U.S. patent application number 13/907742 was filed with the patent office on 2014-12-04 for tensile stressed doped amorphous silicon.
The applicant listed for this patent is Novellus Systems, Inc.. Invention is credited to Keith Fox, Dong Niu, Joseph L. Womack.
Application Number | 20140357064 13/907742 |
Document ID | / |
Family ID | 51901748 |
Filed Date | 2014-12-04 |
United States Patent
Application |
20140357064 |
Kind Code |
A1 |
Fox; Keith ; et al. |
December 4, 2014 |
TENSILE STRESSED DOPED AMORPHOUS SILICON
Abstract
The method and apparatus disclosed herein relate to preparing a
stack structure for an electronic device on a semiconductor
substrate. A particularly beneficial application of the method is
in reduction of internal stress in a stack containing multiple
layers of silicon. Typically, though not necessarily, the internal
stress is a compressive stress, which often manifests as wafer bow.
In some embodiments, the method reduces the internal stress of a
work piece by depositing phosphorus doped silicon layers having low
internal compressive stress or even tensile stress. The method and
apparatus disclosed herein can be used to reduce compressive bow in
stacks containing silicon.
Inventors: |
Fox; Keith; (Tigard, OR)
; Niu; Dong; (West Linn, OR) ; Womack; Joseph
L.; (Tigard, OR) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Novellus Systems, Inc. |
Fremont |
CA |
US |
|
|
Family ID: |
51901748 |
Appl. No.: |
13/907742 |
Filed: |
May 31, 2013 |
Current U.S.
Class: |
438/485 ;
118/697 |
Current CPC
Class: |
H01L 21/02532 20130101;
H01L 21/02507 20130101; H01L 21/02694 20130101; H01L 21/02123
20130101; H01L 21/0217 20130101; H01L 27/11582 20130101; H01L
21/02274 20130101; H01L 21/02211 20130101 |
Class at
Publication: |
438/485 ;
118/697 |
International
Class: |
H01L 21/02 20060101
H01L021/02 |
Claims
1. A method of preparing a stack structure for an electronic device
on a semiconductor substrate, the method comprising: (a) depositing
a doped silicon layer on the substrate by introducing a first
process gas comprising a silicon hydride and a dopant precursor
into a chemical vapor deposition (CVD) apparatus having a plasma,
wherein the substrate is held at a temperature of about 450.degree.
C. to 550.degree. C., the silicon hydride in the first process gas
is introduced at a flow rate of at least about 0.01 sccm/cm.sup.2
of substrate deposition surface area, and the doped silicon layer
has a compressive stress of less than about 100 MPa or a tensile
stress, and a thickness of less than about 500 .ANG.; (b)
depositing a dielectric layer on the substrate by introducing a
second process gas comprising a precursor of the dielectric layer
into the CVD apparatus having a plasma, wherein the dielectric
layer has a compressive stress of at least about 200 MPa and a
thickness of less than about 500 .ANG.; and (c) repeating
operations (a) and (b) for two or more times to form a stack
structure.
2. The method of claim 1, wherein the silicon hydride comprises
silane.
3. The method of claim 2, wherein the dopant precursor comprises
phosphine.
4. The method of claim 3, wherein the silicon hydride flow rate in
(a) is at least about 1200 sccm or greater for a 300 mm-diameter
substrate, or at an equivalent flow rate scaled by substrate
deposition surface area.
5. The method of claim 3, wherein the silicon hydride flow rate in
(a) is at least about 0.038 sccm/cm.sup.2 of substrate deposition
surface area.
6. The method of claim 3, wherein the first process gas further
comprises helium, and the volume ratio of phosphine to helium is
about 5:95.
7. The method of claim 3, wherein depositing the doped silicon
layer in (a) is conducted at a pressure of between about 1 and 10
Torr.
8. The method of claim 3, wherein the plasma is provided by
powering a high frequency radio power source at about 500 watt or
higher.
9. The method of claim 1, wherein the roughness of the doped
silicon layer is about 5 .ANG. or less as measured by atomic force
microscopy.
10. The method of claim 1, wherein the stack structure comprises
amorphous or microcrystalline silicon, and the method further
comprises activating the amorphous or microcrystalline silicon by
heating, thereby converting the silicon to a polycrystalline
state.
11. The method of claim 1, wherein the bulk resistivity of the
doped silicon layer after activation is less than about 0.001
ohmcentimeter.
12. The method of claim 1, wherein the CVD apparatus is a
plasma-enhanced chemical vapor deposition (PECVD) apparatus.
13. The method of claim 1, wherein the stack structure has a total
thickness of between about 1 and 6 micrometers.
14. The method of claim 1, wherein the stack structure comprises at
least 50 layers.
15. The method of claim 1, wherein the stack structure is
configured to be part of a vertically integrated memory device.
16. The method of claim 1, wherein the stack structure comprises
alternating layers of doped silicon and a dielectric material.
17. The method of claim 1, wherein the stack structure comprises
alternating layers of doped and undoped silicon.
18. A system for preparing a stack structure for an electronic
device on a semiconductor substrate, the system comprising: (a) a
multi-chamber apparatus comprising at least one chemical vapor
deposition (CVD) chamber having a plasma for depositing at least
one layer of the stack over the semiconductor substrate; and (b) a
controller comprising instructions for: depositing a doped silicon
layer on the substrate by introducing a first process gas
comprising a silicon hydride and a dopant precursor into a CVD
chamber, wherein the substrate is held at a temperature of about
450.degree. C. to 550.degree. C., the silicon hydride in the
process gas is introduced at a flow rate of at least about 0.01
sccm/cm.sup.2 of substrate deposition surface area, and the doped
silicon layer has a compressive stress of less than about 100 MPa
or a tensile stress, and a thickness of less than about 500 .ANG.;
depositing a dielectric layer on the substrate by introducing a
second process gas comprising a precursor of the dielectric layer
into the CVD apparatus having a plasma, wherein the dielectric
layer has a compressive stress of at least about 200 MPa and a
thickness of less than about 500 .ANG..
19. The system of claim 18, the multi-chamber apparatus further
comprising a rapid thermal annealing chamber.
20. The system of claim 18, wherein the controller comprises
instructions for receiving the semiconductor substrate from outside
the system and transferring the substrate from one chamber to
another in the multi-chamber apparatus.
Description
BACKGROUND
[0001] Various thin film layers for semiconductor devices may be
deposited by chemical vapor deposition (CVD) and/or plasma-enhanced
chemical vapor deposition (PECVD) processes. Some memory devices,
such as NAND flash memory, are arranged in two-dimensional arrays.
Because such memory devices are limited to a planar arrangement,
die size and memory density parameters may constrain the total
memory capacity of the device. In turn, moving to larger die sizes
to expand memory capacity may comparatively increase the cost of
the memory device, which may delay adoption of larger capacity
memory devices. Certain approaches for arranging memory gates into
three-dimensional (3D) arrays have been proposed. Some of these
approaches incorporate transistors formed by patterning stacks of
alternating film composition. FIG. 1 schematically shows an example
film stack 1000 including alternating layers of first film 1020 and
second film 1040 on a substrate 1060. These 3D arrays are sometimes
utilized to form vertically integrated memory (VIM) structures.
[0002] Oftentimes, the unit layers used to form a VIM structure on
a substrate have an as-deposited internal stress. In many cases the
as-deposited internal stress is compressive. When many layers are
positioned on top of one another, this internal stress may build
up, thereby causing the coated substrate to bow instead of being
flat. This bowing is undesirable because it can make subsequent
processing more difficult. When the bow of a substrate exceeds a
certain level, the substrate may be unusable.
[0003] In addition to internal stress, the roughness of a stack has
an impact on the ability to conduct lithography reliably. If the
surface of a stack is too rough, the incident radiation employed in
photolithography will scatter too much. Also, the roughness of the
stack is determined by the maximum roughness of any given layer in
the stack. Therefore, it is important that each and every layer in
the stack have a relatively low roughness.
SUMMARY
[0004] The disclosure concerns stacks of oxide and silicon layers
such as those employed in vertically integrated memory stacks.
Certain embodiments herein relate to methods and apparatus for
reducing the bow in multilayer stacks fabricated on semiconductor
substrates. The methods produce silicon layers having tensile,
neutral, or low compressive internal stress. By comparison to other
methods, the disclosed methods reduce the internal stress in one or
more silicon layers of a work piece. A particularly beneficial
application of the process is in reduction of internal stress in a
stack containing two or more layers of dielectric material (e.g.,
silicon oxide) and two or more layers of doped silicon. Typically,
though not necessarily, the internal stress of the dielectric
layers is a compressive stress, which often manifests as wafer bow.
The method and apparatus can be used to reduce compressive bow in
stacks containing dielectric layers.
[0005] Some embodiments disclosed herein provide the benefit of
improved smoothness of the layers and surface of multilayer stacks
fabricated on semiconductor substrates, which improves the
reliability of lithography on the substrate.
[0006] In some embodiments, each of the layers, whether an oxide
layer or a silicon layer, is about 500 .ANG. thick or less (e.g.,
about 250 to 350 .ANG. thick). In some vertically integrated memory
applications, the stacks contain at least about 50 layers of
alternating dielectric and silicon (e.g., about 70 to 75 layers).
In some embodiments, the overall thickness of the stack may be in
the neighborhood of about 2 to 4 .mu.m in height.
[0007] In one aspect of the embodiments herein, a method of
preparing a stack structure for an electronic device on a
semiconductor substrate is provided. The method includes (a)
depositing a doped silicon layer on the substrate by introducing a
first process gas comprising a silicon hydride and a dopant
precursor into a chemical vapor deposition (CVD) apparatus having a
plasma, (b) depositing a dielectric layer on the substrate by
introducing a second process gas comprising a precursor of the
dielectric layer into the CVD apparatus having a plasma; and (c)
repeating operations (a) and (b) for two or more times to form a
stack structure.
[0008] In some embodiments, the CVD apparatus is a plasma-enhanced
chemical vapor deposition (PECVD) apparatus. In some embodiments,
during deposition of a doped silicon layer according to (a) above,
the substrate is held at a temperature of about 450.degree. C. to
550.degree. C., the silicon hydride in the first process gas is
introduced at a flow rate of at least about 0.01 sccm/cm.sup.2 of
substrate deposition surface area, and the doped silicon layer has
a compressive stress of less than about 100 MPa or a tensile
stress, and a thickness of less than about 500 .ANG..
[0009] In some embodiments, during deposition of a dielectric layer
according to (b) above, the dielectric layer has a compressive
stress of at least about 200 MPa and a thickness of less than about
500 .ANG..
[0010] In some embodiments, the silicon hydride used in the process
described above comprises silane. In some embodiments, the silicon
hydride is introduced into the CVD apparatus at a flow rate of at
least about 1200 sccm or greater for a 300 mm-diameter substrate,
or at an equivalent flow rate scaled by substrate deposition
surface area. In other embodiments, the silicon hydride is supplied
at a flow rate of at least about 0.038 sccm/cm.sup.2 of substrate
deposition surface area.
[0011] In some embodiments, the first process gas comprises
phosphine as the dopant precursor. In some embodiments, the first
process gas further comprises helium, and the volume ratio of
phosphine to helium is about 5:95.
[0012] In some embodiments, depositing the doped silicon layer in
(a) is conducted at a pressure of between about 1 and 10 Torr. In
some embodiments, the roughness of the doped silicon layer is about
5 .ANG. or less as measured by atomic force microscopy. In some
embodiments, the plasma is provided by powering a high frequency
radio power source at about 500 watts or higher. In some
embodiments, the bulk resistivity of the doped silicon layer after
activation is less than about 0.001 ohmcentimeter.
