U.S. patent application number 11/189896 was filed with the patent office on 2005-11-24 for semiconductor device and method for manufacturing the same.
This patent application is currently assigned to Matsushita Electric Industrial Co., Ltd.. Invention is credited to Shibata, Yoshiyuki.
Application Number | 20050259457 11/189896 |
Document ID | / |
Family ID | 32767575 |
Filed Date | 2005-11-24 |
United States Patent
Application |
20050259457 |
Kind Code |
A1 |
Shibata, Yoshiyuki |
November 24, 2005 |
Semiconductor device and method for manufacturing the same
Abstract
A first DRAM section including a first memory cell having a
first capacitance and a second DRAM section including a second
memory cell having a second capacitance different from the first
capacitance are provided on the same semiconductor substrate.
Inventors: |
Shibata, Yoshiyuki; (Shiga,
JP) |
Correspondence
Address: |
MCDERMOTT WILL & EMERY LLP
600 13TH STREET, N.W.
WASHINGTON
DC
20005-3096
US
|
Assignee: |
Matsushita Electric Industrial Co.,
Ltd.
Osaka
JP
|
Family ID: |
32767575 |
Appl. No.: |
11/189896 |
Filed: |
July 27, 2005 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11189896 |
Jul 27, 2005 |
|
|
|
10694891 |
Oct 29, 2003 |
|
|
|
Current U.S.
Class: |
365/51 ;
257/E21.013; 257/E21.018; 257/E21.647; 257/E21.648; 257/E21.66 |
Current CPC
Class: |
H01L 28/84 20130101;
H01L 27/1085 20130101; Y10S 257/906 20130101; H01L 27/10852
20130101; H01L 28/90 20130101; H01L 27/10894 20130101 |
Class at
Publication: |
365/051 |
International
Class: |
G11C 005/02 |
Foreign Application Data
Date |
Code |
Application Number |
Jan 31, 2003 |
JP |
2003-023542 |
Claims
1-5. (canceled)
6. A semiconductor device, comprising a first DRAM section
including a first memory cell having a first capacitive element and
a second DRAM section including a second memory cell having a
second capacitive element, the first DRAM section and the second
DRAM section being provided on the same substrate, wherein: the
semiconductor substrate and the first capacitive element are
connected to each other by a first plug, and the semiconductor
substrate and the second capacitive element arc connected to each
other by a second plug; and different materials are used for the
first plug and for the second plug.
7. The semiconductor device of claim 6, wherein a contact
resistance between the first plug and the semiconductor substrate
is different from that between the second plug and the
semiconductor substrate.
8-12. (canceled)
13. The semiconductor device of claim 6, wherein the first plug is
made of tungsten, and the second plug is made of silicon.
14. The semiconductor device of claim 6, wherein the first
capacitive element has a first capacitor lower electrode, a first
capacitor insulating film, and a first capacitor upper electrode,
the second capacitive element has a second capacitor lower
electrode, a second capacitor insulating film, and a second
capacitor upper electrode, the first plug is connected to the first
capacitor lower electrode, and the second plug is connected to the
second capacitor lower electrode.
15. The semiconductor device of claim 6, wherein the first memory
cell has, on the semiconductor substrate, a first gate insulating
film, a first gate electrode formed on the first gate insulating
film, a first diffusion layer formed in the semiconductor substrate
beside the first gate electrode, and a silicide layer formed on the
first diffusion layer, the second memory cell has, on the
semiconductor substrate, a second gate insulating film, a second
gate electrode formed on the second gate insulating film, and a
second diffusion layer formed in the semiconductor substrate beside
the second gate electrode, the first plug is connected to the
silicide layer, and the second plug is connected to the second
diffusion layer.
Description
BACKGROUND OF THE INVENTION
[0001] The present invention relates to a semiconductor device in
which a CMOS logic section and a plurality of DRAM sections used
for different applications are formed together on the same
semiconductor substrate, and more particularly to a DRAM-embedded
system LSI having a reduced power consumption and an increased
processing speed while ensuring a sufficient signal holding
characteristic.
[0002] In recent years, DRAM-embedded system LSI chips have been
attracting public attention. In a DRAM-embedded system LSI chip, a
CMOS (complementary metal oxide semiconductor) logic section and a
general-purpose DRAM (dynamic random access memory) section as a
memory device, which used to be formed on separate chips, are
formed together on the same chip.
[0003] For example, a DRAM-embedded system LSI chip used for an
image processing application, or the like, includes a DRAM section
as a memory device for storing an image information signal, and a
CMOS logic section for retrieving necessary information from the
DRAM section and performing an arithmetic operation based on the
retrieved information, and the DRAM section and the CMOS logic
section are formed together on the same chip.
[0004] A DRAM-embedded system LSI chip as described above realizes
a higher communication speed than that realized by older techniques
where data or information is exchanged between a CMOS logic section
and a DRAM section that are formed on separate chips. A
semiconductor device in which a CMOS logic section and a plurality
of DRAM sections are formed together on the same chip includes a
CMOS logic section formed on a silicon substrate, for example, and
a DRAM section including cell capacitors (memory cell capacitors)
of a particular type formed in the silicon substrate ("trench
capacitors") or on the silicon substrate ("stacked
capacitors").
[0005] It was technically difficult to form a CMOS logic section
and a DRAM section together on the same chip because they had large
areas. However, with recent miniaturization techniques, a
DRAM-embedded system LSI chip having a chip size less than or equal
to 100 mm.sup.2 has been realized. Now, a plurality of CMOS logic
sections and a plurality of application-specific DRAM sections are
formed on the same chip, whereas a single chip accommodated only
one CMOS logic section and one application-specific DRAM section
with older techniques.
[0006] However, the DRAM sections formed on a conventional
DRAM-embedded system LSI chip all have cell capacitors of the same
structure. Therefore, it is difficult to provide, on the same chip,
a DRAM section that needs to be accessed at a high speed and a DRAM
section that requires a sufficient signal holding characteristic.
This problem will be further discussed below.
[0007] FIG. 12A is a plan view illustrating a general configuration
of a conventional DRAM-embedded system LSI chip.
[0008] As illustrated in FIG. 12A, a CMOS logic section 11 is
formed on a chip 10, and a first DRAM section 12 and a second DRAM
section 13 having the same cell capacitor structure are formed on
the same chip 10. The first DRAM section 12 is associated with the
CMOS logic section 11 and needs to operate at a high speed. On the
other hand, the second DRAM section 13 is intended to operate at a
low power consumption with a sufficient signal holding
characteristic.
[0009] The value of a signal stored in each memory cell of a DRAM
is determined based on the charge stored in the capacitor of the
memory cell. FIG. 12B is a graph illustrating the relationship
between the reference charge and the charge stored in a cell
capacitor of the first DRAM section 12 and the second DRAM section
13 illustrated in FIG. 12A. Referring to FIG. 12B, where the
reference charge, based on which the value of a signal is
determined, is set to Qs, the value of a signal stored in a memory
cell of the first DRAM section 12 and the second DRAM section 13 is
determined to be "high" when a charge Qh, which is larger than the
reference charge Qs, is stored in the cell capacitor and to be
"low" when a charge Ql, which is smaller than the reference charge
Qs, is stored in the cell capacitor.
[0010] However, even if the charge Qh is stored in the cell
capacitor (whereby the signal stored in the memory cell is supposed
to be determined to be "high"), the state of the signal being held
may change over time due to, for example, a leak current through
the capacitor insulating film of the cell capacitor, an OFF-state
leak current through the transfer gate, or a leak current from the
substrate contact portion (a portion where the cell capacitor and
the semiconductor substrate are connected to each other) to the
semiconductor substrate. As a result, even if a charge that is
large enough for the signal to be determined "high" is initially
stored in a cell capacitor of the DRAM section, the stored charge
may gradually leak out over time, whereby the signal of the cell
capacitor is no longer determined to be "high". Where Qh is the
charge initially stored in a cell capacitor, t is the charge
holding time, which is the amount of time elapsed since the charge
Qh is initially stored, Q' is the amount of charge stored in the
cell capacitor after the charge holding time t, and I.sub.leak is
the amount of leak current, i.e., the amount of charge that leaks
out of the cell capacitor, Q' can be expressed as in Expression (1)
below. For the sake of simplicity, it is assumed that the leak
current is constant over time.
Q'=Qh-I.sub.leak.multidot.t (1)
[0011] As can be seen from Expression (1), the charge stored in the
cell capacitor of the DRAM section decreases over time. Therefore,
where a sufficient signal holding characteristic is required (i.e.,
where it is required that the charge holding time t, which is the
amount of time until the charge Q' decreases to be less than or
equal to Qs, is long), the charge Qh initially stored in the cell
capacitor needs to be increased. Thus, the capacitance of the cell
capacitor needs to be as large as about 30 fF.
[0012] FIG. 12C is an equivalent circuit diagram illustrating a
memory cell of the second DRAM section 13 (the DRAM section that is
intended to hold a signal written thereto sufficiently long)
illustrated in FIG. 12A. In the figure, "C" denotes a cell
capacitor, and "T" denotes a transfer gate. Where the capacitance
of the cell capacitor C is set to 30 fF, a charge of 30 fC is
stored in the cell capacitor C when a voltage VD applied across the
cell capacitor C is set to 1.0 V.
[0013] On the other hand, for the first DRAM section 12 (the DRAM
section to which signals are written at a high speed), a signal
written to a memory cell needs to be transitioned from "low" to
"high" and from "high" to "low" at a high speed, whereby the charge
stored in the cell capacitor needs to be moved at a high speed.
Therefore, a large amount of charge being stored in the cell
capacitor means a large amount of charge that needs to be moved at
a high speed. Thus, as the capacitance of the cell capacitor is
increased, a larger amount of charge will need to be moved, thereby
failing to realize a reduced power consumption and an increased
processing speed.
[0014] Moreover, for devices that are intended for use with
portable terminals, or the like, a plurality of DRAM sections
formed on the same semiconductor substrate have different operating
voltages. For example, the operating voltage of a DRAM section that
is required to provide a high processing speed is 2.5 V or 3.3 V,
while the operating voltage of a DRAM section that is required to
operate at a low power consumption is 1.5 V or 1.2 V. Accordingly,
the cell capacitors of these DRAM sections are controlled by the
different operating voltages of 2.5 V or 3.3 V and 1.5 V or 1.2 V.
If the cell capacitors of the memory cells are of the same
structure, a cell capacitor of the DRAM section of the lower
operating voltage stores a smaller amount of charge therein.
Therefore, the capacitance of the cell capacitor needs to be set so
that the amount of charge stored in a cell capacitor of the DRAM
section of the lower operating voltage is larger than the reference
charge. Thus, the capacitance of the cell capacitors forming the
memory cells of the other DRAM section of the higher operating
voltage is set equally. Therefore, a cell capacitor of the DRAM
section that is controlled by the higher operating voltage stores
more charge than necessary, thereby lowering the operating speed
and increasing the power consumption.
SUMMARY OF THE INVENTION
[0015] In view of the above, it is an object of the present
invention to provide a DRAM-embedded device in which a CMOS logic
section and a plurality of DRAM sections used for different
applications are formed together on the same semiconductor
substrate, and which realizes both a reduced power consumption and
an increased processing speed while ensuring a sufficient signal
holding characteristic.
[0016] A first semiconductor device of the present invention is a
semiconductor device, including a first DRAM section including a
first memory cell having a first capacitance and a second DRAM
section including a second memory cell having a second capacitance
different from the first capacitance, the first DRAM section and
the second DRAM section being provided on the same semiconductor
substrate.
[0017] In the first semiconductor device, the first DRAM section
and second DRAM section are provided on the same semiconductor
substrate, and the first capacitance set for the first memory cell
of the first DRAM section is different from the second capacitance
set for the second memory cell of the second DRAM section. If the
first capacitance is set to be smaller than the second capacitance,
the charge stored in the capacitor of the first memory cell is
smaller than that stored in the capacitor of the second memory
cell. Therefore, in the first DRAM section, it is not necessary to
move a large amount of charge for signal processing, and thus it is
possible to reduce the power consumption and to quickly move the
charge, thereby realizing high-speed signal processing. On the
other hand, in the second DRAM section, a large amount of charge is
stored, thereby realizing a sufficient signal holding
characteristic.
