 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to monolithic ceramic electronic components,
such as a monolithic ceramic capacitor provided with internal electrodes
composed of a base metal, for example, nickel or a nickel alloy.
2. Description of the Related Art
Various types of monolithic ceramic electronic components provided with a
plurality of ceramic layers and internal electrodes formed between the
ceramic layers have been commercially available. A typical example thereof
is a monolithic ceramic capacitor in which a ceramic dielectric material
is used for the ceramic layers.
Conventionally, a noble metal such as palladium or platinum or an alloy
thereof is used for internal electrodes in such a monolithic ceramic
capacitor because the dielectric material must be fired in air at a
temperature as high as approximately 1,300.degree. C. However, such
materials for electrodes are very expensive, resulting in an increase in
production cost.
In order to reduce the production cost, the use of base metals as materials
for internal electrodes in monolithic ceramic capacitors has been
implemented, and various types of nonreducing dielectric materials which
can be fired in a neutral or reducing atmosphere in order to prevent the
electrodes from oxidizing during firing have been developed. Examples of
the base metal used for internal electrodes include cobalt, nickel and
copper. In view of cost and oxidation resistance, nickel is predominantly
used.
There is now a demand for further reduction in size and larger capacitance
with respect to monolithic ceramic capacitors, and an increase in
dielectric constant and a decrease in thickness have been studied with
respect to ceramic dielectric materials, and simultaneously, a decrease in
thickness has been studied with respect to materials for electrodes.
In general, internal electrodes of monolithic ceramic capacitors are formed
by a printing method, such as screen-printing, using a paste containing a
metallic powder. When nickel powder is used as the metallic powder to be
incorporated in such a paste, a nickel powder having an average particle
diameter of more than 0.25 .mu.m, which is produced by a liquid phase
method or a chemical vapor method, is used in many cases. However, with
such a large particle size, it is difficult to decrease the thickness of
internal electrodes.
When a nickel powder having an average particle diameter as large as 0.25
.mu.m is used, in order to make the dielectric ceramic exhibit dielectric
properties, the thickness of the electrodes must be set at 0.8 .mu.m or
more.
Although a decrease in the thickness of the ceramic dielectric layer is the
most effective means to increase the capacitance of the monolithic ceramic
capacitor, for example, if the thickness of the ceramic layer is 3 .mu.m
or less against the thickness of 0.8 .mu.m of the internal electrode,
delamination, which is a fatal structural defect in the monolithic
capacitor, frequently occurs due to a difference in shrinkage factor
between the electrode and the ceramic.
With respect to high dielectric constant type monolithic ceramic capacitors
satisfying F-level and E-level characteristics and
temperature-compensating type monolithic ceramic capacitors satisfying
SL-level and CG-level characteristics stipulated in the Japanese
Industrial Standard (JIS), if the thickness of the ceramic layer is as
thin as 3 Mm or less, the electrical characteristics may degrade,
resulting in difficulty in obtaining a high-performance monolithic ceramic
capacitor.
SUMMARY OF THE INVENTION
Accordingly, it is an object of the present invention to provide a
monolithic ceramic electronic component, such as a monolithic ceramic
capacitor, in which the thickness of inner electrodes and ceramic layers
can be decreased without the occurrence of structural defects, enabling
miniaturization and high reliability.
A monolithic ceramic electronic component in accordance with the present
invention includes a laminate including a plurality of ceramic layers
obtained by sintering a ceramic raw material powder, and a plurality of
internal electrodes located between the ceramic layers and obtained by
sintering a metallic powder. The ceramic layers have a thickness of about
3 .mu.m or less and contain ceramic grains having an average particle
diameter of more than about 0.5 .mu.m. The particle diameter of the
ceramic grains in the thickness direction of the ceramic layers is smaller
than the thickness of the ceramic layers. The internal electrodes have a
thickness of about 0.2 to 0.7 .mu.m.
Preferably, the monolithic ceramic electronic component further includes an
external electrode formed on each of the opposing ends of the laminate,
the ceramic layers are composed of a ceramic dielectric material, and each
of the plurality of internal electrodes is formed with an edge being
exposed to either one of the opposing ends of the laminate so as to be
electrically connected to either one of the external electrodes to form a
monolithic ceramic capacitor.
