U.S. patent application number 13/596623 was filed with the patent office on 2012-12-20 for method for operating a semiconductor device.
This patent application is currently assigned to INFINEON TECHNOLOGIES AG. Invention is credited to Markus ZUNDEL.
Application Number | 20120319761 13/596623 |
Document ID | / |
Family ID | 40719573 |
Filed Date | 2012-12-20 |
United States Patent
Application |
20120319761 |
Kind Code |
A1 |
ZUNDEL; Markus |
December 20, 2012 |
METHOD FOR OPERATING A SEMICONDUCTOR DEVICE
Abstract
A semiconductor device includes a semiconductor substrate having
at least a pn-junction arranged in the semiconductor substrate. At
least a field electrode is arranged at least next to a portion of
the pn-junction, wherein the field electrode is insulated from the
semiconductor substrate. A switching device is electrically
connected to the field electrode and adapted to apply selectively
and dynamically one of a first electrical potential and a second
electrical potential, which is different to the first electrical
potential, to the field electrode to alter the avalanche breakdown
characteristics of the pn-junction.
Inventors: |
ZUNDEL; Markus; (Egmating,
DE) |
Assignee: |
INFINEON TECHNOLOGIES AG
Neubiberg
DE
|
Family ID: |
40719573 |
Appl. No.: |
13/596623 |
Filed: |
August 28, 2012 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11964292 |
Dec 26, 2007 |
8274109 |
|
|
13596623 |
|
|
|
|
Current U.S.
Class: |
327/530 |
Current CPC
Class: |
H01L 2924/0002 20130101;
H01L 29/8611 20130101; H01L 29/8613 20130101; H01L 29/7813
20130101; H01L 29/4236 20130101; H01L 29/41766 20130101; H01L
29/407 20130101; H01L 29/7397 20130101; H01L 29/66727 20130101;
H01L 2924/00 20130101; H01L 29/42368 20130101; H01L 29/66734
20130101; H01L 29/402 20130101; H01L 2924/0002 20130101 |
Class at
Publication: |
327/530 |
International
Class: |
G05F 3/02 20060101
G05F003/02 |
Claims
1.-17. (canceled)
18. A method for operating a semiconductor device, comprising:
providing a semiconductor device which comprises a semiconductor
substrate comprising at least a pn-junction arranged in the
semiconductor substrate, at least a field electrode arranged at
least next to a portion of the pn-junction, the field electrode
being insulated from the semiconductor substrate; operating the
pn-junction at least temporarily in reverse mode; applying a first
electrical potential to the field electrode; and temporarily
disconnecting the field electrode from the first electrical
potential and applying a second electrical potential to the field
electrode so that the avalanche breakdown characteristics of the
pn-junction are altered.
19. The method of claim 18, wherein the second electrical potential
is repeatedly temporarily applied to the field electrode.
20. The method of claim 18, wherein the second electrical potential
is applied to the field electrode shortly before an avalanche
breakdown occurs.
21.-25. (canceled)
Description
FIELD OF THE INVENTION
[0001] This description refers to embodiments of semiconductor
devices and particularly power semiconductor devices with improved
avalanche breakdown characteristics. Further embodiments refer to a
method for operating a semiconductor device and a method for
manufacturing a semiconductor device.
BACKGROUND OF THE INVENTION
[0002] Improvement of reliability and robustness of semiconductor
devices is an ongoing aim in the development of semiconductor
device and particularly power semiconductor devices. To ensure that
devices which were delivered to customer meet the demands, the
semiconductor devices are subjected to extended reliability tests
before delivery. For example, the properties of the semiconductor
device during an avalanche breakdown influence the reliability and
are therefore subject for further developments. Transient avalanche
breakdowns frequently occur in demanding applications and can
render a device useless.
[0003] Two main device concepts having different avalanche
properties are commonly used--the field electrode concept and the
dense trench concept. In semiconductor devices of the field
electrode concept, the avalanche breakdown occurs close to the
contact plug for contacting the body region. Different thereto, the
location of the breakdown is arranged close to the bottoms of the
trenches in semiconductor devices formed according to the dense
trench concept. While semiconductor devices of the trench dense
concept exhibit improved properties regarding on-state resistance
in comparison with standard devices of the field electrode concept,
dense trench devices are more susceptible to avalanche
breakdowns.
BRIEF SUMMARY OF THE INVENTION
[0004] According to an embodiment, a semiconductor device is
provided. The semiconductor device includes a semiconductor
substrate having at least a pn-junction arranged in the
semiconductor substrate. At least a field electrode is arranged at
least next to a portion of the pn-junction, wherein the field
electrode is insulated from the semiconductor substrate. A
switching device is electrically connected to the field electrode
and adapted to apply selectively and dynamically one of a first
electrical potential and a second electrical potential, which is
different to the first electrical potential, to the field electrode
to alter the avalanche breakdown characteristics of the
pn-junction.
[0005] During operation, the semiconductor device can be brought
into different modes of operation by selectively switching between
two different electrical potentials applied to the field electrode.
This influences the location of a possible avalanche breakdown so
that the breakdown behaviour of semiconductor device can be
switched.
BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS
[0006] A full and enabling disclosure of the present invention,
including the best mode thereof, to one of ordinary skill in the
art, is set forth more particularly in the remainder of the
specification, including reference to the accompanying figures.
Therein:
[0007] FIGS. 1A and 1B show embodiments of semiconductor
devices.
[0008] FIG. 2 shows another embodiment of a semiconductor
device.
[0009] FIG. 3 shows yet another embodiment of a semiconductor
device.
[0010] FIGS. 4A and 4B show two different modes of operation of a
semiconductor device.
[0011] FIG. 5 shows a further embodiment of a semiconductor
device.
[0012] FIG. 6 shows yet a further embodiment of a semiconductor
device.
[0013] FIG. 7 shows another embodiment of a semiconductor
device.
[0014] FIG. 8 shows a further embodiment of a semiconductor
device.
[0015] FIG. 9 shows a top view on a semiconductor device.
[0016] FIG. 10 shows a reference semiconductor device.
[0017] FIG. 11 shows another reference semiconductor device.
[0018] FIGS. 12A to 12D show steps of a method for manufacturing a
semiconductor device according to an embodiment.