[0018] Thus, in the first semiconductor device, a plurality of DRAM
sections are provided by using a plurality of types of memory cells
having different capacitances, whereby the capacitance of the
memory cell of each DRAM section can be optimized according to the
application of the DRAM section. Therefore, in a DRAM-embedded
device in which a CMOS logic section and a plurality of DRAM
sections are formed together on the same chip, it is possible to
realize both a reduced power consumption and an increased
processing speed while ensuring a sufficient signal holding
characteristic.
[0019] In the first semiconductor device, it is preferred that an
operating voltage of the first DRAM section is higher than that of
the second DRAM section, and the first capacitance is smaller than
the second capacitance.
[0020] In this way, the first capacitance of the first memory cell
in the first DRAM section of the higher operating voltage is set to
be relatively small, while the second capacitance of the second
memory cell in the second DRAM section of the lower operating
voltage is set to be relatively large. Thus, those effects provided
by the first semiconductor device can reliably be obtained. The
reason therefor will now be described. Where a plurality of DRAM
sections of different operating voltages are provided on the same
semiconductor substrate with the memory cells of all the DRAM
sections having an equal capacitance, a larger amount of charge is
stored in the capacitive element of the memory cell of a DRAM
section of a higher operating voltage than in that of the memory
cell of another DRAM section of a lower operating voltage. However,
the capacitance of the memory cell in a DRAM section of a lower
operating voltage needs to be set so that a charge greater than the
reference charge is stored in the capacitive element. In such a
case, more charge than necessary is stored in the capacitive
element of the memory cell in the DRAM section of the higher
operating voltage. As a result, it is necessary to move the stored
charge, which is more than necessary, for signal processing,
thereby reducing the operating speed and increasing the power
consumption. Thus, even when a plurality of DRAM sections of
different operating voltages are provided, those effects provided
by the first semiconductor device can reliably be obtained by
optimizing the capacitance of the capacitive element for each DRAM
section according to the application thereof by setting the first
capacitance of the first memory cell in the first DRAM section of
the higher operating voltage to be relatively small, while setting
the second capacitance of the second memory cell in the second DRAM
section of the lower operating voltage to be relatively large.
[0021] In the first semiconductor device, it is preferred that a
capacitor lower electrode of the first memory cell is provided in
the semiconductor substrate, whereas neither of a capacitor upper
electrode and a capacitor lower electrode of the second memory cell
is provided in the semiconductor substrate.
[0022] Thus, the capacitor lower electrode of the first memory cell
is provided in the semiconductor substrate, whereas the capacitor
lower electrode and the capacitor upper electrode of the second
memory cell are not provided in the semiconductor substrate,
whereby the area where the capacitive element of the second memory
cell is formed can be larger than that where the capacitive element
of the first memory cell is formed. As a result, those effects
provided by the first semiconductor device can reliably be obtained
by setting the first capacitance of the first memory cell to be
smaller than the second capacitance of the second memory cell.
[0023] Moreover, since the capacitor lower electrode of the first
memory cell is provided in the semiconductor substrate, there is no
plug for connecting the capacitive element of the first memory cell
to the semiconductor substrate. Accordingly, the resistance
(contact resistance) at the junction between the plug and the
semiconductor substrate and the resistance of the plug itself are
absent. As a result, in the first DRAM section including the first
memory cell, the resistance to the charge movement for signal
processing is low, thereby realizing high-speed processing. In
contrast, the capacitor lower electrode and the capacitor upper
electrode of the second memory cell are not provided in the
semiconductor substrate, whereby it is necessary to provide a plug
for connecting the capacitive element of the second memory cell to
the semiconductor substrate. Thus, there is a contact resistance
between the plug and the semiconductor substrate, and there is a
resistance of the plug itself. As a result, the leak current, which
occurs due to the charge stored in the capacitive element of the
second memory cell leaking to the semiconductor substrate via the
plug, is reduced. Thus, in the second DRAM section including the
second memory cell, the charge stored in the capacitive element is
unlikely to leak, whereby it is possible to hold the signal stored
therein sufficiently long and to reduce the amount of power that is
used for recharging the capacitive element.
[0024] Thus, in a DRAM-embedded device in which a CMOS logic
section and a plurality of DRAM sections are formed together on the
same chip, it is possible to realize both a reduced power
consumption and an increased processing speed while ensuring a
sufficient signal holding characteristic, by optimizing the
resistance occurring due to the connection between the capacitive
element of the memory cell and the semiconductor substrate for each
DRAM section according to the application thereof, as described
above.
[0025] Moreover, since the capacitor lower electrode of the first
memory cell is provided in the semiconductor substrate, it is
possible to save the distance over which a charge would otherwise
have to be moved between the capacitive element of the first memory
cell and the semiconductor substrate. Therefore, the charge
movement can be completed quickly, and thus a signal can be stored
in the first memory cell at a high speed. Thus, in the first DRAM
section including the first memory cell, it is possible to realize
high-speed signal processing.
[0026] In the first semiconductor device, it is preferred that a
capacitor lower electrode of the first memory cell is made of
silicon and a surface thereof is not roughened, whereas a capacitor
lower electrode of the second memory cell is made of silicon and a
surface thereof is roughened.
[0027] In this way, the surface of the capacitor lower electrode of
the second memory cell is selectively roughened, whereby the
surface area of the capacitor lower electrode of the second memory
cell is larger than that of the capacitor lower electrode of the
first memory cell. Thus, the second capacitance of the second
memory cell can be set to be larger than the first capacitance of
the first memory cell. Therefore, those effects provided by the
first semiconductor device can reliably be obtained.
[0028] Moreover, effects as described above can reliably be
obtained while employing the same structure and the same material
for the first memory cell and the second memory cell, as long as
the surface of the capacitor lower electrode of the second memory
cell is roughened.
[0029] In the first semiconductor device, it is preferred that
different materials are used for a capacitor insulating film of the
first memory cell and for a capacitor insulating film of the second
memory cell.
[0030] Thus, different materials are used for the capacitor
insulating film of the first memory cell and for that of the second
memory cell, whereby the second capacitance of the second memory
cell can be set to be larger than the first capacitance of the
first memory cell, for example, by using the difference between the
dielectric constants of these materials. Therefore, those effects
provided by the first semiconductor device can reliably be
obtained.
[0031] A second semiconductor device of the present invention is a
semiconductor device, including a first DRAM section including a
first memory cell having a first capacitive element and a second
DRAM section including a second memory cell having a second
capacitive element, the first DRAM section and the second DRAM
section being provided on the same substrate, wherein: the
semiconductor substrate and the first capacitive element are
connected to each other by a first plug, and the semiconductor
substrate and the second capacitive element are connected to each
other by a second plug; and different materials are used for the
first plug and for the second plug.
[0032] In the second semiconductor device, the first DRAM section
and the second DRAM section, which are different from each other,
are provided on the same semiconductor substrate, and different
materials are used for the first plug in the first DRAM section for
connecting the first capacitive element to the semiconductor
substrate and for the second plug in the second DRAM section for
connecting the second capacitive element to the semiconductor
substrate. Therefore, the resistances (contact resistances) at the
junction between the first and second plugs and the semiconductor
substrate and the resistances of the first and second plugs can be
optimized individually according to the applications of the first
DRAM section and the second DRAM section. Consider a case where the
total resistance of the contact resistance between the first plug
and the semiconductor substrate and the resistance of the first
plug itself is smaller than the total resistance of the contact
resistance between the second plug and the semiconductor substrate
and the resistance of the second plug itself In such a case, in the
first DRAM section including the first memory cell having the first
plug, the resistance to the charge movement for signal processing
is low, thereby realizing high-speed signal processing. In
contrast, in the second DRAM section including the second memory
cell having the second plug, the resistance is high as described
above, whereby the leak current, which occurs due to the charge
stored in the second capacitive element leaking to the
semiconductor substrate via the second plug, is reduced. Therefore,
in the second DRAM section, it is possible to hold the signal
stored therein sufficiently long and to reduce the amount of power
that is used for recharging the capacitive element.
[0033] Thus, in the second semiconductor device, a plurality of
DRAM sections are provided using memory cells whose plugs for
connecting the capacitive element to the semiconductor substrate
are different from each other, whereby it is possible to optimize
the resistance occurring due to the connection between the
capacitive element and the semiconductor substrate for each DRAM
section according to the application thereof Therefore, in a
DRAM-embedded device in which a CMOS logic section and a plurality
of DRAM sections are formed together on the same chip, it is
possible to realize both a reduced power consumption and an
increased processing speed while ensuring a sufficient signal
holding characteristic.
[0034] In the second semiconductor device, it is preferred that a
contact resistance between the first plug and the semiconductor
substrate is different from that between the second plug and the
semiconductor substrate.
[0035] In this way, the contact resistances between the first and
second plugs and the semiconductor substrate can be optimized
individually. For example, when a silicide layer is formed on the
semiconductor substrate connected to one of the first plug and the
second plug, the contact resistance can be reduced. Therefore,
those effects provided by the second semiconductor device can
reliably be obtained.
[0036] A first method for manufacturing a semiconductor device of
the present invention is a method for manufacturing a semiconductor
device, including the steps of: successively forming a first
capacitor insulating film and a first capacitor upper electrode on
a semiconductor substrate in a first memory region, thereby forming
a first capacitive element; forming an interlayer insulating film
on the semiconductor substrate, on which the first capacitive
element has been formed; forming a plug connected to the
semiconductor substrate in the interlayer insulating film in a
second memory region; and successively forming a second capacitor
lower electrode connected to the plug, a second capacitor
insulating film and a second capacitor upper electrode on the
interlayer insulating film in the second memory region, thereby
forming a second capacitive element.
[0037] According to the first method for manufacturing a
semiconductor device, the first capacitive element and the second
capacitive element having different structures are formed on the
same semiconductor substrate in the first memory region and the
second memory region, respectively. In the first memory region, the
first capacitive element is formed directly on the semiconductor
substrate. In contrast, in the second memory region, the second
capacitive element is formed on a plug, wich is formed in the
interlayer insulating film formed on the semiconductor substrate,
whereby the area where the second capacitive element is formed can
be larger than that where the first capacitive element is formed in
the first memory region. Thus, the capacitance of the first
capacitive element in the first memory region can be set to be
smaller than that of the second capacitive element in the second
memory region. Therefore, where the DRAM section including the
memory cell having the first capacitive element is formed in the
first memory region, it is not necessary in that DRAM section to
move a large amount of charge for signal processing and thus it is
possible to reduce the power consumption and to quickly move the
charge, thereby realizing high-speed signal processing. On the
other hand, where the DRAM section including the memory cell having
the second capacitive element is formed in the second memory
region, a large amount of charge is stored in that DRAM section,
thereby satisfying a sufficient signal holding characteristic.
[0038] Thus, according to the first method for manufacturing a
semiconductor device, a plurality of DRAM sections are provided
using a plurality of types of capacitive elements having different
capacitances, whereby it is possible to optimize the capacitance of
the capacitive element for each DRAM section according to the
application thereof. Therefore, in a DRAM-embedded device in which
a CMOS logic section and a plurality of DRAM sections are formed
together on the same chip, it is possible to realize both a reduced
power consumption and an increased processing speed while ensuring
a sufficient signal holding characteristic.
[0039] Moreover, in the first memory region, the first capacitive
element is formed directly on the semiconductor substrate, whereby
it is not necessary to form a plug for connecting the first
capacitive element to the semiconductor substrate. Accordingly, the
contact resistance between the plug and the semiconductor substrate
and the resistance of the plug itself are absent. As a result,
where the DRAM section including the memory cell having the first
capacitive element is formed in the first memory region, the
resistance to the charge movement for signal processing is low in
that DRAM section, thereby realizing high-speed processing. In
contrast, in the second memory region, the second capacitive
element is formed on a plug, which is formed in the interlayer
insulating film on the semiconductor substrate, whereby there is a
contact resistance between the plug and the semiconductor
substrate, and there is a resistance of the plug itself. As a
result, the leak current, which occurs due to the charge stored in
the first capacitive element leaking to the semiconductor substrate
via the plug, is reduced. Thus, where the DRAM section including
the memory cell having the second capacitive element is formed in
the second memory region, the charge is unlikely to leak from the
second capacitive element in that DRAM section, whereby it is
possible to hold the signal stored therein sufficiently long and to
reduce the amount of power that is used for recharging the
capacitive element.
