U.S. patent number RE47,208 [Application Number 15/582,014] was granted by the patent office on 2019-01-15 for manufacturing method of solid-state image sensor.
This patent grant is currently assigned to Conversant Intellectual Property Management Inc.. The grantee listed for this patent is Conversant Intellectual Property Management Inc.. Invention is credited to Akira Tsukamoto.
![](/patent/grant/RE047208/USRE047208-20190115-D00000.png)
![](/patent/grant/RE047208/USRE047208-20190115-D00001.png)
![](/patent/grant/RE047208/USRE047208-20190115-D00002.png)
![](/patent/grant/RE047208/USRE047208-20190115-D00003.png)
![](/patent/grant/RE047208/USRE047208-20190115-D00004.png)
![](/patent/grant/RE047208/USRE047208-20190115-D00005.png)
![](/patent/grant/RE047208/USRE047208-20190115-D00006.png)
![](/patent/grant/RE047208/USRE047208-20190115-D00007.png)
![](/patent/grant/RE047208/USRE047208-20190115-D00008.png)
![](/patent/grant/RE047208/USRE047208-20190115-D00009.png)
![](/patent/grant/RE047208/USRE047208-20190115-D00010.png)
View All Diagrams
United States Patent |
RE47,208 |
Tsukamoto |
January 15, 2019 |
Manufacturing method of solid-state image sensor
Abstract
A single crystal silicon layer is formed on a principal surface
of a first wafer by epitaxial growth. A silicon oxide layer is
formed on the single crystal silicon layer. Next, a defect layer is
formed inside the single crystal silicon layer by ion implantation,
and then, the second wafer is bonded to the silicon oxide layer on
the first wafer. After that, an SOI wafer including the silicon
oxide layer formed on the second wafer and the single crystal
silicon layer formed on the silicon oxide layer is formed by
separating the first wafer including the single crystal silicon
layer from the second wafer including the single crystal silicon
layer in the defect layer. Then, a photodiode is formed in the
single crystal silicon layer. An interconnect layer is formed on a
surface of the single crystal silicon layer which is opposite to
the silicon oxide layer.
Inventors: |
Tsukamoto; Akira (Osaka,
JP) |
Applicant: |
Name |
City |
State |
Country |
Type |
Conversant Intellectual Property Management Inc. |
Ottawa |
N/A |
CA |
|
|
Assignee: |
Conversant Intellectual Property
Management Inc. (Ottawa, CA)
|
Family
ID: |
42739265 |
Appl.
No.: |
15/582,014 |
Filed: |
April 28, 2017 |
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
13220079 |
Aug 29, 2011 |
|
|
|
|
PCT/JP2009/005509 |
Oct 21, 2009 |
|
|
|
Reissue of: |
14061750 |
Oct 23, 2013 |
9018031 |
Apr 28, 2015 |
|
|
Foreign Application Priority Data
|
|
|
|
|
Mar 18, 2009 [JP] |
|
|
2009-066877 |
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H01L
27/14643 (20130101); H01L 27/14632 (20130101); H01L
27/14689 (20130101); H01L 27/1464 (20130101); H01L
31/18 (20130101); H01L 27/146 (20130101); H01L
27/14689 (20130101); H01L 31/18 (20130101); H01L
31/1892 (20130101); H01L 27/14687 (20130101); H01L
31/1892 (20130101); Y10S 438/977 (20130101) |
Current International
Class: |
H01L
31/18 (20060101); H01L 27/146 (20060101) |
Field of
Search: |
;438/455,459 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
2959704 |
|
Oct 1996 |
|
JP |
|
10-308355 |
|
Nov 1998 |
|
JP |
|
10-321548 |
|
Dec 1998 |
|
JP |
|
2000-049063 |
|
Feb 2000 |
|
JP |
|
3048201 |
|
Mar 2000 |
|
JP |
|
3048201 |
|
Mar 2000 |
|
JP |
|
3385972 |
|
Jan 2003 |
|
JP |
|
3385972 |
|
Mar 2003 |
|
JP |
|
2005-259828 |
|
Sep 2005 |
|
JP |
|
2008-103668 |
|
May 2008 |
|
JP |
|
2008-514011 |
|
May 2008 |
|
JP |
|
Other References
Office Action issued in U.S. Appl. No. 13/220,079 dated Oct. 1,
2012. cited by applicant .
Office Action issued in U.S. Appl. No. 13/220,079 dated Mar. 14,
2013. cited by applicant .
Pfann, W.G., "Principles of Zone-Melting", Transactions of the
American Institute of Mining and Metallurgical Engineers, vol. 194,
pp. 747-753, 1952. cited by applicant .
Wagner, Carl, "Theoretical Analysis of Diffusion of Solutes During
the Solidification of Alloys", Transactions of the American
Institute of Mining and Metallurgical Engineers, vol. 200, pp.
154-160, 1954. cited by applicant .
Burton, J.A. et al., "The Distribution of Solute in Crystals Grown
from the Melt. Part 1. Theoretical", The Journal of Chemical
Physics, vol. 21, No. 11, pp. 1987-1991, Nov. 1953. cited by
applicant .
Office Action issued in U.S. Appl. No. 13/220,079 dated Jul. 24,
2013. cited by applicant .
Shimura, Fumio "Semiconductor Silicon Crystal Technology", partial
English translation, pp. 62-69 and pp. 76-85, 1993. cited by
applicant .
W. G. Pfann "Principles of Zone-Melting" Transactions of the
American Institue of Mining and Metallurgical Engineers. vol. 194,
pp. 747-753, 1952. cited by applicant .
Carl Wagner "Theoretical Analysis of Diffusion of Solutes During
the Solidification of Alloys" Transactions of the American Institue
of Mining and Metallurgical Engineers. vol. 200, pp. 154-160, 1954.
cited by applicant .
J. A. Burton et al. "The Distribution of Solute in Crystals Grown
from the Melt. Part I. Theoretical" The Journal of Chemical
Physics, vol. 21, No. 11, pp. 1987-1991, Nov. 1953. cited by
applicant.
|
Primary Examiner: Nguyen; Tuan H
Attorney, Agent or Firm: Conversant IP Management Inc.
Parent Case Text
CROSS-REFERENCE TO RELATED APPLICATION
.[.This application is a Divisional of application Ser. No.
13/220,079, filed Aug. 29, 2011, which is a continuation of PCT
International Application PCT/JP2009/005509, filed on Oct. 21,
2009, which in turn claims benefit of Japanese Patent Application
No. 2009-066877, filed on Mar. 18, 2009, the entire contents of
each of which are incorporated herein by reference..]. .Iadd.This
application is a Reissue of U.S. patent application Ser. No.
14/061,750, filed on Oct. 23, 2013 now U.S. Pat. No. 9,018,031,
which is a divisional application of U.S. patent application Ser.
No. 13/220,079, filed on Aug. 29, 2011, which is a continuation of
PCT International Application PCT/JP2009/005509, filed on Oct. 21,
2009, which in turn claims benefit of Japanese Patent Application
No. 2009-066877, filed on Mar. 18, 2009, the entire contents of
each of which are incorporated herein by reference. .Iaddend.