DETAILED DESCRIPTION OF THE INVENTION
[0019] Reference will now be made in detail to various embodiments,
one or more examples of which are illustrated in the figures. Each
example is provided by way of explanation, and is not meant as a
limitation of the invention. For example, features illustrated or
described as part of one embodiment can be used on or in
conjunction with other embodiments to yield yet a further
embodiment. It is intended that the present invention includes such
modifications and variations. The examples are described using
specific language which should not be construed as limiting the
scope of the appending claims. The drawings are not scaled and are
for illustrative purposes only.
[0020] The term "lateral" as used in this specification intends to
describe an orientation parallel to the main surface of a
semiconductor wafer or die.
[0021] The term "vertical" as used in this specification intends to
describe an orientation which is arranged perpendicular to the main
surface of the semiconductor wafer or die.
[0022] The term "above" as used in this specification describes a
location of a structural feature which is arranged closer to the
first surface in relation to another structural feature.
[0023] Consequently, the term "below" as used in this specification
describes a location of a structural feature which is arranged
closer to the second surface in relation to another structural
feature.
[0024] Specific embodiments described in this specification pertain
to, without being limited thereto, power semiconductor devices such
as diodes and particularly to devices which are controlled by
field-effect such as power field-effect transistors (FETs) and
insulated gate bipolar transistors (IGBTs).
[0025] FIG. 1A shows a semiconductor device 80 according to a first
embodiment. The semiconductor device 80 includes an appropriate
semiconductor substrate 81 such as silicon or silicon carbide
having a pn-junction 89 formed therein. The pn-junction 89 is
formed between two regions of opposite conductivity type. In the
embodiment shown in FIG. 1A the pn-junction 89 is formed between a
doping region 82 arranged in the semiconductor substrate 81 and an
adjoining region of the semiconductor substrate 81 of opposite
type. Doping region 82 is arranged on a surface of the
semiconductor substrate 81. For example, semiconductor substrate 81
can be n-doped while doping region 82 can be p-doped. Doping region
82 is contacted by a contact layer 83 arranged on one side of the
semiconductor substrate 82 and having a terminal 84. On the
opposite side of the semiconductor substrate 81 a further contact
layer 85 is formed for contacting the semiconductor substrate 81.
Further contact layer 85 also has a terminal 86. Embodiment of FIG.
1A may represent a diode such as a power rectifier.
[0026] Pn-junction 89 has in this embodiment at least two regions.
Region 88, circled by a dotted line, forms a flat or straight
junction while region 87, circled by a dotted line, is curved.
Curved pn-junction is formed on a peripheral portion, more
specifically at a laterally peripheral portion, of doping region
82. The region 87, where the pn-junction 89 is curved, is
susceptible to an avalanche breakdown when operated in reverse
mode, since the electrical field under reverse condition is
inhomogeneous there. Therefore, region 87 is likely to experience
an avalanche breakdown in reverse mode. Due to the curvature of
region 87, the avalanche breakdown may occur at a voltage which is
smaller than the volume breakdown voltage of the semiconductor
substrate 81.
[0027] To avoid an avalanche breakdown at low reverse voltages a
field electrode 90 is arranged next to the curved region 87 of
pn-junction 89. Field electrode 90 is insulated from the
semiconductor substrate 81 by an insulating layer 97 to avoid
short-circuiting of pn-junction 89.
[0028] The field electrode 90 is electrically connected to an
output terminal 91 of a switching device 93, which has two input
terminals 94 and 95 and a control terminal 92. The switching device
93 allows a selective connection of the field electrode 90 with one
of two different electrical potentials fed to the respective input
terminals 94, 95. For example, one electrical potential can be the
electrical potential of the doping region 82, and the other
electrical potential can be the electrical potential of the
semiconductor substrate 81 under reverse conditions. In certain
embodiments, input terminal 94 can be connected to terminal 84 of
contact layer 83 for contacting the doping region 82, while input
terminal 95 can be connected to terminal 86 of contact layer
85.
[0029] For example, when applying the electrical potential fed to
doping region 82 to the field electrode 90 under reverse condition,
the electrical field formed between the doping region 82 and the
semiconductor substrate 81 is at least partially pushed out of the
breakdown-susceptible region 87 and the likelihood is reduced that
a breakdown will occur in region 87. When appropriately selecting
the layout and arrangement of the field electrode 90 and the
electrical potential applied thereto, the electrical breakdown
characteristics of the semiconductor device 80 can be altered such
that the likelihood, that a breakdown will occur in curved region
87, is lower than the likelihood for a breakdown in the straight or
flat region 88. Switchable field plate 90 therefore allows a change
of the electrical breakdown characteristics of the semiconductor
device 80 when operated under reverse condition. In other words,
the semiconductor device 80 can be switched between two different
breakdown regimes. The respective electrical potentials can be
applied in an alternating manner with a given duty cycle.
[0030] FIG. 1B shows another embodiment of semiconductor device 80
having a trench 96 arranged in the semiconductor substrate 81 next
to pn-junction 89. Trench 96 accommodates the field electrode 90.
As in the embodiment shown in FIG. 1A, field plate 90 is switchable
between two electrical potentials by switching device 93.
[0031] Field electrode 90 can be used to switch between different
avalanche breakdown regions which are circled in FIG. 1B by dotted
lines 98 and 99, respectively. A breakdown would occur in region 98
if source potential is applied to field electrode 90. Different
thereto, a breakdown would occur in region 99 if drain potential is
applied to field electrode 90 since this increases the electrical
field density close to region 99.
[0032] Switching device 93 can be integrated in the semiconductor
substrate 81 or formed separately thereto. In the latter case, a
control contact pad, which is in electrical connection with field
electrode 90, is formed on the semiconductor substrate 81 to allow
an external connection of field electrode 90 with the output
terminal 91 of the switching device 93. Terminals 84 and 86 are
also typically formed by appropriate contact pads.
[0033] When the switching device 93 is integrated in the
semiconductor substrate 81, a separate contact pad can be provided
to allow external connection to control terminal 92 of the
switching device 93. In this case, input terminals 94, 95 of the
switching device 93 are typically connected to terminals 84 and 86,
respectively. This provides for an external control of the
switching device 93 and the field electrode 90 without the need of
providing separately the electrical potentials. It would, however,
also be possible to provide contact pads in electrical connections
with the input terminals 94, 95 of the switching device 93 to apply
externally respective electrical potentials.
