|
|
|
|
Back
to AlN filler Page 
Back
to Filler-property’ Contribution to Rheology
“Research Outline
(Enumeration of Current Projects)”(1) Research Interests “Focal area of investigation is
spherical AlN filler, semiconductor package, composite particle”
[ Contents ]
[1.1
Next Generations of High Density Semiconductor Integrated Circuit]
[1.2
Semiconductor Packaging Technology]
[1.3
Need to fill the silica powder at more high-density]
[1.5
Need to develop the new filler powder with high thermal conductivity]
[1.6
Another byproduct of above researches “Understanding Filler-material Property”]
2. Our Another
Work for System In Package
[ Research Examples ]
Flame Synthesis of Aluminum Nitride Filler-powder
Functions of New
Aerosol Multilayered Gas Sensor
High-density semiconductor
integrated technology play a dominant role in recent information technology
(IT). Its improvement is still
important to enable the further miniaturization of personal digital assistance
(PDA) device. New integrated
devices, such as the System in Package, Multi Chip Module or Chip Sized
Package, are one of key-technology to make the high-density.
System In Package “Example of New High Density
Semiconductor Integrated Circuit”
In the high-fidelity VLSI electric devices (System In Package etc.), a “semiconductor packaging material” is one of
key-material. The material
encapsulates IC and other components (likely as resister and capacitor) to
disconnect electrically and reinforce mechanically, chemically. It was initially made from ceramics in
the 1960’s, and eventually from resin polymer composite filled with silica (SiO2)
filler-powder. The material
necessitates a high thermal conductivity, a low thermal expansion and a good moldability. Higher the packing content of filler,
higher the thermal conductivity (lower the thermal expansion); but then, the
moldability degrades. There is a
yearning demand, especially, “how to mold the
material into the new integrated devices”.
In the recent trend of miniaturization of PDA devices, the width to
encapsulate the packaging material becomes narrower than ever. However, the number of devices used in
the PDA is increasing. There is a yearning demand in high thermal conductivity. Theoretical thermal conductive
coefficients of SiO2 is larger than resin polymer, so then a
high-density packing of SiO2 filler is one of promising approach to
obtain it. Furthermore, there is
an environmentally conscious reason.
In the recent semiconductor integrated technology, Pb free bonding is an indispensable requirement. Sn-Ag system, a alternative for Pb
bonding, necessitates a high temperature for the treatment. It means a low reliability of products,
because of decreasing the Reflow property. High-density packing of SiO2 filler also
contributes the reliability at high temperature.
Semiconductor packaging material encapsulates IC chips etc., and necessitates a high thermal conductivity, a low thermal expansion and a good moldability. Higher the packing content of filler, higher the thermal conductivity (lower the thermal expansion); but then, the moldability degrades.
There is a keen interest to clarify the relevance of primary properties of silica particles onto rheological characteristics of the resin polymer composite filled with SiO2 particles. As a rule of thumb, a largish SiO2 primary particle size or a smallish specific surface area decreases the viscosity of the composite system, because the resin supplied on the particle surface per unit weight apparently increases. These works could empirically show some intrinsic SiO2 filler primary-properties for lowering viscosity; such as the broader size distribution, the lower viscosity. However, that was highly stereotyped, and there might be remained other intrinsic silica-properties which affected the viscosity, as below;.
(1) Previous works used several void fraction theory for the packaging materials with particulate fillers. However, it is reported those theories do not agree with new material systems, which uses smaller fillers. There was still limited on the decisive understanding of rheological characteristics solely from these SiO2 primary properties.
(2) Measurement of filler powder, such as particle size distribution, is very entry-level. Laser diffraction method is widely used because of its convenience for measurement. However, it can not detect the morphological information such as roundness of silica filler. The parameter should be important at new material systems, which uses smaller fillers.