Claims
What is claimed is:
1. A manufacturing method of a solid-state image sensor comprising:
forming a first single crystal silicon layer .Iadd.having a first
impurity concentration .Iaddend.on a principal surface of a first
wafer by epitaxial growth; forming a silicon oxide layer on the
first single crystal silicon layer; forming a defect layer inside
the first single crystal silicon layer by ion implantation; bonding
a second wafer to the silicon oxide layer on the first wafer
.Iadd.to form a combined wafer.Iaddend.; .[.forming.].
.Iadd.separating the combination wafer at the defect layer to form
.Iaddend.an SOI wafer including the silicon oxide layer .Iadd.that
was .Iaddend.formed on the second wafer and .Iadd.a portion of
.Iaddend.the first single crystal silicon layer .Iadd.that was
.Iaddend.formed on the silicon oxide layer .[.by separating the
first wafer including the first single crystal silicon layer from
the second wafer including the first single crystal silicon layer
in the defect layer.].; forming a second single crystal silicon
layer .Iadd.having a second impurity concentration .Iaddend.on the
.Iadd.portion of the .Iaddend.first single crystal silicon layer by
epitaxial growth; forming a photodiode in .[.the first single
crystal silicon layer or.]. the second single crystal silicon
layer; and forming an interconnect layer including a photodiode
charge read-out structure on a surface of the second single crystal
silicon layer which is opposite to the first single crystal silicon
layer.
2. The method of claim 1, further comprising after forming the
interconnect layer, .[.selectively etching part of or the entire.].
.Iadd.etching the .Iaddend.second wafer with respect to the silicon
oxide layer, wherein in forming the photodiode, the light-receiving
section of the photodiode is formed to face the silicon oxide
layer.
3. A manufacturing method of a solid-state image sensor comprising:
forming a first single crystal silicon layer having .Iadd.a first
.Iaddend.impurity concentration .[.of 1.times.10.sup.17 cm.sup.-3
or more.]. on a principal surface of a first wafer by epitaxial
growth; forming a silicon oxide layer on the first single crystal
silicon layer; forming a defect layer inside the first single
crystal silicon layer by ion implantation; bonding a second wafer
to the silicon oxide layer on the first wafer .Iadd.to form a
combined wafer.Iaddend.; .[.forming.]. .Iadd.separating the
combination wafer at the defect layer to form .Iaddend.an SOI wafer
including the silicon oxide layer .Iadd.that was .Iaddend.formed on
the second wafer and .Iadd.a portion of .Iaddend.the first single
crystal silicon layer .Iadd.that was .Iaddend.formed on the silicon
oxide layer .[.by separating the first wafer including the first
single crystal silicon layer from the second wafer including the
first single crystal silicon layer in the defect layer.].; forming
a second single crystal silicon layer .Iadd.on the portion of the
first single crystal silicon layer by epitaxial growth, the second
single crystal silicon layer .Iaddend.having .Iadd.a second
impurity concentration .Iaddend.lower .Iadd.than the first
.Iaddend.impurity concentration .[.than the first single crystal
silicon layer on the first single crystal silicon layer by
epitaxial growth.].; forming a photodiode in the second single
crystal silicon layer so that a light-receiving section faces the
silicon oxide layer; forming an interconnect layer including a
photodiode charge read-out structure on a surface of the second
single crystal silicon layer which is opposite to the first single
crystal silicon layer; and .[.selectively etching part of or the
entire.]. .Iadd.etching the .Iaddend.second wafer with respect to
the silicon oxide layer.
4. The method of claim 3, wherein the first single crystal silicon
layer is of a first conductivity type, the second single crystal
silicon layer is of a second conductivity type, and the photodiode
is of the second conductivity type.
5. The method of claim 3, wherein the first single crystal silicon
layer is of a first conductivity type, the second single crystal
silicon layer is of a second conductivity type, the forming of the
photodiode includes forming a well of the first conductivity type
in the second single crystal silicon layer, and the photodiode is
of the second conductivity type, and is formed in the well.
.Iadd.6. A manufacturing method of a solid-state image sensor
comprising: forming a first single crystal silicon layer having an
impurity concentration on a principal surface of a first wafer by
epitaxial growth; forming a silicon oxide layer on the first single
crystal silicon layer; forming a defect layer inside the first
single crystal silicon layer by ion implantation; bonding a second
wafer to the silicon oxide layer on the first wafer to form a
combined wafer; separating the combination wafer at the defect
layer to form an SOI wafer including the silicon oxide layer that
was formed on the second wafer and a portion of the first single
crystal silicon layer that was formed on the silicon oxide layer;
forming a well of a first conductivity type in the portion of the
first single crystal silicon layer; forming a photodiode of a
second conductivity type in the well of the first conductivity type
so that a light-receiving section faces the silicon oxide layer;
forming an interconnect layer including a photodiode charge
read-out structure on a surface of the first single crystal silicon
layer which is opposite to the first single crystal silicon layer;
and etching the second wafer with respect to the silicon oxide
layer. .Iaddend.
Description
BACKGROUND
The present disclosure relates to solid-state image sensors and
manufacturing methods of the sensors, and more particularly to
solid-state image sensors using silicon-on-insulator (SOI)
substrates and manufacturing methods of the sensors.
Solid-state image sensors such as charge coupled devices (CCDs),
CMOS image sensors, etc., are generally used for digital cameras,
video cameras, etc. In recent years, with an improvement in
solid-state image sensors, high-definition images can be obtained
by a solid-state image sensor with highly dense pixels.
A conventional solid-state image sensor includes a transfer gate, a
photoelectric conversion element, a MOS transistor, various
interconnects, etc. on a semiconductor substrate. However, a
light-receiving region of a photoelectric conversion element needs
to be provided avoiding the transfer gate, the MOS transistor, the
interconnects, etc. There is thus the problem that the aperture
ratio of the light-receiving region reduces with a reduced pixel
size due to reduction in the area of the semiconductor substrate
and an increase in the number of pixels.
Thus, in recent years, increasing attention has been given to a
solid-state image sensor of back surface irradiation including a
transfer gate, a MOS transistors, and an interconnect layer on a
first (front) surface of a semiconductor substrate, and a
photoelectric conversion element on a second (back) surface so that
the back surface serves as a light-receiving region.
A substrate of a solid-state image sensor of back surface
irradiation needs to be thinned to a thickness of 3-10 .mu.m. As a
means of thinning, polishing or etching of a conventional silicon
substrate from a back surface is considered, but the means is less
controllable in uniformly reducing the thickness of an initial
substrate, which is usually 500 .mu.m, to 10 .mu.m or less.
For example, Japanese Translation of PCT International Application
No. 2008-514011 shows a manufacturing method of a solid-state image
sensor of back surface irradiation using an SOI wafer including a
single crystal silicon layer on a silicon oxide layer.
In the manufacturing method shown in Japanese Translation of PCT
International Application No. 2008-514011, the SOI wafer including
the silicon oxide layer formed on a base wafer, and the single
crystal silicon layer formed on the silicon oxide layer is used.
The method includes forming a photoelectric conversion element on
the single crystal silicon layer with a light-receiving section
facing the silicon oxide layer of the single crystal silicon layer,
forming an interconnect layer on a surface of the single crystal
silicon layer which is opposite to the surface closer to the
silicon oxide layer, and selectively removing the base wafer under
the silicon oxide layer.