[0040] Thus, according to the first method for manufacturing a
semiconductor device, the resistance occurring due to the
connection between the semiconductor substrate and the capacitive
element can be optimized for each DRAM section according to the
application thereof. Therefore, in a DRAM-embedded device in which
a CMOS logic section and a plurality of DRAM sections are formed
together on the same chip, it is possible to realize both a reduced
power consumption and an increased processing speed while ensuring
a sufficient signal holding characteristic.
[0041] Moreover, in the first memory region, the first capacitive
element is formed directly on the semiconductor substrate, whereby
it is possible to save the distance over which a charge would
otherwise have to be moved between the first capacitive element and
the semiconductor substrate. Therefore, the charge movement can be
completed quickly, and thus a signal can be stored in the first
capacitive element at a high speed. Thus, in the DRAM section
including the memory cell having the first capacitive element, it
is possible to realize high-speed signal processing.
[0042] A second method for manufacturing a semiconductor device of
the present invention is a method for manufacturing a semiconductor
device, including the steps of: forming an interlayer insulating
film on a semiconductor substrate; forming a first plug and a
second plug connected to the semiconductor substrate in the
interlayer insulating film in a first memory region and in a second
memory region, respectively; forming a first capacitor lower
electrode, which is made of silicon and connected to the first
plug, on the interlayer insulating film in the first memory region,
and forming a second capacitor lower electrode, which is made of
silicon and connected to the second plug, on the interlayer
insulating film in the second memory region; selectively roughening
only a surface of the second capacitor lower electrode; and forming
a first capacitor upper electrode on the first capacitor lower
electrode with a first capacitor insulating film interposed
therebetween, thereby forming a first capacitive element, and
forming a second capacitor upper electrode on the roughened second
capacitor lower electrode with a second capacitor insulating film
interposed therebetween, thereby forming a second capacitive
element.
[0043] According to the second method for manufacturing a
semiconductor device, the first capacitive element and the second
capacitive element, which are different from each other, are formed
on the same semiconductor substrate in the first memory region and
the second memory region, respectively. The surface of only the
second capacitor lower electrode in the second memory region is
selectively roughened. Therefore, the surface area of the second
capacitor lower electrode in the second memory region is larger
than that of the first capacitor lower electrode in the first
memory region. Thus, the capacitance of the second capacitive
element is larger than that of the first capacitive element. As a
result, where the DRAM section including the memory cell having the
first capacitive element is formed in the first memory region, it
is not necessary in that DRAM section to move a large amount of
charge for signal processing and thus it is possible to reduce the
power consumption and to quickly move the charge, thereby realizing
high-speed signal processing. On the other hand, where the DRAM
section including the memory cell having the second capacitive
element is formed in the second memory region, a large amount of
charge is stored in that DRAM section, thereby satisfying a
sufficient signal holding characteristic.
[0044] Thus, according to the second method for manufacturing a
semiconductor device, a plurality of DRAM sections are provided by
using a plurality of types of capacitive elements having different
capacitances, whereby it is possible to optimize the capacitance of
the capacitive element for each DRAM section according to the
application thereof Therefore, in a DRAM-embedded device in which a
CMOS logic section and a plurality of DRAM sections are formed
together on the same chip, it is possible to realize both a reduced
power consumption and an increased processing speed while ensuring
a sufficient signal holding characteristic.
[0045] Moreover, effects as described above can reliably be
obtained while employing the same structure and the same material
for the memory cell having the first capacitive element and for the
memory cell having the second capacitive element, as long as the
surface of the second capacitor lower electrode in the second
memory region is roughened.
[0046] In the second method for manufacturing a semiconductor
device, it is preferred that a step of removing a native oxide film
covering the surface of the second capacitor lower electrode by
using hydrofluoric acid is performed between the step of forming
the first and second capacitor lower electrodes and the step of
roughening the surface of the second capacitor lower electrode; and
the step of roughening the surface of the second capacitor lower
electrode includes a step of subjecting the semiconductor substrate
to a heat treatment in a silicon-containing gas atmosphere.
[0047] In this way, the native oxide film covering the surface of
the second capacitor lower electrode can be removed by using
hydrofluoric acid before the surface of the second capacitor lower
electrode is roughened. Therefore, particles in a
silicon-containing gas can reliably be attached to the surface of
the second capacitor lower electrode through the heat treatment on
the semiconductor substrate in a silicon-containing gas atmosphere.
As a result, hemispherical grains are reliably formed on the
surface of the second capacitor lower electrode, whereby the
surface area of the second capacitor lower electrode can reliably
be larger than that of the first capacitor lower electrode.
Therefore, those effects provided by the second method for
manufacturing a semiconductor device can reliably be obtained.
[0048] A third method for manufacturing a semiconductor device of
the present invention is a method for manufacturing a semiconductor
device, including the steps of: forming an interlayer insulating
film on a semiconductor substrate; forming a first plug and a
second plug connected to the semiconductor substrate in the
interlayer insulating film in a first memory region and in a second
memory region, respectively; forming a first capacitor lower
electrode, which is made of silicon and connected to the first
plug, on the interlayer insulating film in the first memory region,
and forming a second capacitor lower electrode, which is made of
silicon and connected to the second plug, on the interlayer
insulating film in the second memory region; forming a first
capacitor insulating film made of a first material on the first
capacitor lower electrode, and forming a second capacitor
insulating film made of a second material different from the first
material on the second capacitor lower electrode; and forming a
first capacitor upper electrode on the first capacitor insulating
film, thereby forming a first capacitive element, and forming a
second capacitor upper electrode on the second capacitor insulating
film, thereby forming a second capacitive element.
[0049] According to the third method for manufacturing a
semiconductor device, the first capacitive element and the second
capacitive element, which are different from each other, are formed
on the same semiconductor substrate in the first memory region and
the second memory region, respectively. The first material and the
second material, which are different from each other, are used for
the first capacitor insulating film of the first capacitive element
and for the second capacitor insulating film of the second
capacitive element, respectively. Therefore, the capacitance of the
first capacitive element can be set to be different from that of
the second capacitive element. Assuming that the capacitance of the
first capacitive element is smaller than that of the second
capacitive element, where the DRAM section including the memory
cell having the first capacitive element is formed in the first
memory region, it is not necessary in that DRAM section to move a
large amount of charge for signal processing and thus it is
possible to reduce the power consumption and to quickly move the
charge, thereby realizing high-speed signal processing. On the
other hand, where the DRAM section including the memory cell having
the second capacitive element is formed in the second memory
region, a large amount of charge is stored in that DRAM section,
thereby satisfying a sufficient signal holding characteristic.
[0050] Thus, according to the third method for manufacturing a
semiconductor device, a plurality of DRAM sections are provided by
using a plurality of types of capacitive elements having different
capacitances, whereby it is possible to optimize the capacitance of
the capacitive element for each DRAM section according to the
application thereof Therefore, in a DRAM-embedded device in which a
CMOS logic section and a plurality of DRAM sections are formed
together on the same chip, it is possible to realize both a reduced
power consumption and an increased processing speed while ensuring
a sufficient signal holding characteristic.
[0051] A fourth method for manufacturing a semiconductor device of
the present invention is a method for manufacturing a semiconductor
device, including the steps of: forming an interlayer insulating
film on a semiconductor substrate; forming a first hole reaching
the semiconductor substrate in the interlayer insulating film in a
first memory region; forming a first plug made of a metal film in
the first hole; forming a second hole reaching the semiconductor
substrate in the interlayer insulating film in a second memory
region; forming a second plug made of a silicon film in the second
hole; and successively forming a first capacitor lower electrode
connected to the first plug, a first capacitor insulating film and
a first capacitor upper electrode on the interlayer insulating film
in the first memory region, thereby forming a first capacitive
element, and successively forming a second capacitor lower
electrode connected to the second plug, a second capacitor
insulating film and a second capacitor upper electrode on the
interlayer insulating film in the second memory region, thereby
forming a second capacitive element.
[0052] According to the fourth method for manufacturing a
semiconductor device, the first capacitive element and the second
capacitive element are formed on the same semiconductor substrate
in the first memory region and the second memory region,
respectively. The first plug, which is made of a metal film, is
formed for connecting the first capacitive element in the first
memory region to the semiconductor substrate, and the second plug,
which is made of a silicon film, is formed for connecting the
second capacitive element in the second memory region to the
semiconductor substrate. Thus, the resistance of the first plug can
be made lower than that of the second plug. Moreover, the contact
resistance between the first plug and the semiconductor substrate
can be made lower than that between the second plug and the
semiconductor substrate. As a result, where the DRAM section
including the memory cell having the first capacitive element is
formed in the first memory region, the resistance to the charge
movement for signal processing is low in that DRAM section, thereby
realizing high-speed signal processing. In contrast, where the DRAM
section including the memory cell having the second capacitive
element is formed in the second memory region, the resistance
between the second capacitive element and the semiconductor
substrate is relatively high in that DRAM section, whereby the leak
current, which occurs due to the charge stored in the second
capacitive element leaking to the semiconductor substrate via the
second plug, is reduced. As a result, the charge stored in the
second capacitive element is unlikely to leak, whereby it is
possible to hold the signal stored therein sufficiently long and to
reduce the power consumption.
[0053] Thus, according to the fourth method for manufacturing a
semiconductor device, the resistance occurring due to the
connection between the capacitive element and the semiconductor
substrate via the plug can be optimized for each DRAM section
according to the application thereof Therefore, in a DRAM-embedded
device in which a CMOS logic section and a plurality of DRAM
sections are formed together on the same chip, it is possible to
realize both a reduced power consumption and an increased
processing speed while ensuring a sufficient signal holding
characteristic.
[0054] Note that in the fourth method for manufacturing a
semiconductor device, a silicide layer may be formed on the
semiconductor substrate in advance before the formation of the
first plug. Then, the contact resistance between the first plug and
the semiconductor substrate is reduced, whereby those effects
provided by the fourth method for manufacturing a semiconductor
device can reliably be obtained.
[0055] Moreover, in the fourth method for manufacturing a
semiconductor device, a step of forming a barrier metal film in the
first hole may be performed between the step of forming the first
hole and the step of forming the first plug.
BRIEF DESCRIPTION OF THE DRAWINGS
[0056] FIG. 1A is a plan view illustrating a general configuration
of a semiconductor device according to first to fifth embodiments
of the present invention, FIG. 1B and FIG. 1C are equivalent
circuit diagrams illustrating memory cells of a first DRAM section
and a second DRAM section, respectively, illustrated in FIG. 1A,
and FIG. 1D and FIG. 1E are graphs illustrating the charge storing
characteristics of cell capacitors of the first DRAM section and
the second DRAM section, respectively, illustrated in FIG. 1A.
[0057] FIG. 2 is a cross-sectional view illustrating a
semiconductor device according to the second embodiment of the
present invention.
[0058] FIG. 3A to FIG. 3D are cross-sectional views illustrating
steps in a method for manufacturing the semiconductor device
according to the second embodiment of the present invention.
[0059] FIG. 4A and FIG. 4B are cross-sectional views illustrating
steps in a method for manufacturing the semiconductor device
according to the second embodiment of the present invention.
[0060] FIG. 5 is a cross-sectional view illustrating a
semiconductor device according to the third embodiment of the
present invention.
[0061] FIG. 6A to FIG. 6D are cross-sectional views illustrating
steps in a method for manufacturing the semiconductor device
according to the third embodiment of the present invention.
[0062] FIG. 7 is a cross-sectional view illustrating a
semiconductor device according to the fourth embodiment of the
present invention.