[0013] In some embodiments, the stack structure manufactured by the
method described above comprises amorphous or microcrystalline
silicon, and the method further comprises activating the amorphous
or microcrystalline silicon by heating, thereby converting the
silicon to a polycrystalline state. In some embodiments, the stack
structure has a total thickness of between about 1 and 6
micrometers. In some embodiments, the stack structure comprises at
least 50 layers. In some embodiments, the stack structure is
configured to be part of a vertically integrated memory device. In
some embodiments, the stack structure comprises alternating layers
of doped silicon and a dielectric material. In some embodiments,
the stack structure comprises alternating layers of doped and
undoped silicon.
[0014] In another aspect of the embodiments herein, a system is
provided for preparing a stack structure for an electronic device
on a semiconductor substrate. The system include a system
controller and a multi-chamber apparatus comprising at least one
vapor deposition (CVD) chamber having a plasma for depositing at
least one layer of the stack over the semiconductor substrate. The
system controller include instructions for depositing a doped
silicon layer on the substrate by introducing a first process gas
comprising a silicon hydride and a dopant precursor into a CVD
apparatus, wherein the substrate is held at a temperature of about
450.degree. C. to 550.degree. C., the silicon hydride in the
process gas is introduced at a flow rate of at least about 0.01
sccm/cm.sup.2 of substrate deposition surface area, and the doped
silicon layer has a compressive stress of less than about 100 MPa
or a tensile stress, and a thickness of less than about 500 .ANG..
The system controller also includes instructions for depositing a
dielectric layer on the substrate by introducing a second process
gas comprising a precursor of the dielectric layer into the CVD
apparatus having a plasma, wherein the dielectric layer has a
compressive stress of at least about 200 MPa and a thickness of
less than about 500 .ANG..
[0015] In some embodiments, the multi-chamber apparatus of the
system also includes a rapid thermal annealing chamber. In some
embodiments, the controller also includes instructions for
receiving the semiconductor substrate from outside the system and
transferring the substrate from one chamber to another in the
multi-chamber apparatus.
[0016] These and other features will be described below with
reference to the associated drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0017] FIG. 1 shows an example film stack having alternating layers
of material.
[0018] FIG. 2A shows a flowchart illustrating a method of
processing a semiconductor substrate in accordance with an
embodiment disclosed herein; FIGS. 2B-2C show some of the relevant
processing conditions for the embodiment.
[0019] FIG. 3 shows an example of an annealing chamber that may be
used in accordance with the embodiments herein.
[0020] FIG. 4 depicts an example of a PECVD deposition processing
station that may be used in accordance with an embodiment
herein.
[0021] FIGS. 5-6 show multi-station semiconductor processing
apparatuses in accordance with the embodiments herein.
[0022] FIGS. 7A-7P show cross-sectional views illustrating a method
for fabricating a nonvolatile memory device in accordance with a
method disclosed herein.
[0023] FIGS. 8A-8C show the bow, internal stress, and thickness
non-uniformity as functions of silane flow rate (left panel) and
the high-frequency RF power (right panel) in a series of
experiments producing a phosphine doped silicon layer.
[0024] FIG. 9 shows the impact of high-frequency RF power on the
bulk resistivity of the deposited silicon film in a series of
experiments producing a phosphine doped silicon layer.
[0025] FIGS. 10A and 10B show the impact of phosphine to silane
flow rate ratio on the bulk resistivity and sheet resistance of the
deposited silicon films after activation.
[0026] FIGS. 11A-11B show the impact of temperature and silane flow
rate on internal stress of the deposited film
[0027] FIGS. 12A-12E show how various processing parameters affect
the bow and smoothness of a 73-layer stack comprising alternating
oxide layers and silane doped silicon layers.
DETAILED DESCRIPTION
[0028] In this application, the terms "semiconductor wafer,"
"wafer," "substrate," "wafer substrate," and "partially fabricated
integrated circuit" are used interchangeably. One of ordinary skill
in the art would understand that the term "partially fabricated
integrated circuit" can refer to a silicon wafer during any of many
stages of integrated circuit fabrication thereon. Further, the
terms "electrolyte," "plating bath," "bath," and "plating solution"
are used interchangeably. The following detailed description
assumes the invention is implemented on a wafer. However, the
invention is not so limited. The work piece may be of various
shapes, sizes, and materials. In addition to semiconductor wafers,
other work pieces that may take advantage of this invention include
various articles such as printed circuit boards and the like.
[0029] When silicon is discussed herein without qualification, it
is intended that the silicon may have any morphological form, e.g.,
amorphous or polycrystalline silicon. Frequently, the as-deposited
silicon does not initially possess a polycrystalline morphology.
Rather its initial, as-deposited, morphology is amorphous or
microcrystalline (i.e., a structure having small grains of
crystalline Si within an amorphous phase). In the context of device
fabrication, the as-deposited silicon is sometimes referred to as
"unactivated." It may be subsequently "activated" to convert it
into polycrystalline silicon or "polysilicon." Activation is
typically accomplished by heating in a furnace or heating via a
rapid thermal annealing (RTA) process. In many device fabrication
schemes, unactivated silicon is subjected to some post-deposition
processing before it is activated. Activation may reduce internal
stress in silicon layers and stacks containing silicon layers.
[0030] In the following description, numerous specific details are
set forth in order to provide a thorough understanding of the
presented embodiments. The disclosed embodiments may be practiced
without some or all of these specific details. In other instances,
well-known process operations have not been described in detail to
not unnecessarily obscure the disclosed embodiments. While the
disclosed embodiments will be described in conjunction with the
specific embodiments, it will be understood that it is not intended
to limit the disclosed embodiments
[0031] VIM structures, due to their 3D arrangement, allow for an
increased total memory capacity per unit of die area as compared to
2D memory structures. VIM structures are typically formed from
alternating layers of material. As used herein, the multilayer
stacks include alternating layers of two or more materials, at
least one of which includes silicon. In certain cases these
materials may be doped and undoped silicon, or doped silicon and a
dielectric material, or a combination of these materials. One
non-limiting example of a dielectric material is an oxide such as
silicon oxide. In some embodiments, a repeating unit of three
different layers may be used in a stack. For example, in a
particular embodiment, the stack comprises repeating layers of
doped silicon, undoped silicon, and dielectric.
[0032] Oftentimes, the unit layers used to form a VIM structure on
a substrate have an as-deposited internal stress. Wafer bow can be
a problem for the stack and stress compensation is needed. In VIM
and similar applications, stack bow and smoothness need to be
managed. When depositing two alternating films back-to-back to
build a stack structure for an electronic device on a semiconductor
substrate, stacks containing more than 30, 60, or even 90 layers
are deposited on a substrate, and are then patterned. The bow of
the wafer cannot exceed an upper limit allowed by follow-on
processing steps. In embodiments with a stack made up of oxide and
amorphous silicon (a-Si) layers, the layers, or unit films,
typically have compressive internal stress. The layers contribute
to the overall compressive bow of the wafer as the number of layers
increases and/or thickness of the layer(s) increase. The bow of the
full stack is the sum of the combined thickness and layers of the
unit film. Because both the films are very compressive
as-deposited, the full stack structure will also have very
compressive bow that is above the maximum limits for post wafer
processing steps after the VIM stack structure is deposited.
[0033] In addition to a low value of stack bow, VIM stacks should
have minimum surface roughness. This requires unit film with smooth
surfaces and interfaces. The individual film layers have the
largest impact on the stack structure's final surface roughness and
planarization. Typically PECVD deposited film become rougher as
thicknesses increase. Because the VIM 3D stacks are microns thick,
the unit films making the layers need to be extremely smooth. In
some implementations, VIM stacks with surface roughness of more
than 10 to 15 .ANG. will exceed the upper limits of the photo
patterning follow-on process steps.
[0034] Furthermore, uniformity of deposited film layers is a
desired property for layers in stacks for electronic devices,
because non-uniformity in the layers leads to irregular surfaces,
weakened structure, and deformation or lamination of the stack.
[0035] Some embodiments disclosed herein relate to processes
involving high SiH.sub.4 flow, which produce doped silicon films
that have one or more of the following properties: natural or even
tensile stress, smooth surfaces, uniform layers, and excellent
sheet and bulk resistivity even at very thin film thickness upon
RTA. The processes allow films to be deposited with a full range of
adjustable stresses ranging from compressive to tensile. Because
the processes provide tunable stress and the ability to make the
film tensile, they can be used to tune the full stack bow and lower
the compressive stress of the stack.
[0036] Typically, pairs of alternating layers of a VIM stack (e.g.,
a layer pair of doped silicon and dielectric, or a layer pair of
undoped silicon and doped silicon) are provided as discrete pairs
which may be interleaved with one another to the extent appropriate
for the particular application under consideration. Thus, for
example, a multilayer stack may primarily contain alternating
layers of dielectric and doped silicon, with some inserted or
interleaved pairs of layers of doped silicon and undoped silicon.
In another example, the multilayer stack primarily contains pairs
of undoped silicon and doped silicon, with one or more pairs of
dielectric and doped silicon interleaved in the overall stack.
[0037] In some embodiments, the silicon layers in the stack are not
sacrificial. In other words, they are preserved through the
process. In a particular application, the silicon layer and the
oxide layer are both retained in the final VIM stack product. The
silicon layers may be used as gate electrodes, capacitor
electrodes, etc.
[0038] In other embodiments, some silicon layers are sacrificial.
That is, the silicon layers are removed by wet etching or other
process prior to fabrication of the final product. The disclosure
should not be limited to embodiments in which the silicon layers
are non-sacrificial.
[0039] Typically, one or more or all of the silicon layers of the
multilayer stack are deposited by a chemical vapor deposition
process such as a plasma-enhanced chemical vapor deposition (PECVD)
process. In some embodiments, all layers are deposited by
PECVD.
[0040] In some embodiments, the dielectric layer contains silicon
and oxygen. For example, the dielectric layer may be a silicon
oxide layer. In one example, a PECVD film may be made by reaction
of a silicon-containing precursor and a source of oxygen.
Non-limiting examples include silane and tetra-ethylorthosilicate
(TEOS) as silicon containing precursors, and molecular oxygen,
O.sub.2, and N.sub.2O as oxygen sources. Other chemical elements
may be included to modulate a film's composition to enhance certain
film properties. For example, elements such as carbon, hydrogen, or
nitrogen may be introduced.
[0041] In some embodiments, the thickness of the individual layers
is typically between about 10-1,000 .ANG., or between about 200-550
.ANG.. These thicknesses are suitable for VIM and other
applications. In such embodiments, the overall height of the
complete multilayer stack may be between about 1-6 .mu.m, or
between about 1.5-4 .mu.m. In certain cases, the layers are each of
substantially similar thickness (e.g., within about 50%), and in
some cases, the thickness of the individual layers is even more
similar (e.g., within about 20%).
[0042] In VIM applications, the dielectric layers generally serve
as insulating layers between the vertically stacked devices. In
stacks employing the dielectric and doped silicon pairs, the doped
silicon layers serve as gate electrodes. In stacks containing the
undoped and doped silicon pairs, the undoped silicon layers serve
as sacrificial layers which are subsequently etched and replaced
with an oxide or other dielectric and the doped silicon layers,
which remain after etch, serve as insulating layers.
[0043] In various embodiments disclosed herein, the silicon layers
comprise doped silicon layers. That is, the silicon in the layer
contains an incorporated dopant such as phosphorus, boron, or
arsenic. In various embodiments described herein, phosphorus is
typically identified as the dopant. Additionally, in the
description of these embodiments, the dopant is frequently
identified as phosphine, a phosphorus precursor gas. However, the
disclosure should not be limited to phosphorus as the dopant or to
phosphine as the dopant precursor for phosphorus.
[0044] As mentioned above, silicon, whether doped or undoped, is
typically deposited in an amorphous or microcrystalline state and
is only subsequently activated to assume its final polycrystalline
state (sometimes referred to as polysilicon). In various
applications, it is intended that the pre-activated state remains
intact, without conversion to the polycrystalline state, during
multiple steps following deposition. Therefore, it may be important
in some applications (such as VIM applications) to not increase the
temperature or otherwise apply thermal energy to a level that would
activate the as-deposited silicon prior to patterning and
etching.