[0034] When considering the planar device as shown in FIG. 1A, the
field electrode 90 is typically formed after formation of the
pn-junction 89, while in the trench device as shown in FIG. 1B,
field electrode 90 is typically formed before forming the
pn-junction 89. Typically, both devices can include a back
metallization formed by contact layer 85 and a front metallization
formed by contact layer 84.
[0035] With reference to FIG. 2, a further embodiment of a
semiconductor device is described. The semiconductor device 1
includes a semiconductor substrate 2 which can be made of silicon,
silicon carbide, I11-V semiconductor material, II-VI semiconductor
material, heterojunction semiconductor material or any other
suitable semiconductor material. The semiconductor substrate 2 can
include a single crystal material and at least one epitaxial layer
formed thereon. In a typical application, the free surface of the
epitaxial layer forms a first surface 51 and the free surface of
the single crystal material forms a second surface 52 of the
semiconductor substrate 2. The first and second surfaces 51 and 52
are arranged opposite to each other and run substantially lateral.
The term single crystal material as used herein refers to a wafer
formed by cutting a slice for instance from a single-crystal ingot
or boule. On the single crystal material, at least an epitaxial
layer can be formed by any suitable deposition process.
Alternatively, a wafer without any additional epitaxial layer can
be used as semiconductor substrate 2. Furthermore, a semiconductor
substrate 2 formed by bonding two wafers can also be used.
Particularly for, but not limited to, power semiconductor devices
the semiconductor substrate 2 includes a single-crystal material
and an epitaxial layer formed thereon which allows fine adjustment
of the doping concentration of the epitaxial layer during
deposition according to specific needs. In most applications
silicon or silicon carbide are used as material for the
semiconductor substrate 2.
[0036] A first doping region 31 is arranged in the semiconductor
substrate 2 at the first surface 51. The first doping region 31
typically serves as source region and is of a first conductivity
type. In many applications, the first or source region 31 is highly
n-doped, typically in the range from about 1*10.sup.19/cm.sup.3 to
about 1*10.sup.21/cm.sup.3. In the following description, the first
doping region 31 is referred to as source region 31 without being
limited thereto.
[0037] At the second surface 52, a second doping region 32 is
arranged in the semiconductor substrate 2. In case of a
FET-transistor, the second doping region 32 is a drain region
having the same conductivity type as the source region 31. Contrary
thereto, in case of an IGBT the second doping region 32 forms an
emitter region which is of opposite conductivity to the source
region 31. The power semiconductor devices as described herein such
as power FETs and IGBTs are designed to have a characteristic
vertical load current which flows from the source region 31 to the
drain or to the emitter region 32, respectively. In a typical
application, the drain or emitter region 32 is highly doped in the
range from about 1*10.sup.19/cm.sup.3 to about
1*10.sup.21/cm.sup.3. In the following description, the second
doping region 32 is referred to as drain region 32 without being
limited thereto. In case of an IGBT, the second doping region 32 is
referred to as emitter region.
[0038] A third doping region 33 is arranged in the semiconductor
substrate 2 in contact with the source region 31. The third doping
region 33 typically forms a body region and has a conductivity type
opposite to the source region 31 so that a first pn-junction 35 is
formed between the source region 31 and the third doping region 33.
The doping concentration of the third doping region 33 is typically
p-doped in the range from about 5*10.sup.16/cm.sup.3 to about
5*10.sup.18/cm.sup.3. In the following description, the third
doping region 33 is referred to as body region 33 without being
limited thereto.
[0039] A fourth doping region 34 is arranged between the body
region 33 and the drain region 32 and typically forms a drift
region having the same conductivity type as the source region 31.
The doping concentration of the fourth doping region 34
substantially corresponds to the background doping concentration of
the semiconductor substrate 2 or the epitaxial layer if one is
used. However, the doping concentration of the fourth doping region
34 can also exhibit a doping profile having a maximum or a minimum
at a desired location or an increasing or decreasing doping
concentration in vertical direction. Tailoring the doping profile
of the fourth doping region 34 helps to improve the breakdown and
on-state characteristics of the semiconductor device 1. Typically,
the doping concentration of the fourth doping region 34 is in the
range from about 1*10.sup.15/cm.sup.3 to about
5*10.sup.17/cm.sup.3. The fourth doping region 34 is in direct
contact with the body region 33 and forms with it a second
pn-junction 36. In the following description, the fourth doping
region 34 is referred to as drift region 34 without being limited
thereto.
[0040] At least a first trench 10 is arranged in the semiconductor
substrate 2 and extends from the first surface 51 into the
semiconductor substrate 2 towards the second surface 52 with the
bottom of the trench 10 being spaced to the second surface 52 and
particularly to the drain region 32 in case of an FET or to the
emitter region 32 in case of an IGBT. Laterally spaced to the first
trench 10 is arranged a second trench 20 which also extends from
the first surface 51 into the semiconductor substrate 2. In this
embodiment, a further first trench 10 is formed in the
semiconductor substrate 2 in spaced relation to the second trench
20 so that the second trench 20 is arranged between the two first
trenches 10. The portions of the semiconductor substrate 2 left
between adjacent trenches are referred to as mesa structures 54.
The first and second trenches 10, 20 run vertically and
substantially perpendicularly to the first surface 51.
[0041] In some embodiments, the semiconductor device 1 includes a
first and a second trench. In other embodiments the semiconductor
device includes two first trenches 10 and a second trench 20
arranged between the two first trenches 10. In further embodiments,
at least three first trenches 10 and at least one second trench 20
are provided with the second trench being arranged between two of
the three first trenches 10. In many applications, selective
trenches are formed as second trenches 20, for example every
second, third, fourth or every fifth trench. Other ratios would
also be possible.
[0042] In this embodiment, first trenches 10 are shown to have a
greater vertical extension than the second trench 20. As becomes
more apparent in connection with other embodiments, all trenches
10, 20 can also have the same vertical extension.
[0043] The first trench 10 includes a gate electrode 11 and a field
electrode 12 which both are arranged in the first trench 10 with
the gate electrode 11 being arranged above the field electrode 12
in proximity to the first surface 31. The gate electrode 11 extends
vertically, i.e. parallel to the vertical extension of the first
trench 10, from the source region 31 to the drift region 34. Since
the body region 33 is arranged between the source region 31 and the
drift region 34, the gate electrode 11 of the first trench 10
extends completely through the body region 33.