[0063] FIG. 8A to FIG. 8D are cross-sectional views illustrating
steps in a method for manufacturing the semiconductor device
according to the fourth embodiment of the present invention.
[0064] FIG. 9 is a cross-sectional view illustrating a
semiconductor device according to the fifth embodiment of the
present invention.
[0065] FIG. 10A to FIG. 10E are cross-sectional views illustrating
steps in a method for manufacturing the semiconductor device
according to the fifth embodiment of the present invention.
[0066] FIG. 11A and FIG. 11B are cross-sectional views illustrating
steps in a method for manufacturing the semiconductor device
according to the fifth embodiment of the present invention.
[0067] FIG. 12A is a plan view illustrating a general configuration
of a conventional DRAM-embedded system LSI chip, FIG. 12B is a
graph illustrating the relationship between the reference charge
and the charge stored in a cell capacitor of a first DRAM section
and a second DRAM section illustrated in FIG. 12A, and FIG. 12C is
an equivalent circuit diagram illustrating a memory cell of the
second DRAM section illustrated in FIG. 12A.
DETAILED DESCRIPTION OF THE INVENTION
First Embodiment
[0068] A semiconductor device according to the first embodiment of
the present invention will now be described with reference to the
drawings.
[0069] FIG. 1A is a plan view illustrating a general configuration
of the semiconductor device according to the first embodiment of
the present invention.
[0070] As illustrated in FIG. 1A, a CMOS logic section 101 is
formed on a chip 100, and a first DRAM section 102 and a second
DRAM section 103 are formed on the same chip 100. The first DRAM
section 102 is intended to provide temporary storage of data being
processed at a high speed, i.e., to store a charge that is moved at
a high speed. In contrast, the second DRAM section 103 is intended
to hold a signal written thereto sufficiently long.
[0071] A feature of the present embodiment is that the capacitances
of the memory cells of the first DRAM section 102 and the second
DRAM section 103 are individually set for their respective
applications.
[0072] FIG. 1B and FIG. 1C are equivalent circuit diagrams
illustrating memory cells of the first DRAM section 102 and the
second DRAM section 103, respectively, illustrated in FIG. 1A.
[0073] As illustrated in FIG. 1B, a memory cell of the first DRAM
section 102 includes a first transfer gate 104A and a first
capacitor 105A. As illustrated in FIG. 1C, a memory cell of the
second DRAM section 103 includes a second transfer gate 104B and a
second capacitor 105B.
[0074] The charge storing characteristics of the first capacitor
105A and the second capacitor 105B will now be described, assuming
that the capacitance of the first capacitor 105A is C.sub.a, the
power supply voltage of the first capacitor 105A is V.sub.DDa, the
plate voltage of the first capacitor 105A is V.sub.Pa, the charge
stored in the first capacitor 105A is Q.sub.a, the capacitance of
the second capacitor 105B is C.sub.b, the power supply voltage of
the second capacitor 105B is V.sub.DDb, the plate voltage of the
second capacitor 105B is V.sub.Pb, and the charge stored in the
second capacitor 105B is Q.sub.b.
[0075] FIG. 1D is a graph illustrating the relationship between the
capacitance C.sub.a of the first capacitor 105A illustrated in FIG.
1B and the operating voltage (V.sub.DDa-V.sub.Pa) applied between
the electrodes of the first capacitor 105A. FIG. 1E is a graph
illustrating the relationship between the capacitance C.sub.b of
the second capacitor 105B illustrated in FIG. 1C and the operating
voltage (V.sub.DDb-V.sub.Pb) applied between the electrodes of the
second capacitor 105B. For example, C.sub.a=6 fF and
(V.sub.DDa-V.sub.Pa)=1.2 V for the first capacitor 105A of the
first DRAM section 102, whereas C.sub.b=20 fF and
(V.sub.DDb-V.sub.Pb)=0.75 V for the second capacitor 105B of the
second DRAM section 103. The amount of charge stored in the
capacitors can be represented as a physical quantity whose value is
equal to the area of the hatched portion in FIG. 1D and FIG. 1E.
Therefore, the amount of charge stored in the first capacitor 105A
is 7.2 fC based on Expression (2) below.
Q.sub.a=C.sub.a(V.sub.DDa-V.sub.Pa)=6 fF.multidot.1.2 V=7.2 fC
(2)
[0076] On the other hand, the amount of charge stored in the second
capacitor 105B is 15 fC based on Expression (3) below.
Q.sub.b=C.sub.b.multidot.(V.sub.DDb-V.sub.Pb)=20 fF.multidot.0.75
V=15 fC (3)
[0077] Thus, the amount of charge stored in the first capacitor
105A is smaller than that in the second capacitor 105B by 7.8 fC.
Therefore, according to the present embodiment, in the first DRAM
section 102 of the higher operating voltage, it is not necessary to
move a large amount of charge for signal processing, and thus it is
possible to reduce the power consumption and to quickly move the
charge, thereby realizing high-speed signal processing. On the
other hand, in the second DRAM section 103, a large amount of
charge is stored, thereby realizing a sufficient signal holding
characteristic. Thus, it is possible to realize a DRAM-embedded
device that realizes both a reduced power consumption and an
increased processing speed while ensuring a sufficient signal
holding characteristic, by optimizing the capacitance of a memory
cell for each DRAM section according to the application thereof, as
described above. Note that according to the present embodiment, the
amount of charge stored in the first capacitor 105A is smaller than
that where C.sub.a=C.sub.b=20 fF as in the conventional
semiconductor device by 16.8 fC. Therefore, it is possible to save
the power that would otherwise be used for moving the extra 16.8 fC
charge. Moreover, since it is not necessary to move the extra
amount of charge, it is possible to realize high-speed signal
processing.
[0078] Note that in the first embodiment, the capacitance and
applied voltage of each of the first capacitor 105A and the second
capacitor 105B are arbitrarily determined within a certain range
according to the purpose thereof. Note however that the charge
Q.sub.a stored in the first capacitor 105A needs to be smaller than
the charge Q.sub.b stored in the second capacitor 105B.
[0079] Moreover, the operating voltages of the first DRAM section
102 and the second DRAM section 103 are different from each other
in the first embodiment, they may be equal to each other and
similar effects can still be obtained, as long as the capacitance
of the first capacitor 105A is set to be smaller than that of the
second capacitor 105B.
Second Embodiment
[0080] A semiconductor device according to the second embodiment
and a method for manufacturing the same will now be described with
reference to the drawings. Also in the present embodiment, the
general configuration of the semiconductor device is as shown in
the plan view of FIG. 1A.
[0081] FIG. 2 is a cross-sectional view illustrating the
semiconductor device according to the second embodiment. More
specifically, FIG. 2 shows memory cells of the first DRAM section
102 and the second DRAM section 103 formed on the chip 100
illustrated in FIG. 1A. Note that the two regions where the first
DRAM section 102 and the second DRAM section 103 are formed are
referred to as "first memory region R.sub.A" and "second memory
region R.sub.B", respectively.
[0082] As illustrated in FIG. 2, a first gate electrode 203A is
formed on a semiconductor substrate 200 in the first memory region
R.sub.A with a first gate insulating film 202A interposed between
the semiconductor substrate 200 and the first gate electrode 203A.
Thus, a transfer gate is formed in the first memory region R.sub.A.
A first capacitor upper electrode 205 is formed on the
semiconductor substrate 200 in the first memory region R.sub.A with
a first capacitor insulating film 204 interposed between the
semiconductor substrate 200 and first capacitor upper electrode
205. A second gate electrode 203B is formed on the semiconductor
substrate 200 in the second memory region R.sub.B with a second
gate insulating film 202B interposed between the semiconductor
substrate 200 and the second gate electrode 203B. Thus, a transfer
gate is formed in the second memory region R.sub.B. A first
diffusion layer 206A is formed in the semiconductor substrate 200
beside the first gate electrode 203A. The first capacitor upper
electrode 205 is formed on a diffusion layer 201 (first capacitor
lower electrode), which is connected to the first diffusion layer
206A, with the first capacitor insulating film 204 interposed
between the diffusion layer 201 and the first capacitor upper
electrode 205. Thus, a cell capacitor of the first DRAM section 102
is formed in the first memory region R.sub.A. A second diffusion
layer 206B is formed in the semiconductor substrate 200 beside the
second gate electrode 203B. A first interlayer insulating film 207
is formed on the semiconductor substrate 200, the first gate
electrode 203A, the first capacitor upper electrode 205 and the
second gate electrode 203B. A contact plug 208 connected to the
second diffusion layer 206B is formed through the first interlayer
insulating film 207 in the second memory region R.sub.B. A second
capacitor lower electrode 209 is formed on the first interlayer
insulating film 207 in the second memory region R.sub.B. The second
capacitor lower electrode 209 has a tubular shape with a bottom
surface and is connected to the contact plug 208. A second
capacitor insulating film 210 is formed so as to cover the second
capacitor lower electrode 209, and a second capacitor upper
electrode 211 is formed so as to cover the second capacitor
insulating film 210. Thus, a cell capacitor of the second DRAM
section 103 is formed in the second memory region R.sub.B. A second
interlayer insulating film 212 is formed on the first interlayer
insulating film 207 and the second capacitor upper electrode 211. A
wiring layer 213 is formed on the second interlayer insulating film
212. The wiring layer 213 is formed so that portions thereof are
connected to the first capacitor upper electrode 205 and the second
capacitor upper electrode 211 through contact holes formed in the
second interlayer insulating film 212.
[0083] Next, a method for manufacturing the semiconductor device
according to the second embodiment will be described with reference
to the drawings.
[0084] FIG. 3A to FIG. 3D, FIG. 4A and FIG. 4B are cross-sectional
views illustrating steps in the method for manufacturing the
semiconductor device according to the second embodiment. Note that
in FIG. 3A to FIG. 3D, FIG. 4A and FIG. 4B, the same members as
those of the semiconductor device of the second embodiment
illustrated in FIG. 2 will be denoted by the same reference
numerals.
[0085] First, as illustrated in FIG. 3A, an impurity is implanted
into a cell capacitor formation region of the semiconductor
substrate 200 in the first memory region R.sub.A so as to form the
diffusion layer 201 to be the first capacitor lower electrode.
Then, an insulating film 202, whose main component is silicon
oxide, for example, and an electrode film 203, which is a silicon
film or a silicon/metal layered film, for example, are successively
formed across the entire surface of the semiconductor substrate
200.
[0086] Then, as illustrated in FIG. 3B, the electrode film 203 and
the insulating film 202 are patterned by photolithography and
etching. Thus, in the first memory region R.sub.A, the first gate
electrode 203A is formed on the semiconductor substrate 200 with
the first gate insulating film 202A interposed between the
semiconductor substrate 200 and the first gate electrode 203A, and
the first capacitor upper electrode 205 is formed on the
semiconductor substrate 200 with the first capacitor insulating
film 204 interposed between the semiconductor substrate 200 and the
first capacitor upper electrode 205. At the same time, in the
second memory region R.sub.B, the second gate electrode 203B is
formed on the semiconductor substrate 200 with the second gate
insulating film 202B interposed between the semiconductor substrate
200 and the second gate electrode 203B. Thus, a transfer gate is
formed in each of the first memory region R.sub.A and the second
memory region R.sub.B, and a cell capacitor of the first DRAM
section 102 is formed in the first memory region R.sub.A.
[0087] Then, as illustrated in FIG. 3C, an impurity is implanted
into the semiconductor substrate 200 in the first memory region
R.sub.A and the second memory region R.sub.B, thereby forming the
first diffusion layer 206A and the second diffusion layer 206B,
respectively. In this process, the first diffusion layer 206A is
connected to the diffusion layer 201.
[0088] Then, as illustrated in FIG. 3D, the first interlayer
insulating film 207 is formed across the entire surface of the
semiconductor substrate 200. Then, in the second memory region
R.sub.B, a contact hole reaching the second diffusion layer 206B is
formed in the first interlayer insulating film 207 by
photolithography and etching. Then, a conductive film is formed
across the entire surface of the semiconductor substrate 200 so as
to completely fill the contact hole. Then, an unnecessary portion
of the conductive film on the first interlayer insulating film 207
is removed, thereby forming the contact plug 208 connected to the
second diffusion layer 206B through the contact hole.