As representative manufacturing methods of an SOI wafer, roughly
two types of separation by implanted oxygen (SIMOX) and bonding are
known. The SIMOX is a manufacturing method of an SOI wafer
utilizing the feature that a silicon oxide layer is formed inside a
silicon substrate and a recrystallized silicon layer is formed near
the surface of the silicon substrate by ion-implanting highly
concentrated oxygen into the silicon substrate with highly
accelerated energy and performing heat treatment. This method
accurately controls the depth for ion implantation, thereby
providing excellent uniformity of the thickness of the
recrystallized silicon layer formed near the surface of the silicon
substrate. However, a non-single crystal silicon oxide layer occurs
in the heat treatment, and the recrystallized silicon layer is
formed on the non-single crystal silicon oxide layer, thereby
causing a large number of crystal defects in the recrystallized
silicon layer. Although considerable efforts have been made to
reduce the crystal defects, the problem is not yet overcome, since
the mechanism of occurrence of the defects is intrinsic. The
defects are considered difficult to overcome in the future.
Then, wafer bonding was suggested. A Unibond technique (Smart Cut
(registered trademark)) is practically the current mainstream.
In the Unibond technique, a silicon oxide film is formed on the
surface of a silicon substrate, hydrogen ions are implanted via the
formed silicon oxide film, and then the silicon oxide film of the
silicon substrate is bonded to a base wafer (supporting substrate).
After that, heat treatment is performed and the silicon substrate
is separated in the implanted position, thereby providing a method
(hydrogen ion implantation separation) of forming an SOI wafer.
This technique is shown in, e.g., Japanese Patent No. 2959704, or
Japanese Patent No. 3385972.
As the method of separating the hydrogen ions, a silicon oxide film
is formed on at least one of two silicon wafers. Hydrogen ions or
noble gas ions are implanted from above a first silicon wafer,
thereby forming a defect layer (a sealing layer) inside the silicon
wafer. Then, the silicon wafer is adhered to the other silicon
wafer via the silicon oxide film and heat treatment (separation
heat treatment) is performed to separate one of the wafers to be a
thin film using the defect layer as a surface to be cleaved
(surface to be separated). With further heat treatment (bonding
heat treatment) is performed to strengthen the bonding, thereby
forming an SOI wafer.
SUMMARY
A silicon wafer used in manufacturing an SOI wafer is usually
formed by Czochralski (CZ) growth. In the CZ growth, massive
polycrystalline silicon is put into a crucible made of quartz and
melted by resistance heating in an argon atmosphere. A silicon
ingot comes into contact with single crystal silicon as a seed and
is pulled up while gradually rotating to manufacture the silicon
ingot. The resistivity of silicon is controlled by the
concentration of a dopant, but the resistivity in the axial
direction and the in-plane direction is difficult to uniform.
Variations in concentric resistivity, oxygen concentration, etc.
caused by the pull-up of the silicon wafer produced by the CZ
growth are referred to as "swirls." FIG. 9 is a schematic view of a
swirl representing the magnitude of resistivity by concentration.
Even with such a problem, single crystals can be formed with a
great opening size of 300 mm or more depending on a pull-up
machine, and thus, wafers with a great opening size used in
manufacture of semiconductors in recent years are all produced by
the CZ growth.
SOI wafers manufactured by hydrogen ion implantation separation
using silicon wafers produced by the CZ growth are in practical use
mainly in logic LSI manufactures. In usually manufactured SOI
wafers, variations in resistivity caused by the CZ growth are not
problematic.
A manufacturing method of a solid-state image sensor of back
surface irradiation using an SOI wafer, which is formed by hydrogen
ion implantation separation using a silicon wafer produced by CZ
growth as a material, will be described below.
FIGS. 10A-10E and 11A-11E are cross-sectional views illustrating a
manufacturing method of a solid-state image sensor of back surface
irradiation using a conventional SOI wafer.
As shown in FIG. 10A, a bond wafer 5 made of single crystal silicon
is prepared. The bond wafer 5 is made of single crystal silicon
produced by CZ growth. P-type impurities are implanted into the
bond wafer 5 so that the bond wafer 5 will be a p-well 6 of a
completed solid-state image sensor 3 shown in FIG. 11E. The bond
wafer 5 is shown in the figure so that the principal surface faces
downwards.
Then, as shown in FIG. 10B, the bond wafer 5 is thermally oxidized
to form a silicon oxide layer 7. While only the silicon oxide layer
7 on the principal surface is shown in the figure, silicon oxide
films are actually formed on the back and side surfaces as
well.
Next, as shown in FIG. 10C, hydrogen ions are implanted from the
principal surface of the bond wafer 5. Accelerated energy for
implanting the hydrogen ions is controlled to adjust the depth of a
defect layer 8, which is caused by the hydrogen ion implantation,
from the principal surface of the bond wafer 5. This determines the
thickness of a single crystal silicon layer 9 being an upper layer
when an SOI wafer 1 is completed, which will be described later
with reference to FIG. 11A.
Then, as shown in FIG. 10D, the surfaces of the bond wafer 5 and a
base wafer 2 for maintaining the strength of the SOI wafer 1 are
cleaned and bonded. The base wafer 2 is made of single crystal
silicon produced by CZ growth.
After that, as shown in FIG. 10E, the base wafer 2 and the bond
wafer 5 are separated from the back surface, thereby separating
part of the bond wafer 5 from the base wafer 2 in the defect layer
8. The part of the bond wafer 5 separated from the base wafer 2 is
returned to the oxidation shown in FIG. 10B, and may be reused in
manufacture of another SOI wafer.
Then, heat treatment is performed to strengthen the bonding of the
bonded surfaces. Normally, it is said that bonding of bonded
surfaces is strengthen by heat treatment at a temperature of
1000.degree. C. A method of lowering the temperature of the heat
treatment by pretreatment with plasma, etc. is also considered.
With the above-described processes, as shown in FIG. 11A, the SOI
wafer 1 is completed, which includes the base wafer 2, the silicon
oxide layer 7, and the single crystal silicon layer 9.
FIG. 11B illustrates forming photodiodes 4 and drains 11 in the
single crystal silicon layer 9 which is the uppermost layer of the
SOI wafer 1, and forming read-out gates 10 and interconnects 12 on
the single crystal silicon layer 9 to form a solid-state MOS image
sensor. In the case of a solid-state CCD image sensor, photodiodes,
charge transfer sections, gate electrodes and interconnects are
used instead. The photodiodes 4 are formed by ion-implanting
impurities such as arsenic (As), properties (P), etc. into a p-type
substrate.
Then, as shown in FIG. 11C, the base wafer 2 is removed from the
SOI wafer 1. When the base wafer 2 is made of single crystal
silicon, it can be easily removed by etching with an alkali
solution. The intermediate silicon oxide layer 7 is not etched with
the alkali solution, it can be thus processed with thickness
accuracy equal to that in manufacturing the SOI wafer 1. This is
the greatest advantage of using the SOI wafer 1. At this time, when
removing the base wafer 2 from the entire surface of the SOI wafer
1, the strength of the wafer is insufficient, and thus, an extra
support wafer for reinforcement is to be bonded on the front
surface provided with the interconnects 12, etc. The support wafer
is however, not shown in the figure.