Preferably, the internal electrodes are formed of a paste containing the
metallic powder, and the metallic powder in the paste has an average
particle diameter of about 10 to 200 nm.
Preferably, the metallic powder is composed of a base metal, and the base
metal preferably contains nickel.
Preferably, the internal electrodes are formed by a method including a step
of applying the paste containing the metallic powder by a printing method.
Preferably, the ceramic raw material powder before sintering has an average
particle diameter of about 25 to 250 nm.
Preferably, each of the ceramic grains constituting the ceramic layers has
a uniform composition and a uniform crystal system, and the individual
ceramic grains have the same composition and the same crystal system.
Preferably, each of the ceramic grains constituting the ceramic layers has
a uniform composition and a uniform crystal system, and the ceramic layers
are composed of at least 2 types of ceramic grains having different
compositions.
BRIEF DESCRIPTION OF THE DRAWING
FIG. 1 is a sectional view showing a monolithic ceramic capacitor in
accordance with an embodiment of the present invention.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
An embodiment of the present invention will be described, which is applied
to a monolithic ceramic capacitor 1 having the structure shown in FIG. 1.
The monolithic ceramic capacitor 1 includes a laminate 3 including a
plurality of ceramic layers 2 composed of a ceramic dielectric material
which are laminated, and first and second external electrodes 6 and 7
which are provided on first and second ends 4 and 5, respectively. The
monolithic ceramic capacitor 1 constitutes a chip-type monolithic ceramic
electronic component in a rectangular parallelepiped shape.
First internal electrodes 8 and second internal electrodes 9 are
alternately placed in the laminate 3. The first internal electrodes 8 are
formed at a plurality of specific interfaces between the ceramic layers 2
with an edge being exposed to the first end 4 so as to be electrically
connected to the first external electrode 6. The second internal
electrodes 9 are formed at a plurality of specific interfaces between the
ceramic layers 2 with an edge being exposed to the second end 5 so as to
be electrically connected to the second external electrode 7.
In order to fabricate the monolithic ceramic capacitor 1, a principal raw
material such as barium titanate, i.e., a ceramic raw material powder, and
additives for improving characteristics, etc., are prepared as starting
materials,. Preferably, for the reason described below, the ceramic raw
material powder to be used has an average particle diameter of about 25 to
250 nm, for example, by adjusting the calcining temperature or by
employing wet synthesis. The ceramic raw material powder is produced by
wet-mixing oxides or carboxides, which is known as a solid-phase method,
or by wet synthesis, which is known as hydrothermal synthesis or
hydrolysis, so as to satisfy a predetermined composition, followed by
drying and calcining.
Predetermined amounts of the raw material powder and the additives are
weighed, and a mixed powder is formed by wet mixing. More specifically,
the individual additives are mixed to the ceramic raw material powder in
the form of oxide powders or carboxide powders, followed by wet mixing. At
this stage, in order to make the individual additives soluble in a
solvent, alkoxides or compounds such as acetylacetates or metallic soaps
may be formed. Alternatively, a solution containing the individual
additives may be applied to the surface of the ceramic raw material
powder, followed by heat treatment.
Next, a ceramic slurry is prepared by adding an organic binder and a
solvent to the mixed powder. Ceramic green sheets for forming the
dielectric ceramic layers 2 are formed by using the ceramic slurry. The
thickness of the green sheets is set so that the thickness after firing is
about 3 .mu.m or less for the reason described below.
Conductive paste films for forming internal electrodes 8 and 9 are then
formed on the specific ceramic green sheets by a printing method, such as
screen-printing. The thickness of the conductive paste films is set so
that the thickness after firing ranges from about 0.2 to 0.7 .mu.m.
A paste constituting the conductive paste films contains a metallic powder,
a binder and a solvent. The metallic powder preferably has an average
particle diameter of about 10 to 200 nm for the reason described below.