[0089] Then, as illustrated in FIG. 4A, the second capacitor lower
electrode 209 is formed on the first interlayer insulating film 207
in the second memory region R.sub.B. The second capacitor lower
electrode 209 has a tubular shape with a bottom surface and is
connected to the contact plug 208. The second capacitor lower
electrode 209 is made of phosphorus-containing silicon, for
example. Then, the second capacitor insulating film 210, which is a
layered film of a silicon oxide film and a silicon nitride film (ON
film), for example, is formed so as to cover the second capacitor
lower electrode 209, and the second capacitor upper electrode 211,
which is made of phosphorus-containing silicon, for example, is
formed so as to cover the second capacitor insulating film 210.
Thus, a cell capacitor of the second DRAM section 103 is formed in
the second memory region R.sub.B.
[0090] Then, as illustrated in FIG. 4B, the second interlayer
insulating film 212 is formed across the entire surface of the
semiconductor substrate 200. Then, contact holes reaching the first
capacitor upper electrode 205 and the second capacitor upper
electrode 211 are formed in the second interlayer insulating film
212. Then, the wiring layer 213 is formed on the second interlayer
insulating film 212. In this process, the wiring layer 213 is
formed so that portions thereof are connected to the first
capacitor upper electrode 205 and the second capacitor upper
electrode 211. Thus, the first DRAM section 102 and the second DRAM
section 103 are formed in the first memory region R.sub.A and the
second memory region R.sub.B, respectively.
[0091] According to the second embodiment, the cell capacitor
formed in the first memory region R.sub.A includes the first
capacitor lower electrode, which is formed from the diffusion layer
201 connected to the first diffusion layer 206A, the first
capacitor insulating film 204, and the first capacitor upper
electrode 205. Thus, the cell capacitor is a so-called "planar
capacitor" formed directly on the semiconductor substrate 200. In
contrast, the cell capacitor formed in the second memory region
R.sub.B is a so-called "stacked capacitor" formed on the contact
plug 208, which is formed in the first interlayer insulating film
207. Thus, in the second memory region R.sub.B, a larger area can
be used for forming the cell capacitor, as compared to the planar
capacitor formed in the first memory region R.sub.A. Furthermore,
since the second capacitor lower electrode 209 formed in the second
memory region R.sub.B has a tubular shape, the second capacitor
lower electrode 209 has a larger surface area. Thus, the
capacitance of the cell capacitor formed in the second memory
region R.sub.B is larger than that of the cell capacitor formed in
the first memory region R.sub.A. Therefore, in the first DRAM
section 102 where the cell capacitor formed in the first memory
region R.sub.A is used as the capacitive element, it is not
necessary to move a large amount of charge for signal processing,
and thus it is possible to reduce the power consumption and to
quickly move the charge, thereby realizing high-speed processing.
On the other hand, in the second DRAM section 103 where the cell
capacitor formed in the second memory region R.sub.B is used as the
capacitive element, a large amount of charge can be stored, thereby
realizing a sufficient signal holding characteristic.
[0092] Thus, it is possible to realize a DRAM-embedded device that
realizes both a reduced power consumption and an increased
processing speed while ensuring a sufficient signal holding
characteristic, by optimizing the capacitance of the cell capacitor
for each DRAM section according to the application thereof, as in
the present embodiment.
[0093] Moreover, according to the second embodiment, the cell
capacitor of the first memory region R.sub.A is formed directly on
the semiconductor substrate 200, whereby there is no contact plug
for connecting the cell capacitor to the first diffusion layer
206A. Accordingly, the contact resistance between the contact plug
and the first diffusion layer 206A and the resistance of the
contact plug itself are absent. Thus, in the first memory region
R.sub.A, the resistance from the transfer gate to the first
capacitor lower electrode (the diffusion layer 201) of the cell
capacitor is only the diffusion resistance in the first diffusion
layer 206A. Specifically, the diffusion resistance is as low as
about 5 .OMEGA.. As a result, in the first DRAM section 102 where
the cell capacitor formed in the first memory region R.sub.A is
used as the capacitive element, high-speed processing can be
realized because of the low resistance to the charge movement for
signal processing. In contrast, the cell capacitor of the second
memory region R.sub.B is formed on the contact plug 208, which is
formed in the first interlayer insulating film 207, whereby there
is a total resistance of the contact resistance between the contact
plug 208 and the second diffusion layer 206B and the resistance of
the contact plug 208 itself. Specifically, the total resistance is
as high as about 10 k.OMEGA.. Therefore, the leak current, which
occurs due to the charge stored in the cell capacitor formed in the
second memory region R.sub.B leaking from the second diffusion
layer 206B to the semiconductor substrate 200 via the contact plug
208, is reduced to be as small as 1 fA or less. Thus, in the second
DRAM section 103 where the cell capacitor formed in the second
memory region R.sub.B is used as the capacitive element, the charge
stored in the cell capacitor is unlikely to leak, whereby it is
possible to hold the signal stored therein sufficiently long and to
reduce the amount of power that is used for recharging the cell
capacitor.
[0094] Thus, it is possible to realize a DRAM-embedded device that
realizes both a reduced power consumption and an increased
processing speed while ensuring a sufficient signal holding
characteristic, by optimizing the resistance occurring due to the
connection between the cell capacitor and the semiconductor
substrate 200 for each DRAM section according to the application
thereof, as in the present embodiment.
[0095] Moreover, according to the second embodiment, the capacitor
lower electrode of the cell capacitor formed in the first memory
region R.sub.A is replaced with the diffusion layer 201 connected
to the first diffusion layer 206A. Therefore, in the first memory
region R.sub.A, it is possible to save the distance over which a
charge would otherwise have to be moved between the cell capacitor
and the semiconductor substrate 200. Therefore, in the first DRAM
section 102 where the cell capacitor formed in the first memory
region R.sub.A is used as the capacitive element, the charge
movement can be completed quickly, and thus a signal can be stored
in the cell capacitor at a high speed, thereby realizing high-speed
signal processing.
[0096] Note that while the inner surface and the outer surface of
the second capacitor lower electrode 209, which is formed in a
tubular shape in the second memory region R.sub.B, are both used as
a cell capacitor in the second embodiment, only the inner surface
thereof may alternatively be used as a cell capacitor.
[0097] Moreover, while the cell capacitor of the second memory
region R.sub.B has a tubular shape in the second embodiment, it may
alternatively have a cylindrical shape or any other suitable
shape.
[0098] Moreover, while the cell capacitor of the first memory
region R.sub.A is a planar capacitor in the second embodiment, it
may alternatively be a so-called "trench capacitor", which is
formed in a hole provided in the semiconductor substrate 200.
[0099] Moreover, while an insulating film whose main component is
silicon oxide is used as the insulating film 202 in the method for
manufacturing the semiconductor device according to the second
embodiment, any other suitable insulating film may alternatively be
used.
[0100] Moreover, while a silicon film or a silicon/metal layered
film is used as the electrode film 203 in the method for
manufacturing the semiconductor device according to the second
embodiment, any other suitable conductive film may alternatively be
used.
[0101] Moreover, while phosphorus-containing silicon is used for
the second capacitor lower electrode 209 and the second capacitor
upper electrode 211 in the second memory region R.sub.B in the
method for manufacturing the semiconductor device according to the
second embodiment, silicon containing any other suitable impurity
or any other suitable metal may alternatively be used.
[0102] Moreover, while an ON film is used as the second capacitor
insulating film 210 in the second memory region R.sub.B in the
method for manufacturing the semiconductor device according to the
second embodiment, any other suitable capacitor insulating film may
alternatively be used.
Third Embodiment
[0103] A semiconductor device according to the third embodiment and
a method for manufacturing the same will now be described with
reference to the drawings. Also in the present embodiment, the
general configuration of the semiconductor device is as shown in
the plan view of FIG. 1A.
[0104] FIG. 5 is a cross-sectional view illustrating the
semiconductor device according to the third embodiment. More
specifically, FIG. 5 shows memory cells of the first DRAM section
102 and the second DRAM section 103 formed on the chip 100
illustrated in FIG. 1A. Note that the two regions where the first
DRAM section 102 and the second DRAM section 103 are formed are
referred to as "first memory region R.sub.A" and "second memory
region R.sub.B", respectively.
[0105] As illustrated in FIG. 5, a first gate electrode 302A is
formed on a semiconductor substrate 300 in the first memory region
R.sub.A with a first gate insulating film 301A interposed between
the semiconductor substrate 300 and the first gate electrode 302A.
Thus, a transfer gate is formed in the first memory region R.sub.A.
A second gate electrode 302B is formed on the semiconductor
substrate 300 in the second memory region R.sub.B with a second
gate insulating film 301B interposed between the semiconductor
substrate 300 and the second gate electrode 302B. Thus, a transfer
gate is formed in the second memory region R.sub.B. A first
diffusion layer 303A is formed in the semiconductor substrate 300
beside the first gate electrode 302A. A second diffusion layer 303B
is formed in the semiconductor substrate 300 beside the second gate
electrode 302B. A first interlayer insulating film 304 is formed on
the semiconductor substrate 300, the first gate electrode 302A and
the second gate electrode 302B. A first contact plug 305A made of
silicon, for example, which is connected to the first diffusion
layer 303A, is formed in the first interlayer insulating film 304
in the first memory region R.sub.A. A second contact plug 305B made
of silicon, for example, which is connected to the second diffusion
layer 303B, is formed in the first interlayer insulating film 304
in the second memory region R.sub.B. A first capacitor lower
electrode 306A made of silicon is formed on the first interlayer
insulating film 304 in the first memory region R.sub.A. The first
capacitor lower electrode 306A has a tubular shape with a bottom
surface and is connected to the first contact plug 305A. A second
capacitor lower electrode 306B made of silicon is formed on the
first interlayer insulating film 304 in the second memory region
R.sub.B. The second capacitor lower electrode 306B has a tubular
shape with a bottom surface and is connected to the second contact
plug 305B. The surface of the second capacitor lower electrode 306B
in the second memory region R.sub.B is a roughened surface 308
having hemispherical grains. A first capacitor insulating film
309A, which is an ON film or a metal oxide film having a high
dielectric constant, for example, is formed so as to cover the
first capacitor lower electrode 306A, and a first capacitor upper
electrode 310A made of silicon, for example, is formed so as to
cover the first capacitor insulating film 309A. Thus, a cell
capacitor of the first DRAM section 102 is formed in the first
memory region R.sub.A. A second capacitor insulating film 309B,
which is an ON film or a metal oxide film having a high dielectric
constant, for example, is formed so as to cover the second
capacitor lower electrode 306B including the roughened surface 308,
and a second capacitor upper electrode 310B made of silicon, for
example, is formed so as to cover the second capacitor insulating
film 309B. Thus, a cell capacitor of the second DRAM section 103 is
formed in the second memory region R.sub.B. A second interlayer
insulating film 311 is formed on the first interlayer insulating
film 304, the first capacitor upper electrode 310A and the second
capacitor upper electrode 310B. A wiring layer 312 is formed on the
second interlayer insulating film 311. The wiring layer 312 is
formed so that portions thereof are connected to the first
capacitor upper electrode 310A and the second capacitor upper
electrode 310B.
[0106] A feature of the semiconductor device according to the third
embodiment is that the surface of the second capacitor lower
electrode 306B in the second memory region R.sub.B is the roughened
surface 308. Therefore, the surface area of the second capacitor
lower electrode 306B in the second memory region R.sub.B is larger
than that of the first capacitor lower electrode 306A in the first
memory region R.sub.A. Thus, the capacitance of the cell capacitor
formed in the second memory region R.sub.B is larger than that of
the cell capacitor formed in the first memory region R.sub.A.
Specifically, where the surface of the second capacitor lower
electrode 306B is formed into the roughened surface 308 by the HSG
(hemi spherical grained) technique, the capacitance of the cell
capacitor formed in the second memory region R.sub.B is about 20
fF, while the capacitance of the cell capacitor formed in the first
memory region R.sub.A is about 10 fF.