[0044] It is worth mentioning that gate electrode and field
electrode are different from each other and serve different
purposes. A gate electrode is arranged close to two different
pn-junctions separated by a channel forming region often referred
to as body region. The gate electrode is for generating a
conductive channel in the channel forming region to provide an
electrical path across the two pn-junctions. Different thereto, a
field electrode is arranged close to only one pn-junction or even
remote to a pn-junction so that it is not adapted to form a
conductive channel. The main purpose of a field electrode is to
influence the distribution of the electrical field, typically in a
drift region, to change and define the breakdown characteristic of
the device.
[0045] A gate insulating layer 13, sometimes referred to as gate
oxide layer (GOX), is arranged between the gate electrode 11 and
the semiconductor substrate 2 and particularly between the gate
electrode 11 and the body region 33.
[0046] A field insulating layer 14, typically a field oxide (FOX),
is arranged between the field electrode 12 and the semiconductor
substrate 2, particularly the drift region 34, and insulates the
field electrode 12 form the drift region 34. The field oxide layer
14 has a significant greater thickness in comparison with the gate
insulating layer 13 to withstand high electrical field strengths
occurring during operation of the semiconductor device 1 and to
avoid electrical breakdown between the field electrode 12 and the
drift region 34 as it becomes more apparent from the description
below.
[0047] A gate-field insulating layer 15, sometimes referred to as
POLOX (Polyoxide), is arranged between the gate electrode 11 and
the field electrode 12.
[0048] The second trench 20 also includes a gate electrode 21 and a
field electrode 22 with the gate electrode 21 being arranged above
the field electrode 22. Similar to the first trench 10, the gate
electrode 21 is laterally insulated from the semiconductor
substrate 2 by a gate insulating layer 23 while the field electrode
22 is insulated from the semiconductor substrate 2 by a field
insulating layer 24. Furthermore, a gate-field insulating layer 25
insulates the gate electrode 21 from the field electrode 22.
[0049] The trench 10, 20 define respective separate cells of the
semiconductor device 1 which are electrically connected in parallel
to each other to increase the available cross-section for the load
current and to reduce the on-state resistance.
[0050] A contact region 40 is formed in the semiconductor substrate
2 at the first surface 51 between adjacent trenches 10, 20.
Typically, the contact region 40 is a groove filled with a highly
conductive material. Contact region 40 can be manufactured by
etching a groove, forming an optional barrier layer (not shown)
such as a silicide layer on the groove walls, and filling the
groove with the conductive material, which can be highly doped
polysilicon or a metal. Before filling the groove or forming the
optional barrier layer, a highly doped region 37 having the same
conductivity type as the body region 33 can be formed by shallow
implantation. Highly doped region 37 can have a vertical extension
of about a 10 to 120 nm. In the Figures, the vertical extension of
region 37 is exaggerated. The groove accommodating the contact
region can have a vertical extension of about 200 to 1500 nm
depending on the actual design of the semiconductor device. The
contact region 40 is sometimes referred to as poly plug since it is
typically made of highly doped polysilicon. The highly doped region
37 typically defines the lower end 41 of the contact region 40.
[0051] The semiconductor device 1 further includes a switching
device 60 having an output terminal 63 and a first and a second
input terminal 61 and 62, respectively. Furthermore, the switching
device 60 includes a control terminal 64. The first input terminal
61 is connected to the source region 31 while the second input
terminal 62 is connected to the drain region 32. The output
terminal 63 is connected to the field electrode 22 of the second
trench 20. The operation of the switching device 60 can be
controlled by applying a suitable control signal to the control
terminal 64.
[0052] The purpose of the switching device 60 is to change
dynamically the electrical potential of the second trench's field
electrode 22. Typically, the electrical potential applied to the
field electrode 22 is switched between a first electrical potential
and a second electrical potential to change the electrical
breakdown characteristics of the semiconductor device 1. The first
electrical potential typically substantially corresponds to the
electrical potential which is applied to the source region 31. The
second electrical potential typically substantially corresponds to
the electrical potential which is applied to the drain or emitter
region 32 during operation. Although FIG. 2 shows that there is a
direct electrical connection between the switching device 60 and
the source region 31 and the drain region 32, respectively, a
skilled person will appreciate that each electrical connection has
a given resistivity and that at least for this reason the
electrical potential applied to the source region 31 and to the
field electrode 22 can differ from each other. Moreover, the
switching device 60, which can be for instance designed as an
integrated circuit, also exhibits an internal resistance.
Furthermore, the electrical connections and the field electrode 22
have a given capacitance which results in a given time constant for
charging the electrical connections and the field electrodes.
Depending on the time constant and the duration of the temporal
application of the second electrical potential, the field electrode
of the second trench may not fully reach the first or second
electrical potential. In certain embodiments the switching device
can therefore be electrically connected to the field electrode of
the second trench and adapted to selectively and dynamically switch
the field electrode of the second trench between two different
electrical potentials.
[0053] By changing the electrical potential of the field electrode
22 of the second trench 20, the semiconductor device 1 can be
switched between two modes of operation with respect to the
avalanche breakdown characteristics. During normal mode of
operation, the field electrode 22 is roughly at source potential,
and the second trench 20 has the same electrical properties as the
adjacent first trenches 10, the field electrodes of which 12 are
typically permanently clamped at source potential. Therefore, field
electrode of the first trench is typically permanently electrically
connected to the first doping region in certain embodiments.
[0054] To illustrate the electrical properties of the semiconductor
device 1 and the influence of the electrical potential applied to
the field electrodes 12, 22 the course of iso-electrical potential
lines can be considered. FIGS. 4A and 4B qualitatively exemplify
this for different modes of operation when operated under reverse
conditions.
[0055] FIG. 4A shows the semiconductor device 1 in reverse mode, at
which all field electrodes 12, 22 are substantially at source
potential. The drain region 32 is at drain potential, and the main
drop of the electrical field between the source and the drain
regions 31, 32 occurs in the drift region 34. For the sake of
simplicity, the operation of the semiconductor device is explained
with reference to a Power FET, but would be in principle the same
for an IGBT and other field controlled devices. The applied
electrical potentials are indicated by letters S and D with S
referring to source potential and D to drain potential.
