electrical power. Conduit 2730 is shown to
extend entirely through the reactor assembly 2610 and is secured at
the bottom thereof by a respective fitting 2733. In this manner,
anode assemblies 2785 effectively urge the anode support member
2697 downward to clamp the flow guide 2550, nozzle member 2530,
mid-chamber member 2690, and drain cup member 2627 against the
bottom portion 2737 of the housing assembly 2605. This allows for
easy assembly and disassembly of the reactor 2610.
[0168] The illustrated embodiment also includes a weir member 2739
that detachably snaps or otherwise easily secures to the upper
exterior portion of anode support member 2697. As shown, weir
member 2739 includes a rim 2742 that forms a weir over which the
processing solution flows into the helical flow chamber 2640. Weir
member 2739 also includes a transversely extending flange 2744 that
extends radially inward and forms an electric field shield over all
or portions of one or more of the anodes 2585. Since the weir
member 2739 may be easily removed and replaced, the reactor
assembly 2610 may be readily reconfigured and adapted to provide
different electric field shapes. Such differing electrical field
shapes are particularly useful in those instances in which the
reactor must be configured to process more than one size or shape
of a workpiece.
[0169] The anode support member 2697, with the anodes 2727 in
place, forms the contoured wall 2560 and slanted wall 2565 that is
illustrated in FIG. 29. As noted above, the lower region of anode
support member 2697 is contoured to define the upper interior wall
of antechamber 2510 and preferably includes one or more gas outlets
2665 that are disposed therethrough to allow gas bubbles to exit
from the antechamber 2510 to the exterior environment.
[0170] With particular reference to FIG. 32, inlet 2515 is defined
by an inlet fluid guide, shown generally at 2810, that is secured
to the floor of drain cup member 2627 by one or more fasteners
2815. Inlet fluid guide 2810 includes a plurality of open channels
2817 that guide fluid received at inlet 2515 to an area beneath
mid-chamber member 2690. Channels 2817 of the illustrated
embodiment are defined by upwardly angled walls 2819.
Electroplating solution exiting channels 2817 flows therefrom to
one or more further channels 2821 that are likewise defined by
walls that angle upward.
[0171] Central anode 2580 includes an electrical connection rod
2581 that proceeds to the exterior of the reactor assembly 2610
through central apertures formed in nozzle member 2550, drain cup
member 2627 and inlet fluid guide 2810. The small fluid regions
shown at 2590 in FIG. 29 are formed in FIG. 32 by vertical channels
2823 that proceed through drain cup member 2627 and the bottom wall
of nozzle member 2550. The vertical channels 2823 of the drain cup
member 2627 are separated from the vertical channels 2823 at the
bottom wall 2825 of nozzles member 2550 by an intermediate chamber
2827 that is defined by the exterior portion of bottom wall 2825
and the interior wall at the bottom of drain cup member 2627. As
illustrated, the exterior portion of bottom wall 2825 extends at a
downward angle from a central region thereof. This construction
assists in preventing bubbles from entering the main chamber 2505
since any bubbles reaching vertical channel 2823 of drain cup
member 2627 will proceed into chamber 2827 and flow to the upper
portions thereof proximate connection rod 2581 without entering
main chamber 2505.
[0172] Numerous modifications may be made to the foregoing system
without departing from the basic teachings thereof. Although the
present invention has been described in substantial detail with
reference to one or more specific embodiments, those of skill in
the art will recognize that changes may be made thereto without
departing from the scope and spirit of the invention as set forth
in the appended claims.
[0173] While the preferred embodiment of the invention has been
illustrated and described, it will be appreciated that various
changes can be made therein without departing from the spirit and
scope of the invention.
[0174] The embodiments of the invention in which an exclusive
property or privilege is claimed are defined as follows:
* * * * *