[0045] In the application of preparing an integrated circuit, after
a stack structure is prepared, an optional soft anneal process may
be applied in some embodiments to reduce internal stress of the
stack. Methods and apparatus including soft annealing are disclosed
in U.S. patent application Ser. No. 13/857,566, files on Apr. 5,
2013, which is incorporated by reference in its entirety.
[0046] In the application of preparing an integrated circuit, after
a stack structure is prepared, a patterning of the stack structure
typically takes place after a hard mask (e.g., an ashable hard mask
such as an amorphous carbon layer) is deposited over the stack. The
mask may then be lithographically patterned. In certain cases,
patterning is accomplished through a photolithographic process. In
other cases, patterning is achieved through an electron beam
lithography process. The resulting vertically etched stack
structures will have dimensions appropriate for the particular
application for which they will be used (e.g., VIM
applications).
[0047] In certain embodiments, after patterning, the stack is
etched to partially or fully remove one or more layers of the stack
(e.g., sacrificial layers of undoped silicon) while substantially
preserving the doped silicon. This selective etching may be
primarily horizontal in direction and produces an undercut
structure such as a fishbone structure. A wet etch may be employed
to selectively remove (wholly or partially) a layer or layers from
the stack. Examples of wet etchants include hydrofluoric acid
(including buffered versions of the acid), potassium hydroxide
(KOH) and tetramethyl ammonium hydroxide (TMAOH). In vertical
memory applications, a suitable dielectric, for example an oxide,
may be deposited in the openings made by the removal of the undoped
Si layers.
[0048] After patterning and optional etching, the amorphous or
microcrystalline silicon is activated to convert it to its
polycrystalline state. The polycrystalline state of doped silicon
has significantly better electrical properties, most notably lower
resistivity, than the pre-activation silicon.
[0049] Activation is typically accomplished by heating the
amorphous or microcrystalline silicon to a controlled, elevated
temperature for a controlled duration. In many cases activation
occurs at a temperature of about 750.degree. C. or higher. While
lower temperatures may be used for the activation anneal, these
lower temperatures typically require much longer times to complete
the transition. This heating may be performed for a few seconds or
longer, as necessary for the particular temperature and
application. In some cases activation is accomplished using a laser
crystallization process. Polysilicon often has an average grain
size of about 10 nanometers to about 10 micrometers, although other
grain sizes may be achieved.
[0050] A significant challenge encountered when depositing
multilayer stacks (e.g., stacks containing many layers of doped and
undoped silicon or layers of doped silicon and dielectric) is that
these layers exhibit significant internal stress as-deposited. This
internal stress causes the entire wafer to change shape through
bowing. The bow of a wafer is measured as the difference in
elevation between the center of the wafer and the perimeter of the
wafer when the wafer is positioned horizontally. This bowing occurs
at least in part because each of these three materials (dielectric
and undoped/doped silicon) has a significant compressive internal
stress. Because each layer in a stack contributes to the overall
stress and bow of the wafer, stacks having more layers exhibit more
stress and are more prone to bowing. Single layer films and stacks
having only a few layers are much less likely to exhibit bowing
problems. As a consequence, a wafer containing a stack of these
materials deposited on a single side of the wafer will have a dome
shape in which the stack is on the top or outer surface of the dome
and the underlying substrate is on the bottom or inner surface of
the dome. If the bow shift is sufficiently large in a given wafer,
it can result in a number of problems. One of these problems is
that it is difficult to use a conventional wafer processing
apparatus, such as a wafer chuck, to conduct subsequent processing
on a bowed wafer. A severely bowed wafer is very difficult to
properly register to the extent required for photolithography, as
this process requires precise optical focusing at precise locations
on the substrate. When a wafer is bowed, the lithographic beam may
try to focus on one part of the wafer but end up exposing a
different part of the wafer due to the wafer's irregular shape.
Further, the individual dies on the bowed wafer may not be properly
processed in post stack deposition operations. For many
applications, a wafer with a bow of greater than about 350 .mu.m
and more typically 150 .mu.m or even 90 .mu.m is not deemed
suitable for processing.
[0051] A relatively low internal stress of the silicon layer is
desirable. In some implementations, the silicon oxide layers
produced as described herein have an internal stress of about 300
to 350 MPa compressive. Therefore, the silicon film ideally has an
internal stress that offsets, at least partially, the high
compressive stress of the oxide layers. In certain embodiments, the
internal stress of the silicon layers is neutral to somewhat
tensile to at least partially offset the intrinsic compressive
stress of the oxide layers.
[0052] Hydrogen has been found to play a role in controlling
internal stress in the film. It has been found that increasing the
amount of hydrogen in a silicon film reduces its internal
compressive stress. The amount of hydrogen present in a silicon
film is a strong function of the mass flow rate of a silicon
hydride precursor employed in the deposition process. Greater flow
rates of silicon hydride during deposition produce silicon films
with greater concentrations of hydrogen.
[0053] However, the amount of hydrogen can only be increased to a
certain point. The activation process which converts the amorphous
silicon and to polycrystalline silicon naturally dehydrogenates the
amorphous silicon. If the amorphous silicon and contains too much
hydrogen, the hydrogen gas leaving the film during the activation
process will cause the film to blister or delaminated. Therefore, a
careful balance is required to ensure that a sufficient amount of
hydrogen is included in the film to reduce the compressive stress
but the amount hundred and is limited to a level that is not cause
delamination or blistering during the activation process.
[0054] It should also be noted that regardless of the amount of
hydrogen in the film, the activation process naturally makes
silicon and film more tensile, in other words, the internal
compressive stress is reduced. However, certain lithography steps
need to be performed prior to activation. Therefore, methods are
desired to reduce the internal stress of the film prior to
activation. In various embodiments disclosed herein, the internal
stress of a stack containing silicon oxide layers and doped silicon
layers is controlled prior to activation. In certain embodiments,
it is controlled immediately prior to activation. This is because
at this stage in certain fabrication process flows such as those
employed to form vertically integrated memory (VIM), lithographic
steps are employed prior to activation.
[0055] In addition to internal stress, the roughness of a stack has
an impact on the ability to conduct lithography reliably. If the
surface of a stack is too rough, the incident radiation employed in
photolithography will scatter too much. Also, the roughness of the
stack is determined by the maximum roughness of any given layer in
the stack. Therefore, it is important that each and every layer in
the stack have a relatively low roughness.
[0056] The methods and apparatus described herein provide the
benefits of reduced bowing and roughness, thereby allowing
manufacturers to meet stringent bow tolerances while maintaining
extremely high quality multilayer stacks.
Methods
[0057] A challenge arises when forming stacks of alternating layers
of oxide and silicon. For many applications including VIM
applications, the oxide must be of very high quality. For example,
it must have an appropriately low dielectric constant comparable to
thermal oxide, a high breakdown voltage, and appropriate resistance
to wet etching. Unfortunately, oxide meeting these requirements
frequently has a relatively high internal stress. For oxide layers
deposited at a thickness of about 150 to 250 .ANG., for example,
the internal stress of a stack can be about 300 and 350 MPa
compressive. This internal stress can accumulate layer-by-layer
within the stack. If unchecked, the resulting wafer can have a
significant bow, which can make it unsuitable for lithography and
other process steps. For some applications, such as VIM
applications, a wafer having a bow of about 90 .mu.m or greater
cannot be used.
[0058] Therefore, it becomes necessary to manage the internal
stress in the layers of the stack. Unfortunately, the
above-mentioned tight constraints on the electrical properties of
the oxide do not afford significant latitude for reducing the
internal stress in the deposit oxide layer. Compounding the
problem, the silicon layer typically has an intrinsic compressive
internal stress. Therefore the total compressive stress within the
stack quickly grows to a level producing an unsuitable bow.
[0059] One aspect of the present disclosure provides methods for
depositing silicon layers in the stack that contain relatively low
compressive stress or even tensile stress, while meeting other
requirements of the silicon layer such as low surface roughness and
low resistivity.
[0060] The methods disclosed herein reduce the internal stress
and/or roughness of one or more silicon layers of a work piece.
Typically, though not necessarily, the internal stress is a
compressive stress. A particularly beneficial application of the
methods is in reduction of internal stress in a stack containing
two or more layers of silicon. Often, the internal stress of a
layer or group of layers in a stack is manifest as wafer bow, as
described above. The methods can be used to reduce compressive bow
in stacks containing silicon. Furthermore, in some embodiments, the
methods may reduce the roughness of the surface of the work
piece.
[0061] Some embodiments provide methods with the benefits of
reduced internal stress (e.g., compressive stress) without
activating the as-deposited amorphous or microcrystalline silicon.
Therefore, the as-deposited stack has acceptable bow, which allows
the required processing (e.g., patterning and selective etching)
prior to activation to convert the as-deposited silicon in the
stack to polysilicon.
[0062] FIG. 2A shows a flowchart of an embodiment in accordance
with the methods described herein; FIGS. 2B and 2C illustrate the
processing conditions required in the process. The method 3000 for
preparing a stack structure on a substrate involves depositing a
doped silicon layer on a substrate, as indicated by block 3030. The
method 3000 also involves depositing a dielectric layer on the
substrate as indicated by block 3050. These steps are repeated for
at least two times, thereby forming a stack structure over the
substrate, see block 3070.
[0063] As described herein, the stack typically includes
alternating layers of material. In one embodiment, the alternating
layers include doped and undoped silicon. In another embodiment,
the alternating layers include doped silicon and dielectric
material. Other types of layers may also be used. In certain
embodiments, each of the layers is deposited by a chemical vapor
deposition process such as a plasma-enhanced chemical vapor
deposition (PECVD) process.
[0064] As illustrated in FIG. 2B, depositing a doped silicon layer
according to the method disclosed herein, as indicated by block
3030, requires a set of processing conditions shown in blocks
3031-3037. Typically, though not necessarily, these conditions are
employed concurrently over the course of the silicon layer
deposition process. The deposition of a doped silicon layer
involves introducing a first process gas comprising a silicon
hydride (e.g., silane) and a dopant precursor (e.g., phosphine)
into a chemical vapor deposition (CVD) apparatus having a plasma,
block 3031. The deposition of the doped silicon layer in this
example requires the substrate to be held at a temperature of about
450.degree. C. to 550.degree. C. as indicated by block 3033.
Furthermore, some embodiments require the silicon hydride in the
first process gas to be introduced at flow rate of at least about
0.01 standard cubic centimeter per minute per square centimeter of
depositing surface (sccm/cm.sup.2), see block 3035. In some
embodiments, the silicon hydride in the first process gas is
introduced at flow rate of at least about 0.017 sccm/cm.sup.2, or
at least about 0.038 sccm/cm.sup.2. In some embodiments, the
silicon hydride is introduced at a flow rate of least about 1200
sccm or greater for a 300 mm-diameter substrate, or at an
equivalent flow rate scaled by substrate deposition surface area.
The deposition method disclosed above forms a doped silicon layer
having a compressive stress of less than about 100 MPa or a tensile
stress and a thickness of less than about 500 .ANG., see block
3037.
[0065] In some embodiments, as illustrated in FIG. 2C, depositing a
dielectric layer on a substrate, as indicated by block 3050,
requires a set of processing conditions shown in blocks 3051-3053.
The deposition of a doped silicon layer involves introducing a
second process gas comprising a precursor of the dielectric layer
into a CVD apparatus having a plasma, block 3051, thereby forming a
dielectric layer having a compressive stress of at least about 200
MPa and a thickness of less than about 500 .ANG., as shown in block
3053.
[0066] According to the example disclosed in FIG. 2A, a stack
structure is formed by repeating the steps of blocks 3030 and 3050,
see block 3070. In some embodiments not shown in the figure, a soft
anneal process is performed after the stack structure is formed,
which is achieved by providing a stack-coated substrate in an
anneal chamber and subjecting the substrate to an elevated
temperature for a particular duration. In many implementations, the
soft anneal temperature should be below the temperature at which
silicon transitions to polysilicon. This is especially important
where the soft anneal process takes place before the etching
operation, as the etching operation is most successful (i.e., etch
selectivity is greatest between the two materials) when performed
on unactivated silicon.