For example, a paste containing a nickel powder, an ethyl cellulose binder
and a solvent, such as triphenol, may be used. The paste is elaborately
prepared by a three-roller mill or the like so that the aggregation of the
nickel powder having a very small average particle diameter of about 10 to
200 nm is loosened or avoided and the nickel powder is satisfactorily
dispersed.
The metallic powder, more specifically, the nickel powder, may be
advantageously produced, for example, by a chemical vapor method, a
hydrogen arc discharge method, or a gas evaporation method.
In the chemical vapor method, nickel chloride is vaporized by heating, and
the resulting nickel chloride vapor is brought into contact with hydrogen
at a predetermined temperature while being transported by an inert gas,
thus causing a reaction to produce a nickel powder. The nickel powder is
recovered by cooling the reactant gas containing the nickel powder.
In the hydrogen arc discharge method, arc discharge is performed in an
atmosphere containing hydrogen gas to melt and vaporize nickel, and thus a
fine nickel powder is produced from the vapor phase. By dissolving
supersaturated hydrogen in molten nickel by means of arc or plasma heat, a
high-temperature state locally occurs when hydrogen is released from the
molten nickel, and the evaporation of nickel is accelerated, and thus a
nickel vapor is released. By condensing and cooling the nickel vapor, the
fine nickel powder is produced.
In the gas evaporation method, a nickel ingot is melted in a vessel filled
with an inert gas, such as Ar, He or Xe, by heating means, such as
high-frequency induction heating, so that a nickel vapor is produced. The
resulting nickel vapor is cooled and solidified by being brought into
contact with the inert gas in the atmosphere, and thus the fine nickel
powder is produced.
Next, a plurality of ceramic green sheets including the ceramic green
sheets provided with the conductive paste films as described above are
laminated and pressed, followed by cutting as required. In this way, a
green laminate 3 is fabricated, in which the plurality of ceramic green
sheets and the conductive paste films for forming the plurality of
internal electrodes 8 and 9 located between the ceramic green sheets are
laminated, and an edge of each conductive paste film for forming the
internal electrode 8 or 9 is exposed to the end 4 or 5, respectively.
Next, the laminate 3 is fired in a reducing atmosphere. At this stage, for
the reason described below, the firing conditions are set so that ceramic
grains constituting the ceramic layers 2 after firing have an average
particle diameter of more than about 0.5 .mu.m.
The first and second external electrodes 6 and 7 are formed on the first
and second ends 4 and 5 of the laminate 3, respectively, so as to be
electrically connected to the exposed edges of the first and second
internal electrodes 8 and 9.
The material composition of the external electrodes 6 and 7 is not
particularly limited. Specifically, the same material as that for the
internal electrodes 8 and 9 may be used. Alternatively, a sintered layer
composed of a conductive metallic powder, such as Ag, Pd, Ag--Pd, Cu or a
Cu alloy, or a sintered layer composed of the conductive metallic powder
added with glass frit, such as B.sub.2 O.sub.3 --Li.sub.2 O--SiO.sub.2
--BaO-based glass, B.sub.2 O.sub.3 --SiO.sub.2 --BaO-based glass, Li.sub.2
O--SiO.sub.2 --BaO-based glass or B.sub.2 O.sub.3 --SiO.sub.2 --ZnO-based
glass, may be used. An appropriate material is selected depending on the
intended application of the monolithic ceramic capacitor 1, the operating
environment of the monolithic ceramic capacitor 1, etc.
Additionally, the external electrodes 6 and 7 may be formed by applying a
metallic powder paste to the fired laminate 3 followed by baking, or may
be formed by applying the metallic powder paste to the green laminate 3
and by firing simultaneously with the laminate 3.
As required, the external electrodes 6 and 7 are coated with plating layers
10 and 11 composed of Ni, Cu, a Ni--Cu alloy or the like, respectively.
Furthermore, second plating layers 12 and 13 composed of solder, tin or
the like may be formed on the plating layers 10 and 11, respectively.