(3) We are studying about a new parameter, “particulate
surface structural contribution”, such as fine-particle amounts
adhered onto a core-particle. It
was referred as representing the strategy for improvement of polymeric
materials, which was a mandatory-clause in resin composite systems filled with
SiO2 particles; not to mention ordinary ceramic materials. One of important differences in
ordinary ceramic materials and resin composite systems was as to the sintering
of constituent particles being carried out or not. Primary properties of raw powder could conceivably affect
rather in resin system characteristics directly, but they were backward
region. To pack densely the
commercialized raw powder in resin system, various particle treatments to make
more spherical shape are commonly applied for commercialized powder. The flame conditions, e.g.,
fuel/oxidizer ratio, flame temperature, gas flow rate and precursor
constituents, should affect the morphology. The treatment could also affect the particulate surface
structure. Thus far, there were
very little concerns on it, and few linkage studies to connect microscopically
the particulate surface structural contribution and rheological properties in
resin systems; although many reports were presented on particle size
distribution at seeing in broad perspective, likely as Hosfield model of the
distribution.
(4)
A cause of the lack of understanding
was presumably a deficiency of appropriate
observation method of this material system having the liquid
typed-matrix of epoxy resin.
Observing directly the SiO2 internal particle-aggregated
structures must be promising to consolidate the scientific foundation. Thus far, that was very few, although
some polymeric materials without silica filler, having much rather solidified
matrix, were investigated with transmission microscopy. Our previous papers were described the
principle of transparent microscopy using the photo-elasticity of resin polymer
covered around SiO2 agglomerates, which locally provided a stress
toward the resin distributed in the surrounding areas, and structured the
rearrangement of polymer molecules.
Also concerned its validity to clarify the particle-size distribution or
coupling treatment influence on rheological property.
Semiconductor packaging material encapsulates IC chips etc., and necessitates a high thermal conductivity, a low thermal expansion and a good moldability. Higher the packing content of filler, higher the thermal conductivity (lower the thermal expansion); but then, the moldability degrades.
“New filler-powder“ having a high thermal conductive coefficient and a very spherical morphology is one of highly promising approach to improve the semiconductor-packaging properties. Presently, a spherical SiO2 particle flame-fused natural quartz ingredient has provided for the most part of practical filler-powder, having a bimodal size distribution and a roughly 10μm in average size (Ogawa et al., 1990; Otsuka, 1993; Unger, 1994; Abe et al., 1996; Kitano, 1998), e.g., as shown in Fig. 1 ( Abe et al., 1996). AlN, initially, has received much interest in the print-circuit board as a substitute of conventional ceramic substrate. It has following advantages; (1) theoretical thermal conductive coefficients of AlN, metal Si, Al2O3 and SiO2 are roughly 300, 90, 20 and 2 Wm-1K-1 at 400K; (2) the thermal expansion coefficient of AlN is close to that of Si substrate, 5×10-6 K-1 (Sheppard, 1990; Nakajima, 1998). The common methods for commercial AlN powder are; (1) the direct-nitridation of metallic Al (Weimer et al., 1994; Nagai et al., 1997) and (2) the reduction-nitridation of Al2O3 with carbon reductant in the presence of nitrogen (Kuramoto et al.,, 1989; Komeya et al., 1993). Whereas, the direct-nitridation necessitates a crushing/milling procedure, and the resultant powder has an angular shape mostly, such as shown in Fig. 2 ( Nagai et al., 1997). The main target of commercially-circulated powder via reduction-nitridation is for the raw powder of ceramic sintered body, which has a monomodal size distribution and far often smaller in mean diameter than the practical SiO2 filler-powder (e.g., submicron). The process, furthermore, is an endoergic reaction and needs a pretty high heating temperature in powder preparation, around 1700 ℃. It is not certain to be possible to nitride the large-sized raw Al2O3 necessitated to get the large-sized filler-powder. Several attempts in new aerosol processes such as CVD, plasma and fluidized-bed, were suggested too as a new production route of AlN, but all in all, they were likely to be insufficient for filler-powder needs or economical powder supply (Hotta et al., 1994; Wakimura et al., 1995; Pratsinis et al., 1995).