[0067] In some applications of the methods described above, the
stack structure is used to prepare an integrated circuit. In some
applications, after the stack structure is formed, a hard mask is
deposited on the stack. The mask may be an ashable hard mask in
certain embodiments. The hard mask is then patterned. This
patterning may be achieved through photolithographic techniques.
The patterned stack is then etched to selectively remove (partially
or wholly) one of the materials in the stack. A wet etch is often
used for this purpose. The as-deposited silicon may then be
activated to form polysilicon in layers of the integrated
circuit.
[0068] The stack may have a composition and arrangement as
described above (i.e., it may include layers of doped polysilicon
with alternating layers of one or more other materials). In some
embodiments, those other materials are a dielectric (e.g., a
silicon oxide or a variant thereof such as an oxicarbide or an
oxinitride), undoped silicon, and the like. In some cases, the
alternating layers are deposited in a single vacuum controlled
apparatus with multiple stations or variable process conditions to
effect deposition of the layers of different compositions. In
certain embodiments, the layers of the stack are deposited by a
PECVD process.
Example Properties of the Stack Structure
[0069] In some embodiments, the stack structure produced by the
methods disclosed herein has properties desirable for applications
in integrated circuits.
[0070] Thickness.
[0071] In some embodiments, each layer of doped silicon deposited
in the stack is about 250 to 350 .ANG. in thickness. The total
stack height, including all oxide and silicon layers is about to
2-4 .mu.m in height.
[0072] Resistivity.
[0073] The conductivity of the deposited silicon layers should be
as high as possible. Stated another way, the resistivity or sheet
resistance should be as low as possible. The resistivity of the
silicon layers is most appropriately measured after activation.
Activation converts the as deposited amorphous silicon and the
polysilicon. The polysilicon is present in the final product, and
therefore its electrical properties are relevant. After activation,
the bulk resistivity of the polysilicon may be at most about 0.001
ohm-centimeter or more specifically at most about 0.0085
ohmcentimeter in some embodiments. For a polysilicon film having a
thickness of about 250 to 350 .ANG., the sheet resistance should be
below about 300 ohms per square in some embodiments. The
resistivity of the silicon layer is a strong function of the dopant
concentration.
[0074] Roughness.
[0075] In certain embodiments, the as deposited and activated
silicon layer may be characterized as "ultra-smooth." In certain
embodiments, the roughness of the silicon layer is about 5 .ANG. or
less as measured by atomic force microscopy. In certain
embodiments, the roughness of the silicon layer is about 3 .ANG. or
less as measured by atomic force microscopy. The roughness of a
single layer in a stack propagates throughout stack so that the
roughness of the entire is dictated by the roughness of the
roughest layer in the stack. As indicated, the activation process
crystallizes the film; that is activation converts the amorphous
silicon matrix to silicon crystallites. Typically, these
crystallites have a side-to-side dimension of about 85 to 125
.ANG.. This size suggests that the activated film may have a
roughness considerably greater than the 5 .ANG. cutoff mentioned
below. However, it has been found that when depositing silicon
using the deposition conditions described herein, the activated
polysilicon film has a roughness that is no greater than
approximately the roughness of the as-deposited amorphous silicon
film. Without wishing to be bound by a particular theory or
mechanism, it is believed that under appropriate deposition
conditions, the as-deposited amorphous silicon is not only smooth
but contains nucleation sites that promote crystal growth in a
sidewise or horizontal direction with little or no upward or
vertical growth that would increase the roughness of the
as-deposited film.
[0076] Internal Stress.
[0077] In some embodiments, a doped silicon layer of a stack
deposited using the methods described herein has a compressive
stress of less than about 100 MPa or even a tensile stress. In
certain embodiments, the internal stress of the silicon layer is
about neutral (i.e., neither tensile nor compressive) to about 100
MPa tensile. In certain embodiments, the internal stress of the
silicon layer is between about neutral and 50 MPa tensile. If left
unmitigated, the oxide layers in the stack of 70 to 75 layers will
produce a poem about 135 .mu.m. One embodiment of the disclosed
process employs silane as the silicon hydride of the process gas,
and phosphine as the dopant precursor in the process gas. This
embodiment involves a method for manufacturing 300 mm wafers in a
four showerhead four wafer station reaction chamber. It employs a
silane flow rate of 1450 sccm, a phosphine flow rate of 725 sccm
(phosphine is provided as a 5% composition and 95% helium),
pressure of 6 Torr, temperature of 500.degree. C., an RF power of
870 W (at 13.57 MHz), and the carrier gas flow rate of 15,000 sccm
(helium).
Parameters for Silicon Layer Deposition
[0078] To achieve desirable properties of the stack structure
disclosed above, relevant parameters are required for the
deposition of silicon layers. The relevant parameters include but
are not limited to flow rate of a silicon and hydride process gas,
the temperature of the substrate, the power delivered by the plasma
source, and the chamber pressure.
[0079] Process Gas.
[0080] The principal process gas should contain silicon. It
typically also includes hydrogen. While silane is identified as the
precursor gas in most examples disclosed herein, other silicon
hydride process gases may be used. These include di-silane,
tri-silane, and various substituted silanes, di-silanes, and
tri-silanes. In the substituted compounds, one or more of the
hydrogen atoms in the base silane, di-silane, or tri-silane is
replaced by a hydrocarbon moiety or other substituent. One of the
roles of the silicon and hydride precursor is to provide hydrogen
atoms in the as-deposited silicon and film.
[0081] The mass flow rate of the silane delivered to the reactor is
relatively high by conventional silicon deposition standards. In
various embodiments, the flow rate of the silane process gas is at
least about 1200 sccm in a four station PECVD reactor such as the
Vector family of PECVD reactors available from Lam Research
Corporation of Fremont Calif.
[0082] In some embodiments, the dopant precursor gas is delivered
in conjunction with the silane. For example, phosphine may be
provided along with the silane process gas. In one example where
phosphine is used as the dopant precursor, phosphine is provided at
a flow rate of about 600 to 1000 sccm. In this example, phosphine
is provided in a gas mixture 5% phosphine by volume in helium
process gas. In some cases, the amount of phosphine (provided as
the 5% composition) is between about 650 and 850 sccm (e.g., about
725 sccm). Of course, phosphine or other dopant precursor can be
used in other concentrations, in which case the listed flow rates
should be adjusted to provide equivalent amounts of the dopant
precursor. In various embodiments, the phosphine or other dopant
precursor is provided with silane in a particular ratio of dopant
precursor with respect to the silane process gas. FIGS. 10A-10B
show the effect of varying the phosphine-silane ratio on the
resistivity of the deposited silicon and film, which figures are
further discussed below.
[0083] Typically, a carrier gas is used to deliver the silane and
dopant precursor to the reaction chamber. In various embodiments,
the carrier gas is a noble gas such as argon or helium. In certain
embodiments, where helium is employed, the helium provided at a
flow rate of about 10,000 to 20,000 sccm for a four station PECVD
reactor such as any of the Vector family of reactors available from
Lam Research Corporation. In specific embodiments, the flow rate of
helium is between about 13,000 and 17,000 sccm (e.g., about 15,000
sccm).
[0084] Pressure. In certain embodiments, the pressure in the
reactor is between about 4 and 7 Torr; or between 1 and 10 Torr,
e.g., about 6 Torr.
[0085] Power.
[0086] In certain embodiments, the power delivered to a plasma
generator is between about 500 and 1000 W or between about 750 and
950 W (e.g., about 870 W) for an inductively coupled plasma source
for a four station reactor such as a Vector PECVD reactor available
from Lam Research Corporation. In some embodiments, the RF power
should be applied at the frequency of 13.57 MHz. Other sources of
plasma such as capacitively coupled plasma generator may be
employed.
[0087] Temperature.
[0088] In certain embodiments, the substrate temperature during the
silicon deposition process is between about 400 and 600.degree. C.
In some cases, the substrate temperature is between about 450 and
575.degree. C. In some examples, the temperature of the substrate
is between about 475 and 525.degree. C. (e.g., about 500.degree.
C.). Typically, though not necessarily, temperature of the
substrate is maintained by applying heating and/or cooling through
a pedestal or chuck on which the wafer is held during the silicon
deposition process.
[0089] Some parameters such as those for the plasma power and the
flow rates of process gases are presented in values appropriate for
use in a four station deposition chamber such as one of the Vector
PECVD reactors available from Lam Research Corporation. It should
be noted, that these parameters assume that the substrates
processed in such chambers are 300 mm wafers. The values set forth
herein for the power and process gas flow rates can be scaled to
other reactors that use other substrate sizes or by multiplying the
parameter value by a ratio of the surface areas of the wafer under
consideration to that of the 300 mm wafer. For example, if a four
station reactor having a volume and dimensions similar to that of
Lam Research Vector system, but instead uses wafer pedestals for
450 mm wafers, the above process parameters may be scaled by the
surface area of a 450 mm wafer divided by the surface area of a 300
mm wafer, or a factor of 2.25.
[0090] The reduction in wafer bow as described herein can be
achieved without delaminating, peeling or blistering. Blistering is
often evidenced by small bubbles or divots where material popped
out of the stack. Peeling involves separation between layers at an
intermediate position in the stack, i.e., between two different
layers of the stack. Delamination results when the entire stack
separates from the substrate.
Apparatus
[0091] The methods described herein may be performed by any
suitable apparatus. A suitable apparatus includes hardware for
accomplishing the process operations and a system controller having
instructions for controlling process operations in accordance with
the present embodiments.
[0092] Some embodiments relate to a system for preparing a stack
structure for an electronic device on a semiconductor substrate. In
some embodiments, the system includes a multi-chamber apparatus
comprising at least one PECVD chamber for depositing at least one
layer of the stack over the semiconductor substrate. The system
also includes a controller comprising instructions for depositing a
doped silicon layer on the substrate by introducing a first process
gas comprising a silicon hydride and a dopant precursor into a
chemical vapor deposition (CVD) apparatus having a plasma. In some
embodiments, the substrate is held at a temperature of about
450.degree. C. to 550.degree. C. In some embodiments, the silicon
hydride in the process gas is introduced at a flow rate of at least
about 0.01 sccm/cm.sup.2 of substrate deposition surface area. In
some embodiments, the silicon hydride is introduced at a flow rate
of at least about 0.038 sccm/cm.sup.2. In some embodiments, the
doped silicon layer has a compressive stress of less than about 100
MPa or a tensile stress, and a thickness of less than about 500
.ANG.. The controller also comprises instructions for depositing a
dielectric layer on the substrate by introducing a second process
gas comprising a precursor of the dielectric layer into the CVD
apparatus having a plasma. In some embodiments, the dielectric
layer has a compressive stress of at least about 200 MPa and a
thickness of less than about 500 .ANG..
[0093] In some embodiments, the multi-chamber apparatus also
includes a rapid thermal annealing (RTA) chamber. In some
embodiments, the RTA chamber is a soft anneal chamber integrated
with an apparatus for depositing a stack containing layers of doped
silicon and layers of at least one other material such as silicon
oxide (or other dielectric) or undoped silicon. Such deposition
apparatus may include hardware for accomplishing the deposition
operations and its own system controller having instructions for
controlling deposition operations. In some embodiments, the
deposition apparatus is a PECVD apparatus.