After that, as shown in FIG. 11D, boron (B.sup.+) ions are
implanted from the surface provided with the silicon oxide layer 7
to form a p.sup.+ type depletion barrier layer 13 near the
interface in the single crystal silicon layer 9 between the single
crystal silicon layer 9 and the silicon oxide layer 7. The
depletion barrier layer 13 blocks spread of the depletion layers of
the photodiodes 4 to the interface with the silicon oxide film and
prevents noise electrons generated at interface states from being
accumulated at the photodiodes 4 and becoming dark signals.
Next, as shown in FIG. 11E, color filters 14 and on-chip
microlenses 15 are formed on the silicon oxide layer 7 being the
back surface of the SOI wafer 1. Then, the solid-state image sensor
3 of back surface irradiation is completed.
The manufacturing method of the solid-state image sensor of back
surface irradiation using the conventional SOI wafer is excellent
in uniformly and controllably thinning the silicon wafer. However,
slight variations in the resistivity of the substrate influence
characteristics of the individual photodiodes, and lead to
variations in sensitivity and the amount of saturated signals. As a
result, fixed pattern noise occurs in an obtained image due to a
swirl shown in FIG. 12.
In view of the problem, it is an objective of the present
disclosure to reduce variations in resistivity of a substrate and
to reduce degradation in the quality of an obtained image in a
solid-state image sensor using an SOI substrate.
In order to achieve the objective, the present disclosure provides
a manufacturing method of a solid-state image sensor implemented by
a structure using an SOI wafer made of single crystal silicon
formed by epitaxial growth.
Specifically, a manufacturing method of a first solid-state image
sensor according to the present disclosure includes forming a
single crystal silicon layer on a principal surface of a first
wafer by epitaxial growth; forming a silicon oxide layer on the
single crystal silicon layer; forming a defect layer inside the
single crystal silicon layer by ion implantation; bonding the
second wafer to the silicon oxide layer on the first wafer; forming
an SOI wafer including the silicon oxide layer formed on the second
wafer and the single crystal silicon layer formed on the silicon
oxide layer by separating the first wafer including the single
crystal silicon layer from the second wafer including the single
crystal silicon layer in the defect layer; forming a photodiode in
the single crystal silicon layer; and forming an interconnect layer
including a photodiode charge read-out structure on a surface of
the single crystal silicon layer which is opposite to the silicon
oxide layer.
In the manufacturing method of the first solid-state image sensor
according to the present disclosure, the first solid-state image
sensor according to the present disclosure does not include single
crystal silicon produced by CZ growth, which inevitably has
concentric variations in impurity concentration. This prevents
fixed pattern noise occurring in the solid-state image sensor
manufactured by a conventional manufacturing method. This enables
manufacture of a solid-state image sensor with reduced degradation
in the quality of an obtained image.
The manufacturing method of the first solid-state image sensor
according to the present disclosure may further include, after
forming the interconnect layer, selectively etching part of or the
entire second wafer with respect to the silicon oxide layer. In
forming the photodiode, the light-receiving section of the
photodiode is formed to face the silicon oxide layer.
A manufacturing method of a second solid-state image sensor
according to the present disclosure includes forming a first single
crystal silicon layer on a principal surface of a first wafer by
epitaxial growth; forming a silicon oxide layer on the first single
crystal silicon layer; forming a defect layer inside the first
single crystal silicon layer by ion implantation; bonding the
second wafer to the silicon oxide layer on the first wafer; forming
an SOI wafer including the silicon oxide layer formed on the second
wafer and the first single crystal silicon layer formed on the
silicon oxide layer by separating the first wafer including the
first single crystal silicon layer from the second wafer including
the first single crystal silicon layer in the defect layer; forming
a second single crystal silicon layer on the first single crystal
silicon layer by epitaxial growth; forming a photodiode in the
first single crystal silicon layer or the second single crystal
silicon layer; and forming an interconnect layer including a
photodiode charge read-out structure on a surface of the second
single crystal silicon layer which is opposite to the first single
crystal silicon layer.
In the manufacturing method of the second solid-state image sensor
according to the present disclosure, the second solid-state image
sensor does not include single crystal silicon produced by CZ
growth, which inevitably has variations in impurity concentration.
This prevents fixed pattern noise occurring in the solid-state
image sensor manufactured by a conventional manufacturing method.
This enables manufacture of a solid-state image sensor with reduced
degradation in the quality of an obtained image. Furthermore,
read-out gates, drains, interconnects, etc. of the solid-state
image sensor can be formed on the surface of the defect free second
single crystal silicon layer, thereby further improving the image
quality of the solid-state image sensor.
The manufacturing method of the second solid-state image sensor
according to the present disclosure may further include, after
forming the interconnect layer, selectively etching part of or the
entire second wafer with respect to the silicon oxide layer. In
forming the photodiode, the light-receiving section of the
photodiode is formed to face the silicon oxide layer.
A manufacturing method of a third solid-state image sensor
according to the present disclosure includes forming a first single
crystal silicon layer having impurity concentration of
1.times.10.sup.17 cm.sup.-3 or more on a principal surface of a
first wafer by epitaxial growth; forming a silicon oxide layer on
the first single crystal silicon layer; forming a defect layer
inside the first single crystal silicon layer by ion implantation;
bonding the second wafer to the silicon oxide layer on the first
wafer; forming an SOI wafer including the silicon oxide layer
formed on the second wafer and the first single crystal silicon
layer formed on the silicon oxide layer by separating the first
wafer including the first single crystal silicon layer from the
second wafer including the first single crystal silicon layer in
the defect layer; forming a second single crystal silicon layer
having lower impurity concentration than the first single crystal
silicon layer on the first single crystal silicon layer by
epitaxial growth; forming a photodiode in the second single crystal
silicon layer so that a light-receiving section faces the silicon
oxide layer; forming an interconnect layer including a photodiode
charge read-out structure on a surface of the second single crystal
silicon layer which is opposite to the first single crystal silicon
layer; and selectively etching part of or the entire second wafer
with respect to the silicon oxide layer.
In the manufacturing method of the third solid-state image sensor
according to the present disclosure, the third solid-state image
sensor does not include single crystal silicon produced by CZ
growth, which inevitably has variations in impurity concentration.
This prevents fixed pattern noise occurring in the solid-state
image sensor manufactured by a conventional manufacturing method.
This enables manufacture of a solid-state image sensor with reduced
degradation in the quality of an obtained image. Furthermore,
read-out gates, drains, interconnects, etc. of the solid-state
image sensor can be formed on the surface of the defect free second
single crystal silicon layer, thereby further improving the image
quality of the solid-state image sensor. There is no need to form a
depletion barrier layer by ion implantation to perform activation
annealing after forming the read-out gates, interconnects, etc.
Thus, a negative influence of heat treatment on the interconnects,
etc. can be avoided.
In the manufacturing method of the third solid-state image sensor
according to the present disclosure, the first single crystal
silicon layer is preferably of a first conductivity type. The
second single crystal silicon layer is preferably of a second
conductivity type. The photodiode is preferably of the second
conductivity type.
In the manufacturing method of the third solid-state image sensor
according to the present disclosure, the first single crystal
silicon layer may be of a first conductivity type. The second
single crystal silicon layer may be of a second conductivity type.
The forming of the photodiode may include forming a well of the
first conductivity type in the second single crystal silicon layer.
The photodiode may be of the second conductivity type, and may be
formed in the well.