[0056] The common electrical potential of all field electrodes 12,
22 causes the electrical field to be "pushed" out of the mesa
structures 54 towards the drain region 32 so that the main drop of
the electrical field occurs remote to the body region 33 and the
contact regions 40. This situation is schematically illustrated in
FIG. 4A, and an avalanche breakdown would likely to occur close to
the bottoms of the trenches 10, 20 as explained in more detail
below in connection with a reference structure as shown FIG. 11. An
exemplified iso-electrical potential line is denoted by 55.
[0057] To avoid an avalanche breakdown close to the bottoms of the
trenches 10, 20 the electrical `behaviour` of the second trench 20
is changed by temporarily applying the drain potential D to the
field electrode 22. By doing so, second trench 20 `disappears` in
relation to its influence to push out the electrical field. In
fact, since the field electrode 20 of the second trench 20 is now
at drain potential D, the drain potential and therefore the
electrical field can deeply penetrate the drift region 34 and the
mesa structures 54 adjacent to the second trench 20 and comes close
to the contact region 40 so that an avalanche would now likely to
occur close to the contact region 40 and remote to the field oxides
14, 24 of the first and second trenches 10 and 20, respectively. On
the other hand, the electrical potential of the field electrodes 12
of the first trenches 10 remain at source potential S since the
field electrode 12 is permanently electrically connected to the
source region 31. FIG. 4B schematically shows the situation when
the drain potential D is applied to the field electrode 20 while
keeping the field electrodes 12 at source potential S.
[0058] In a typical application, the switching device 60
selectively connects the field electrode 22 of the second trench 20
with one of the source and the drain regions 31 and 32,
respectively. More particularly, the field electrode 22 is
temporarily connected to the drain region 32 in the event of an
avalanche breakdown. For this reason, the second trench 20 is
referred to as switchable trench in this description.
[0059] In many power applications, so-called repetitive avalanche
breakdowns occur which are caused by the external load to which the
power semiconductor device 1 is connected. These avalanche
breakdowns or avalanche pulses regularly happen with a given
frequency. The switching device 60 can be arranged to apply the
drain voltage at the repetitive avalanche breakdown frequency.
Alternatively, a suitable periodic control signal can be applied to
the control terminal 64 of the switching device 60 to effect the
switching.
[0060] Another option would be to monitor the drain voltage and to
cause the switching device 60 to apply the drain potential to the
field electrode 22 of the second trench 20 when the drain voltage
exceeds a given threshold since the drain voltage rises shortly
before occurrence of a breakdown. To this end, a suitably fast
monitoring device (not shown) should be used which is capable of
accurately detecting approaching and recurring breakdowns.
[0061] In order to switch at the repetitive avalanche frequency,
the switching device 60 is adapted to apply the drain potential for
a period between about some hundred nanoseconds to about some tens
milliseconds. Furthermore, when monitoring the drain voltage, a
threshold voltage can be defined which triggers the switching
device 60 to apply the drain potential to the second trench's field
electrode 22.
[0062] Using a switching device 60 also provides for a modification
of the dependency of the gate-drain capacitance (Cgd) from the
drain-source voltage (Vds). When the pulse rate of the avalanche
breakdown pulses are known for a specific application, the Cgd(Vds)
dependency can be influenced by controlled switching of the
electrical potential. For example, if a slow drop of the gate-drain
capacitance Cgd is desired, the drain potential should be applied
at a very early stage of a repetitive avalanche pulse so that
depletion of the mesa structures 54 between the trenches 10, 20 can
be delayed.
[0063] The switching device 60 can be arranged and formed
separately to the semiconductor device 1. In this case, the
semiconductor device 1 includes a control contact pad 73 which is
electrically connected to the field electrode 22 so that the
desired electrical potential can be externally applied to the field
electrode. FIG. 9 shows a top view of a semiconductor device 1
having a source contact 71 forming a contact pad and a control
contact pad 73 being separate thereto on an upper side of the
semiconductor device 1. The drain or emitter contact would be on
the not shown lower side of the semiconductor device 1. The output
terminal 63 of the switching device 60 is electrically connected to
the control contact pad 73 for instance by bond wire 75.
[0064] FIG. 9 shows a switching device 60 which is formed as a
separate integrated circuit, i.e. the switching device 60 is not
integrated in the semiconductor substrate 2. Both the semiconductor
substrate 2 and the switching device 60 can be arranged on a common
substrate or board 70. There are many options for arranging the
switching device 60, for instance as chip-on-chip or chip-by-chip
arrangement. In the alternative case of an integrated solution, the
switching device 60 can be formed as a circuit or device which is
integrated in the semiconductor substrate 2, for instance at its
periphery.
[0065] The improved properties of the semiconductor device 1 as
described herein become even more apparent when referring to
reference devices shown in FIGS. 10 and 11. These reference devices
do not include a switchable trench. FIG. 10 shows a reference
device 101 with largely spaced trenches 110 each having a gate and
a field electrode 111 and 112, respectively. Similar to the
semiconductor device 1, the reference device 101 includes a source
region 131, a drain region 132, a body region 133 and a drift
region 134. Body region 133 is contacted by contact region 140
arranged in the mesa structure 154 and having a highly doped region
137. Due to the large spacing between adjacent trenches 110, which
significantly exceed the lateral extension of each trench 110, the
electrical field can deeply penetrate the drift region 134 and
extends close to the contact region 140 where an avalanche
breakdown would occur. The location of a breakdown is indicated by
a flash.
[0066] A semiconductor reference device 201 with closely spaced
trenches 211 is shown in FIG. 11. The lateral spacing between
adjacent trenches 210 is significantly smaller than the lateral
spacing of the trenches 110 in FIG. 10. The semiconductor device
201 also includes a source region 231, a body region 233, a drift
region 234 and a drain region 232, respectively, similar to the
semiconductor device 101 of FIG. 10. Furthermore, gate electrodes
211 and field electrodes 212 are arranged in each trench 210. The
electrical potential of the field electrodes 212 is permanently
kept at source potential during operation so that, as explained
above, the electrical field is pushed out of the mesa-structures
254. An avalanche breakdown, which is indicated by a flash, would
therefore occur close to any of the bottoms of the trenches 210
where the iso-electrical potential lines are concentrated.
