U.S. patent application number 15/016527 was filed with the patent office on 2017-08-10 for preconditioning to enhance hydrophilic fusion bonding.
The applicant listed for this patent is Taiwan Semiconductor Manufacturing Co., Ltd.. Invention is credited to Chih-Hang Chang, Chien-Ning Hsin, Jen-Hao Liu, I-Shi Wang.
Application Number | 20170225948 15/016527 |
Document ID | / |
Family ID | 59410655 |
Filed Date | 2017-08-10 |
United States Patent
Application |
20170225948 |
Kind Code |
A1 |
Hsin; Chien-Ning ; et
al. |
August 10, 2017 |
PRECONDITIONING TO ENHANCE HYDROPHILIC FUSION BONDING
Abstract
A method for fusion bonding a pair of substrates together with
silane preconditioning is provided. A surface of a first oxide
layer or a surface of a second oxide layer is preconditioned with
silane. The first and second oxide layers are respectively arranged
on first and second semiconductor substrates. Water is applied to
the surface of the first or second oxide layer. The surfaces of the
first and second oxide layers are brought in direct contact. The
first and second oxide layers are annealed. A method for
manufacturing a microelectromechanical systems (MEMS) package using
the fusion bonding is also provided.
Inventors: |
Hsin; Chien-Ning; (Hsinchu
City, TW) ; Wang; I-Shi; (Sanxia Township, TW)
; Liu; Jen-Hao; (Zhunan Township, TW) ; Chang;
Chih-Hang; (Taoyuan City, TW) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Taiwan Semiconductor Manufacturing Co., Ltd. |
Hsin-Chu |
|
TW |
|
|
Family ID: |
59410655 |
Appl. No.: |
15/016527 |
Filed: |
February 5, 2016 |
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
B81C 2203/036 20130101;
B81C 2203/037 20130101; H01L 21/02334 20130101; H01L 21/02337
20130101; B81C 1/00357 20130101; B81C 1/00269 20130101; H01L
21/2007 20130101; H01L 21/3105 20130101 |
International
Class: |
B81C 1/00 20060101
B81C001/00 |
Claims
1. A method for bonding a pair of semiconductor substrates
together, the method comprising: preconditioning a surface of a
first oxide layer or a surface of a second oxide layer with silane,
wherein the first and second oxide layers are respectively arranged
on first and second semiconductor substrates; applying water to the
surface of the first or second oxide layer; bringing the surfaces
of the first and second oxide layers in direct contact; and
annealing the first and second oxide layers.
2. The method according to claim 1, further comprising:
preconditioning the surfaces of the first and second oxide layers
with silane prior to bringing the surfaces in direct contact.
3. The method according to claim 1, wherein the second oxide layer
has a thickness less than about 5 nanometers, and wherein the
method further comprises: forming the first oxide layer on the
second semiconductor substrate with a thickness exceeding about 50
nanometers.
4. The method according to claim 1, wherein the second oxide layer
is a native oxide.
5. The method according to claim 1, wherein preconditioning the
surface of the first or second oxide layer comprises applying
monosilane (SiH4) to the surface of the first or second oxide
layer.
6. The method according to claim 1, wherein preconditioning the
surface of the first or second oxide layer comprises applying heat
to the first or second oxide layer.
7. The method according to claim 1, further comprising: activating
the surface of the first or second oxide layer by a plasma
treatment that applies a plasma to the surface of the first or
second oxide layer.
8. The method according to claim 7, further comprising: cleaning
the surface of the first or second oxide layer before activating
the surface of the first or second oxide layer; and rinsing the
surface of the first or second oxide layer after activating the
surface of the first or second oxide layer.
9. The method according to claim 7, further comprising: activating
the surfaces of the first and second oxide layers by the plasma
treatment.
10. A method for forming a microelectromechanical systems (MEMS)
package, the method comprising: forming an oxide layer on a capping
substrate; preconditioning a surface of the oxide layer by exposing
the surface of the oxide layer to silane; fusion bonding a device
substrate to the capping substrate through the oxide layer after
exposing the oxide layer to the silane, wherein the fusion bonding
is performed at the surface of the oxide layer; forming a movable
element in the device substrate; and bonding the device substrate
to a support device on an opposite side of the device substrate as
the capping substrate.
11. The method according to claim 10, wherein fusion bonding the
device substrate to the capping substrate comprises:
preconditioning a surface of a second oxide layer with the silane,
the second oxide layer arranged on the device substrate; and
arranging the surfaces of the oxide layer and the second oxide
layer in direct contact.
12. The method according to claim 11, wherein fusion bonding the
device substrate to the capping substrate comprises: applying water
to the surface of the oxide layer or the surface of the second
oxide layer; and annealing the oxide layer and the second oxide
layer.
13. The method according to claim 11, wherein the second oxide
layer has a thickness less than about 5 nanometers, and wherein the
method further comprises: forming the oxide layer with a thickness
exceeding about 50 nanometers.
14. The method according to claim 11, wherein the second oxide
layer is a native oxide.
15. The method according to claim 11, wherein preconditioning the
surface of the oxide layer or the surface of the second oxide layer
comprises: applying monosilane (SiH4) to the surface of the oxide
layer or the surface of the second oxide layer.
16. The method according to claim 11, wherein preconditioning the
surface of the oxide layer or the surface of the second oxide layer
comprises: applying heat to the oxide layer or the second oxide
layer.
17. The method according to claim 11, further comprising:
activating the surface of the oxide layer or the surface of the
second oxide layer by plasma treatment.
18. A method for bonding a pair of semiconductor substrates
together, the method comprising: forming an oxide layer on a first
semiconductor substrate; preconditioning a surface of a native
oxide layer with silane, the native oxide layer arranged on a
second semiconductor substrate; arranging a surface of the oxide
layer and the surface of the native oxide layer in direct contact;
and annealing the oxide layer and the native oxide layer.