A solid-state image sensor according to the present disclosure
includes a plate-like single crystal silicon layer formed by
epitaxial growth, and including a first surface and a second
surface facing the first surface, an interconnect layer provided on
the first surface of the single crystal silicon layer and including
a photodiode charge read-out structure; and a plurality of
photodiodes formed in the single crystal silicon layer so that
light-receiving sections face the second surface.
The solid-state image sensor according to the present disclosure
does not include single crystal silicon produced by CZ growth,
which inevitably has variations in impurity concentration. This
prevents fixed pattern noise occurring in a conventional
solid-state image sensor.
The solid-state image sensor according to the present disclosure
may further include an insulating film provided on the second
surface of the single crystal silicon layer, and color filters
provided on the insulating film to correspond to the
photodiodes.
The solid-state image sensor according to the present disclosure
may further include on-chip microlenses provided on the second
surface of the single crystal silicon layer to correspond to the
photodiodes.
The solid-state image sensor according to the present disclosure
may further include an insulating film provided on the second
surface of the single crystal silicon layer; color filters provided
on the insulating film; and on-chip microlenses provided on the
color filters to correspond to the photodiodes.
According to the solid-state image sensor and the manufacturing
method of the solid-state image sensor of the present disclosure, a
solid-state image sensor does not include single crystal silicon
produced by CZ growth, which inevitably has variations in impurity
concentration. As a result, a solid-state image sensor is obtained,
which is free from fixed pattern noise occurring in a solid-state
image sensor manufactured by a conventional manufacturing
method.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a cross-sectional view of a solid-state image sensor
according to a first embodiment of the present disclosure.
FIGS. 2A-2E are cross-sectional views illustrating a manufacturing
method of the solid-state image sensor according to the first
embodiment in order of steps.
FIGS. 3A-3F are cross-sectional views illustrating the
manufacturing method of the solid-state image sensor according to
the first embodiment in order of steps.
FIGS. 4A-4E are cross-sectional views illustrating a manufacturing
method of the solid-state image sensor according to a second
embodiment in order of steps.
FIGS. 5A-5G are cross-sectional views illustrating the
manufacturing method of the solid-state image sensor according to
the second embodiment in order of steps.
FIGS. 6A-6E are cross-sectional views illustrating a manufacturing
method of the solid-state image sensor according to a third
embodiment in order of steps.
FIGS. 7A-7F are cross-sectional views illustrating the
manufacturing method of the solid-state image sensor according to
the third embodiment in order of steps.
FIG. 8 is a schematic view illustrating fixed pattern noise caused
by variations in activation of laser annealing.
FIG. 9 is a schematic view illustrating variations in impurity
concentration in a semiconductor wafer manufactured by conventional
CZ growth.
FIGS. 10A-10E are cross-sectional views illustrating a
manufacturing method of a solid-state image sensor of back surface
irradiation using a conventional SOI wafer in order of steps.
FIGS. 11A-11E are cross-sectional views illustrating a
manufacturing method of a solid-state image sensor of back surface
irradiation using a conventional SOI wafer in order of steps.
FIG. 12 is a schematic view illustrating fixed pattern noise of a
solid-state image sensor of back surface irradiation using a
conventional SOI wafer.
DETAILED DESCRIPTION
First Embodiment
A solid-state image sensor according to a first embodiment of the
present disclosure will be described hereinafter with reference to
FIG. 1. FIG. 1 illustrates a cross-sectional structure of the
solid-state image sensor according to the first embodiment.
As shown in FIG. 1, the solid-state image sensor according to the
first embodiment includes a single crystal silicon layer 24 formed
by epitaxial growth, and a silicon oxide layer 22 having a surface
formed by heat treatment. Impurities are implanted into the single
crystal silicon layer 24 to have p-type conductivity for
functioning as a p-well 21.
The single crystal silicon layer 24 includes photodiodes 19 and
drains 26. Read-out gates 25 and interconnects 27 are formed on the
single crystal silicon layer 24. An interlayer insulating film 34
is formed on the single crystal silicon layer 24 to cover the
read-out gates 25 and the interconnects 27. The photodiodes 19 are
formed by forming n-type regions by implantation of ions such as
arsenic, phosphorus, antimony, etc. The light-receiving sections of
the photodiodes 19 are formed to face the silicon oxide layer
22.
For example, boron (B.sup.+) ions are implanted from the surface
provided with the silicon oxide layer 22, thereby forming the
p.sup.+ type depletion barrier layer 28 under the single crystal
silicon layer 24. The impurity concentration of the depletion
barrier layer 28 varies depending on the impurity concentration of
the photodiodes 19, and usually preferably ranges from
1.times.10.sup.17 cm.sup.-3 to 1.times.10.sup.19 cm.sup.-3 (both
inclusive). This prevents spread of the depletion layers of the
photodiodes 19 to the interface with the silicon oxide layer 22,
and provides the advantage of reducing dark signals generated by
accumulating noise electrons occurring in the interface states at
the photodiodes 19.
Color filters 29 are formed on the surface of the silicon oxide
layer 22 which is opposite to the single crystal silicon layer 24.
On-chip microlenses 30 are formed on the color filters 29.
The solid-state image sensor according to this embodiment does not
include single crystal silicon formed by Czochralski (CZ) growth,
which inevitably has concentric variations in impurity
concentration. This prevents fixed pattern noise to improve image
quality.
A manufacturing method of a solid-state image sensor including the
above-described structure will be described hereinafter with
reference to FIGS. 2A-2E and 3A-3F. FIGS. 2A-2E and 3A-3F
illustrate a manufacturing method of the solid-state image sensor
according to the first embodiment of the present disclosure in
order of steps.
First, as shown in FIG. 2A, a bond wafer 20 made of single crystal
silicon is prepared. In this embodiment, the conductivity type and
the impurity concentration of the bond wafer 20 are not directly
related to the p-well 21 of the completed solid-state image sensor
18 shown in FIG. 3F. In FIGS. 2A-2E and 3A-3F, the bond wafer 20 is
shown with the principal surface facing downward. The bond wafer 20
may be made of single crystal silicon produced by CZ growth.
Then, as shown in FIG. 2B, the single crystal silicon layer 24,
into which p-type impurities are implanted, is disposed on the
principal surface of the bond wafer 20 by epitaxial growth. In this
embodiment, the single crystal silicon layer 24 needs to have a
thickness greater than the thickness of the surface on which a
solid-state image element is eventually formed. Since a solid-state
image sensor using visible light with sensitivity needs to include
a substrate with a thickness ranging from several micrometers to
ten micrometers, the single crystal silicon layer 24 by epitaxial
growth needs to have the deposition thickness ranging from several
micrometers to ten micrometers.
Next, as shown in FIG. 2C, the bond wafer 20 is thermally oxidized
to form the silicon oxide layer 22. While only the front (lower)
surface of the formed silicon oxide layer 22 is shown in the
figure, the silicon oxide layer 22 is actually formed on the back
and side surfaces of the bond wafer 20 as well.
After that, as shown in FIG. 2D, hydrogen ions are implanted to the
single crystal silicon layer 24 from the silicon oxide layer 22 to
form a defect layer 23 inside the single crystal silicon layer 24.