[0067] Both, the semiconductor device 101 and the semiconductor
device 201, have a fixed location of their avalanche breakdown.
Different thereto, the location of the avalanche breakdown of the
semiconductor device 1 as described herein can be controlled and
shifted. A beneficial aspect of the semiconductor device 1 is that
it can be controlled to behave like a dense trench device under
normal or forward conditions and like a device with largely spaced
trenches during avalanche breakdown. To this end, the lateral
extension of the trenches 10, 20 of the semiconductor device 1 is
in many embodiments equal to or larger than the lateral spacing
(lateral extension of the mesa structures 54) between adjacent
trenches 10, 20 to maintain the dense trench regime.
[0068] By selectively applying different electrical potentials to
field electrodes of selected trenches the electrical properties of
these trenches and thus the characteristics of the whole device can
be changed. In forward mode, the semiconductor device 1 should
function as a device of the dense trench concept to make use of the
reduced on-state resistance of dense trench devices. Particularly,
the doping concentration of the drift region 34 can be increased in
comparison with largely spaced trenches of the field electrode
concept. However, the avalanche breakdown of a dense trench device
as shown in FIG. 11 is located close to the field insulating layer.
During an avalanche breakdown charge carriers generated in close
proximity to the trench 210 flow along the trench-substrate
interface either towards the body region 233 and then to the
contact region 240 or towards the drain region 232 depending on
their conductivity type. During this flow, the charge carriers may
be scattered into the adjacent insulating layers (field insulating
layer or gate insulating layer) and become trapped and accumulated
in the respective insulating layers. The trapped charge carriers
can cause a drift of the electrical properties of the device which
could lead to a changed behaviour and eventually to a malfunction
of the device.
[0069] The semiconductor device 1 as described herein avoids or at
least significantly reduces accumulation of the hot charge carriers
generated during avalanche breakdown since the location of the
avalanche breakdown is shifted towards the contact region 40 away
from the insulating layers of the trench. Particularly, the
location of the avalanche breakdown associated with the repetitive
avalanche breakdown occurring at substantially constant frequency
and caused by external applications can be influenced. Hence,
generated charge carriers flow directly either to the contact
region or to the drain in sufficient distance to the trenches so
that no accumulation takes place.
[0070] Basically, at least one of the trenches should be arranged
to allow application of different electrical potentials. Typically,
the switchable trench is arranged between two adjacent `normal`
trenches having a constant electrical potential applied to their
field electrodes. The number and arrangement of the switchable
trenches can be adjusted according to specific needs. In most
applications, up to 50% of all trenches can be formed as switchable
trenches. In other embodiments, at least 5% of all trenches are
formed as switchable trenches so that a typical range is from about
5% to 50%.
[0071] It is not required that all trenches are formed as
switchable trenches although the iso-electrical potential lines
remain concentrated at the bottoms of all non-switched trenches in
reverse mode. However, the likelihood that the actual avalanche
breakdown occurs close to the contact region arranged adjacent to
the switched trench is significantly larger than the likelihood
that the breakdown occurs at the non-switched trenches.
[0072] For manufacturing the semiconductor device 1 as described
herein and to ensure reliable operation of the device, the
following may be considered.
[0073] The gate-field insulating layer 25 of the switchable trench
20 should have a sufficient thickness to withstand the large
potential difference between the gate electrode 21 and the field
electrode 22 when applying the drain potential to the field
electrode 22. This does not necessarily apply to the gate-field
insulating layer 15 of the first trenches 10. However, to simplify
the manufacturing process, both gate-field insulating layers 15, 25
can be formed with the same thickness. It is, however, also
possible to manufacture the switchable or second trenches 20 with a
gate-field insulating layer 25 which is significantly larger than
the gate-field insulating layer 15 of the first trenches 10. In
case of silicon oxide as material for the gate-field insulating
layer 25 the thickness of the gate-field insulating layer 25
expressed in nanometers should be approximately equal to or higher
than the potential difference expressed in volt which the
gate-field insulating layer 25 should withstand. For example, for a
potential difference of about 50 V the gate-field insulating layer
25 should have a thickness of at least 50 nm to 60 nm.
[0074] Further, the upper edge 27 of the field electrode 22 should
be arranged above the upper edge 17 of the field electrode 10 to
ensure that the density of the iso-electrical potential lines is
higher close to the contact region 40 than close to the bottom of
the first trenches 10 even at moderate source-drain voltages. This
locally increases the electrical field strength at the contact
regions 40 in comparison with the field strength at the bottoms of
the first trenches 10 to ensure that the breakdown will occur close
to the contact region. In certain embodiments, the upper edge 27 of
the field electrode 22 can be arranged above the lower edge (first
pn-junction) of the body region 33.
[0075] More particularly, the upper edge 27 of the field electrode
22 of the second trench 20 should be arranged above the lower end
41 of the contact region 40 to increase even more the density of
the iso-electrical potential lines close to the contact region 40.
In certain embodiments, the field electrode 22 of the second trench
20 can therefore have an upper edge facing the first surface 51 and
can be arranged above the lower end 41 of the contact region 40
formed in the semiconductor substrate 2 between adjacent trenches
for providing an electrical connection to the body region 33.
[0076] Moreover, reducing the lateral distance of adjacent trenches
decreases the density of the iso-electrical potential lines at the
bottoms of the trenches. Closely spaced trenches with field
electrodes at source potential avoid deep penetration of the
electrical field into the mesa structures 54 and thus reduce the
curvature of the electrical field at the trench's bottoms.
[0077] The above described structural features can either by
embodied in separate semiconductor devices or combined in a single
semiconductor device according to specific needs.
[0078] As a consequence of the increased thickness of the
gate-field insulating layer 26 and the arrangement of the field
electrode's upper edge arrangement, the gate electrode 21 of the
second trenches 20 typically does not extend as far as the gate
electrode 11 of the first trenches in vertical direction. More
precisely, each gate electrode 11, 21 has a lower edge 16 and 26,
respectively, which faces towards the second surface 52. The lower
edge 16 of the first gate electrode 11 defines a reference plane
which runs parallel to the first surface 51. The lower edge 26 of
the gate electrode 21 of the second trench 20 is arranged above the
reference plane. More particularly, the lower edge 26 of the gate
electrode 21 is arranged above the second pn-junction 36, which
corresponds to the lower edge of body region 33, and spaced to it
by a distance d2. On the other hand, the lower edge 26 of the
second trench's gate electrodes 21 is disposed by a distance dl
from first pn-junction 35 which corresponds to the upper edge of
the body region 31. The distance dl is typically larger than the
distance d2 and the ratio between d2:d1 can be from about 1:1 to
about 1:100.