With respect to the thickness of the internal electrodes 8 and 9 in the
present invention, and with respect to the average particle diameters of
the Ni powder contained in the paste used for forming the internal
electrodes 8 and 9, the ceramic raw material powder before sinteling for
forming the ceramic layers 2, and the ceramic grains constituting the
ceramic layers, and to the thickness of the ceramic layers 2 in the
embodiment of the present invention, the ranges described above are
defined. Herein, the "average particle diameter" means a diameter of
particles (D.sub.50) corresponding to 50% particles in the number-size
distribution obtained by analyzing electron micrographs of the powders and
the ceramic grains.
In the present invention, the reason for setting the thickness of the
internal electrodes 8 and 9 at about 0.7 .mu.m or less is that if the
thickness exceeds about 0.7 .mu.m, when the thickness of the ceramic layer
2 is as thin as about 3 .mu.m or less, delamination due to a difference in
shrinkage factors between the internal electrodes 8 and 9 containing
nickel and the ceramic layers 2 inevitably occurs. In other words, by
setting the thickness of the internal electrodes 8 and 9 at about 0.7
.mu.m or less, the thickness of the ceramic layers 2 can be reduced to
about 3 .mu.m or less without any problems, thus enabling miniaturization
and an increase in capacitance of the monolithic ceramic capacitor 1.
On the other hand, the thickness of the internal electrodes 8 and 9 is set
at about 0.2 .mu.m or more because, if the thickness is less than about
0.2 .mu.m, nickel contained in the internal electrodes 8 and 9 reacts with
the ceramic contained in the ceramic layers 2 during firing, resulting in
oxidation of nickel or delamination due to the oxidation, and the function
as the internal electrodes may be lost.
The reasons for setting the average particle diameter of the ceramic grains
of the ceramic dielectric material at more than about 0.5 .mu.m and for
defining that the particle diameter of the ceramic grains in the thickness
direction of the ceramic layers be smaller than the thickness of the
ceramic layers are as follows.
That is, when the thickness of the ceramic layers is set at about 3 .mu.m
or less, if the average particle diameter of the ceramic grains is about
0.5 .mu.m or less, the dielectric properties of the ceramic are degraded
due to thermal stress resulting from a difference in thermal shrinkage
factors between the internal electrode layers and the ceramic layers
during firing and cooling of the monolithic ceramic capacitor. When the
average particle diameter of the ceramic is set at more than about 0.5
.mu.m by appropriately selecting the firing temperature and the ceramic
composition, the dielectric properties of the ceramic layers improve, thus
enabling miniaturization and an increase in capacitance of the monolithic
ceramic capacitor.
If the particle diameter of the ceramic grains exceeds the thickness of the
ceramic layers, delamination occurs due to firing, which is
disadvantageous. However, when the particle diameter of the ceramic grains
in the thickness direction of the ceramic layer is not larger than the
thickness of the ceramic layers, even if the particle diameter of the
ceramic grains in the longitudinal direction of the ceramic layers is
equal to or larger than the thickness of the ceramic layers, no problem
arises with respect to the properties.
In the case of the high-dielectric-constant type monolithic ceramic
capacitors satisfying the F-level and E-level characteristics stipulated
in JIS, preferably, each of the ceramic grains constituting the ceramic
layers has a uniform composition and a uniform crystal system, and the
individual ceramic grains have the same composition and the same crystal
system. Thereby, the dielectric constant of the ceramic layers is
increased and a monolithic ceramic capacitor with high reliability can be
obtained.
In the case of the temperature-compensating type monolithic ceramic
capacitors satisfying the SL-level and CG-level characteristics stipulated
in JIS, preferably, each of the ceramic grains constituting the ceramic
layers has a uniform composition and a uniform crystal system, and the
ceramic layers are composed of at least 2 types of ceramic grains having
different compositions. Thereby, the Q factor of the ceramic layers is
increased, and the dielectric constant-temperature characteristics become
planar.
The reason for setting the average particle diameter of the Ni powder used
for the internal electrodes, preferably, at about 10 to 200 nm is as
follows.
When the average particle diameter of the Ni powder is less than about 10
nm, it is difficult to produce a paste with a viscosity that is applicable
to a printing method, such as screen-printing. Even if screen-printing is
performed using a paste having such a high viscosity, it is difficult to
form planar conductive paste films for forming the internal electrodes 8
and 9 because of high viscosity, and thin spots and pinholes occur,
resulting in a decrease in coverage and electrode disconnection.