[0094] The FIG. 3 shows a simplified depiction of a suitable
activation chamber 4000 in accordance with an embodiment disclosed
herein. Wafer 4500 is mounted on a susceptor 4410 supported by
susceptor support 4120. A susceptor position control 4020 may
rotate the wafer 4500 during processing, and may also raise and
lower susceptor 4410 to certain positions in order to load, unload
or process a wafer 4500. Heat control mechanism 4030 controls heat
sources 4040 and 4240, which heats the wafer 4500 to a
substantially uniform temperature during processing. Gas flow
control 4050 regulates the flow of gases into the reaction chamber
4090 of the anneal chamber 4000. This control may be achieved by
controlling the flow of gas in inlet channel 4060 and gas injection
head 4070, as well as exhaust gases from the reaction chamber 4090
through outlet channel 4080. In another embodiment of the
activation chamber (not shown), gases are introduced into the
reaction chamber through a plurality of side jets. In yet other
embodiments of the activation apparatus, the reaction chamber may
be shaped in a non-rectangular fashion (e.g., bell shaped, rounded,
etc.). Any chamber shape may be used.
[0095] Activation can be conducted in any vehicle that permits
delivery of energy to the silicon layers under conditions allowing
adequate temperature control. One example of such a vehicle is a
furnace that includes conventional heating elements such as
resistive heating elements. In some embodiments, the anneal is
conducted in a chamber that employs radiation from one or more
lamps or laser sources to provide annealing energy. Rapid thermal
annealing (RTA) chambers are a widely used type of lamp source
annealing chamber that can be used to perform activation. One
example of a suitable RTA tool is the Applied Materials Vantage
Radiance Plus RTP. In some examples, the RTA tool has multiple
bulbs, all positioned above the wafer. The temperature is monitored
near the wafer bottom to ensure that the entire wafer (not just the
wafer surface) is adequately heated through. This is opposed to a
laser anneal, which locally heats the surface of the wafer. In some
embodiments, the wafer rotates on a platen during RTA. RTA tools
and methods are further discussed and described in U.S. Pat. No.
6,151,447, filed on Nov. 25, 1997, which is incorporated by
reference herein in its entirety.
[0096] FIG. 4 schematically shows a deposition process station 3100
in accordance with the embodiments herein. For simplicity, process
station 3100 is depicted as a standalone process station having a
process chamber body 3172 for maintaining a low-pressure
environment. However, it will be appreciated that a plurality of
process stations 3100 may be included in a common low-pressure
process tool environment. Process station 3100 includes a process
gas delivery line 3174 for providing process gases, such as inert
gases, precursors, reactants, and treatment reactants, for delivery
to process station 3100. In the example shown in FIG. 4, a
showerhead 3178 is included to distribute process gases within
process station 3100. Substrate 3186 is located beneath showerhead
3178, and is shown resting on a holder 3180 supported by a pedestal
3182. In some embodiments, pedestal 3182 may be configured to
rotate about a vertical axis. Additionally or alternatively,
pedestal 3182 may be configured to translate horizontally and/or
vertically.
[0097] In some embodiments, showerhead 3178 may be a dual-plenum or
multi-plenum showerhead having a plurality of sets of gas
distribution holes. For example, a first set of gas distribution
holes may receive gas from a first process gas delivery line and a
second set of gas distribution holes may receive gas from a second
process gas delivery line, etc. Such physical isolation of process
gases may provide an approach to reducing the amount of small
particles generated from reaction of incompatible process gases in
process gas delivery plumbing upstream of showerhead 3178.
[0098] Showerhead 3178 and holder 3180 electrically communicate
with RF power supply 3188 and matching network 3190 for powering a
plasma 3192. Plasma 3192 may be contained by a plasma sheath 3194
located adjacent to showerhead 3178 and holder 3180. While FIG. 4
depicts a capacitively-coupled plasma, plasma 3192 may be generated
by any suitable plasma source. In one non-limiting example, plasma
3192 may include a parallel plate plasma source.
[0099] In the embodiment shown in FIG. 4, RF power supply 3188 may
provide RF power of any suitable frequency. In some embodiments, RF
power supply 3188 may be configured to control high- and
low-frequency RF power sources independently of one another.
Example low-frequency RF powers may include, but are not limited
to, frequencies between 200 kHz and 2000 kHz. Example high
frequency RF powers may include, but are not limited to,
frequencies between 13.56 MHz and 80 MHz. Likewise, RF power supply
3188 and matching network 3190 may be operated at any suitable
power to form plasma 3192. Examples of suitable powers include, but
are not limited to, powers between 250 W and 5000 W for a
high-frequency plasma (assuming a four station reaction chamber)
and powers between 0 W and 2500 W (assuming a four station reaction
chamber) for a low-frequency plasma for a four-station
multi-process tool including four 15-inch showerheads. RF power
supply 3188 may be operated at any suitable duty cycle. Examples of
suitable duty cycles include, but are not limited to, duty cycles
of between 5% and 90%.
[0100] In some embodiments, holder 3180 may be temperature
controlled via heater 3184. Further, in some embodiments, pressure
control for process station 3100 may be provided by butterfly valve
3196 or by any other suitable pressure control device. As shown in
FIG. 4, butterfly valve 3196 throttles a vacuum provided by a
vacuum pump (not shown) fluidly coupled to process station exhaust
line 3198. However, in some embodiments, pressure control of
process station 3100 may also be adjusted by varying a flow rate of
one or more gases introduced to process station 3100.
[0101] It will be appreciated that control of one or more process
parameters may be provided locally (e.g., RF power may be
controlled by a plasma controller communicating with RF power
supply 3188, process station pressure may be controlled by a valve
controller communicating with butterfly valve 3196 or with gas
metering valves or flow controllers included coupled with process
gas delivery line 3174, etc.) or under partial or total control
provided by a system controller (described in more detail below)
communicating with process station 3100 without departing from the
scope of the present disclosure.
[0102] As described above, one or more process stations may be
included in a multi-station processing tool. In some embodiments of
a multi-station process tool, control and/or supply of various
process inputs (e.g., process gases, plasma power, heater power,
etc.) may be distributed from shared sources to a plurality of
process stations included in the process tool. For example, in some
embodiments, a shared plasma generator may supply plasma power to
two or more process stations. In another example, a shared gas
distribution manifold may supply process gases to two or more
process stations.
[0103] FIG. 5 shows a schematic view of an embodiment of another
multi-station processing tool 3300 with an inbound load lock 3302
and an outbound load lock 3304. A robot 3306, at atmospheric
pressure, is configured to move substrates from a cassette loaded
through a pod 3308 into inbound load lock 3302 via an atmospheric
port 3310. Inbound load lock 3302 is coupled to a vacuum source
(not shown) so that, when atmospheric port 3310 is closed, inbound
load lock 3302 may be pumped down. Inbound load lock 3302 also
includes a chamber transport port 3316 interfaced with processing
chamber 3314. Thus, when chamber transport 3316 is opened, another
robot (not shown) may move the substrate from inbound load lock
3302 to a pedestal of a first process station for processing.
[0104] In some embodiments, inbound load lock 3302 may be connected
to a remote plasma source (not shown) configured to supply a plasma
to load lock. This may provide remote plasma treatments to a
substrate positioned in inbound load lock 3302. Additionally or
alternatively, in some embodiments, inbound load lock 3302 may
include a heater (not shown) configured to heat a substrate. This
may remove moisture and gases adsorbed on a substrate positioned in
inbound load lock 3302. While the embodiment depicted in FIG. 5
includes load locks, it will be appreciated that, in some
embodiments, direct entry of a substrate into a process station may
be provided.
[0105] The depicted processing chamber 3314 comprises four process
stations, numbered from 1 to 4 in the embodiment shown in FIG. 5.
In some embodiments, processing chamber 3314 may be configured to
maintain a low pressure environment so that substrates may be
transferred among the process stations without experiencing a
vacuum break and/or air exposure. Each process station depicted in
FIG. 5 includes a process station substrate holder (shown at 3318
for station 1) and process gas delivery line inlets. In some
embodiments, one or more process station substrate holders 3318 may
be heated.
[0106] In some embodiments, each process station may have different
or multiple purposes. For example, a process station may be
switchable between a soft anneal mode and a conventional anneal
mode. Additionally or alternatively, in some embodiments,
processing chamber 3314 may include one or more matched pairs of
soft and conventional annealing stations. In another example, a
process station may be switchable between two or more film types,
so that stacks of different film types may be deposited in the same
process chamber.
[0107] While the depicted processing chamber 3314 comprises four
stations, it will be understood that a processing chamber according
to the present disclosure may have any suitable number of stations.
For example, in some embodiments, a processing chamber may have
five or more stations, while in other embodiments a processing
chamber may have three or fewer stations.
[0108] FIG. 5 also depicts an embodiment of a substrate handling
system 3390 for transferring substrates within processing chamber
3314. In some embodiments, substrate handling system 3390 may be
configured to transfer substrates between various process stations
and/or between a process station and a load lock. It will be
appreciated that any suitable substrate handling system may be
employed. Non-limiting examples include substrate carousels and
substrate handling robots.
[0109] In some embodiments, a low-pressure transfer chamber may be
included in a multi-station processing tool to facilitate transfer
between a plurality of processing chambers. For example, FIG. 6
schematically shows another embodiment of a multi-station
processing tool 3400. In the embodiment shown in FIG. 6,
multi-station processing tool 3400 includes a plurality of
processing chambers 3314 including a plurality of process stations
(numbered 1 through 4). Processing chambers 3314 are interfaced
with a low-pressure transport chamber 3404 including a robot 3406
configured to transport substrates between processing chambers 3314
and load lock 3408. An atmospheric substrate transfer module 3410,
including an atmospheric robot 3412, is configured to facilitate
transfer of substrates between load lock 3408 and pod 3308.
System Controller
[0110] Turning back to FIG. 5, multi-station processing tool 3300
also includes an embodiment of a system controller 3350 employed to
control process conditions and hardware states of processing tool
3300. For example, in some embodiments, system controller 3350 may
control one or more process parameters during a PECVD film
deposition phase to achieve a desired internal stress of the
deposited film. While not shown in FIG. 6, it will be appreciated
that the embodiment of multi-station processing tool 3400 may
include a suitable system controller like the embodiment of system
controller 3350 shown in FIG. 5.
[0111] In some embodiments, a system controller 3350 (which may
include one or more physical or logical controllers) controls some
or all of the operations of a process tool. Instructions for
implementing appropriate control operations are executed on the
processor. These instructions may be stored on the memory devices
associated with the controller or they may be provided over a
network. In certain embodiments, the system controller executes
system control software.
[0112] For example, a controller may control the delivery of
appropriate gases (e.g., nitrogen or inert gases), the receipt of a
wafer from outside the apparatus and/or the transfer a wafer from
one station to the next of a multi-station chamber. The controller
may also control the temperature profile, RF power, etc. during
deposition.
[0113] System controller 3350 may include one or more memory
devices 3356, one or more mass storage devices 3354, and one or
more processors 3352. Processor 3352 may include a CPU or computer,
analog and/or digital input/output connections, stepper motor
controller boards, etc.
[0114] In some embodiments, system controller 3350 controls all of
the activities of processing tool 3300. In some embodiments, system
controller 3350 executes machine-readable system control software
3358 stored in mass storage device 3354, loaded into memory device
3356, and executed on processor 3352 so that the apparatus will
perform a method in accordance with the present embodiments.
Alternatively, the control logic may be hard coded in the
controller. Applications Specific Integrated Circuits, Programmable
Logic Devices (e.g., field-programmable gate arrays, or FPGAs) and
the like may be used for these purposes. In the following
discussion, wherever "software" or "code" is used, functionally
comparable hard coded logic may be used in its place.
[0115] System control software 3358 may include instructions for
controlling the timing, mixture of gases, flow rate of process
gases, chamber and/or station pressure, chamber and/or station
temperature, substrate temperature, target power levels, RF power
levels, substrate pedestal, chuck and/or susceptor position, and
other parameters of a particular process performed by processing
tool 3300. System control software 3358 may be configured in any
suitable way. For example, various process tool component
subroutines or control objects may be written to control operation
of the process tool components for performing various process tool
processes. System control software 3358 may be coded in any
suitable computer readable programming language.