19. The method according to claim 18, wherein preconditioning the
surface of the native oxide layer comprises applying monosilane
(SiH4) to the surface of the native oxide layer.
20. The method according to claim 18, further comprising: applying
water to the surface of the oxide layer or to the surface of the
native oxide layer.
Description
BACKGROUND
[0001] Fusion bonding (also known as direct bonding) is a process
for joining surfaces without intermediate layers. The process is
based on the formation of chemical bonds between the surfaces when
the surfaces are sufficiently flat, clean, and smooth. Fusion
bonding has many applications in the semiconductor manufacturing
industry. For example, fusion bonding may be used to package
microelectromechanical systems (MEMS) devices, such as
accelerometers, pressure sensors, and gyroscopes, or to manufacture
semiconductor substrates, such as silicon-on-insulator (SOI)
substrates.
BRIEF DESCRIPTION OF THE DRAWINGS
[0002] Aspects of the present disclosure are best understood from
the following detailed description when read with the accompanying
figures. It is noted that, in accordance with the standard practice
in the industry, various features are not drawn to scale. In fact,
the dimensions of the various features may be arbitrarily increased
or reduced for clarity of discussion.
[0003] FIGS. 1-9 illustrate a series of cross-sectional views of
some embodiments of a method for fusion bonding a pair of
substrates together with silane preconditioning.
[0004] FIG. 10 illustrates a flowchart of some embodiments of the
method of FIGS. 1-9.
[0005] FIG. 11 illustrates a process tool for performing silane
preconditioning in FIGS. 1-9.
[0006] FIG. 12 illustrates a cross-sectional view of some
embodiments of a microelectromechanical systems (MEMS) package with
a pair of substrates fusion bonded according to the method of FIGS.
1-9.
[0007] FIGS. 13-21 illustrate a series of cross-sectional views of
some embodiments of a method for forming the MEMS package of FIG.
12.
[0008] FIG. 22 illustrates a flowchart of some embodiments of the
method of FIGS. 13-21
DETAILED DESCRIPTION
[0009] The present disclosure provides many different embodiments,
or examples, for implementing different features of this
disclosure. Specific examples of components and arrangements are
described below to simplify the present disclosure. These are, of
course, merely examples and are not intended to be limiting. For
example, the formation of a first feature over or on a second
feature in the description that follows may include embodiments in
which the first and second features are formed in direct contact,
and may also include embodiments in which additional features may
be formed between the first and second features, such that the
first and second features may not be in direct contact. In
addition, the present disclosure may repeat reference numerals
and/or letters in the various examples. This repetition is for the
purpose of simplicity and clarity and does not in itself dictate a
relationship between the various embodiments and/or configurations
discussed.
[0010] Further, spatially relative terms, such as "beneath,"
"below," "lower," "above," "upper" and the like, may be used herein
for ease of description to describe one element or feature's
relationship to another element(s) or feature(s) as illustrated in
the figures. The spatially relative terms are intended to encompass
different orientations of the device in use or operation in
addition to the orientation depicted in the figures. The apparatus
may be otherwise oriented (rotated 90 degrees or at other
orientations) and the spatially relative descriptors used herein
may likewise be interpreted accordingly.
[0011] Some microelectromechanical systems (MEMS) devices are
packaged according to a process that utilizes an epitaxial
deposition. In such a process, a recess is formed in a silicon
substrate, a first oxide layer is formed lining the silicon
substrate and the recess, and the recess is filled with a second
oxide layer. A chemical mechanical polish (CMP) is performed into
the second oxide layer to coplanarize upper surfaces of the oxide
layers, and a silicon layer is formed by epitaxy over the silicon
substrate and the oxide layers. With the silicon layer formed, a
first etch is performed to form an opening exposing the second
oxide layer and a second etch is performed to remove the second
oxide layer through the opening. However, this approach to
packaging MEMS devices is complicated and requires an expensive
photomask for exposing the second oxide layer.
[0012] Other MEMS devices are packaged according to a process that
utilizes fusion bonding. In such a process, a recess is formed in a
first silicon substrate, and a first oxide layer is formed lining
the silicon substrate and the recess. A second silicon substrate is
arranged over the first silicon substrate and the first oxide
layer, and is fusion bonded to the first silicon substrate through
the first oxide layer. During fusion bonding, a surface of a second
oxide layer of native oxide on the second substrate is activated,
and the surface of the second oxide layer or a surface of the first
oxide layer is covered in water. The surfaces of the first and
second oxide layers are then brought into contact and annealed.
[0013] While the fusion-bonding approach can be performed with one
less photomask than the epitaxial approach, the strength of the
bond may be weak due to a low density of silanol groups on the
surface of the first or second oxide layer. This may, in turn, lead
to peeling when thinning the second silicon substrate and/or when
the non-bond area (e.g., the recess) is large. Further, the
wafer-to-wafer (W2W) uniformity of the bond may be low, thereby
leading to poor yields during bulk manufacture.
[0014] The present application is directed to a method for fusion
bonding a pair of substrates together with silane preconditioning.
In some embodiments, a surface of a first oxide layer or a surface
of a second oxide layer is preconditioned with silane. The first
and second oxide layers are respectively arranged on first and
second semiconductor substrates. Water is applied to the surface of
the first or second oxide layer, and the surfaces of the first and
second oxide layers are arranged in direct contact. With the
surfaces of the first and second oxide layers in direct contact,
the first and second oxide layers are annealed. Advantageously, by
preconditioning the surface of the first or second oxide layer with
silane, the density of silanol groups on the surface is high. This,
in turn, results in a strong bond and a low likelihood of peeling
(i.e., bond failure). Further, by preconditioning the surface of
the first or second oxide layer with silane, the density of silanol
groups on the surface is substantially uniform between workpieces
during bulk manufacture. This, in turn, results in good W2W
uniformity and yield.