On the other hand, when the average particle diameter of the Ni powder
exceeds about 200 nm, since the nickel particles are excessively large, it
is difficult to form planar conductive paste films for forming the
internal electrodes 8 and 9, resulting in a decrease in coverage. The
unevenness at the interfaces between the internal electrodes 8 and 9 and
the ceramic layers 2 also increases.
The reason for setting the average particle diameter of the ceramic raw
material powder for forming the ceramic device, preferably, in the range
from about 25 to 250 nm is as follows.
That is, when the average particle diameter of the ceramic raw material
powder is set at less than about 25 nm, the ceramic raw material powder
tends to aggregate, resulting in a difficulty in obtaining a uniform green
sheet, and when the thickness of the device is set at about 3.0 .mu.m,
short-circuiting easily occur. On the other hand, when the average
particle diameter of the ceramic raw material powder exceeds about 250 nm,
the evenness of the surface of the green sheet is deteriorated, resulting
in an increase in unevenness at the interfaces between the internal
electrodes 8 and 9 and the ceramic layers 2.
Although the monolithic ceramic capacitor has been described in the above
embodiment as the monolithic ceramic electronic component, the present
invention is also applicable to other monolithic ceramic electronic
components including substantially the same structure, such as
multilayered ceramic substrates.
With respect to the metallic powder contained in the paste for forming the
internal electrodes, in addition to the nickel powder described above, a
powder of a nickel alloy, a powder of other base metals, such as copper or
a copper alloy, or a powder of a noble metal may be used.
The present invention will be described below in detail based on the
examples. However, it is to be understood that the invention is not
limited to those examples.
EXAMPLE 1
Monolithic ceramic capacitors having the structure as shown in FIG. 1 were
fabricated in this example.
1. Fabrication of Samples
First, as ceramic raw material powders, (Ba, Sr)TiO.sub.3 powders having
different average particle diameters shown in Table 1 were produced by
hydrolysis. Table 2 shows the composition of the ceramics containing the
powders shown in Table 1 as principal raw materials used in the example.
With respect to the additives, a solution containing the additive
components was applied to the surfaces of the (Ba, Sr)TiO.sub.3 powders,
and heat treatment was performed at 500.degree. C. In this case, in order
to make the additives soluble in an organic solvent, alkoxides were
formed, and also compounds such as acetylacetates or metallic soaps were
formed. The ceramic raw material powders having the desired composition
shown in Table 2 were then calcined, and by adjusting the calcining
temperature, the ceramic raw materials having the average particle
diameters of 15 nm, 25 nm, 200 nm, and 300 nm were prepared as shown in
Table 3.
TABLE 1
Average Particle
Ba.sub.0.7 Sr.sub.0.3 TiO.sub.3 Diameter
Powder Type D50 (nm)
A 15
B 25
C 50
TABLE 1
Average Particle
Ba.sub.0.7 Sr.sub.0.3 TiO.sub.3 Diameter
Powder Type D50 (nm)
A 15
B 25
C 50
TABLE 1
Average Particle
Ba.sub.0.7 Sr.sub.0.3 TiO.sub.3 Diameter
Powder Type D50 (nm)
A 15
B 25
C 50
Next, a polyvinyl butyral-based binder and an organic solvent, such as
ethanol, were added to the individual barium titanate-based ceramic raw
material powders shown in Table 3, followed by wet mixing using a ball
mill, to prepare ceramic slurries. The ceramic slurries were formed into
sheets by a doctor blade process. By adjusting the slit width of the
doctor blade, ceramic green sheets having thicknesses of 4.2 .mu.m and 1.4
.mu.m were formed. The individual thicknesses of 4.2 .mu.m and 1.4 .mu.m
corresponded to the thicknesses of 3 .mu.m and 1 .mu.m of the ceramic
layers after lamination and firing, respectively, as is obvious from the
evaluation results described below.