After bonding the bond wafer 20 to a base wafer 17, the bond wafer
20 is separated along the defect layer 23. Thus, accelerated energy
in implanting the hydrogen ions is controlled to adjust the depth
of the ion implantation, thereby determining the thickness of the
single crystal silicon layer 24 in the silicon-on-insulator (SOI)
wafer. Therefore, in this embodiment, the depth of the defect layer
23 eventually becomes the thickness of the single crystal silicon
forming the solid-state image sensor 18. When the solid-state image
sensor 18 of back surface irradiation is formed, it has a thickness
ranging from about several micrometers to about ten micrometers as
described above. Since the deposition thickness of the single
crystal silicon layer 24 by the epitaxial growth shown in FIG. 2B
is larger than the thickness of the single crystal silicon forming
the solid-state image sensor 18, the defect layer 23 formed by
implanting the hydrogen ions is provided in the single crystal
silicon layer 24 formed by epitaxial growth.
Then, as shown in FIG. 2E, a base wafer 17 for maintaining the
strength of an SOI wafer 16 to be completed in FIG. 3B. The
surfaces of the prepared base wafer 17 and the single crystal
silicon layer 24 are cleaned, and then, the surfaces of the base
wafer 17 and the single crystal silicon layer 24 are bonded. At
this time, atomic force at the surfaces promotes the bonding only
by softly pushing the base wafer 17 and the bond wafer 20, on which
the single crystal silicon layer 24 is deposited, since the
surfaces are sufficiently flat. The cleaning aims not only to
remove contaminated materials, particles, etc. and to make the
surfaces hydrophilic to improve bonding effects. In this
embodiment, the base wafer 17 may be made of single crystal silicon
produced by CZ growth. While the base wafer 17 is shown as if it is
thinner than the bond wafer 20 for convenience, the base wafer 17
needs to have a thickness for maintaining the strength of the SOI
wafer 16. Therefore, when the diameter is 200 mm or more, a
thickness of 500 .mu.m is usually needed. The bond wafer 20 also
needs to have the same degree of thickness as the base wafer 17 to
be subjected to the steps to bonding of the bond wafer 20 to the
base wafer 17.
After that, as shown in FIG. 3A, when the base wafer 17 and the
bond wafer 20 are separated from the back surface, the single
crystal silicon layer 24 is separated in the defect layer 23. Part
of the bond wafer 20 and the single crystal silicon layer 24
eparated from the base wafer 17 may be returned to the epitaxial
growth shown in FIG. 2B and reused in manufacture of another SOI
wafer.
Then, heat treatment is performed to strengthen the bonding of
bonded surfaces. The temperature of the heat treatment is
400.degree. C. or more in view of the bonding strength, and is
preferably about 1000.degree. C.
With the above-described steps, the SOI wafer 16 including the base
wafer 17, the silicon oxide layer 22, and the single crystal
silicon layer 24 is completed as shown in FIG. 3B. In the SOI wafer
16 formed in this embodiment, the entire single crystal silicon
layer 24 is formed by epitaxial growth. Thus, the SOI wafer 16
formed in this embodiment has fewer variations in impurity
concentration than a conventional single crystal silicon layer
produced by CZ growth.
Next, as shown in FIG. 3C, the photodiodes 19 and the drains 26 are
formed in the single crystal silicon layer 24, and the read-out
gates 25 and the interconnects 27 are formed on the single crystal
silicon layer 24 to form a solid-state MOS image sensor. In the
case of a solid-state charge coupled device (CCD) image sensor,
photodiodes, charge transfer sections, gate electrodes and
interconnects are used instead. In this embodiment, the photodiodes
19 are formed in the p-type single crystal silicon layer 24. Thus,
in forming the photodiodes 19, impurities such as arsenic,
phosphorus, antimony, etc. are ion-implanted to form an n-type
region. The interlayer insulating film 34 is formed on the single
crystal silicon layer 24 to cover the read-out gates 25 and the
interconnects 27.
Note that, in this embodiment, in FIG. 2B, the single crystal
silicon layer 24 having p-type conductivity is formed by epitaxial
growth. The single crystal silicon layer 24 may be of an n-type,
and the p-well 21 may be formed by ion implantation etc. in the
step shown in FIG. 3C to form the photodiodes 19 etc. in the formed
p-well 21.
Then, as shown in FIG. 3D, the base wafer 17 is removed. It is
efficient to perform etching with an alkali solution after removing
a large part of the base wafer 17 by polishing, dry etching, etc.
However, only etching with an alkali solution suffices. Since the
base wafer 17 is made of single crystal silicon, removal is
facilitated by etching with an alkali solution. Furthermore, the
intermediate silicon oxide layer 22 is not etched by an alkali
solution, and serves as an etching stopper. This enables processing
of thinning films with high accuracy in the case where the
thickness of the SOI wafer 16 is determined mainly by the depth of
hydrogen ion implantation. The greatest advantage of using the SOI
wafer 16 is that the thickness accuracy is dramatically high as
compared to the case where a single layer single crystal silicon
wafer with a thickness of 500 .mu.m or more is processed to a
thickness of about 10 .mu.m by combining polishing and etching.
Removal of the base wafer 17 may be limited to the light-receiving
region of the solid-state image sensor 18 using a masking step.
When the base wafer 17 is removed from the entire surface of the
SOI wafer 16, the strength of the entire wafer is insufficient.
Thus, the SOI wafer 16 is preferably strengthen by bonding a quartz
wafer etc. as a support wafer on the front surface provided with
the interconnects 27 etc. in advance. Note that the support wafer
is not shown in the figure. In this embodiment, an example has been
described where only the base wafer 17 is removed. Even when
selectively removing the silicon oxide layer 22 by etching with
acid after removing the base wafer 17 with an alkali solution etc.,
the processing accuracy of the thickness is not different and the
advantage of using the SOI wafer 16 can be provided.
Then, as shown in FIG. 3E, boron ions are implanted from the
surface provided with the silicon oxide layer 22 to form a p.sup.+
type depletion barrier layer 28 near the interface in the single
crystal silicon layer 24 between the single crystal silicon layer
24 and the silicon oxide layer 22. The concentration of the
depletion barrier layer 28 varies depending on the impurity
concentration of the photodiodes 19, but usually preferably ranges
from 1.times.10.sup.17 cm.sup.-3 to 1.times.10.sup.19 cm.sup.-3
(both inclusive). This blocks spread of the depletion layers of the
photodiodes 19 to the interface with the silicon oxide film and
prevents noise electrons generated at interface states from being
accumulated at the photodiodes 19 and becoming dark signals. Note
that, even when the step is omitted, basic functions as an image
sensor can be obtained although image quality is degraded.
After that, as shown in FIG. 3F, color filters 29 and on-chip
microlenses 30 are formed on the silicon oxide layer 22 as
necessary. As a result, the solid-state image sensor 18 of back
surface irradiation is completed, which is formed of the single
crystal silicon layer 24 by epitaxial growth.
In the manufacturing method of the solid-state image sensor
according to this embodiment, the solid-state image sensor 18 does
not include single crystal silicon produced by CZ growth, which
inevitably has concentric variations in impurity concentration. As
a result, the solid-state image sensor free from fixed pattern
noise as shown in FIG. 12 can be manufactured.
While in this embodiment, the solid-state image sensor of back
surface irradiation has been described, a conventional solid-state
image sensor of front surface irradiation can be manufactured
similarly using an SOI wafer. Thinning a light-receiving region in
a solid-state image sensor of front surface irradiation is
advantageous in for example, allowing specific short wavelength to
have sensitivity, and utilizing transmitted light of the
solid-state image sensor for some purposes. In this case, the
depletion barrier layer 28 shown in FIG. 3E needs to be formed by
implanting boron ions to the surface of the single crystal silicon
layer 24 which is opposite to the above-described surface. The
formation is performed in the step shown in FIG. 3D.