[0015] With reference to FIGS. 1-9, a series of cross-sectional
views 100-900 illustrate some embodiments of a method for fusion
bonding a pair of substrates together with silane preconditioning.
As will be seen, the silane preconditioning advantageously
increases silanol groups for increased hydrophilicity and bond
strength.
[0016] As illustrated by the cross-sectional view 100 of FIG. 1, a
first oxide layer 102 is formed on a first semiconductor substrate
104. In some embodiments, the first oxide layer 102 is formed
directly on the first semiconductor substrate 104 and/or with a
thickness T exceeding about 50 nanometers. The first oxide layer
102 comprises a surface 106 that is on an opposite side of the
first oxide layer 102 as the first semiconductor substrate 104 and
that has vacant silicon bond sites 108, 110 (i.e., dangling silicon
bonds). The first oxide layer 102 may be, for example, silicon
dioxide or some other oxide. The first semiconductor substrate 104
may be, for example, a bulk silicon substrate of elemental or pure
silicon. In some embodiments, the first oxide layer 102 is formed
by thermal oxidation or vapor deposition, such as, for example,
chemical vapor deposition (CVD) or physical vapor deposition
(PVD).
[0017] As illustrated by the cross-sectional view 200 of FIG. 2, a
second semiconductor substrate 202 supporting a second oxide layer
204 is provided. The second semiconductor substrate 202 may be, for
example, a bulk silicon substrate of elemental or pure silicon. The
second oxide layer 204 is a native oxide layer that grew
spontaneously on the second semiconductor substrate 202 by thermal
oxidation. For example, the second oxide layer 204 may have grown
while the second semiconductor substrate 202 was exposed to an
uncontrolled environment. In some embodiments, the second oxide
layer 204 is silicon dioxide or some other oxide, and/or has a
thickness T less than or equal to about 5 nanometers, such as less
than or equal to about 2 nanometers.
[0018] Contaminants 206, 208 are arranged on a surface 210 of the
second oxide layer 204 that is on an opposite side of the second
oxide layer 204 as the second semiconductor substrate 202. The
contaminants 206, 208 inhibit fusion bonding by, for example,
occupying silicon bond sites 212 on the surface 210 of the second
oxide layer 204 that could otherwise be occupied by silanol groups
and/or by spacing surfaces 210 to be fusion bonded together. The
contaminants 206, 208 may be, for example, particle contaminants
206, such as dust particles, organic contaminants 208, ionic
contaminants, or any combination of the foregoing. The organic
contaminants 208 may, for example, be bonded to the surface 210 of
the second oxide layer 204 through organic bonds, such as
carbon-oxygen (C--O) bonds or carbon-hydrogen (C--H) bonds.
Further, the organic contaminants 208 may be, for example,
hydrocarbons, such as methane (CH.sub.4).
[0019] Also illustrated by the cross-sectional view 200 of FIG. 2,
the surface 210 of the second oxide layer 204 is preconditioned to
enhance a hydrophilicity of the surface 210. The surface 210 of the
second oxide layer 204 is preconditioned by exposure of the surface
210 to a preconditioning agent 214 comprising monosilane 216
(SiH.sub.4) while heat is applied to the surface 210. In some
embodiments, the preconditioning agent 214 further comprises oxide
(O.sub.2) 218 mixed with the monosilane (SiH.sub.4) 216, and/or the
surface 210 of the second oxide layer 204 is exposed to the
preconditioning agent 214 at a temperature of about 400-450 degrees
Celsius. For example, the surface 210 of the second oxide layer 204
may be exposed to the preconditioning agent 214 at a temperature of
about 420 degrees Celsius. Even more, in some embodiments, the
preconditioning agent 214 is a gas. During exposure of the surface
210 of the second oxide layer 204 to the preconditioning agent 214,
the monosilane (SiH.sub.4) 216 decomposes into silicon and hydrogen
according to the following reaction: SiH.sub.4.fwdarw.Si+2H.sub.2.
The hydrogen then reacts with silicon 220 on the surface 210 of the
second oxide layer 204 to form silanol groups 222 on the surface
210. In some embodiments, a gas flow rate of the preconditioning
agent 214 is varied to control the reaction and hence the extent of
the preconditioning.
[0020] Advantageously, the surface 210 of the second oxide layer
204 has a high density of silanol groups 222 (i.e., high
hydrophilicity) due to preconditioning the surface 210. This, in
turn, results in a strong fusion bond and a low likelihood of
peeling (i.e., bond failure). Further, the density of silanol
groups 222 on the surface 210 of the second oxide layer 204 is
substantially uniform between workpieces during bulk manufacture
due to preconditioning the surface 210. This, in turn, results in
good W2W uniformity and yield.
[0021] While the preconditioning agent 214 was applied to the
surface 210 of the second oxide layer 204, the preconditioning
agent 214 may additionally or alternatively be applied to the
surface 106 of the first oxide layer 102 in FIG. 1. Further, while
the preconditioning agent 214 was described as comprising
monosilane 216, the preconditioning agent 214 may comprise some
other silane in place of monosilane 216. The other silane may have,
for example, one or more atoms of silicon with each silicon atom
having four bonds, and each of the four bonds bonding to a hydrogen
atom or another silicon atom. Further, the other silane may be, for
example, a linear silane, a branched silane, or a cyclic silane.