In the meantime, spherical Ni powders having average particle diameters of
5 nm, 15 nm, 50 nm, 100 nm, 180 nm and 250 nm were produced. More
specifically, the Ni powders having the average particle diameters of 5 nm
and 15 nm were produced by the gas evaporation method, the Ni powders
having the average particle diameters of 50 nm and 100 nm were produced by
the hydrogen arc discharge method, and the Ni powders having the average
particle diameters of 180 nm and 250 nm were produced by the chemical
vapor method.
Next, 42% by weight of each Ni powder was combined with 44% by weight of an
organic vehicle, which was produced by dissolving 6% by weight of an ethyl
cellulose-based binder in 94% by weight of triphenol, and 14% by weight of
triphenol, and dispersion and mixing treatment was performed using a
three-roller mill to prepare a paste containing the Ni powder
satisfactorily dispersed.
Next, the resultant Ni pastes were screen-printed on the individual ceramic
green sheets to form conductive paste films for forming internal
electrodes. At this stage, by adjusting the thickness of the screen
pattern, samples provided with the conductive paste films having
thicknesses of 1.2 .mu.m, 1.0 .mu.m, 0.6 .mu.m, 0.3 .mu.m and 0.15 .mu.m
were fabricated. The individual thicknesses of 1.2 .mu.m 1.0 .mu.m, 0.6
.mu.m, 0.3 .mu.m, and 0.15 .mu.m of the conductive paste films after
drying corresponded to the individual thicknesses of 0.8 .mu.m, 0.7 .mu.m,
0.4 .mu.m, 0.2 .mu.m and 0.1 .mu.m of the internal electrodes after
lamination and firing, as is obvious from the evaluation results described
below.
Next, a plurality of ceramic green sheets were laminated in such a manner
that the edges at which the conductive paste layers were exposed
alternately faced different ends of the laminate to be formed, followed by
heat pressing for consolidation. The pressed structure was cut into pieces
with a predetermined size, and green chips as green laminates were
obtained. The green chips were heated at 300.degree. C. in a nitrogen
atmosphere, and after the binder was removed, firing was performed for 2
hours at a firing temperature in the range from 1,000.degree. C. to
1,200.degree. C. shown in Table 4 in a reducing atmosphere comprising
gases of H.sub.2, N.sub.2 and H.sub.2 O having an oxygen partial pressure
of 10.sup.-9 to 10.sup.-12 MPa.
A silver paste containing B.sub.2 O.sub.3 --Li.sub.2 O--SiO.sub.2
--BaO-based glass frit was applied to both ends of each sintered ceramic
laminate, and baking was performed at 600.degree. C. in a nitrogen
atmosphere, and thus external electrodes which were electrically connected
to the internal electrodes were formed.
In the individual samples, monolithic ceramic capacitors obtained as
described above had outer dimensions in which the width was 5.0 mm, the
length was 5.7 mm and the thickness was 2.4 mm, and the ceramic layers
interposed between the internal electrodes had a thickness of 3 .mu.m or 1
.mu.m. The total number of the effective ceramic dielectric layers was 5
and the area of the counter electrode per one layer was
16.3.times.10.sup.-6 m.sup.2.
2. Evaluation of Samples
Next, with respect to the samples of the monolithic ceramic capacitors, the
laminated structure, electrical characteristics and reliability were
evaluated. The results thereof are shown in Table 4. The asterisked sample
numbers indicate that the samples were out of the scope of the present
invention.
In order to measure the average particle diameter of the dielectric ceramic
contained in each of the monolithic ceramic capacitors, the ground surface
of the cross section of the monolithic ceramic capacitor was subjected to
chemical etching and observations were carried out using a scanning
electron microscope.
In order to measure the thicknesses of the internal electrode layers and
the ceramic dielectric layers, the ground surface of the cross section of
the monolithic ceramic capacitor was observed by a scanning electron
microscope.
With respect to the delamination in the monolithic ceramic capacitor, the
cross section of each test piece was ground and judgment was made visually
by microscope observation. The rate of test pieces in which delamination
occurred to the total number of test pieces in each sample was calculated.