[0116] In some embodiments, system control software 3358 may
include input/output control (IOC) sequencing instructions for
controlling the various parameters described above. For example,
each phase of a fabrication process may include one or more
instructions for execution by system controller 3350. The
instructions for setting various process conditions for depositing
low stress, doped silicon layers may be included. Examples of
relevant process conditions are described herein.
[0117] Other computer software and/or programs stored on mass
storage device 3354 and/or memory device 3356 associated with
system controller 3350 may be employed in some embodiments.
Examples of programs or sections of programs for this purpose
include a substrate positioning program, a process gas control
program, a pressure control program, a heater control program, and
a plasma control program.
[0118] A substrate positioning program may include program code for
process tool components that are used to load the substrate onto
process station substrate holder 3318 and to control the spacing
between the substrate and other parts of processing tool 3300.
[0119] A process gas control program may include code for
controlling gas composition and flow rates and optionally for
flowing gas into one or more process stations prior to deposition
in order to stabilize the pressure in the process station. A
pressure control program may include code for controlling the
pressure in the process station by regulating, for example, a
throttle valve in the exhaust system of the process station, a gas
flow into the process station, etc.
[0120] A heater control program may include code for controlling
the current to one or more heating units that are used to heat the
substrate and/or processing chamber. Alternatively, the heater
control program may control delivery of a heat transfer gas (such
as helium) to the substrate.
[0121] A plasma control program may include code for setting RF
power levels applied to the process electrodes in one or more
process stations.
[0122] In some embodiments, there may be a user interface
associated with system controller 3350. The user interface may
include a display screen, graphical software displays of the
apparatus and/or process conditions, and user input devices such as
pointing devices, keyboards, touch screens, microphones, etc.
[0123] In some embodiments, parameters adjusted by system
controller 3350 may relate to process conditions. Non-limiting
examples include process gas composition and flow rates,
temperature, pressure, plasma conditions (such as RF bias power
levels), etc. These parameters may be provided to the user in the
form of a recipe, which may be entered utilizing the user
interface.
[0124] Signals for monitoring the process may be provided by analog
and/or digital input connections of system controller 3350 from
various process tool sensors. The signals for controlling the
process may be output on the analog and digital output connections
of processing tool 3300. Non-limiting examples of process tool
sensors that may be monitored include mass flow controllers,
pressure sensors (such as manometers), thermocouples, etc.
Appropriately programmed feedback and control algorithms may be
used with data from these sensors to maintain process
conditions.
[0125] System controller 3350 may provide program instructions for
implementing the above-described deposition processes. The program
instructions may control a variety of process parameters, such as
DC power level, RF bias power level, pressure, temperature,
duration of exposure to processing gases and/or heat and/or other
energy sources, etc. The instructions may control the parameters to
operate in-situ deposition and further processing (e.g.,
patterning, etching, and activating) of film stacks according to
various embodiments described herein.
Applications
[0126] Many types of device can be fabricated using stacks prepared
as described herein. In various embodiments, the stacks have at
least about 10 layers, or 5 pairs of layers. As mentioned, the
layer pairs can contain a dielectric layer and a silicon layer, an
undoped silicon layer and a doped silicon layer, etc. In some
embodiments, a device stack includes at least about 10 pairs of
layers, or at least about 15 pairs of layers, or at least about 20
pairs of layers, or at least about 25 pairs of layers. In some
embodiments, a device stack includes at least about 70 layers. In
some embodiments, the stack is employed in a memory device such as
a VIM device.
[0127] FIGS. 7A-7P depict fabrication of a memory device as
disclosed herein using an exemplary set of steps. The steps
depicted in FIGS. 7A-7P in no way limit the way in which a memory
device as disclosed herein can be fabricated. It should also be
noted that the figures are not necessarily to scale and do not
necessarily depict the article at every state of preparation, i.e.
intermediate stages of the article may not be illustrated in the
sequence of figures. Further, the reference numbers used in FIGS.
7A-7P do not correspond to reference numbers in other figures,
despite any overlap. The materials and processes discussed with
respect to FIGS. 7A-7P also in no way limit materials or processes
that can be utilized herein. This fabrication process is further
discussed and described in U.S. patent application Ser. No.
12/612,125, filed Nov. 4, 2009, which is herein incorporated by
reference in its entirety.
[0128] FIGS. 7A through 7P are cross-sectional views illustrating a
method for manufacturing a nonvolatile memory device according to
an embodiment of the present invention. Referring to FIG. 7A, an
insulating layer 110, a conductive layer 120 and an insulating
layer 130 are sequentially formed on a semiconductor substrate 100.
The semiconductor substrate 100 may be doped with dopants of first
conductivity type, for example, P-type dopants. The semiconductor
substrate 100 may have a region having a different conductive type,
i.e., a well region. The well region may be formed in a pocket well
or a triple well structure. The insulating layers 110 and 130 may
be formed by depositing an insulating material such as silicon
oxide, silicon nitride, silicon oxynitride or the like. The
conductive layer 120 may be used as a lower select gate. The
conductive layer 120 (hereinafter, referred to as the "lower select
gate") may be formed by depositing a conductive material, such as
doped polycrystalline silicon or metal so that the conductive layer
120 can be used as a gate.
[0129] Referring to FIG. 7B, at least one opening 102 exposing the
semiconductor substrate 100 is formed, by using, for example, a
photolithography and etching process. The first opening 102 may
have, for example, a circular or polygonal (e.g., rectangular)
cross section. The first opening 102 is filled with silicon as will
be described later, to form an active bar used as a channel of a
transistor. If sidewalls 102a of the first opening 102 are formed
inclined at a predetermined angle, the size of the active bar is
varied with its height, so that the channel width of the transistor
may not be uniform. Non-uniform channel width may act as a
hindrance factor in realizing uniform electrical characteristics of
a semiconductor memory device. Therefore, in some embodiments, the
first opening 102 may be formed by using an anisotropic etching
technique, for example, a dry etching technique such that the first
opening 102 has the sidewalls 102a that are vertical to the-surface
of the semiconductor substrate 100.
[0130] Referring to FIG. 7C, an insulating layer 104 is formed on
the sidewalls 102a of the first opening 102. The insulating layer
104 may be used as a gate dielectric (hereinafter, referred to as a
"first gate dielectric portion") of the lower select gate 120. In
one example, the first gate dielectric portion 104 may be formed by
depositing silicon oxide. In the deposition process for forming the
first gate dielectric portion 104, silicon oxide may be deposited
on the semiconductor substrate 100 that is exposed by the first
opening 102. The silicon oxide deposited on the semiconductor
substrate 100 may be removed by using an etching technique so that
the semiconductor substrate 100 and an active bar to be described
later are electrically connected to each other. A spacer for
protecting the silicon oxide layer that is deposited on the
sidewalls 102a of the first opening 102 from the etching may also
be formed prior to the above-mentioned etching step.
[0131] Referring to FIG. 7D, a first active bar ACT1 is deposited
in the first opening 102. The first active bar ACT1 may
substantially fill the first opening 102. The first active bar ACT1
may be formed of the same material as the semiconductor substrate
100. In one example, the first active bar ACT1 may be formed by
depositing a silicon layer and planarizing the deposited silicon
layer through a chemical mechanical polishing (CMP) process. The
silicon layer for forming the first active bar ACT1 may be formed
by depositing polycrystalline silicon or amorphous silicon. In
other embodiments, the first active bar ACT1 may be epitaxially
grown from the portion of the semiconductor substrate 100 that is
exposed by the first opening 102. In this case, the semiconductor
substrate 100 and the first active bar ACT1 may be a single
crystalline silicon which is continuous without crystalline
defects. The first active bar ACT may be doped with the same
conductivity type dopants as the semiconductor substrate 100. In
some embodiments, the semiconductor substrate 100 and the first
active bar ACT1 may have a P-type conductivity. Alternatively, the
first active bar ACT1 may be not doped with dopants.
[0132] Through the foregoing processes, a first structure 101
including the semiconductor substrate 100, the lower select gate
120 above the semiconductor substrate 100, and the first active bar
ACT1 used as a channel of the lower select gate 120 may be
manufactured.
[0133] Referring to FIG. 7E, a first insulating layer group 200a
and a first conductive layer group 200b are formed on the first
structure 101. The first insulating layer group 200a may include a
plurality of insulating layers 210, 230, 250, 270, 290. The first
conductive layer group 200b may include a plurality of conductive
layers 220, 240, 260, 280. The plurality of conductive layers 220,
240, 260, 280 and the plurality of insulating layers 210, 230, 250,
270, 290 may be alternatingly disposed to form a stacked sandwich
structure. The insulating layer 210 may be positioned at the bottom
of the stacked structure directly on the first structure 101, and
the insulating layer 290 may be positioned at the top of the
stacked structure to cover the conductive layer 280. Each of the
plurality of conductive layers 220, 240, 260, 280 of the first
conductive layer group 200b may be used as a cell gate pattern. The
plurality of conductive layers 220, 240, 260, 280 may be formed by
depositing doped polycrystalline silicon or metal. As used herein,
the term "the first conductive layer group" 200b may be
interchanged with the term "first cell gate pattern group", and the
term "each of the plurality of conductive layers" 220, 240, 260,
280 may be interchanged with the term "first cell gate
pattern."
[0134] Each of the plurality of conductive layers 220, 240, 260,
280 may have substantially the same thickness. Since thickness
and/or interval of each of the plurality of conductive layers 220,
240, 260, 280 may determine a channel length, the thickness and/or
interval of each of the plurality of conductive layers 220, 240,
260, 280 may be selected in a range that can solve limitations in
the electrical characteristics due to a short channel. Also, since
the plurality of conductive layers 220, 240, 260, 280 can be formed
by deposition, the channel length can be controlled precisely.
[0135] The plurality of insulating layers 210, 230, 250, 270, 290
may be formed by depositing silicon oxide, for example. In further
embodiments, the plurality of insulating layers 210, 230, 250, 270,
290 may be formed of high-k dielectric. In still further
embodiments, each of the plurality of insulating layers 210, 230,
250, 270, 290 may be formed by depositing a material (e.g., silicon
nitride, silicon oxynitride) having a dielectric constant higher
than silicon oxide.
[0136] The number, thickness, material and the like of layers
constituting the first insulating layer group 200a and the first
conductive layer group 200b may be modified in various forms with
consideration of electrical characteristics of a memory transistor
and technical difficulties in the process of patterning these
layers. Each of the first insulating layer group 200a and the first
conductive layer group 200b may be formed in a step form.
[0137] Referring to FIG. 7F, the first insulating layer group 200a
and the conductive layer group 200b are patterned via a
photolithography and etching process to form a second opening 202
that exposes the first active bar ACT1. The first gate dielectric
portion 104 may be exposed by the second opening 202. The second
opening 202 may be formed by using, for example, a dry etching so
as to have vertical sidewalls 202a and thus realize uniform
electrical characteristics for a transistor. The first opening 102
and the second opening 202 can be connected in a vertical
direction.
[0138] The number of the plurality of conductive layers 220, 240,
260, 280 constituting the first conductive layer group 200b and the
number of the plurality of insulating layers 210, 230, 250, 270,
290 constituting the first insulating layer group 200a may be
varied from the four-conductive-layer and five-insulating-layer
embodiment depicted in the figures. In this embodiment, while the
number of conductive layers 220, 240, 260, 280 constituting the
first conductive layer group 200b is set at 4 layers and the number
of insulating layers 210, 230, 250, 270, 290 constituting the first
insulating layer group 200a is set at 5 layers, these numbers are
arbitrary.
[0139] Referring to FIG. 7G, an insulating layer 204 is formed on
the sidewalls 202a of the second opening 202. The insulating layer
204 may be used as a gate dielectric (hereinafter, referred to as
the "second gate dielectric portion") of a first cell gate pattern
group 200b. The second gate dielectric portion 204 may include a
data storage layer. In one example, the second gate dielectric
portion 204 may be formed in a triple layer structure by
sequentially depositing a silicon oxide layer, a silicon nitride
layer or a silicon oxynitride layer, and a silicon oxide layer or
the like. In the second gate dielectric portion 204, the silicon
nitride layer or the silicon oxynitride layer is used as a charge
storage layer for trapping a charge and storing information, and
any one of the two silicon oxide layers is used as a blocking
barrier and the other is used as a tunnel barrier.