The linear silane has a form of Si.sub.nH.sub.2+2, where n is
greater than or equal to one. For example, the linear silane may be
disilane (Si.sub.2H.sub.6) or trisilane (Si.sub.3H.sub.8). The
branched silane may be, for example, n-pentasilane, isopentasilane,
and neopentasilane, where n is greater than or equal to one. Even
more, the other silane may be, for example, a hydrosilane or an
organosilane. The hydrosilane may be, for example, chlorosilane,
iodosilane, or tribromosilane. The organosilane may be, for
example, methylsilane, 3-(Trimethylsilyl)propanoic acid, or
trimethyl(trifluoromethyl)silane. Even more, while the
preconditioning agent 214 was described as comprising a silane, the
preconditioning agent 214 may comprise, for example, a silene, a
heterosilane, an organoheterosilane, a silanol, a silanone, or a
fluorosilane in place of the silane.
[0022] As illustrated by the cross-sectional view 300 of FIG. 3,
the surface 210 of the second oxide layer 204 is activated by
plasma treatment in which the surface 210 of the second oxide layer
204 is exposed to plasma 302. In some embodiments, the plasma
treatment is performed by a dielectric barrier discharge (DBD)
process, a reactive ion etching (RIE) process, a remote plasma, or
a sequential plasma (SPAB) process. Further, in some embodiments,
the plasma 302 is an argon plasma, an oxygen plasma, or a nitrogen
plasma, and/or the plasma is formed from a process gas mixture
comprising argon gas, hydrogen gas, oxygen gas, nitrogen gas, or
any combination of the foregoing gases. The plasma treatment
removes the contaminants 206, 208 (see FIG. 2) from the surface 210
of the second oxide layer 204, which may increase the number of
vacant silicon bond sites 212, 304 on the surface 210 of the second
oxide layer 204. For example, by removing the contaminants 206,
208, silicon bonds sites 212, 304 that were previously occupied by
contaminants 208 are freed, such that additional silanol groups may
form at the silicon bond sites 212, 304.
[0023] While the surface activation was described as being
performed by plasma treatment, the surface activation may be
performed by CMP, chemical cleaning, plasma treatment, some other
approach to surface activation, or any combination of the
foregoing. Further, while the surface activation was limited to the
surface 210 of the second oxide layer 204, the surface 106 of the
first oxide layer 102 (see FIG. 1) may additionally or alternative
be activated. Such activation may be performed as described above
with regard to plasma treatment and/or using one or more other
approaches to surface activation.
[0024] In some embodiments, a cleaning procedure is performed on
the surface(s) 106, 210 of the first and/or second oxide layers
102, 204 between preconditioning the surface(s) 106, 210 and
activating the surface(s) 106, 210. The cleaning procedure may be,
for example, an RCA cleaning procedure comprising the sequential
application of a first cleaning solution and/or a second cleaning
solution. The first cleaning solution may comprise, for example,
ammonium hydroxide, hydrogen peroxide, deionized water, or any
combination of the foregoing. The second cleaning solution may
comprise, for example, hydrogen chloride, hydrogen peroxide,
deionized water, or any combination of the foregoing. Further, in
some embodiments, the surface(s) 106, 210 of the first and/or
second oxide layers 102, 204 may be rinsed after activating the
surface(s) 106, 210. For example, the surface(s) 106, 210 of the
first and/or second oxide layers 102, 204 may be rinsed with
deionized water.
[0025] As illustrated by the cross-sectional view 400 of FIG. 4,
water 402 (i.e., H.sub.2O) is applied to the surface 106 of the
first oxide layer 102. The water 402 reacts with vacant silicon
bond sites 108, 110 (see FIG. 1) on the surface 106 of the first
oxide layer 102 to terminate the vacant silicon bond sites 108, 110
and form silanol groups 404, 406 on the surface 106. In some
embodiments, the water 402 is deionized water.
[0026] As illustrated by the cross-sectional view 500 of FIG. 5,
water 502 (i.e., H.sub.2O) is applied to the surface 210 of the
second oxide layer 204. The water 502 reacts with vacant silicon
bond sites 212, 304 (see, FIG. 3) on the surface 210 of the second
oxide layer 204 to terminate the vacant silicon bond sites 212, 304
and form additional silanol groups 504, 506 on the surface 210. In
some embodiments, the water 502 is deionized water.
[0027] While FIGS. 4 and 5 describe the application of water 402,
502 to both of the surfaces 106, 210 of the first and second oxide
layers 102, 204, the water 402, 502 may alternatively be applied to
one of the surfaces 106, 210 but not the other surface 106, 210.
For example, water may be applied to the surface 106 of the first
oxide layer 102, but not to the surface 210 of the second oxide
layer 204.
[0028] As illustrated by the cross-sectional view 600 of FIG. 6,
the surfaces 106, 210 of the first and second oxide layers 102, 204
are placed in direct contact, thereby forming a weak bond 602
between the surfaces 106, 210 of the first and second oxide layers
102, 204. Silanol groups 222, 504, 506 on the surface 210 of the
second oxide layer 204 bond to silanol groups 404, 406 on the
surface 106 of the first oxide layer 102 through intermediate water
molecules 604, such that the water molecules 604 form a bridge
between the surfaces 106, 210. Further, in some embodiments, the
silanol groups 222, 404, 406, 504, 506 of the first and second
oxide layers 102, 204 begin to bond by van der Waals forces. For
ease of illustration, bonding by van der Waals forces is not shown
in FIG. 6, since it is more prominent during the annealing
described hereafter.
[0029] As illustrated by the cross-sectional view 700 of FIG. 7,
the first and second oxide layers 102, 204 are annealed to
strengthen the bond 602 between the surfaces 106, 210 of the first
and second oxide layers 102, 204. In some embodiments, the
annealing is performed at about 100-500 degrees Celsius, and/or in
a gaseous atmosphere or a vacuum. For example, the annealing may be
performed at about 400 degrees Celsius. During annealing, the water
molecules 604 between the surfaces 106, 210 of the first and second
oxide layers 102, 204 (see FIG. 6) diffuse into the first and
second oxide layers 102, 204. This, in turn, allows the silanol
groups 222, 404, 406, 504, 506 on the surfaces 106, 210 of the
first and second oxide layers 102, 204 to bond by van der Waals
forces. Further, in some embodiments, the water molecules 604 react
with silicon on the surfaces 106, 210 of the first and second oxide
layers 102, 204 to form additional silanol groups on the surfaces
106, 210.