While in this embodiment, a manufacturing method has been described
using a solid-state MOS image sensor as an example, a similar
manufacturing method is applicable using an SOI wafer in a
solid-state CCD image sensor.
While in this embodiment, an example has been described where the
photodiodes 19 are formed in the p-well 21, the manufacturing
method of the solid-state image sensor of the present disclosure is
applicable with a well of the other conductivity type.
While in this embodiment, an example has been described where
hydrogen ion implantation separation is used, similar advantages
can be obtained by other separation methods used in Uni Bond, for
example, ion implantation separation using argon ions etc. other
than hydrogen ions.
Second Embodiment
A manufacturing method of a solid-state image sensor according to a
second embodiment will be described hereinafter with reference to
FIGS. 4A-4E and 5A-5G.
FIGS. 4A-4E and 5A-5G illustrate a manufacturing method of the
solid-state image sensor according to the second embodiment of the
present disclosure in order of steps.
The steps shown in FIGS. 4A-5B are similar to the steps of the
first embodiment shown in FIGS. 2A-3B, and explanation thereof will
be omitted. Note that the single crystal silicon layer 24 of the
first embodiment corresponds to a first single crystal silicon
layer 32 in this embodiment.
As shown in FIG. 5C, a second single crystal silicon layer 31 is
deposited on the first single crystal silicon layer 32 by epitaxial
growth. P-type impurities are implanted into the second single
crystal silicon layer 31. As a result, the sum of the thickness of
the seperated first single crystal silicon layer 32 and the
thickness of the second single crystal silicon layer 31 is the
substrate thickness of the solid-state image sensor 18 shown in
FIG. 5G. As described above, the substrate needs to have a
thickness ranging from about several micrometers to about ten
micrometers in a solid-state image sensor of back surface
irradiation. Therefore, where the depth of the defect layer 23 in
the first single crystal silicon layer 32 is d1, and the thickness
of the deposited second single crystal silicon layer 31 in FIG. 5C
is d2, the sum of d1 and d2 needs to be equal to the substrate
thickness of the solid-state image sensor 18.
Then, as shown in FIG. 5D, photodiodes 19 and drains 26 are formed
on the second single crystal silicon layer 31, and read-out gates
25 and interconnects 27 are formed on the second single crystal
silicon layer 31 to form a solid-state MOS image sensor. In the
case of a solid-state CCD image sensor, photodiodes, charge
transfer sections, gate electrodes and interconnects are used
instead. In this embodiment, since the photodiodes 19 are formed in
the second single crystal silicon layer 31, impurities such as
arsenic, phosphorus, antimony, etc. are ion-implanted in forming
the photodiodes 19 to form an n-type region. Furthermore, the
interlayer insulating film 34 is formed on the single crystal
silicon layer 24 to cover over the read-out gates 25 and the
interconnects 27.
Note that, in this embodiment, the second single crystal silicon
layer 31 of n-type conductivity may be formed by epitaxial growth,
a p-well 21 may be formed by ion implantation etc., and the
photodiodes 19 etc. may be formed in the p-well 21.
Next, as shown in FIG. 5E, the base wafer 17 is removed. It is
efficient to perform etching with an alkali solution after removing
a large part of the base wafer 17 by polishing, dry etching, etc.
However, only etching with an alkali solution suffices. Since the
base wafer 17 is made of single crystal silicon, removal is
facilitated by etching with an alkali solution. Furthermore, the
intermediate silicon oxide layer 22 is not etched by an alkali
solution, and serves as an etching stopper. This enables processing
of thinning films with high accuracy in the case where the
thickness of the SOI wafer 16 is determined mainly by the depth of
hydrogen ion implantation. The greatest advantage of using the SOI
wafer 16 is that the thickness accuracy is dramatically high as
compared to the case where a single layer single crystal silicon
wafer with a thickness of 500 .mu.m or more is processed to a
thickness of about 10 .mu.m by combining polishing and etching.
Removal of the base wafer 17 may be limited to the light-receiving
region of the solid-state image sensor 18 using a masking step.
When the base wafer 17 is removed from the entire surface of the
SOI wafer 16, the strength of the entire wafer is insufficient.
Thus, the SOI wafer 16 is preferably strengthen by bonding a quartz
wafer etc. as a support wafer on the front surface provided with
the interconnects 27 etc. in advance. Note that the support wafer
is not shown in the figure. In this embodiment, an example has been
described where only the base wafer 17 is removed. Even when
selectively removing the silicon oxide layer 22 by etching with
acid after removing the base wafer 17 with an alkali solution etc,
the processing accuracy of the thickness is not different and the
advantage of using the SOI wafer 16 can be provided.
Then, as shown in FIG. 5F, boron ions are implanted from the
surface provided with the silicon oxide layer 22 to form a p.sup.+
type depletion barrier layer 28 near the interface in the first
single crystal silicon layer 32 between the first single crystal
silicon layer 32 and the silicon oxide layer 22. The concentration
of the depletion barrier layer 28 varies depending on the impurity
concentration of the photodiodes 19, but usually preferably ranges
from 1.times.10.sup.17 cm.sup.-3 to 1.times.10.sup.19 cm.sup.-3
(both inclusive). This blocks spread of the depletion layers of the
photodiodes 19 to the interface with the silicon oxide layer 22 and
prevents noise electrons generated at interface states from being
accumulated at the photodiodes 19 and becoming dark signals. Note
that, even when the step is omitted, basic functions as an image
sensor can be obtained although image quality is degraded.
After that, as shown in FIG. 5G, color filters 29 and on-chip
microlenses 30 are formed on the silicon oxide layer 22 as
necessary. As a result, the solid-state image sensor 18 of back
surface irradiation is completed, which is formed of the first
single crystal silicon layer 32 and the second single crystal
silicon layer 31 by epitaxial growth.
In the manufacturing method of the solid-state image sensor
according to this embodiment, the solid-state image sensor 18 does
not include single crystal silicon produced by CZ growth, which
inevitably has concentric variations in impurity concentration. As
a result, the solid-state image sensor free from fixed pattern
noise as shown in FIG. 12 can be manufactured.
In this embodiment, the SOI wafer 16 is formed by hydrogen ion
implantation separation. A large part of the defect layer 23 due to
ion implantation remains on the surface separated by hydrogen ion
implantation. Thus, when the read-out gates 25, the drains 26, the
interconnects 27, etc. are formed on the surface separated by
hydrogen ion implantation; threshold voltages vary at the read-out
gates 25 due to an increase in interface states, dark outputs
increase at the drains 26, and an increase in contact resistance,
variations in resistance, etc. at the interconnects 27, thereby
causing degradation in image equality due to an increase in noise
etc. as a solid-state image sensor. In this embodiment, since the
second single crystal silicon layer 31 is formed on a first single
crystal silicon layer 32 including a surface on which a large part
of the defect layer 23 remains. Thus, the read-out gates 25, the
drains 26, the interconnects 27, etc. of the solid-state image
sensor 18 can be formed on the surface of the defect free second
single crystal silicon layer 31. This improves image quality of the
image sensor. Furthermore, epitaxial growth of single crystal
silicon is usually performed with silane source gas at a high
temperature of 1000.degree. C. Thus, hydrogen atoms ion-implanted
from the silicon oxide layer 22 into the first single crystal
silicon layer 32 desorb to recover defects caused by ion
implantation.