[0140] The second gate dielectric portion 204 may be formed on the
first active bar ACT1 that is exposed by the second opening 202.
The second gate dielectric portion 204 that is deposited on the
first active bar ACT1 during the deposition process may be removed
by using an etching technique so that the first active bar ACT1 and
a second active bar ACT2 may be electrically connected to each
other. Prior to this etching process, a spacer for protecting the
second gate dielectric portion 204 that is deposited on the
sidewalls 202a of the second opening 202 from the etching may be
formed.
[0141] Referring to FIG. 7H, a second active bar ACT2 is formed in
the second opening 202 by using a process that is the same as or
similar to that described with reference to FIG. 7D. The second
active bar ACT2 may be connected to the first active bar ACT1 in a
vertical direction, and may also be connected to the surface of the
semiconductor substrate 100. The second active bar ACT2 may be
formed of the same material as the first active bar ACT1. In some
embodiments, the second active bar ACT2 may be formed by depositing
amorphous or polycrystalline silicon and planarizing the deposited
amorphous or polycrystalline silicon layer through a chemical
mechanical polishing (CMP) process. In other embodiments, the
second active bar ACT2 may be formed by epitaxially growing single
crystalline silicon from the first active bar ACT1. The second
active bar ACT2 may be formed to have the same conductivity type as
the semiconductor substrate 100 and the first active bar ACT1, for
example, P-type conductivity. Alternatively, the semiconductor
substrate 100 and the first and second active bars ACT1 and ACT2
may not be doped with dopants.
[0142] Through the foregoing processes, a second structure 201
including the plurality of cell gate patterns 220, 240, 260, 280 on
the first structure 101, and the second active bar ACT2 that is
used as a channel of the plurality of cell gate patterns 220, 240,
260, 280 can be formed.
[0143] Referring to FIG. 7I, a second insulating layer group 300a
that includes a plurality of insulating layers 310, 330, 350, 370,
390 and a second conductive layer group 300b that includes a
plurality of conductive layers 320, 340, 360, 380 which are
alternatingly disposed with the plurality of insulating layers 310,
330, 350, 370, 390 are formed in a stacked sandwich structure by
using a process that is the same as or similar to that described
with reference to FIG. 7E. The insulating layer 310 is disposed at
the bottom of the stack, directly on the second structure 201, and
the insulating layer 390 is at the top of the stack and covers the
conductive layer 380. Each of the plurality of conductive layers
320, 340, 360, 380 of the second conductive layer group 300b may be
used as a cell gate pattern. For this purpose, the plurality of
conductive layers 320, 340, 360, 380 may be formed by depositing
doped polycrystalline silicon or metal. The plurality of conductive
layers 320, 340, 360, 380 may each have substantially the same
thickness. In this specification, the term "second conductive layer
group 300b" may be exchanged with the term "second cell gate
pattern group", and the term "each of the plurality of conductive
layers" 320, 340, 360, 380 may be exchanged with the term in "each
second cell gate pattern." The plurality of insulating layers 310,
330, 350, 370, 390 may be formed by depositing silicon nitride or
silicon oxynitride having a dielectric constant higher than silicon
oxide. Each of the second insulating layer group 300a and the
second conductive layer group 300b may be formed in a step
form.
[0144] Referring to FIG. 7J, the second insulating layer group 300a
and the second conductive layer group 300b are patterned using a
process that is the same as or similar to that described with
reference to FIG. 7F to form a third opening 302 that exposes the
second active bar ACT2. The second gate dielectric portion 204 may
be exposed by the third opening 302. The third opening 302 may be
formed using, for example, a dry etching process so as to provide
an opening having vertical sidewalls 302 and to thus realize
uniform electrical characteristics for the transistors. The second
opening 202 and the third opening 302 may be connected in a
vertical direction. Although the third opening 302 is formed
inclined with a predetermined angle, the number of conductive
layers 320, 340, 360, 380 constituting the second conductive layer
group 300b and the number of insulating layers 310, 330, 350, 370,
390 constituting the second insulating layer group 300a may be
selected such that a third active bar can be formed at a height
that can neglect an increasing degree of cell dispersion.
[0145] In one example, the second conductive layer group 300b and
the first conductive layer group 200b may be formed in the same
structure and the second insulating layer group 300a and the first
insulating layer group 200a may be formed in the same
structure.
[0146] Referring to FIG. 7K, an insulating layer 304 is formed on
the sidewalls 302a of the third opening 302 using a process that is
the same as or similar to that described with reference to FIG. 7G.
The insulating layer 304 may be used as a gate dielectric
(hereinafter, referred to as the "third gate dielectric") of the
second cell gate pattern group 300b. The third gate dielectric
portion 304 may be formed with the same structure as the second
gate dielectric portion 204. In one example, the third gate
dielectric portion 304 may be formed in a triple layer structure by
sequentially depositing a silicon oxide layer that acts as a
blocking barrier, a silicon nitride layer or a silicon oxynitride
layer that acts as a charge storage layer, and a silicon oxide
layer that acts as a tunnel barrier.
[0147] In the deposition process for forming the third gate
dielectric portion 304, the third gate dielectric portion 304 may
be deposited on the second active bar ACT2 that is exposed by the
third opening 302. The third gate dielectric portion that is
deposited on the second active bar ACT2 may be removed via etching.
A spacer for protecting the third gate dielectric portion 304
deposited on the sidewalls 302a of the third opening 3 02 from the
etching may be formed prior to the etching operation.
[0148] Referring to FIG. 7L, a third active bar ACT3 is deposited
in the third opening 302 using a process that is the same as or
similar to that described with reference to FIG. 7D. The third
active bar ACT3 may be connected to the second active bar ACT2 in a
vertical direction and to the surface of the semiconductor
substrate 100. The third active bar ACT3 may be formed of the same
material as the second active bar ACT2. In some embodiments, the
third active bar ACT3 may be formed by depositing amorphous or
polycrystalline silicon and then planarizing the deposited
amorphous or polycrystalline silicon layer through a chemical
mechanical polishing (CMP) process, or by epitaxially growing a
single crystalline silicon layer from the second active bar ACT2.
The third active bar ACT3 may be formed with the same conductivity
type as the semiconductor substrate 100, the first active bar ACT1
and the second active bar ACT2, for example, P-type conductivity.
In other embodiments, the semiconductor substrate 100, the first,
second and third active bars may not be doped with dopants.
[0149] Through the foregoing processes, a third structure 301
including the plurality of cell gate patterns 320, 340, 360, 380 on
the second structure 201, and the third active bar ACT3 used as a
channel of the plurality of cell gate patterns 320, 340, 360, 380
can be formed.
[0150] Referring to FIG. 7M, an insulating layer 410, a conductive
layer 420, and an insulating layer 430 are sequentially formed on
the third structure 301 by, for example, a process that is the same
as or similar to that described with reference to FIG. 7A. The
insulating layers 410, 430 may be formed by depositing an
insulating material such as silicon oxide, silicon nitride, silicon
oxynitride or the like. The conductive layer 420 may be used as an
upper select gate. The conductive layer 420 (hereinafter, referred
to as the "upper select gate") may be formed by depositing a
conducive material layer, for example, doped polycrystalline
silicon or metal such that the conductive layer 420 can be used as
a gate.
[0151] Referring to FIG. 7N, a fourth opening 402 that exposes the
third active bar ACT3 is formed via a process that is the same as
or similar to that described with reference to FIG. 7B. The fourth
opening 402 may be formed using, for example, a dry etching process
so as to have substantially vertical sidewalls 402a.
[0152] Referring to FIG. 7O, an insulating layer 404 is formed on
sidewalls 402a of the third opening 302 via a process that is the
same as or similar to that described with reference to FIG. 7C. The
insulating layer 404 may be used as a gate dielectric (hereinafter,
referred to as the "fourth gate dielectric") of the upper select
gate 420. The fourth gate dielectric 404 may be formed by
depositing, for example, a silicon oxide layer. In the deposition
process for forming the fourth gate dielectric 404, the silicon
oxide layer may be deposited on the third active bar ACT3 that is
exposed by the fourth opening 402. The silicon oxide layer
deposited on the third active bar ACT3 may be removed using an
etching process. In this case, a spacer for protecting the silicon
oxide layer that is deposited on the sidewalls 402a of the fourth
opening 402 from the etching may be formed prior to the etching
step.
[0153] Referring to FIG. 7P, a fourth active bar ACT4 is formed in
the fourth opening 402 using a process that is the same as or
similar to that described with reference to FIG. 7D. The fourth
active bar ACT4 may be connected to the third active bar ACT3 in a
vertical direction and to the surface of the semiconductor
substrate 100. The fourth active bar ACT4 may be formed of the same
material as the third active bar ACT3. In some embodiments, the
fourth active bar ACT4 may be formed by depositing amorphous or
polycrystalline silicon and then planarizing the deposited
amorphous or polycrystalline silicon layer through a chemical
mechanical polishing (CMP) process. In other embodiments, the
fourth active bar ACT4 may be formed by epitaxially growing a
single crystalline silicon layer from the third active bar ACT3.
The fourth active bar ACT 4 may be formed to have the same
conductivity type as the semiconductor substrate 100, for example,
P-type conductivity. In other embodiments, the semiconductor
substrate 100, the first, second, third and fourth active bars
ACT1, ACT2, ACT3 and ACT 4 may not be doped with dopants.
[0154] Through the foregoing processes, a fourth structure 401 that
includes the upper select gate and the fourth active bar ACT4
(which is used as a channel of the upper select gate) can be formed
on the third structure 301. The active bar which is formed on the
semiconductor substrate 100 may be formed of the same material
(e.g., silicon) as the semiconductor substrate 100, and may have
the same conductivity type as the semiconductor substrate 100.
[0155] In particular, since the active bar ACT is formed by
separately forming the first to fourth active bars ACT1 to ACT 4,
it is possible to minimize a width difference according to height
of each of the first to fourth active bars ACT1 to ACT 4.
Accordingly, since the width of the active bar ACT can be set to
not vary greatly according to its height, deterioration of cell
dispersion characteristics can be reduced and/or minimized.
[0156] A bit line 610 that is electrically connected to the active
bar ACT may be formed on the fourth structure 401. The bit line 610
may be formed to extend in a direction that crosses the upper
select gate 420. In one example, the bit line 610 may be formed by
depositing an aluminum layer on the fourth structure 401 and then
patterning the deposited aluminum layer through a photolithography
and etching process.
[0157] Additionally, the apparatus and methods described herein may
be used in conjunction with lithographic patterning tools or
processes, for example, for the fabrication or manufacture of
semiconductor devices, displays, LEDs, photovoltaic panels and the
like that incorporate stacks as described. Typically, though not
necessarily, such tools/processes will be used or conducted
together in a common fabrication facility. Lithographic patterning
of a film typically comprises some or all of the following steps,
each step enabled with a number of possible tools: (1) application
of photoresist on a workpiece, i.e., substrate, using a spin-on or
spray-on tool; (2) curing of photoresist using a hot plate or
furnace or UV curing tool; (3) exposing the photoresist to visible
or UV or extreme UV light or an electron beam with a tool such as a
wafer stepper; (4) developing the resist so as to selectively
remove resist and thereby pattern it using a tool such as a wet
bench; (5) transferring the resist pattern into an underlying film
or workpiece by using a dry or plasma-assisted etching tool; and
(6) removing the resist using a tool such as an RF or microwave
plasma or wet resist stripper. One embodiment is a method as
described herein further including: applying photoresist to the
wafer to define the locations of silicon containing stacks;
exposing the photoresist to an energy source; patterning the resist
and transferring the pattern to the wafer; and selectively removing
the photoresist from the wafer. One embodiment is system including
an apparatus as described herein further comprising a stepper.