[0030] As illustrated by the cross-sectional view 800 of FIG. 8,
annealing of the first and second oxide layers 102, 204 continues.
With the continued annealing, the silanol groups 222, 404, 406,
504, 506 (see FIG. 7) on the surfaces 106, 210 of the first and
second oxide layers 102, 204 polymerize to form siloxane molecules
802 and water molecules 804 between the surfaces 106, 210. The
silanol groups 222, 404, 406, 504, 506 may, for example, polymerize
according to the following equation:
Si--OH+OH--Si.fwdarw.Si--O--Si+H.sub.2O. Further, the resulting
water molecule 804 between the surfaces 106, 210 of the first and
second oxide layers 102, 204 diffuse from the bond 602 into the
first and second oxide layers 102, 204.
[0031] As illustrated by the cross-sectional view 900 of FIG. 9,
water molecules 604, 804 diffusing through the first and second
oxide layers 102, 204 reach the second semiconductor substrate 202.
When the water molecules 604, 804 reach the second semiconductor
substrate 202, the water molecules 604, 804 react with the second
semiconductor substrate 202 to form silicon dioxide molecules 902
and hydrogen gas molecules 904 according to the following equation:
Si+2H.sub.2O.fwdarw.SiO.sub.2+2H.sub.2. The resulting silicon
dioxide molecules 902 expand the second oxide layer 204 and further
strengthen the bond 602. For example, a thickness of the second
oxide layer 204 may increase. The resulting hydrogen gas molecules
904 do not react with the first and second oxide layers 102, 204
and the first and second semiconductor substrates 104, 202, and
dissolve in the first and second oxide layers 102, 204.
[0032] While the water molecules 604, 804 were described as
diffusing to the second semiconductor substrate 202, the water
molecules 604, 804 may additionally or alternatively diffuse to the
first semiconductor substrate 104. However, due to the increased
thickness of the first oxide layer 102 relative to the second oxide
layer 204, the likelihood of the water molecules 604, 804 reaching
the first semiconductor substrate 104 is less than that of reaching
the second semiconductor substrate 202.
[0033] With reference to FIG. 10, a flowchart 1000 of some
embodiments of the method of FIGS. 1-9 is provided.
[0034] At 1002, a first oxide layer is formed on a first substrate.
See, for example, FIG. 1.
[0035] At 1004, a surface of a second oxide layer on a second
substrate is preconditioned by exposing the surface to silane. The
second oxide layer is native oxide. See, for example, FIG. 2.
Advantageously, the surface of the second oxide layer has a high
density of silanol groups (i.e., a high hydrophilicity) due to
preconditioning the surface. This, in turn, results in a strong
fusion bond and a low likelihood of peeling. Further, the density
of silanol groups on second oxide layer surface is substantially
uniform between workpieces during bulk manufacture due to
preconditioning the surface, such that W2W uniformity and yield are
high.
[0036] At 1006, the surface of the second oxide layer is activated
by plasma treatment. See, for example, FIG. 3. In alternative
embodiments, the surface of the second oxide layer is activated by
other approaches to surface activation.
[0037] At 1008, water is applied to a surface of the first oxide
layer or to the surface of the second oxide layer. See, for
example, FIGS. 4 and 5.
[0038] At 1010, the surfaces of the first and second oxide layers
are arranged in direct contact. See, for example, FIG. 6.
[0039] At 1012, the first and second oxide layers are annealed.
See, for example, FIGS. 7-9.
[0040] While the method described by the flowchart 1000 is
illustrated and described herein as a series of acts or events, it
will be appreciated that the illustrated ordering of such acts or
events are not to be interpreted in a limiting sense. For example,
some acts may occur in different orders and/or concurrently with
other acts or events apart from those illustrated and/or described
herein. Further, not all illustrated acts may be required to
implement one or more aspects or embodiments of the description
herein, and one or more of the acts depicted herein may be carried
out in one or more separate acts and/or phases.
[0041] With reference to FIG. 11, a cross-sectional view 1100 of
some embodiments of a process tool for silane preconditioning is
provided. The process tool may, for example, be used to perform Act
1004 of FIG. 10. As illustrated, upper and lower components 1102,
1104 of a housing define a reactor chamber 1106 that is laterally
surrounded by a first heater 1108. An exhaust outlet 1110 is
arranged in the housing and configured to provide gases from the
reactor chamber 1106 to an exhaust pump 1112 through an exhaust
outlet line 1114. Gas inlets 1116 are arranged in the housing and
configured to provide gases from one or more gas sources 1118
through corresponding gas inlet lines 1120. In some embodiments,
valves 1122 corresponding to the gas inlet lines 1120 are arranged
along the gas inlet lines 1120 and configured to control the flow
(e.g., flow rate) of the gases to the gas inlets 1116. Further, in
some embodiments, the gas is silane (e.g., SiH.sub.4) or a
functional equivalent of silane.
[0042] Gas distribution lines 1124 and corresponding nozzles 1126
are arranged in the reactor chamber 1106, and extend from the gas
inlets 1116. The gas distribution lines 1124 and the nozzles 1126
are configured to introduce the gases into the reactor chamber
1106. Further, a workpiece holder 1128 is arranged in the reactor
chamber 1106, and is configured to support a plurality of
workpieces 1130 at a plurality of different locations. The
workpieces 1130 may, for example, comprise the second semiconductor
substrate 202 of FIG. 2. The workpiece holder 1128 is further
configured to support a second heater 1132 arranged under the
workpieces 1130. The workpiece holder 1128 is arranged over and
connected to an actuator 1134 by a shaft 1136. The actuator 1134 is
configured to move the workpiece holder 1128 into and out of the
reactor chamber 1106 while loading and unloading the reactor
chamber 1106.