Furthermore, the SiO2 filler-powder, practically, is produced by flame fusion thechnology, and the flame aerosol synthesis is an established industrial process bringing sizable profits for large-scale manufacture of spherical-shaped particle, also at a pigmentary titania, a fumed silica for optical fiber and a super-paramagnetic particle (McMillin et al., 1996; Pratsinis et al., 1996). AlN filler-powder via flame synthesis route is reasonable because of its simplification, spherical-shaped particle, large-scale manufacture and cost-effectiveness. It could be also put the hopes to use the pre-existing manufacturing apparatus and scientific foundation built at the SiO2 filler-powder. However, there were very little concerns on it except SiO2 thus far, although only a few patents/articles were reported such as the flame-fused δ-Al2O3 particle (Hiragushi et al., 1982; Ogawa et al., 1986), or the diamond synthesis via the imperfect-combustion flame, which meant an insufficient O2 ratio to LPG (liquefied petroleum gas) (Hirose, 1996). Conceivable reason for the little concern in flame synthesis AlN might be its obscurity in melting point or its apt tendency to make angular shape. Al2O3 also has an angular shape based on their crystalline phases, but very spherical α-Al2O3 was developed by one of gas-phase synthesis (Tanaka et al., 1997). Ease-to-make tendency of spherical shape, which was an essentiality of aerosol synthesis, was utilized in the preparation; AlN is just as anxious to do likewise.
My work concerned with a flame AlN aerosol
synthesis and its effect of flame temperature, which offers a new filler-powder
supply for resin polymer composite system filled with ceramic particles. It appears to be available the AlN
powder having suitable properties as filler-powder, i.e., a spherical morphology and a fairly
large aerosol particle size, roughly 10 μm in average size.
Flame Synthesis of Aluminum
Nitride Filler-powder
Epoxy resin composite filled with
SiO2 particles is ‘also’ one of widely attracted material
systems for recent electrical devices such as insulators, electrorheological
fluids, chemical mechanical polishing (planarization, CMP); i.e. the “filler” material field. There is a keen interest to clarify the
relevance of primary properties of silica particles onto rheological
characteristics of the composite system.
As a rule of thumb, a largish SiO2 primary particle size or a
smallish specific surface area decreases the viscosity of the composite system,
because the resin supplied on the particle surface per unit weight apparently
increases. These works could
empirically show some intrinsic SiO2 filler primary-properties for
lowering viscosity; such as the broader size distribution, the lower
viscosity. However, that was
highly stereotyped, and there might be remained other intrinsic
silica-properties which affected the viscosity. There was still limited on the decisive understanding of
rheological characteristics solely from these SiO2 primary
properties. Several theoretical
approaches were also attempted.
These theoretical works could explain the relevance of aggregation of
simplified particles likely as mono-modal spherical one, but that of
practically used raw powder was still unknown; which commercial had complicated
morphology and particle size distribution, much rather ellipse and
bimodal. Linkage study is
supposedly necessary for connecting the former empirical works of SiO2
primary properties and the later theoretical ones.
The high-fidelity VLSI electric devices (System In Package
etc.) integrates several device components on the solely one board, such as
capacitor, register, filter and sensor etc. A
SnO2 multilayered gas sensor was prepared by the aerosol
electrostatic process. The
multilayer consisted of an internal sub-sensor layer for CH4
detection and a second sub-layer for CO detection, which was coated on the
internal layer. The CH4
sensor had a membrane structure of entangled SnO2 whisker-shaped
particles. The CO sensor comprised
a dendrite structure of SnO2-Pd composite particles. High selectivity of CH4 and
CO was done based on a catalytic ability of exterior SnO2-Pd
composite layer, which worked as the filter for diffusion of CO.
Functions of New Aerosol Multilayered Gas Sensor