In order to measure the coverage, the internal electrodes of the sample
monolithic capacitors were peeled off, and photomicrographs of the states
in which the surfaces of the electrodes had holes were taken, followed by
image analysis for quantification.
With respect to the samples which were judged satisfactory in the
structural evaluation described above, the electrical characteristics
described below were evaluated.
The capacitance (C) and the dielectric loss (tan .delta.) were measured
using an automatic bridge-type meter in accordance with JIS C 5102, and
the relative dielectric constant (.epsilon.) was computed based on the
capacitance measured.
In the high-temperature load test, while a DC field of 10 kV/mm was being
applied at 150.degree. C., the change in insulation resistance with time
was measured for each test piece, and the point at which the insulation
resistance (R) reached 10.sup.5.OMEGA. or less was defined as failure. The
average life to reach failure was computed.
TABLE 4
Material Characteristics
Monolithic Capacitor Structure Ceramic Raw
Monolithic
Grain Ni Material Powder Ceramic
Evaluation of Electrical
Average Average Average Capacitor
Evaluation of Laminated Characteristics
Device Electrode Particle Particle Raw Particle Firing
Structure Average
Sample Thickness Thickness Diameter Diameter Material Diameter
Temperature Delamination Coverage tan .delta. Life
No. (.mu.m) (.mu.m) (.mu.m) (nm) No. (nm) (.degree.
C.) (%) (%) .epsilon. (%) (hour)
*A1 3 0.8 3.0 250 4 300 1200
75 68 -- -- --
*A2 3 0.8 1.5 180 3 200 1170
80 72 -- -- --
*A3 3 0.8 3.0 50 2 25 1100
45 85 -- -- --
*A4 3 0.8 1.5 15 1 15 1050
55 92 -- -- --
*A5 3 0.7 4.0 250 2 25 1200
48 75 -- -- --
A6 3 0.7 1.5 180 3 200 1170
0 80 17800 4.6 45
A7 3 0.7 0.7 50 2 25 1100
0 95 12470 3.7 88
A8 3 0.4 5.0 (3.0) 50 2 25 1200
0 88 18630 4.6 92
A9 3 0.4 1.7 15 2 25 1100
0 94 14350 4.7 88
A10 3 0.2 0.7 100 3 200 1170
0 73 13270 4.3 48
A11 3 0.2 0.7 15 2 25 1100
0 95 13320 3.8 95
A12 3 0.2 0.7 50 4 300 1200
0 74 12650 3.6 22
A13 3 0.2 0.7 5 1 15 1050
10 67 9730 2.7 98
A14 3 0.4 3.0 100 1 15 1050
5 96 19820 4.8 95
*A15 3 0.4 0.4 50 2 25 1050
0 97 5650 1.9 3
*A16 3 0.4 0.2 50 2 25 1000
0 97 3470 1.2 1
*A17 3 0.1 1.5 50 3 200 1170
80 65 -- -- --
*A18 3 0.1 0.4 15 2 25 1100
90 78 -- -- --
*A19 1 0.8 1.0 250 2 25 1100
100 84 -- -- --
*A20 1 0.7 0.4 100 2 25 1050
0 86 4430 1.7 2
A21 1 0.7 1.0 250 2 25 1100
0 65 14330 4.5 27
A22 1 0.7 0.6 180 1 15 1050
5 73 12560 4.6 33
A23 1 0.7 0.6 100 3 200 1170
0 86 13150 3.3 26
A24 1 0.4 1.0 15 2 25 1100
0 94 14620 4.7 48
A25 1 0.2 1.0 50 4 300 1200
0 63 9310 3.9 16
*A26 1 0.1 1.0 15 2 25 1100
100 91 -- -- --
*A27 1 0.1 0.6 5 2 25 1100
100 66 -- -- --
As shown in Table 4 the thickness of the internal electrodes was 0.8 .mu.m
and delamination occurred at a high rate with respect to the asterisked
sample Nos. A1 to A4 and A19. With respect to the asterisked sample Nos.
A17, A18, A26, and A27, the thickness of the internal electrodes was 0.1
.mu.m and delamination also occurred at a high rate. The delamination in
the latter was caused by oxidation of nickel.