[0043] With reference to FIG. 12, a cross-sectional view 1200 of
some embodiments of a MEMS package with a pair of substrates 1202,
1204 fusion bonded according to the method of FIGS. 1-9 and/or FIG.
10 is provided. As illustrated, a support device 1206 comprises a
device region 1208 arranged on an upper surface of a support
substrate 1210, and an interconnect structure 1212 covering the
device region 1208 and the support substrate 1210. The device
region 1208 comprises electronic devices 1214, 1216, 1218, such as
transistors, photodiodes, memory cells, etc. The support substrate
1210 may be, for example, a bulk semiconductor substrate, such as a
bulk silicon substrate, or a silicon-on-insulator (SOI)
substrate.
[0044] The interconnect structure 1212 interconnects the electronic
devices 1214, 1216, 1218 in the device region 1208. A stack of
conductive layers 1220, 1222 are arranged within a dielectric
region 1224 comprising multiple inter-layer dielectric (ILD) layers
and a passivation layer (not individually shown). The conductive
layers 1220, 1222 comprise individual features 1226, 1228, such as
lines and pads, and a topmost conductive layer 1222 comprises a
support bond ring 1230 and a fixed sensing electrode 1232. The
support bond ring 1230 extends laterally to enclose a cavity 1234
overlying the support device 1206, and the fixed sensing electrode
1232 is arranged under the cavity 1234. Vias 1236, 1238 are
arranged in the dielectric region 1224, between the conductive
layers 1220, 1222, to interconnect the conductive layers 1220,
1222. Similarly, contacts 1240 are arranged in the dielectric
region 1224, between a bottommost conductive layer (not labeled)
and the device region 1208, to connect the device region 1208 to
the bottommost conductive layer. The conductive layers 1220, 1222,
the vias 1236, 1238, and the contacts 1240 may be, for example, a
metal, such as copper, aluminum copper, or tungsten, or some other
conductive material. The dielectric region 1224 may be or otherwise
include, for example, an oxide, a low .kappa. dielectric (i.e., a
dielectric with a dielectric constant less than about 3.9), or some
other dielectric material.
[0045] A support ISD layer 1242 and a device bond ring 1244 are
stacked over the support device 1206, between the support device
1206 and a MEMS device substrate 1202. The support ISD layer 1242
and the device bond ring 1244 laterally enclose the cavity 1234 and
bond the MEMS device substrate 1202 to the support device 1206 at
an interface between the support bond ring 1230 and the device bond
ring 1244. The support ISD layer 1242 is arranged over the device
bond ring 1244, between the device bond ring 1244 and the MEMS
device substrate 1202. The support ISD layer 1242 may be, for
example, a thermal oxide or some other dielectric material. The
device bond ring 1244 underlies the support ISD layer 1242 and
interfaces with the support bond ring 1230. In some embodiments,
the device bond ring 1244 interfaces with the support bond ring
1230 at a eutectic bond. Further, the device bond ring 1244
overlaps with the support bond ring 1230 and has a ring-shaped
footprint of similar or equal size (e.g., in terms of area) as that
of the support bond ring 1230. The device bond ring 1244 may be,
for example, germanium or some other conductive material.
[0046] The MEMS device substrate 1202 is arranged over and bonded
to the support device 1206 through the support ISD layer 1242 and
the device bond ring 1244. The MEMS device substrate 1202 comprises
a fixed region 1246 and a movable element 1248 (i.e., a movable
region), and may be, for example, a bulk semiconductor substrate of
elemental or pure silicon. The fixed region 1246 corresponds to a
region of the MEMS device substrate 1202 that is fixed relative to
the movable element 1248. The movable element 1248 corresponds to a
region of the MEMS device substrate 1202 that is suspended within
the cavity 1234, over the fixed sensing electrode 1232, and is
configured to move within the cavity 1234 in proportion to external
stimuli, such as acceleration. The movable element 1248 is
suspended by one or more cantilever beams or springs (not shown)
that connect the movable element 1248 to the fixed region 1246.
Further, the movable element 1248 comprises a movable sensing
electrode (not shown) that is electrically coupled to the support
device 1206 by through substrate vias (TSVs) 1250, 1252 extending
through the MEMS device substrate 1202 and the support ISD layer
1242 to, for example, the device bond ring 1244. The TSVs 1250,
1252 may, for example, comprise a metal or some other conductive
material.
[0047] A capping substrate 1204 is arranged over and bonded to the
MEMS device substrate 1202 through a capping ISD layer 1254 lining
a lower surface of the capping substrate 1204. In some embodiments,
the capping ISD layer 1254 conformally lines the lower surface. The
capping substrate is bonded to the MEMS device substrate 1202 by a
fusion bond 1256 at an interface between the capping ISD layer 1254
and the MEMS device substrate 1202. The fusion bond 1256 may, for
example, be formed according to the method of FIGS. 1-9 and/or of
FIG. 10, such that the fusion bond 1256 has high strength and W2W
uniformity. The capping substrate 1204 defines an upper surface of
the cavity 1234 and, in some embodiments, comprises a lower recess
1258 defining an upper region of the cavity 1234. The capping
substrate 1204 may be, for example, a bulk semiconductor substrate
of elemental or pure silicon.
[0048] In operation, the movable element 1248 moves within the
cavity 1234 in proportion to external stimuli, such as
acceleration. For example, as the MEMS package is accelerated, the
movable element 1248 moves within the cavity 1234 in proportion to
the acceleration. Capacitive coupling between the fixed sensing
electrode 1232 and the movable sensing electrode (not shown) is
then used to measure the movement of the movable element 1248 and
to indirectly measure the external stimuli.