[0079] This arrangement leads to an increased overlap of the
gate-field insulating layer 25 of the second trench 20 and the body
region 33 in vertical direction which results in an increased
threshold voltage of the second trench 20. The increased threshold
voltage causes a delayed formation of a conductive channel between
source and drift regions 31 and 34 in comparison with the channel
formation of the first trench 10 when the same voltage is applied
to the gate electrodes 11, 21, respectively. Delaying channel
formation improves the safe operation area (SOA) characteristic of
the semiconductor device 1.
[0080] Alternatively, the thickness of the gate insulating layer 23
of the second trench 20 can be increased to even avoid channel
formation when the gate voltage is applied. For example, the gate
insulating layer 23 can have the same thickness as the field
insulating layer 24 which would simplify the manufacturing method.
When using a thick gate insulating layer 23, the gate electrode 21
of the second trench 20 can also be operated at source potential or
at any other potential such as drain or gate potential or at an
intermediate potential. In this case, the second trench 20 is
`inactivated` with respect to its ability to cause channel
formation. In other words, each of the first and second trenches
can include a gate insulating layer for electrically insulating the
respective gate electrodes from the semiconductor substrate by a
gate insulating layer, wherein the gate insulating layer of the
second trench is thicker than the gate insulating layer of the
first trench.
[0081] In both cases, the first and the second trenches 10, 20 have
different threshold voltages with the threshold voltage of the
second trench 20 being higher than the threshold voltage of the
first trench 10. In other words the first and second trenches can
include respective threshold voltages for forming a conductive
channel in the body region adjacent to the respective gate
electrodes, wherein the threshold voltages of the second trench is
higher than the threshold voltage of the first trench.
[0082] In forward mode, the inactivated second trench 20 does not
contribute to the load current between source and drain regions 31,
32. This slightly increases the overall on-state resistance defined
by all cells of the power semiconductor device 1. To reduce the
influence of the inactivated second trenches 20, the proportion of
the second trenches 20 with respect to the first trenches 10 should
be kept small.
[0083] Second trenches 20 having a structure which results in a
delayed channel formation or a channel formation at higher voltages
also slightly increase the overall on-state resistance but to a
lesser degree than inactivated second trenches 20.
[0084] In certain embodiments, for operating the semiconductor
device a first electrical potential is applied to the first doping
region 31 and each of the field electrodes 12, 22 of the first and
second trenches. A second electrical potential is applied to the
second doping region 32. At least the field electrode 22 of the
second trench 20 is temporarily disconnected from the first
electrical potential and the second electrical potential is applied
to the field electrode 22 of the second trench 20 while keeping the
field electrode 12 of the first trench 10 at the first electrical
potential. Hence, the field electrode 22 of the second trench is
alternatingly connected to the first and the second electrical
potential. The duty ratio for temporarily connecting the second
electrical potential to field electrode 22 of the second trench 20
can be selected according to specific needs. In certain
embodiments, the duty ratio is selected such to avoid avalanche
breakdown near the bottoms of the trenches 10, 20 as explained
above.
[0085] According to another embodiment, a semiconductor device
include a semiconductor substrate 2 having a first surface 51 and a
second surface 52 arranged opposite to the first surface 51, a
first doping region 31 formed in the semiconductor substrate 2 at
the first surface 51, a second doping region 32 formed in the
semiconductor substrate 2 at the second surface 52, and a third
doping region 33 in contact with the first doping region 31. The
third doping region 33 is of opposite conductivity type to the
first doping region 31. First doping region 31 can be a source
region, second doping region 32 a drain region and third doping
region 33 a body region, respectively. At least a first and a
second trench 10, 20 are formed in the semiconductor substrate 2
and extend through the third doping region 33. The first and second
trenches 10, 20 are laterally spaced to each other to form a
mesa-structure 54 therebetween. First and second trenches 10, 20
include at least a field electrode 12, 22. A contact region 40 is
formed in the mesa-structure 54 between adjacent trenches 10, 20
for providing an electrical connection to the third doping region
33. The field electrode 12 of the second trench has an upper edge
27 facing the first surface 51 and is arranged above a lower end 41
of the contact region 40. The first trench 10 can include a gate
electrode 11 arranged above the field electrode 12. In addition to
that, second trench can have an optional gate electrode 21 arranged
above the field electrode 22. Gate electrode 21 can be electrically
inactivated by arranging an at least partially thick gate
insulating layer 23 between gate electrode 21 and the semiconductor
substrate. Different thereto, gate electrode 11 of the first trench
10 is typically electrically active, i.e. capable of forming a
conductive channel in the third semiconductor region 33 when
applying an appropriate voltage to gate electrode 11. To this end,
gate electrode 11 is insulated from third semiconductor region 33
by a comparable thin gate insulating layer 13. In this case, both
gate electrodes 11, 21 define different threshold voltages with the
threshold voltage of gate electrode 21 being large or even
significantly larger than the threshold voltage of gate electrode
11. Alternatively, the gate insulating layers 13, 23 of both
trenches 10, 20 can have substantially the same thickness.
[0086] FIG. 3 shows an embodiment having a second trench 20 with an
inactivated second trench 20 by using a thick gate insulating layer
23. Furthermore, the source region 31 is formed to be arranged in
spaced relation to the second trench 20. The switching device 60 of
this embodiment is a Zener diode with a Zener breakdown voltage at
about 90% of the nominal drain voltage or the avalanche breakdown
voltage.
[0087] The field electrodes 22 of all switchable trenches 20 can be
connected to a common control contact pad 73 to which an integrated
device forming the switching device 60 is connected which allows a
short temporal connection to the drain potential.
[0088] Another embodiment is shown in FIG. 5 which includes a
second trench 20 with a thin gate insulating layer 23 having
substantially the same thickness as the gate insulating layer 13 of
the first trench 10. However, since the drain region 31 does not
laterally extend to the second trench 20 in this embodiment, the
second trench 20 does not contribute to the load current during
forward mode.