In contrast, with respect to the sample Nos. A6 to A16 and A20 to A25 in
which the thickness of the internal electrodes was in the range from 0.2
to 0.7 .mu.m, delamination did not occur or did not substantially occur.
With respect to the asterisked sample Nos. A15, A16, and A20, the average
particle diameter of ceramic grains was about 0.5 .mu.m or less. When the
average particle diameter of ceramic grains was decreased, the dielectric
constant was significantly lower than that of the above, and reliability
was also decreased. It has been confirmed that in a thin layer with a
device thickness of about 3 .mu.m or less, when the particle diameter of
ceramic grains is decreased, electrical characteristics are degraded.
In the sample No. A5, the average particle diameter of ceramic grains
constituting the ceramic layers was larger than the thickness of the
ceramic layers, and delamination occurred with a high percentage. In
contrast, in the sample No. A8, the average particle diameter of ceramic
grains constituting the ceramic layers was 3 .mu.m in the thickness
direction of the ceramic layers, which was the same as the thickness of
the ceramic layers, and the average particle diameter of ceramic grains in
the longitudinal direction was 5 .mu.m. As is shown in the sample No. A8,
even when the particle diameter of ceramic grains of the ceramic layers in
the longitudinal direction was large, if the particle diameter of ceramic
grains in the thickness direction of the ceramic layers is not more than
the thickness of the ceramic layers, delamination does not occur and
electrical characteristics are not degraded.
As is obvious from the results described above, when the ceramic layers
have a thickness of about 3 .mu.m or less, if the internal electrodes have
a thickness of about 0.2 to 0.7 .mu.m, the average particle diameter of
ceramic grains exceeds about 0.5 .mu.m, and the particle diameter of
ceramic grains in the thickness direction of the ceramic layers is smaller
than the thickness of the ceramic layers, delamination is prevented and
superior electrical characteristics are exhibited.
Next, characteristics of the nickel powder, in particular, the average
particle diameter, that can set the thickness of the internal electrodes
at about 0.2 to 0.7 .mu.m without any problems, will be described. In the
sample No. A21, the average particle diameter of the nickel powder was 250
nm, the coverage was decreased and reliability was decreased. In the
sample No. A13, the average particle diameter of the nickel powder was 5
nm, the coverage was decreased and delamination occurred slightly.
In contrast, in the sample Nos. A6 to A12 and A22 to A25, by setting the
average particle diameter of the nickel powder in the range from about 10
to 200 nm, a decrease in coverage was reduced and superior reliability was
obtained.
Next, the average particle diameter of the ceramic raw material powder for
forming the ceramic layers before firing will be described. In the sample
Nos. A12 and A25, the average particle diameter of the ceramic raw
material powder was 300 nm and the coverage as well as reliability was
decreased. In the sample Nos. A13, A14, and A22, the average particle
diameter of the ceramic raw material powder was 15 nm and delamination was
slightly observed.
In contrast, in the sample Nos. A6 to A11, A23, and A24, by setting the
average particle diameter of the ceramic raw material powder in the range
from about 25 to 250 nm, no delamination occurred and superior dielectric
properties were exhibited.
Furthermore, the ceramic grains constituting the ceramic layers of the
monolithic ceramic capacitors were observed by a transmission electron
microscope, and analysis was carried out. The ceramic constituting the
ceramic layers was pulverized and X-ray powder diffraction analysis was
carried out. The resultant diffraction patterns were analyzed by the
Rietveld method and the crystalline phase was identified. As a result, it
was confined that each of the ceramic grains had a uniform composition and
a uniform crystal system, and the individual ceramic grains had the same
composition and the same crystal system.
EXAMPLE 2
First, as ceramic raw material powders, the barium titanate-based raw
material composition shown in Table 5 was prepared by wet synthesis. That
is, solutions of BaCl.sub.2, SrCl.sub.2, CaCl.sub.2, MgCl.sub.2 and
CeCl.sub.3 were mixed, sodium carbonate (Na.sub.2 CO.sub.3) w | | |