[0158] It is to be understood that the configurations and/or
approaches described herein are exemplary in nature, and that these
specific embodiments or examples are not to be considered in a
limiting sense, because numerous variations are possible. The
specific routines or methods described herein may represent one or
more of any number of processing strategies. As such, various acts
illustrated may be performed in the sequence illustrated, in other
sequences, in parallel, or in some cases omitted. Likewise, the
order of the above described processes may be changed.
[0159] The subject matter of the present disclosure includes all
novel and nonobvious combinations and sub-combinations of the
various processes, systems and configurations, and other features,
functions, acts, and/or properties disclosed herein, as well as any
and all equivalents thereof.
EXPERIMENTAL
[0160] The following section and figures relate the results of
experiments that were performed to demonstrate the effects of
processing parameters discussed above on the characteristics of the
stack structure and layers therein.
[0161] The experiments shown in FIGS. 8-11 produced a single layer
of doped silicon of 1000 .ANG. thick on a 1000 .ANG. thick thermal
oxide. In contrast, the experiments shown in FIG. 12 produced a
stack comprising 73 layers, which stack includes alternating layers
of oxide (200 .ANG. thick) and doped silicon (300 .ANG. thick),
starting and ending with oxide.
[0162] The experiments were conducted in a 300 mm VECTOR.RTM.
STRATA.TM. module. Phosphine was provided as a 5% composition and
95% helium by volume. The reaction pressure was 6 Torr. In these
experiments, a process gas comprising silane and phosphine was used
to deposit a layer of phosphorus doped silicon film on a 300 mm
diameter wafer in a four showerhead four wafer station reaction
chamber. Rapid thermal anneal conditions were used to accomplish
activation.
[0163] Table 1 shows data of a series of experiments demonstrating
the impact of silane mass flow rate, high-frequency radiofrequency
power, deposition temperature, and phosphine mass flow rate on the
beneficial properties of the deposited film including internal
stress, smoothness, and conductivity. The experiments were
conducted in blocks, each block containing three to four
experiments. Both the silane flow rate and the phosphine flow rate
were constant within a block, and they varied between blocks. For
instance, for block one, silane flow rate was 915 sccm and
phosphine flow rate was 457.5 sccm (phosphine is provided as a 5%
composition and 95% helium). For block six, silane flow rate was
1450 sccm and phosphine flow rate was 725 sccm (phosphine is
provided as a 5% composition and 95% helium), and so on. The CP
block data were obtained using the SiH.sub.4 flow set-point center
point of the experiments. All experiments were conducted while
maintaining a substrate pedestal temperature of 525.degree. C. Each
of the three or four experiments in a block was conducted at a
different high-frequency RF power, ranging from 250 to 1000.
[0164] Table 1 also shows data of the deposition rate
(.ANG./minute), the mean and range of the thickness of the layer as
deposited (.ANG.), thickness non-uniformity and half-radius
thickness non-uniformity (%), and stress (MPa) and bow (.mu.m) of
the layer.
TABLE-US-00001 TABLE 1 HDR DOE3 Pedestal PH3 DepRate THK THK THK
THK R/2 Stress Bow Block Recipe Temp .degree. C. SiH4 HF RF Trans
Flow (A/min) Mean Range NU % NU % (MPa) (um) One 1 525 915 500 7
457.5 1469 996 29 0.6 1.46 -116 -9 2 525 915 750 7 457.5 1694 978
25 0.7 1.3 -112 -8 3 525 915 250 7 457.5 989 1001 36 1.1 1.8 -131
-9 Two 4 525 1022 500 7 511 1492 979 26 0.7 1.34 -97 -6 5 525 1022
750 7 511 1761 976 24 0.6 1.21 -149 -10 6 525 1022 250 7 511 1008
1014 41 1.3 2.04 -135 -9 Three 7 525 1129 500 7 564.5 1537 977 23
0.7 1.17 -94 -7 8 525 1129 750 7 564.5 1827 979 20 0.5 1.04 -138
-10 9 525 1129 250 7 564.5 1018 1010 52 1.6 2.55 -109 -8 10 525
1129 1000 7 564.5 2090 1008 26 0.5 1.27 -179 -12 CP CP1 525 1182.5
500 7 591.3 1556 976 24 0.7 1.2 -88 -6 Block CP2 525 1182.5 750 7
591.3 1860 986 25 0.6 1.26 -125 -9 CP3 525 1182.5 250 7 591.3 1018
1006 64 1.9 3.19 -130 -9 CP4 525 1182.5 1000 7 591.3 2055 978 20
0.5 1.02 -179 -12 Four 11 525 1236 500 7 618 1571 993 33 1.1 1.65
-79 -6 12 525 1236 750 7 618 1883 997 24 0.8 1.19 -122 -9 13 525
1236 250 7 618 1019 1010 74 2.3 3.64 -126 -9 Five 14 525 1343 500 7
671.5 1604 1000 43 1.4 2.15 -77 -6 15 525 1343 750 7 671.5 1940 979
22 0.8 1.13 -99 -7 16 525 1343 250 7 671.5 1053 1044 62 1.7 2.99
-126 -9 Six 17 525 1450 500 7 725 1655 1030 40 1.1 1.95 -75 -5 18
525 1450 750 7 725 2003 1017 30 0.7 1.46 -80 -6 19 525 1450 250 7
725 1038 1013 83 2.5 4.12 -128 -9
[0165] Data shown in Table 1 were used to generate plots in FIGS.
8A-8C. The plots show the bow, internal stress, and thickness
non-uniformity as functions of silane flow rate (left panel) and
the high-frequency RF power (right panel). Each data point in the
left panel of a figure is based on a value averaged across multiple
high-frequency RF voltages. A smooth curve is fitted to data points
in a plot and shown as a solid curve, while corresponding estimated
errors are shown by dash lines. Each data point in the right panel
is based on a value averaged across multiple silane flow rates.
[0166] FIGS. 8A-8C show the same data, but they identify
independent variable (IV) values (silane flow rate and HF RF power)
for 3 specific dependent variable (DV) values based on the fitted
curves: center point of stress (FIG. 8A), lowest stress (FIG. 8B),
and best uniformity (FIG. 8C). Based on the identified IV values,
associated DV other than the 3 specific DV values are also
identified in each subplot of FIGS. 8A-8C. Specifically, the center
point of stress corresponds to a silane flow rate of 1200 sccm, and
HF RF power of 500 W, see FIG. 8A. The lowest stress occurs at a
silane flow rate of 1450 sccm, and HF RF power of 570 W, see FIG.
8B. The best non-uniformity also occurs at a silane flow rate of
1450 sccm, and HF RF power of 570 W, see FIG. 8C. Importantly,
increasing silane flow rates correspond to reduced bow and
compressive internal stress, and improved uniformity of the silicon
layer (see respective subplots from bottom to top, left panels of
FIGS. 8A-8C).
[0167] FIG. 9 depicts the impact of high-frequency RF power on the
bulk resistivity of the deposited silicon film after it was
annealed or activated using rapid thermal annealing. It shows that
relatively high power is required in order to reduce the bulk
resistivity of the activated film below 0.0011 ohm-centimeter. This
figure suggests to me that the high-frequency RF power should be at
about 500 W or higher, or 1000 W or higher, in order to obtain
suitably low resistivity in the silicon film. FIG. 9 also shows
bulk resistivity as a function of silane flow rate. However, silane
flow rate does not seem to consistently modulate resistivity under
different high-frequency RF power.
[0168] FIGS. 10A and 10B show the impact of phosphine to silane
flow rate ratio on the bulk resistivity and sheet resistance of the
deposited silicon films after activation. As can be seen from the
plots, the percentage of phosphine in the phosphine silane mixture
lower than about 3.5% leads to relatively low bulk resistivity
(<0.00115 ohmcm in FIG. 10A) and sheet resistance (<400
ohmsquare in FIG. 10B). Here, the percentage of phosphine is
calculated as 100*(PH.sub.3/SiH.sub.4) per volume.
[0169] FIGS. 11A and 11B illustrate data at different deposition
temperatures. It also shows the data as a function of silane flow
rate. Basically they show the impact of temperature and silane flow
rate on internal stress of the deposited film. The diamonds shown
on the lower left side of the graphs represent data taken at
550.degree. C. and for relatively low silane flow rates. The data
shown by the circular dots further up and to the right were
generated under experimental condition of 525.degree. C. and at
relatively higher silane flow rates. The triangles on the far right
represent data obtained at 525.degree. C. and with a relatively
high silane flow rate. At that same silane flow rate, data was
taken at a deposition temperature of 500.degree. C. This is
reflected by the square in the upper right. As can be seen, the
data obtained at lower deposition temperatures and higher silane
flow rates show the most desirable internal stress values for the
deposited silicon film. In fact, the data taken at a flow rate of
1450 sccm and 500.degree. C. actually had a tensile internal
stress.
[0170] FIG. 11B shows a subset of data in FIG. 11A, showing only
data produced with a 500 W of high-frequency RF power. However, the
same basic trends can be seen. At relatively high deposition
temperatures (e.g., 550.degree. C.), the internal compressive
stress decreases with increasing silane flow rate. Further
improvements in internal stress can be achieved by lowering the
deposition temperature from 550.degree. C. to 525.degree. C. and
onward to 500.degree. C.
[0171] FIG. 12 show data for a 73-layer stack comprising
alternating oxide layers and silane doped silicon layers, starting
and ending with oxide. Silicon oxide was deposited to a thickness
of 200 .ANG. from a TEOS precursor. Silane doped silicon was
deposited at 300 .ANG. from silane and phosphine under various
process conditions as described herein.
[0172] In FIG. 12A, the bow of the 73 layer stack is shown against
silane flow rate as indicated by diamond-shaped data points. As can
be seen, improvement (reduction) in stack bow was achieved with
higher silane flow rates. In FIG. 12B, the bow of the 73 layer
stack is shown against silane flow rate as open circle in addition
to the data in FIG. 12A, but the experimental condition differs
from that of FIG. 12A in that here the PosA is longer, wherein PosA
is the time that is post deposition with the HF RF power staying on
and silane and dopant flows turned off. As apparent from comparison
of the conditions, longer post deposition HF RF power led to
reduced bow of the stack.
[0173] In FIG. 12C, the bow of the 73 layer stack is shown against
silane flow rate of about 915 sccm as solid circles in addition to
the data in FIG. 12B. Data show that lowering the processing
temperature from 575.degree. C. to 525.degree. C. caused the bow to
shrink from about -135 .mu.m to about -105 .mu.m.
[0174] FIG. 12D shows the same data but with additional data points
illustrating the impact of even higher silane flow rates at
525.degree. C. As shown by hatched data points, the stack bow
improved further with this increased silane flow rate at about 1450
sccm.
[0175] Finally, FIG. 12E depicts the impact of silane flow rate
increase and deposition temperature decrease on roughness of the
deposited 73 layer stack. As shown, the increase in silane flow
rate from 340 to 915 sccm shown by diamond-shaped data points
reduced roughness of the stack surface as indicated by haze
measured by a KLA-Tencor metrology SP1 tool. Haze is a
manifestation of the roughness of the surface. A lower haze value
indicates a smoother surface. In addition, longer PosA (see the
open-circle data point) and lower processing temperature (see the
solid circle data points) also led to lower haze and roughness.
Further increasing silane flow rate to 1450 sccm at 525.degree. C.
did not further reduce roughness.
[0176] The present invention may be embodied in other specific
forms without departing from its spirit or essential
characteristics. The described embodiments are to be considered in
all respects only as illustrative and not restrictive. The scope of
the invention is, therefore, indicated by the appended claims
rather than by the foregoing description. All changes which come
within the meaning and range of equivalency of the claims are to be
embraced within their scope.
* * * * *