[0089] Close to and in contact with the drain region 32 is formed a
field stop region 35 having a doping concentration which is between
the doping concentration of the drift region 34 and the doping
concentration of the drain region 32. The field stop region 35 is
of the same conductivity as the drift region 34. Field stop layers
are also used for IGBTs and form than a further pn-junction with
the emitter region 32 of opposite conductivity.
[0090] FIG. 6 shows a further embodiment having an inactivated
second trench 20 by using a gate insulating layer 23 having a
thickness which is larger than the thickness of the gate insulating
layer 13 of the first trench 10. This embodiment also shows a drain
contact 72 formed at the second surface 52 and a source contact 71
formed at the first surface 51. Both, the drain and the source
contacts 71, 72 are typically a metallization layer made of a
suitable metal or an alloy. The source contact 71 can also be in
electrical connection to the contact regions 40 and is insulated
from the trenches 10, 20 by an insulating layer 74.
[0091] It goes without saying that the field stop layer 35 and the
metallization layers 71 and 72 can also be combined with any other
embodiments described herein.
[0092] A further embodiment is shown in FIG. 7. Every third trench
20 of this semiconductor device 1 is adapted to be switchable so
that about 33% of all trenches correspond to the second trenches
20. The structure of the first and second trenches 10, 20 and of
the source region 31 corresponds to the respective structures as
described in connection with FIG. 2. The optional highly doped
regions 37 are not formed in this embodiment but can be if desired.
In case that no highly doped regions 37 are formed, the lower face
of the contact regions forms its lower edge 41.
[0093] FIG. 8 shows a semiconductor device 1 in which about 25% of
all trenches are adapted to be switchable (every fourth trench).
The switchable trenches (second trenches) 20 include a thick gate
insulating layer which inactivates the second trenches 20 so that
the second trenches 20 will not contribute to the load current in
the forward mode.
[0094] For sake of simplicity, drain region 32, source contact 71,
drain contact 72, control contact pad 73 and switching device 60
are not shown in FIGS. 7 and 8.
[0095] A method for manufacturing a semiconductor device will now
be described in connection with FIGS. 12A to 12D.
[0096] A semiconductor substrate 2 is provided and trenches 10, 20
are formed by anisotropic etching (FIG. 12A).
[0097] Then, the field insulating layers 14, 24 are formed by
depositing an insulating material followed by formation of the
field electrodes 12 and 22 respectively. The field electrodes 12,
22 can be made of highly doped polysilicon. The location of the
upper edge 17, 27 of the respective field electrodes 12, 22 can be
defined by different etching times used for back etching of the
deposited polysilicon.
[0098] In a further step, the respective gate-field insulating
layers 15, 25 are formed by depositing an insulating layer or
oxidising upper portions of the field electrodes 12, 22,
respectively. The gate-field insulating layers 15, 25 can have the
same thickness, or the gate-field insulating layer 25 of the second
trench 20 can be formed to have a thickness which exceeds the
thickness of the gate-field insulating layer 15 of the first
trenches 10. Different thicknesses can be achieved by separate
deposition processes or by depositing an additional insulating
layer into the second trenches 20.
[0099] In further steps, the gate insulating layers 13, 23 and the
gate electrodes 11, 21 are formed.
[0100] As shown in FIG. 12C the source, body and drain (emitter)
regions 31, 33 and 32, respectively, are formed by implantation and
thermal annealing. Also, the contact regions 40 are formed by
etching openings into the mesa structures 54 and filling them with
highly doped polysilicon. Highly doped regions 37 are formed by
implantation before depositing the highly doped silicon.
[0101] The order of the above steps can be changed according to
specific needs. For instance, source and body regions 31, 33 can be
formed prior to formation of the trenches 10, 20 or during
intermediate steps at which the field electrodes 12, 22 or the gate
electrode 11, 21 are not yet completed.
[0102] Finally, the drain contact 72, source contact 71 and control
contact 73 are formed by depositing respective metal layers (FIG.
12D).
[0103] To summarise, the semiconductor device as described in some
embodiments allows the dynamic change of its electrical properties
by temporarily applying an electrical potential, which can
substantially correspond to the drain potential, to field
electrodes of selected trenches. These field electrodes are
connected to a switching device which is adapted for switching
between two electrical potentials. By doing so, the location of an
avalanche breakdown can be shifted from the bottoms of the trenches
to the contact regions. In other words, the semiconductor device
can be switched between dense trench regime and field electrode
regime.
[0104] In more general terms, a switchable field electrode is
arranged close to a pn-junction to alter the breakdown
characteristics of that pn-junction. The field electrode is
switched between the electrical potentials which can be applied
through a switching device.
[0105] A control contact pad, which is electrically connected to
the field electrode arranged next to the pn-junction or the field
electrodes of the switchable trenches and which is arranged on the
semiconductor device, allows application of different electrical
potentials through the switching device.
[0106] To improve reliability of the semiconductor device and to
ensure that the avalanche breakdown occurs close to the contact
region, the upper edge of the field electrode of the switchable
trenches should be arranged above the lower edge of the contact
region.
[0107] Furthermore, the gate-field insulating layer of the
switchable trenches should be adjusted to withstand the high
electrical field occurring during application of the drain
potential to the field electrode. Typically, the gate-field
insulating layer of the switchable trenches is thicker than the
gate-field insulating layer of the other trenches.
[0108] The semiconductor device as described herein is not
restricted to power applications but can be used for any
application for which a change of the electrical properties of
selective trenches is of desired.
[0109] The written description above uses specific embodiments to
disclose the invention, including the best mode, and also to enable
any person skilled in the art to make and use the invention. While
the invention has been described in terms of various specific
embodiments, those skilled in the art will recognise that the
invention can be practiced with modifications within the spirit and
scope of the claims. Especially, mutually non-exclusive features of
the embodiments described above may be combined with each other.
The patentable scope is defined by the claims, and may include
other examples that occur to those skilled in the art. Such other
examples are intended to be within the scope of the claims if they
have structural elements that do not differ from the literal
language of the claims, or if they include equivalent structural
elements with insubstantial differences from the literal languages
of the claims.
* * * * *