[0107] Next, a method for manufacturing the semiconductor device
according to the third embodiment will be described with reference
to the drawings.
[0108] FIG. 6A to FIG. 6D are cross-sectional views illustrating
steps in the method for manufacturing the semiconductor device
according to the third embodiment. Note that in FIG. 6A to FIG. 6D,
the same members as those of the semiconductor device of the third
embodiment illustrated in FIG. 5 will be denoted by the same
reference numerals.
[0109] As illustrated in FIG. 6A, the first gate electrode 302A is
formed on the semiconductor substrate 300 in the first memory
region R.sub.A with the first gate insulating film 301A interposed
between the semiconductor substrate 300 and the first gate
electrode 302A, and the second gate electrode 302B is formed on the
semiconductor substrate 300 in the second memory region R.sub.B
with the second gate insulating film 301B interposed between the
semiconductor substrate 300 and the second gate electrode 302B.
Thus, a transfer gate is formed in the first memory region R.sub.A
and in the second memory region R.sub.B. Then, the first diffusion
layer 303A is formed in the semiconductor substrate 300 beside the
first gate electrode 302A, and the second diffusion layer 303B is
formed in the semiconductor substrate 300 beside the second gate
electrode 302B. Then, the first interlayer insulating film 304 is
formed across the entire surface of the semiconductor substrate
300. Then, the first contact plug 305A is formed in the first
interlayer insulating film 304 in the first memory region R.sub.A.
The first contact plug 305A is made of silicon, for example, and is
connected to the first diffusion layer 303A. At the same time, the
second contact plug 305B is formed in the first interlayer
insulating film 304 in the second memory region R.sub.B. The second
contact plug 305B is made of silicon, for example, and is connected
to the second diffusion layer 303B. Then, the first capacitor lower
electrode 306A is formed on the first interlayer insulating film
304 in the first memory region R.sub.A. The first capacitor lower
electrode 306A has a tubular shape with a bottom surface and is
connected to the first contact plug 305A. At the same time, the
second capacitor lower electrode 306B is formed on the first
interlayer insulating film 304 in the second memory region R.sub.B.
The second capacitor lower electrode 306B has a tubular shape with
a bottom surface and is connected to the second contact plug 305B.
The first capacitor lower electrode 306A and the second capacitor
lower electrode 306B are made of amorphous silicon, for
example.
[0110] Then, as illustrated in FIG. 6B, a photoresist 307 is formed
on the first interlayer insulating film 304 in the first memory
region R.sub.A so as to completely mask the first capacitor lower
electrode 306A. Then, the second capacitor lower electrode 306B in
the second memory region R.sub.B is exposed to a chemical solution
containing hydrofluoric acid, for example, so as to remove a
silicon oxide film (not shown) formed through native oxidization
covering the surface of the second capacitor lower electrode
306B.
[0111] Then, as illustrated in FIG. 6C, the surface of the second
capacitor lower electrode 306B in the second memory region R.sub.B
is formed into the roughened surface 308 having hemispherical
grains by using, for example, the HSG technique, i.e., by
subjecting the semiconductor substrate 300 to a heat treatment at a
temperature near 620.degree. C. in a silicon-containing gas
atmosphere containing silane or disilane, for example. Then, the
photoresist 307 formed in the first memory region R.sub.A is
removed.
[0112] Then, as illustrated in FIG. 6D, an insulating film, which
is an ON film or a metal oxide film having a high dielectric
constant, for example, and an electrode film made of silicon, for
example, are successively formed across the entire surface of the
semiconductor substrate 300. Then, the insulating film and the
electrode film are patterned, whereby the first capacitor
insulating film 309A is formed so as to cover the first capacitor
lower electrode 306A, and the first capacitor upper electrode 310A
is formed so as to cover the first capacitor insulating film 309A.
At the same time, the second capacitor insulating film 309B is
formed so as to cover the second capacitor lower electrode 306B,
and the second capacitor upper electrode 310B is formed so as to
cover the second capacitor insulating film 309B. Thus, a cell
capacitor of the first DRAM section 102 is formed in the first
memory region R.sub.A, and a cell capacitor of the second DRAM
section 103 is formed in the second memory region R.sub.B. Then,
the second interlayer insulating film 311 is formed across the
entire surface of the semiconductor substrate 300. Then, the wiring
layer 312 is formed on the second interlayer insulating film 311.
In this process, the wiring layer 312 is formed so that portions
thereof are connected to the first capacitor upper electrode 310A
and the second capacitor upper electrode 310B through contact holes
formed in the second interlayer insulating film 311. Thus, the
first DRAM section 102 and the second DRAM section 103 are formed
in the first memory region R.sub.A and the second memory region
R.sub.B, respectively.
[0113] According to the third embodiment, the surface of the second
capacitor lower electrode 306B in the second memory region R.sub.B
is formed into the roughened surface 308, while the surface of the
first capacitor lower electrode 306A in the first memory region
R.sub.A is flat. Therefore, the surface area of the second
capacitor lower electrode 306B in the second memory region R.sub.B
is larger than that of the first capacitor lower electrode 306A in
the first memory region R.sub.A. Thus, the capacitance of the cell
capacitor formed in the second memory region R.sub.B is larger than
that of the cell capacitor formed in the first memory region
R.sub.A. Specifically, the capacitance of the cell capacitor formed
in the first memory region R.sub.A can be set to 10 fF, while the
capacitance of the cell capacitor formed in the second memory
region R.sub.B can be set to 20 fF. In this way, where the
operating voltage of the cell capacitors of the first memory region
R.sub.A is equal to that of the second memory region R.sub.B, the
charge stored in the cell capacitor formed in the first memory
region R.sub.A is about one half of that stored in the cell
capacitor formed in the second memory region R.sub.B. Therefore, in
the first DRAM section 102 where the cell capacitor formed in the
first memory region R.sub.A is used as the capacitive element, it
is not necessary to move a large amount of charge for signal
processing, and thus it is possible to reduce the power consumption
and to quickly move the charge, thereby realizing high-speed
processing. On the other hand, in the second DRAM section 103 where
the cell capacitor formed in the second memory region R.sub.B is
used as the capacitive element, a large amount of charge can be
stored, thereby realizing a sufficient signal holding
characteristic.
[0114] Thus, it is possible to realize a DRAM-embedded device that
realizes both a reduced power consumption and an increased
processing speed while ensuring a sufficient signal holding
characteristic, by optimizing the capacitance of the cell capacitor
for each DRAM section according to the application thereof, as in
the present embodiment.
[0115] Moreover, effects as described above can reliably be
obtained while employing the same structure and the same material
for the cell capacitors of the first DRAM section 102 and for those
of the second DRAM section 103.
[0116] Note that in the third embodiment, similar effects can be
obtained when the capacitance of the cell capacitor formed in the
first memory region R.sub.A is set to about 5 to 20 fF and the
capacitance of the cell capacitor formed in the second memory
region R.sub.B is set to 8 to 50 fF, while the capacitance of the
cell capacitor formed in the first memory region R.sub.A is set to
be smaller than that of the cell capacitor formed in the second
memory region R.sub.B.
[0117] Moreover, while silicon is used for each of the first
contact plug 305A, the second contact plug 305B, the first
capacitor upper electrode 310A and the second capacitor upper
electrode 310B in the third embodiment, any other suitable metal
may alternatively be used.
[0118] Moreover, an ON film or a metal oxide film having a high
dielectric constant is used as the first capacitor insulating film
309A and the second capacitor insulating film 309B in the third
embodiment, any other suitable capacitor insulating film may
alternatively be used.
[0119] Moreover, while the inner surface and the outer surface of
the first capacitor lower electrode 306A and the second capacitor
lower electrode 306B, which are formed in a tubular shape in the
first memory region R.sub.A and the second memory region R.sub.B,
respectively, are both used as a cell capacitor in the third
embodiment, only the inner surface thereof may alternatively be
used as a cell capacitor.
[0120] Moreover, while the inner surface and the outer surface of
the second capacitor lower electrode 306B, which is formed in a
tubular shape in the second memory region R.sub.B, are both
roughened in the third embodiment, only the inner surface thereof
may alternatively be roughened.
[0121] Moreover, while the cell capacitor of the first memory
region R.sub.A and that of the second memory region R.sub.B have a
tubular shape in the third embodiment, they may alternatively have
a cylindrical shape or any other suitable shape.
[0122] Moreover, while the cell capacitor of the first memory
region R.sub.A and that of the second memory region R.sub.B are
formed on the first interlayer insulating film 304 in the third
embodiment, they may alternatively be formed in the first
interlayer insulating film 304. In such a case, the inner surface
of the tubular portion (the portion other than the bottom surface)
of the first capacitor lower electrode 306A and the second
capacitor lower electrode 306B may be used as a cell capacitor.
[0123] Moreover, while a silicon oxide film covering the surface of
the second capacitor lower electrode 306B in the second memory
region R.sub.B is removed by using hydrofluoric acid in the method
for manufacturing the semiconductor device according to the third
embodiment, the silicon oxide film may alternatively be removed by
any other suitable method.
[0124] Moreover, in the method for manufacturing the semiconductor
device according to the third embodiment, the second capacitor
lower electrode 306B in the second memory region R.sub.B is formed
in such an atmosphere that the surface of the second capacitor
lower electrode 306B is covered with a native oxide film.
Alternatively, the second capacitor lower electrode 306B may be
formed in such an atmosphere that the surface of the second
capacitor lower electrode 306B is not oxidized. In such a case, it
is possible to eliminate the step of removing a silicon oxide film
formed on the surface of the second capacitor lower electrode
306B.
[0125] Moreover, while a silicon-containing gas containing silane
or disilane is used as the silicon-containing gas for roughening
the surface of the second capacitor lower electrode 306B in the
second memory region R.sub.B in the method for manufacturing the
semiconductor device according to the third embodiment, any other
suitable silicon-containing gas may alternatively be used.
[0126] Moreover, the surface of the second capacitor lower
electrode 306B in the second memory region R.sub.B is roughened by
performing a heat treatment at a temperature near 620.degree. C. in
a silicon-containing gas atmosphere in the method for manufacturing
the semiconductor device according to the third embodiment, any
other suitable method may be used for roughening the second
capacitor lower electrode 306B.
Fourth Embodiment
[0127] A semiconductor device according to the fourth embodiment
and a method for manufacturing the same will now be described with
reference to the drawings. Also in the present embodiment, the
general configuration of the semiconductor device is as shown in
the plan view of FIG. 1A.
[0128] FIG. 7 is a cross-sectional view illustrating the
semiconductor device according to the fourth embodiment. More
specifically, FIG. 7 shows memory cells of the first DRAM section
102 and the second DRAM section 103 formed on the chip 100
illustrated in FIG. 1A. Note that the two regions where the first
DRAM section 102 and the second DRAM section 103 are formed are
referred to as "first memory region R.sub.A" and "second memory
region R.sub.B", respectively.
[0129] As illustrated in FIG. 7, a first gate electrode 402A is
formed on a semiconductor substrate 400 in the first memory region
R.sub.A with a first gate insulating film 401A interposed between
the semiconductor substrate 400 and the first gate electrode 402A.
Thus, a transfer gate is formed in the first memory region R.sub.A.