[0049] While the position of the movable element 1248 is
illustrated and described as being determined by capacitive
coupling between sensing electrodes in FIGS. 12, other approaches
may be employed to sense the position of the movable element 1248
within the cavity 1234. For example, optical approaches may be
employed to sense the position of the movable element 1248 with the
cavity 1234. Further, while the movable element 1248 is illustrated
and described as being suspended within the cavity 1234 between the
support device 1206 and the capping substrate 1204, the movable
element 1248 may be a flexible membrane covering the cavity 1234 in
other embodiments. In such embodiments, the movable element 1248
defines an upper surface of the cavity 1234 and is configured to
deflect in proportion to external stimuli, such as pressure.
Further, the capping substrate 1204, the capping ISD layer 1254,
and the support and device bond rings 1230, 1244 are omitted, and
the MEMS device substrate 1202 is fusion bonded to the support
device 1206 at an interface between the support device 1206 and the
support ISD layer 1242. The fusion bond may, for example, be formed
according to the method of FIGS. 1-9 and/or of FIG. 10, such that
the fusion bond has high strength and W2W uniformity.
[0050] With reference to FIGS. 13-21, a series of cross-sectional
views 1300-2100 of some embodiments of a method for forming the
MEMS package of FIG. 12 is provided.
[0051] As illustrated by the cross-sectional view 1300 of FIG. 13,
a first etch is performed into a capping substrate 1204 to form a
recess 1258 in a surface of the capping substrate 1204. The capping
substrate 1204 may be, for example, a bulk silicon substrate of
elemental or pure silicon. In some embodiments, the process for
performing the first etch comprises depositing and patterning a
first photoresist layer 1302 over the capping substrate 1204. For
example, the first photoresist layer 1302 may be patterned to mask
regions of the capping substrate 1204 that are along a periphery of
the capping substrate 1204. Thereafter, one or more etchants 1304,
such as wet or dry etchants, are applied to the capping substrate
1204 while using the first photoresist layer 1302 as a mask. After
performing the first etch, the first photoresist layer 1302 is
removed or otherwise stripped.
[0052] As illustrated by the cross-sectional view 1400 of FIG. 14,
a capping ISD layer 1254 is formed lining the surface of the
capping substrate 1204, including the recess 1258. In some
embodiments, the capping ISD layer 1254 is silicon dioxide or some
other oxide. Further, in some embodiments, the capping ISD layer
1254 is formed by thermal oxidation or vapor deposition, such as,
for example, CVD or PVD.
[0053] As illustrated by the cross-sectional view 1500 of FIG. 15,
a MEMS device substrate 1202 is arranged over and bonded to the
capping substrate 1204 through the capping ISD layer 1254. The MEMS
device substrate 1202 is bonded to the capping substrate 1204 by
fusion bonding performed with silane preconditioning according to
the method of FIGS. 1-9 and/or 10. For example, the capping
substrate 1204 may correspond to the first semiconductor substrate
104 in FIGS. 1-9, the capping ISD layer 1254 may correspond to the
first oxide layer 102 in FIGS. 1-9, the MEMS device substrate 1202
may correspond to the second semiconductor substrate 202 in FIGS.
1-9, and a native oxide layer (not shown) on the MEMS device
substrate 1202 may correspond to the second oxide layer 204. The
MEMS device substrate 1202 may be, example, a bulk silicon
substrate of elemental or pure silicon.
[0054] As illustrated by the cross-sectional view 1600 of FIG. 16,
the MEMS device substrate 1202 is thinned to a desired thickness T
after forming a fusion bond 1256 at an interface between the MEMS
device substrate 1202 and the capping ISD layer 1254. In some
embodiments, the MEMS device substrate 1202 is thinned by a CMP
and/or an etch back. Further, a support ISD layer 1242 is formed
over the MEMS device substrate 1202. In some embodiments, the
support ISD layer 1242 is silicon dioxide or some other oxide.
Further, in some embodiments, the support ISD layer 1242 is formed
by thermal oxidation.
[0055] As illustrated by the cross-sectional view 1700 of FIG. 17,
a second etch is performed into the support ISD layer 1242 to
remove a region of the support ISD layer 1242 that overlies the
recess 1258 of the capping substrate 1204 (i.e., to form an
opening). In some embodiments, the region is restricted to directly
over the recess 1258. Further, in some embodiments, the process for
performing the second etch comprises depositing and patterning a
second photoresist layer 1702 over the support ISD layer 1242. For
example, the second photoresist layer 1702 may be patterned to mask
regions of the support ISD layer 1242 that laterally surround the
recess 1258. Thereafter, one or more etchants 1704 are applied to
the support ISD layer 1242 while using the second photoresist layer
1702 as a mask. After performing the second etch, the second
photoresist layer 1702 is removed or otherwise stripped.
[0056] As illustrated by the cross-sectional view 1800 of FIG. 18,
a bond ring layer 1802 is formed over the support ISD layer 1242
and the MEMS device substrate 1202. The bond ring layer 1802 may be
formed of, for example, germanium or some other material capable of
eutectic bonding. In some embodiments, the bond ring layer 1802 is
formed using a vapor deposition technique or atomic layer
deposition.
[0057] As illustrated by the cross-sectional view 1900 of FIG. 19,
a third etch is performed into the bond ring layer 1802 (see FIG.
18) to form a device bond ring 1244 laterally enclosing the recess
1258. In some embodiments, the process for performing the third
etch comprises depositing and patterning a third photoresist layer
1902 over the bond ring layer 1802. For example, the third
photoresist layer 1902 may be patterned to mask regions of the bond
ring layer 1802 that correspond to the device bond ring 1244.