The interface between the first single crystal silicon layer 32 and
the second single crystal silicon layer 31 does not preferably
exist in the depletion layers of the photodiodes 19. The crystal
defects remaining inside the photodiodes 19 are the source of dark
currents, and cause fixed pattern noise called "white defects." In
this embodiment, the thickness d1 of the separated first single
crystal silicon layer 32 is set to be less than the thickness of
the depletion barrier layer 28, thereby positioning the interface
between the first single crystal silicon layer 32 and the second
single crystal silicon layer 31 inside the depletion barrier layer
28. As a result, the structure is formed in which the interface
between the first single crystal silicon layer 32 and the second
single crystal silicon layer 31 does not exist in the depletion
layers of the photodiodes 19. This structure further improves image
quality of a solid-state image sensor.
While in this embodiment, an example has been described where
hydrogen ion implantation separation is used, similar advantages
can be obtained by other separation methods used in Uni Bond, for
example, ion implantation separation using argon ions etc. other
than hydrogen ions.
Third Embodiment
A manufacturing method of a solid-state image sensor according to a
third embodiment will be described hereinafter with reference to
FIGS. 6A-6E and 7A-7F.
FIGS. 6A-6E and 7A-7F illustrate a manufacturing method of the
solid-state image sensor according to the third embodiment of the
present disclosure in order of steps. FIG. 8 is a schematic view
illustrating fixed pattern noise caused by variations in activation
of laser annealing.
The steps shown in FIGS. 6A-7E are similar to the steps of the
second embodiment shown in FIGS. 4A-5E, and explanation thereof
will be omitted. Note that the first single crystal silicon layer
32 according to the second embodiment corresponds to a first
depletion barrier single crystal silicon layer 33 in this
embodiment. The first depletion barrier single crystal silicon
layer 33 is made of single crystal silicon having the same
conductivity type as the depletion barrier layer 28 shown in FIG.
5F and impurity concentration equal to that of the depletion
barrier layer 28, i.e., having P.sup.+ type conductivity and
impurity concentration ranging from 1.times.10.sup.17 cm.sup.-3 to
1.times.10.sup.19 cm.sup.-3 (both inclusive). In the hydrogen ion
implantation shown in FIG. 6D, the defect layer 23 formed by
hydrogen ion implantation has a depth equal to the thickness of the
depletion barrier layer 28 of the first and second embodiments,
thereby allowing the first depletion barrier single crystal silicon
layer 33 to function as the depletion barrier layer 28 of the first
and second embodiments. As a result, implantation of B.sup.+ ions
is unnecessary. In this respect, this embodiment differs from the
other embodiments.
Specifically, in this embodiment, after removing the base wafer 17
by etching shown in FIG. 7E, color filters 29, on-chip microlenses
30, etc. are formed the silicon oxide layer 22 as necessary, as
shown in FIG. 7F. As a result, the solid-state image sensor 18 of
back surface irradiation made only of single crystal silicon by
epitaxial growth is completed.
In order to allow the single crystal silicon to have electrical
characteristics of a desired conductivity type by implantation of
impurity ions such as boron, impurity atoms need to be located in
stable positions in the single crystal silicon by heat treatment
generally called "activation annealing." The activation annealing
needs to be performed by heat treatment at a temperature of
800.degree. C. or more. However, when the read-out gates 25, the
interconnects 27, etc. are already formed, and particularly, when
the interconnects 27 are made of metal such as aluminum or copper,
the temperature applied to the entire wafer is considered based on
the melting point of the interconnects 27. Heating at a temperature
of 500.degree. C. or more is difficult. As a result, only part of
ion-implanted impurities can be activated.
As a method of solving this problem, heating called "laser
annealing" can be used. In this method, a wafer is scanned with
intense laser light to heat the entire surface of the wafer. This
locally heats one of the surfaces of the wafer. There are however
two problems in this method. The first problem is heat variations
caused by scanning with laser light. The maximum radius of laser
light ranges from hundreds of micrometers to several millimeters,
which is smaller than a solid-state image sensor and greater than a
pixel size of the solid-state image sensor. The heat variations
caused by scanning of the laser light lead to variations in
activation of impurities to cause variations in conductivity
characteristics. This results in variations in characteristics of
individual pixels of the solid-state image sensor so that fixed
pattern noise caused by scanning variations of the laser light
shown in FIG. 8 occurs in the obtained image, thereby causing
degradation in image quality. The second problem is that the front
surface (upper surface) of the wafer 16 has a high temperature,
even when the SOI wafer 16 is irradiated with laser light from the
back surface, since the thickness of the formation region of the
photodiodes 19 ranges from several micrometers to ten micrometers.
In particular, as one of the advantages of a solid-state image
sensor of back surface irradiation, the layout of the interconnects
27 at the front surface is not limited by arrangement of pixels.
However, when the wafer 16 transmits part of laser light from the
back surface and the interconnects 27 reflects the part of laser
light from the back surface, the temperature of the irradiated part
with the reflected light rises more than the other parts. This may
cause variations in activation of impurities reflecting the layout
of the interconnects 27, and fixed pattern noise of other types
than the fixed pattern noise caused by scanning variations may be a
concern. While the above-described problems may be solved, when
laser annealing techniques with a short wavelength is developed,
the heat treatment in processing the back surface is an essential
problem of a solid-state image sensor of back surface
irradiation.
In the manufacturing method of the solid-state image sensor
according to this embodiment, there is no need to form the
depletion barrier layer 28 by ion implantation to perform
activation annealing, after forming the read-out gates 25 and the
interconnects 27. Therefore, fixed pattern noise caused by the
activation annealing can be avoided.
When the impurity concentration of the first depletion barrier
single crystal silicon layer 33 is sufficiently high, depletion
does not occur at the interface between the first depletion barrier
single crystal silicon layer 33 and the second single crystal
silicon layer 31, and the interface is not included inside the
photodiodes 19. As a result, white defects do not occur to obtain
excellent image quality.
While in this embodiment, an example has been described where
hydrogen ion implantation separation is used, similar advantages
can be obtained by other separation methods used in Uni Bond, for
example, ion implantation separation using argon ions etc. other
than hydrogen ions, similar to the first embodiment.
While in this embodiment, the second single crystal silicon layer
31 is of p-type conductivity and the photodiodes 19 is of n-type
conductivity, the second single crystal silicon layer 31 may be of
the n-type conductivity. In this case, a p-well is formed in the
second single crystal silicon layer 31 by ion implantation etc.,
and the photodiodes 19 of n-type conductivity may be provided in
the formed p-well.
The solid-state image sensor and the manufacturing method of the
solid-state image sensor according to the present disclosure
provide a solid-state image sensor not including single crystal
silicon produced by CZ growth which inevitably has concentric
variations in impurity concentration. This prevents fixed pattern
noise, and thus, the solid-state image sensor and the manufacturing
method of the solid-state image sensor according to the present
disclosure are particularly useful as a solid-state image sensor
and a manufacturing method etc. of the solid-state image sensor
using an SOI substrate.
* * * * *