The second gate electrode 402B is formed on the semiconductor
substrate 400 in the second memory region R.sub.B with the second
gate insulating film 401B interposed between the semiconductor
substrate 400 and the second gate electrode 402B. Thus, a transfer
gate is formed in the second memory region R.sub.B. A first
diffusion layer 403A is formed in the semiconductor substrate 400
beside the first gate electrode 402A. A second diffusion layer 403B
is formed in the semiconductor substrate 400 beside the second gate
electrode 402B. A first interlayer insulating film 404 is formed on
the semiconductor substrate 400, the first gate electrode 402A and
the second gate electrode 402B. A first contact plug 405A connected
to the first diffusion layer 403A is formed in the first interlayer
insulating film 404 in the first memory region R.sub.A. A second
contact plug 405B connected to the second diffusion layer 403B is
formed in the first interlayer insulating film 404 in the second
memory region R.sub.B. A first capacitor lower electrode 406A is
formed on the first interlayer insulating film 404 in the first
memory region R.sub.A. The first capacitor lower electrode 406A has
a tubular shape with a bottom surface and is connected to the first
contact plug 405A. A second capacitor lower electrode 406B is
formed on the first interlayer insulating film 404 in the second
memory region R.sub.B. The second capacitor lower electrode 406B
has a tubular shape with a bottom surface and is connected to the
second contact plug 405B. A first low-dielectric-constant film
407A, which is an ON film, for example, is formed so as to cover
the first capacitor lower electrode 406A, and a first capacitor
upper electrode 410A is formed so as to cover the first
low-dielectric-constant film 407A. Thus, a cell capacitor of the
first DRAM section 102 is formed in the first memory region
R.sub.A. A second high-dielectric-constant film 408B, which is a
tantalum oxide (Ta.sub.2O.sub.5) film having a high dielectric
constant, for example, is formed so as to cover the second
capacitor lower electrode 406B, and a second capacitor upper
electrode 410B is formed so as to cover the second
high-dielectric-constant film 408B. Thus, a cell capacitor of the
second DRAM section 103 is formed in the second memory region
R.sub.B. A second interlayer insulating film 411 is formed on the
first interlayer insulating film 404, the first capacitor upper
electrode 410A and the second capacitor upper electrode 410B. A
wiring layer 412 is formed on the second interlayer insulating film
411. The wiring layer 412 is formed so that portions thereof are
connected to the first capacitor upper electrode 410A and the
second capacitor upper electrode 410B through contact holes formed
in the second interlayer insulating film 411. Note that the terms
"low-dielectric-constant film" and "high-dielectric-constant film"
are used in the present embodiment to indicate that these films
have different dielectric constants.
[0130] A feature of the semiconductor device according to the
fourth embodiment is that the material of the first
low-dielectric-constant film 407A in the first memory region
R.sub.A is different from that of the second
high-dielectric-constant film 408B in the second memory region
R.sub.B. Consider a case where the relative dielectric constant and
the equivalent oxide thickness of the ON film used as the first
low-dielectric-constant film 407A in the first memory region
R.sub.A are set to 3.9 and 4.7 nm, respectively, and the relative
dielectric constant and the equivalent oxide thickness of the
Ta.sub.2O.sub.5 film used as the second high-dielectric-constant
film 408B in the second memory region R.sub.B are set to 20 and 3.2
nm, respectively. In such a case, if the capacitance of the cell
capacitor formed in the first memory region R.sub.A is 10 fF, the
capacitance of the cell capacitor formed in the second memory
region R.sub.B is 14.7 fF based on Expression (4) below.
10 fF.multidot.4.7 nm/3.2 nm=14.7 fF (4)
[0131] Therefore, according to the present embodiment, the
capacitance of the cell capacitor formed in the first memory region
R.sub.A can be set to be smaller than that of the cell capacitor
formed in the second memory region R.sub.B.
[0132] Next, a method for manufacturing the semiconductor device
according to the fourth embodiment will be described with reference
to the drawings.
[0133] FIG. 8A to FIG. 8D are cross-sectional views illustrating
steps in the method for manufacturing the semiconductor device
according to the fourth embodiment. Note that in FIG. 8A to FIG.
8D, the same members as those of the semiconductor device of the
fourth embodiment illustrated in FIG. 7 will be denoted by the same
reference numerals.
[0134] As illustrated in FIG. 8A, the first gate electrode 402A is
formed on the semiconductor substrate 400 in the first memory
region R.sub.A with the first gate insulating film 401A interposed
between the semiconductor substrate 400 and the first gate
electrode 402A, and the second gate electrode 402B is formed on the
semiconductor substrate 400 in the second memory region R.sub.B
with the second gate insulating film 401B interposed between the
semiconductor substrate 400 and the second gate electrode 402B.
Thus, a transfer gate is formed in the first memory region R.sub.A
and in the second memory region R.sub.B. Then, the first diffusion
layer 403A is formed in the semiconductor substrate 400 beside the
first gate electrode 402A, and the second diffusion layer 403B is
formed in the semiconductor substrate 400 beside the second gate
electrode 402B. Then, the first interlayer insulating film 404 is
formed across the entire surface of the semiconductor substrate
400. Then, the first contact plug 405A connected to the first
diffusion layer 403A is formed in the first interlayer insulating
film 404 in the first memory region R.sub.A, and the second contact
plug 405B connected to the second diffusion layer 403B is formed in
the first interlayer insulating film 404 in the second memory
region R.sub.B. Then, the first capacitor lower electrode 406A is
formed on the first interlayer insulating film 404 in the first
memory region R.sub.A, and the second capacitor lower electrode
406B is formed on the first interlayer insulating film 404 in the
second memory region R.sub.B. The first capacitor lower electrode
406A has a tubular shape with a bottom surface and is connected to
the first contact plug 405A, and the second capacitor lower
electrode 406B has a tubular shape with a bottom surface and is
connected to the second contact plug 405B. Then, a
low-dielectric-constant film, which is a silicon nitride film, for
example, and a high-dielectric-constant film, which is a
Ta.sub.2O.sub.5 film, for example, are formed across the entire
surface of the semiconductor substrate 400. Then, the
low-dielectric-constant film and the high-dielectric-constant film
are patterned, whereby the first low-dielectric-constant film 407A
is formed so as to cover the first capacitor lower electrode 406A,
and a first high-dielectric-constan- t film 408A is formed so as to
cover the first low-dielectric-constant film 407A. At the same
time, a second low-dielectric-constant film 407B is formed so as to
cover the second capacitor lower electrode 406B, and the second
high-dielectric-constant film 408B is formed so as to cover the
second low-dielectric-constant film 407B.
[0135] Then, as illustrated in FIG. 8B, a resist mask 409 is formed
on the second high-dielectric-constant film 408B in the second
memory region R.sub.B so as to completely mask the second
high-dielectric-constant film 408B.
[0136] Then, as illustrated in FIG. 8C, the first
high-dielectric-constant film 408A in the first memory region
R.sub.A is selectively removed by wet etching by using a chemical
solution containing hydrofluoric acid, for example. Then, the
semiconductor substrate 400 is subjected to a heat treatment in an
oxygen-containing atmosphere, for example, whereby the first
low-dielectric-constant film 407A in the first memory region
R.sub.A is oxidized into a layered film of a silicon nitride film
and a silicon oxide film (ON film), and the second
high-dielectric-constant film 408B in the second memory region
R.sub.B is crystallized while adding oxygen to the second
high-dielectric-constant film 408B.
[0137] Then, as illustrated in FIG. 8D, a metal film is formed
across the entire surface of the semiconductor substrate 400. Then,
the metal film is patterned by photolithography and dry etching,
whereby the first capacitor upper electrode 410A is formed so as to
cover the first low-dielectric-constant film 407A, and the second
capacitor upper electrode 410B is formed so as to cover the second
high-dielectric-constant film 408B. Thus, a cell capacitor of the
first DRAM section 102 is formed in the first memory region
R.sub.A, and a cell capacitor of the second DRAM section 103 is
formed in the second memory region R.sub.B. Then, the second
interlayer insulating film 411 is formed across the entire surface
of the semiconductor substrate 400. Then, the wiring layer 412 is
formed on the second interlayer insulating film 411. In this
process, the wiring layer 412 is formed so that portions thereof
are connected to the first capacitor upper electrode 410A and the
second capacitor upper electrode 410B through contact holes formed
in the second interlayer insulating film 411. Thus, the first DRAM
section 102 and the second DRAM section 103 are formed in the first
memory region R.sub.A and the second memory region R.sub.B,
respectively.
[0138] According to the fourth embodiment, the capacitor insulating
film of the cell capacitor formed in the first memory region
R.sub.A is an ON film. On the other hand, the capacitor insulating
film of the cell capacitor formed in the second memory region
R.sub.B is formed from a silicon nitride film and a Ta.sub.2O.sub.5
film having a high dielectric constant (note however that in the
semiconductor device of the present embodiment, the capacitor
insulating film of the cell capacitor formed in the second memory
region R.sub.B is formed only from a Ta.sub.2O.sub.5 film).
Therefore, the capacitance of the cell capacitor of the first
memory region R.sub.A is smaller than that of the cell capacitor of
the second memory region R.sub.B. Specifically, when the ON film of
the first low-dielectric-constant film 407A has a relative
dielectric constant of 5 while the Ta.sub.2O.sub.5 film of the
second high-dielectric-constant film 408B has a relative dielectric
constant of 10, the capacitance of the cell capacitor formed in the
first memory region R.sub.A can be set to 10 fF, for example, while
the capacitance of the cell capacitor formed in the second memory
region R.sub.B can be set to 20 fF, for example. Therefore, the
capacitance of the cell capacitor formed in the first memory region
R.sub.A is about one half of that of the cell capacitor formed in
the second memory region R.sub.B. Thus, where the voltage applied
across the cell capacitor formed in the first memory region R.sub.A
is equal to that applied across the cell capacitor formed in the
second memory region R.sub.B, the charge stored in the cell
capacitor formed in the first memory region R.sub.A is about one
half of that stored in the cell capacitor formed in the second
memory region R.sub.B. Therefore, in the first DRAM section 102
where the cell capacitor formed in the first memory region R.sub.A
is used as the capacitive element, it is not necessary to move a
large amount of charge for signal processing, and thus it is
possible to reduce the power consumption and to quickly move the
charge, thereby realizing high-speed processing. On the other hand,
in the second DRAM section 103 where the cell capacitor formed in
the second memory region R.sub.B is used as the capacitive element,
a large amount of charge can be stored, thereby realizing a
sufficient signal holding characteristic.
[0139] Thus, it is possible to realize a DRAM-embedded device that
realizes both a reduced power consumption and an increased
processing speed while ensuring a sufficient signal holding
characteristic, by optimizing the capacitance of the cell capacitor
for each DRAM section according to the application thereof, as in
the present embodiment.
[0140] Note that while an ON film is used for the first
low-dielectric-constant film 407A, a silicon nitride film for the
second low-dielectric-constant film 407B, and a Ta.sub.2O.sub.5
film for the second high-dielectric-constant film 408B in the
fourth embodiment, any other suitable capacitor insulating films
may alternatively be used. Specifically, a film having a high
dielectric constant, such as an aluminum oxide film or a BST
(barium strontium titanate oxide) film, for example, is preferably
used as the second high-dielectric-constant film 408B. Note however
that the capacitance of the cell capacitor formed in the first
memory region R.sub.A needs to be smaller than that of the cell
capacitor formed in the second memory region R.sub.B.
[0141] Moreover, while the inner surface and the outer surface of
the first capacitor lower electrode 406A and the second capacitor
lower electrode 406B, which are formed in a tubular shape in the
first memory region R.sub.A and the second memory region R.sub.B,
respectively, are both used as a cell capacitor in the fourth
embodiment, only the inner surface thereof may alternatively be
used as a cell capacitor.
[0142] Moreover, while the cell capacitor of the first memory
region R.sub.A and that of the second memory region R.sub.B have a
tubular shape in the fourth embodiment, they may alternatively have
a cylindrical shape or any other suitable shape.
[0143] Moreover, while a stacked capacitor having a tubular shape
with a bottom surface is used as the cell capacitor of the first
memory region R.sub.A and that of the second memory region R.sub.B
in the fourth embodiment, a simple stacked capacitor may
alternatively be used.
[0144] Moreover, while the cell capacitor of the first memory
region R.sub.A and that of the second memory region R.sub.B are
formed on the first interlayer insulating film 404 in the fourth
embodiment, they may alternatively be formed in the first
interlayer insulating film 404. In such a case, the inner surface
of the tubular portion (the portion other than the bottom surface)
of the first capacitor lower electrode 406A and the second
capacitor lower electrode 406B may be used as a cell capacitor.