Thereafter, one or more etchants 1904 are applied to the bond ring
layer 1802 while using the third photoresist layer 1902 as a mask.
After performing the third etch, the third photoresist layer 1902
is removed or otherwise stripped.
[0058] As illustrated by the cross-sectional view 2000 of FIG. 20,
a fourth etch is performed into the MEMS device substrate 1202 to
form a movable element 1248 over the recess 1258. In some
embodiments, the process for performing the fourth etch comprises
depositing and patterning a fourth photoresist layer 2002 over the
support ISD layer 1242 and the MEMS device substrate 1202. For
example, the fourth photoresist layer 2002 may be patterned to mask
regions of the MEMS device substrate 1202 that correspond to the
movable element 1248 and regions of the MEMS device substrate 1202
and the support ISD layer 1242 laterally surrounding the recess
1258. Thereafter, one or more etchants 2004 are applied to the MEMS
device substrate 1202 while using the fourth photoresist layer 2002
as a mask. After performing the fourth etch, the fourth photoresist
layer 2002 is removed or otherwise stripped.
[0059] As illustrated by the cross-sectional view 2100 of FIG. 21,
the MEMS device substrate 1202 and the capping substrate 1204 are
rotated 180 degrees and bonded to a support device 1206. The
support device 1206 may be, for example, an IC or a bulk
semiconductor substrate. In some embodiments, the MEMS device
substrate 1202 is eutectically bonded to the support device 1206
through a support bond ring 1230 of the support device 1206. The
support bond ring 1230 has the same or a similar footprint as the
device bond ring 1244, and may be, for example, aluminum copper or
some other material capable of eutectic bonding.
[0060] With reference to FIG. 22, a flowchart 2200 of some
embodiments of the method of FIGS. 13-21 is provided.
[0061] At 2202, a first etch is performed into a capping substrate
to form a recess in a surface of the capping substrate. See, for
example, FIG. 13.
[0062] At 2204, a first oxide layer is formed lining the surface of
the capping substrate. See, for example, FIG. 14.
[0063] At 2206, the capping substrate is fusion bonded to a MEMS
device substrate with silane preconditioning. See, for example,
FIG. 15. The fusion bonding and/or the silane preconditioning may,
for example, be performed as described in the method of FIGS.
1-10.
[0064] At 2208, the MEMS device substrate is thinned and a second
oxide layer is formed on an opposite side of the MEMS device
substrate as the capping substrate. See, for example, FIG. 16.
[0065] At 2210, a second etch is performed into the second oxide
layer to remove a region of the second oxide layer overlying the
recess. See, for example, FIG. 17.
[0066] At 2212, a device bond ring is formed on the second oxide
layer. See, for example, FIGS. 18 and 19.
[0067] At 2214, a third etch is performed into the MEMS device
substrate to form a movable element over the recess. See, for
example, FIG. 20.
[0068] At 2216, the capping substrate and the MEMS device substrate
are eutectically bonded to a support device through the second
oxide layer and the device bond ring. See, for example, FIG.
21.
[0069] While the method described by the flowchart 2200 is
illustrated and described herein as a series of acts or events, it
will be appreciated that the illustrated ordering of such acts or
events are not to be interpreted in a limiting sense. For example,
some acts may occur in different orders and/or concurrently with
other acts or events apart from those illustrated and/or described
herein. Further, not all illustrated acts may be required to
implement one or more aspects or embodiments of the description
herein, and one or more of the acts depicted herein may be carried
out in one or more separate acts and/or phases.
[0070] In some embodiments, the Acts of FIG. 22 are augmented to
form a pressure sensor. An oxide layer is formed on a MEMS device
substrate, as described by Act 2208, and etched as described by Act
2210. The MEMS device substrate is then fusion bonded to a support
device through the oxide layer, as described in Act 2206. The
fusion bonding may be performed, for example, with silane
pre-conditioning according to the method of FIGS. 1-10. Further,
the MEMS device substrate is thinned as described in Act 2208.
[0071] Thus, as can be appreciated from above, the present
disclosure provides a method for bonding a pair of semiconductor
substrates together. A surface of a first oxide layer or a surface
of a second oxide layer is preconditioned with silane. The first
and second oxide layers are respectively arranged on first and
second semiconductor substrates. Water is applied to the surface of
the first or second oxide layer. The surfaces of the first and
second oxide layers are brought in direct contact. The first and
second oxide layers are annealed.
[0072] In other embodiments, the present disclosure provides a
method for forming a MEMS package. An oxide layer is formed on a
capping substrate. A device substrate is preconditioned by exposing
the device substrate to silane. The device substrate is fusion
bonded to the capping substrate through the oxide layer after
exposing the device substrate to the silane. A movable element is
formed in the device substrate. The device substrate is bonded to a
support device on an opposite side of the device substrate as the
capping substrate.
[0073] In yet other embodiments, the present disclosure provides a
method for bonding a pair of semiconductor substrates together. An
oxide layer is formed on a first semiconductor substrate. A surface
of a native oxide layer is preconditioned with silane. The native
oxide layer is arranged on a second semiconductor substrate. A
surface of the oxide layer and the surface of the native oxide
layer are arranged in direct contact. The oxide layer and the
native oxide layer are annealed.
[0074] The foregoing outlines features of several embodiments so
that those skilled in the art may better understand the aspects of
the present disclosure. Those skilled in the art should appreciate
that they may readily use the present disclosure as a basis for
designing or modifying other processes and structures for carrying
out the same purposes and/or achieving the same advantages of the
embodiments introduced herein. Those skilled in the art should also
realize that such equivalent constructions do not depart from the
spirit and scope of the present disclosure, and that they may make
various changes, substitutions, and alterations herein without
departing from the spirit and scope of the present disclosure.
* * * * *