[0145] Moreover, in the method for manufacturing the semiconductor
device according to the fourth embodiment, the capacitor insulating
film of the cell capacitor of the second DRAM section 103 is formed
from the second low-dielectric-constant film 407B and the second
high-dielectric-constant film 408B. Alternatively, the capacitor
insulating film of the cell capacitor of the second DRAM section
103 may be formed only from the second high-dielectric-constant
film 408B. In such a case, the first low-dielectric-constant film
407A and the second high-dielectric-constant film 408B may be
formed separately in the first memory region R.sub.A and in the
second memory region R.sub.B, respectively.
Fifth Embodiment
[0146] A semiconductor device according to the fifth embodiment and
a method for manufacturing the same will now be described with
reference to the drawings. Also in the present embodiment, the
general configuration of the semiconductor device is as shown in
the plan view of FIG. 1A.
[0147] FIG. 9 is a cross-sectional view illustrating the
semiconductor device according to the fifth embodiment. More
specifically, FIG. 9 shows memory cells of the first DRAM section
102 and the fifth DRAM section 103 formed on the chip 100
illustrated in FIG. 1A. Note that the two regions where the first
DRAM section 102 and the second DRAM section 103 are formed are
referred to as "first memory region R.sub.A" and "second memory
region R.sub.B", respectively.
[0148] As illustrated in FIG. 9, a first gate electrode 502A is
formed on a semiconductor substrate 500 made of silicon, for
example, in the first memory region RA with a first gate insulating
film 501A interposed between the semiconductor substrate 500 and
the first gate electrode 502A. Thus, a transfer gate is formed in
the first memory region R.sub.A. A second gate electrode 502B is
formed on the semiconductor substrate 500 in the second memory
region R.sub.B with a second gate insulating film 501B interposed
between the semiconductor substrate 500 and the second gate
electrode 502B. Thus, a transfer gate is formed in the second
memory region R.sub.B. A first diffusion layer 503A is formed in
the semiconductor substrate 500 beside the first gate electrode
502A. Moreover, a silicide layer 516 is formed on the first gate
electrode 502A and the first diffusion layer 503A. A second
diffusion layer 503B is formed in the semiconductor substrate 500
beside the second gate electrode 502B. A first interlayer
insulating film 504 is formed on the semiconductor substrate 500,
the first gate electrode 502A and the second gate electrode 502B. A
first contact plug 510 made of tungsten, for example, which is
connected to the first diffusion layer 503A, is formed in the first
interlayer insulating film 504 in the first memory region R.sub.A.
A second contact plug 507 made of silicon, for example, which is
connected to the second diffusion layer 503B, is formed in the
first interlayer insulating film 504 in the second memory region
R.sub.B. A first capacitor lower electrode 511A is formed on the
first interlayer insulating film 504 in the first memory region
R.sub.A. The first capacitor lower electrode 511A has a tubular
shape with a bottom surface and is connected to the first contact
plug 510. A second capacitor lower electrode 551B is formed on the
first interlayer insulating film 504 in the second memory region
R.sub.B. The second capacitor lower electrode 511B has a tubular
shape with a bottom surface and is connected to the second contact
plug 507. A first capacitor insulating film 512A is formed so as to
cover the first capacitor lower electrode 511A, and a first
capacitor upper electrode 513A is formed so as to cover the first
capacitor insulating film 512A. Thus, a cell capacitor of the first
DRAM section 102 is formed in the first memory region R.sub.A. A
second capacitor insulating film 512B is formed so as to cover the
second capacitor lower electrode 511B, and a second capacitor upper
electrode 513B is formed so as to cover the second capacitor
insulating film 512B. Thus, a cell capacitor of the second DRAM
section 103 is formed in the second memory region R.sub.B. A second
interlayer insulating film 514 is formed on the first interlayer
insulating film 504, the first capacitor upper electrode 513A and
the second capacitor upper electrode 513B. A wiring layer 515 is
formed on the second interlayer insulating film 514. The wiring
layer 515 is formed so that portions thereof are connected to the
first capacitor upper electrode 513A and the second capacitor upper
electrode 513B through contact holes formed in the second
interlayer insulating film 514.
[0149] A feature of the semiconductor device according to the fifth
embodiment is that the material of the first contact plug 510
connected to the first diffusion layer 503A in the first memory
region R.sub.A is different from that of the second contact plug
507 connected to the second diffusion layer 503B in the second
memory region R.sub.B. Specifically, the material of the first
contact plug 510 is tungsten, while the material of the second
contact plug 507 is silicon.
[0150] Next, a method for manufacturing the semiconductor device
according to the fifth embodiment will be described with reference
to the drawings.
[0151] FIG. 10A to FIG. 10E, FIG. 11A and FIG. 11B are
cross-sectional views illustrating steps in the method for
manufacturing the semiconductor device according to the fifth
embodiment. Note that in FIG. 10A to FIG. 10E, FIG. 11A and FIG.
11B, the same members as those of the semiconductor device of the
fifth embodiment illustrated in FIG. 9 will be denoted by the same
reference numerals.
[0152] As illustrated in FIG. 10A, the first gate electrode 502A is
formed on the semiconductor substrate 500 made of silicon, for
example, in the first memory region R.sub.A with the first gate
insulating film 501A interposed between the semiconductor substrate
500 and the first gate electrode 502A, and the second gate
electrode 502B is formed on the semiconductor substrate 500 in the
second memory region R.sub.B with the second gate insulating film
501B interposed between the semiconductor substrate 500 and the
second gate electrode 502B. Thus, a transfer gate is formed in the
first memory region R.sub.A and in the second memory region
R.sub.B. Then, the first diffusion layer 503A is formed in the
semiconductor substrate 500 beside the first gate electrode 502A,
and the second diffusion layer 503B is formed in the semiconductor
substrate 500 beside the second gate electrode 502B. (Note that the
silicide layer 516 can then be formed selectively on the first
diffusion layer 503A and the first gate electrode 502A by using a
salicide technique in order to manufacture the semiconductor device
according to the fifth embodiment as described above.) Then, the
first interlayer insulating film 504, which is a silicon oxide film
containing boron or phosphorus as an impurity, for example, is
formed across the entire surface of the semiconductor substrate
500. Then, a second contact hole 505 reaching the second diffusion
layer 503B is formed by photolithography and etching in the first
interlayer insulating film 504 in the second memory region
R.sub.B.
[0153] Then, as illustrated in FIG. 10B, a silicon film 506 made of
phosphorus-containing silicon, for example, is formed across the
entire surface of the semiconductor substrate 500 so as to
completely bury the second contact hole 505.
[0154] Then, as illustrated in FIG. 10C, an unnecessary portion of
the silicon film 506 on the first interlayer insulating film 504 is
removed by chemical-mechanical polishing, for example, thereby
forming the second contact plug 507 in the second contact hole
505.
[0155] Then, as illustrated in FIG. 10D, a first contact hole 508
reaching the first diffusion layer 503A is formed by
photolithography and etching in the first interlayer insulating
film 504 in the first memory region R.sub.A.
[0156] Then, as illustrated in FIG. 10E, a metal film 509 made of
tungsten, for example, is formed across the entire surface of the
semiconductor substrate 500 so as to completely bury the first
contact hole 508.
[0157] Then, as illustrated in FIG. 11A, an unnecessary portion of
the metal film 509 on the first interlayer insulating film 504 is
removed by chemical-mechanical polishing, for example, thereby
forming the first contact plug 510 in the first contact hole
508.
[0158] Then, as illustrated in FIG. 11B, the first capacitor lower
electrode 511A is formed on the first interlayer insulating film
504 in the first memory region R.sub.A. The first capacitor lower
electrode 511A has a tubular shape with a bottom surface and is
connected to the first contact plug 510. At the same time, the
second capacitor lower electrode 511B is formed on the first
interlayer insulating film 504 in the second memory region R.sub.B.
The second capacitor lower electrode 511B has a tubular shape with
a bottom surface and is connected to the second contact plug 507.
Then, the first capacitor insulating film 512A is formed so as to
cover the first capacitor lower electrode 511A, and the first
capacitor upper electrode 513A is formed so as to cover the first
capacitor insulating film 512A. At the same time, the second
capacitor insulating film 512B is formed so as to cover the second
capacitor lower electrode 511B, and the second capacitor upper
electrode 513B is formed so as to cover the second capacitor
insulating film 512B. Thus, a cell capacitor of the first DRAM
section 102 is formed in the first memory region R.sub.A, and a
cell capacitor of the second DRAM section 103 is formed in the
second memory region R.sub.B. Then, the second interlayer
insulating film 514 is formed across the entire surface of the
semiconductor substrate 500. Then, the wiring layer 515 is formed
on the second interlayer insulating film 514. In this process, the
wiring layer 515 is formed so that portions thereof are connected
to the first capacitor upper electrode 513A and the second
capacitor upper electrode 513B. Thus, the first DRAM section 102
and the second DRAM section 103 are formed in the first memory
region R.sub.A and the second memory region R.sub.B,
respectively.
[0159] According to the fifth embodiment, the first contact plug
510 in the first memory region R.sub.A is made of tungsten, whereas
the second contact plug 507 in the second memory region R.sub.B is
made of phosphorus-containing silicon. Therefore, in the first
memory region R.sub.A, the contact resistance between the first
contact plug 510 and the first diffusion layer 503A is low, since
the contact is a tungsten-silicon contact. In addition, since the
resistance of tungsten is low, whereby even the total of these
resistances combined together is as low as on the order of Q. As a
result, in the first DRAM section 102 formed in the first memory
region R.sub.A using the first contact plug 510, the resistance to
the charge movement for signal processing is low, thereby realizing
high-speed processing. In contrast, in the second memory region
R.sub.B, the contact resistance between the second contact plug 507
and the second diffusion layer 503B is relatively high since the
contact is a silicon-silicon contact. In addition, the resistance
of silicon is also relatively high, whereby the total of these
resistances combined together is as high as on the order of 10
k.OMEGA.. Therefore, the leak current, i.e., the charge stored in
the cell capacitor formed in the second memory region R.sub.B
leaking from the second diffusion layer 503B to the semiconductor
substrate 500 via the second contact plug 507, is reduced. Thus, in
the second DRAM section 103 formed in the second memory region
R.sub.B using the second contact plug 507, the charge stored in the
cell capacitor is unlikely to leak, whereby it is possible to hold
the signal stored therein sufficiently long and to reduce the
amount of power that is used for recharging the cell capacitor.
[0160] Thus, it is possible to realize a DRAM-embedded device that
realizes both a reduced power consumption and an increased
processing speed while ensuring a sufficient signal holding
characteristic, by optimizing the resistance, which occurs due to
the connection between a cell capacitor and the semiconductor
substrate via a plug, for each DRAM section according to the
application thereof, as in the present embodiment.
[0161] Furthermore, in the semiconductor device according to the
present embodiment, the silicide layer 516 is formed in the first
diffusion layer 503A connected to the first contact plug 510,
whereby it is possible to reduce the contact resistance between the
first diffusion layer 503A and the first contact plug 510. Thus,
effects as described above can reliably be obtained.
[0162] Note that while tungsten is used as the material of the
first contact plug 510 in the fifth embodiment, any other suitable
low-resistance metal may alternatively be used.
[0163] Moreover, while a silicon substrate is used as the
semiconductor substrate 500 in the fifth embodiment, any other
suitable semiconductor substrate may alternatively be used. In such
a case, the material of the contact plug needs to be selected in
view of the contact resistance between the diffusion layer formed
in the semiconductor substrate and the contact plug and the
resistance of the contact plug itself.
[0164] Moreover, while the inner surface and the outer surface of
the first capacitor lower electrode 511A and the second capacitor
lower electrode 511B, which are formed in a tubular shape in the
first memory region R.sub.A and the second memory region R.sub.B,
respectively, are both used as a cell capacitor in the fifth
embodiment, only the inner surface thereof may alternatively be
used as a cell capacitor.
[0165] Moreover, while the cell capacitor of the first memory
region R.sub.A and that of the second memory region R.sub.B have a
tubular shape in the fifth embodiment, they may alternatively have
a cylindrical shape or any other suitable shape.
[0166] Moreover, in the fifth embodiment, a barrier metal made of
TiN or Ti, for example, may be formed on the surface of the first
contact hole 508, between the step of forming the metal film 509
and the step of forming the first contact hole 508.
* * * * *