Gary, I found the article. Here is some background on the metal matrix composite sector....
OPTIONS
Doug Fant tells how more powerful electronic devices create design problems
Doug Fant recently completed an extensive study of issues involved in the design of modern electronic
components - and some invesment opportunities.
Doug's family has over 17 patents related to semiconductor design, and he writes on technical and invesment
issues for the Lone Star Growth Investor. His writeup is as follows, and the interview can be accessed using
the button at the bottom of this page.
INVESTMENT OPPORTUNITIES IN ELECTRONIC MATERIALS
Consumer demand for smaller and lighter products with higher performance is driving electronic component
design and manufacturing in a number of ways.
Smaller, more-powerful computers, multifunction cellular phones and "smart" electronic devices are just some
examples of electronic devices impacting this trend. The move toward miniaturization and multiple features is in
turn spawning a generation of high-speed, high-power computer chips such as microprocessors, Application
Specific Integrated Circuits (ASICs), and voice and graphics chips to meet this consumer demand.
However, high-speed, high-power chips consume electricity and thus generate significant amounts of heat. Basic
laws of physics require package designers to consider the effects of this heat. Excess heat may threaten a chip's
reliability and even cause it to fail. For CMOS-integrated circuits for example, each 10-degree Celsius rise in
temperature above a chip's normal operating range doubles its failure rate.
* Most objects when heated expand in size
Additionally, the amount of heat removed is generally equal to the amount of electricity that the chip consumes in
watts. Thus the need for packaging or housing that conducts and removes the excess heat from the chips or
other electronic devices. However, high thermal conductivity is but one of the requirements of materials used in
electronic packaging.
In addition, most objects when heated expand in size. Thus it is important to try to match a packaging material
that expands similarly to the underlying chip or electronic device. Referred to as the Coefficient of Thermal
Expansion (CTE), the CTE of the packaging material should be compatible with the chip or device so as not to
induce mechanical stresses that could cause component failure.
Traditional high-thermal conductivity packaging material such as aluminum and copper have CTE values much
higher than the underlying integrated circuit devices. So the chip-package designers may also utilize an
intervening substrate that functions as the CTE interface, such as Al2O3. The substrate reduces stress between
the differing expansion rates of the chip device and copper or aluminum base or lid, but unfortunately also
suffers from high thermal resistivity, and indeed may also physically fail (crack) itself.
There are more CTE-compatible, low-expansion packaging materials, including nickel ferrous alloys or kovar.
These materials eliminate problems related to differing expansion rates, but they offer poor thermal conductivity.
Aluminum nitride also has excellent properties in this regard; however, it is difficult to manufacture in an
unalloyed state.
* Current Packaging
Current Packaging Solutions. The relentless, consumer-driven move to miniaturize electronic products gives rise
to a third issue as one shrinks product dimensions and size and creates higher- density systems. That is the
strength of the packaging material. Thus these three factors - thermal expansion, conductivity and materials
strength - impact evolving chip-packaging solutions and thermal management for related electronic components.
In response to the demand, electronic component manufacturers have developed a series of composite
materials such as copper-molybdenum, copper-tungsten, Silvar (a specialty product manufactured from
silver/invar composite material TM), aluminum silicon carbide (AlSiC) and nitral, an aluminum/aluminum nitride
composite. These items, when combined with advanced powdered metallurgy and infiltration manufacturing
techniques, allow manufacturers to custom manufacture heat-dissipation components for specific applications.
These metal matrix composites can be tailored to give designers properties that cannot be achieved using
homogeneous materials. For example, you cannot find a metal that would give you 160 w/mC thermal
conductivity along with a CTE of 8 ppm/C. Copper and aluminum are readily available and have good thermal
conductivity, but the CTE of copper is 16 ppm/C while aluminum's CTE is 24 ppm/C. The metal matrix
composites perform within similar low CTEs and high thermal- conductivity ranges and can be tailored to
specific applications. With the lower CTE values than copper or aluminum, many of these composite
heat-dissipation devices may also be directly attached to the electronic device itself.
Copper-Tungsten. Copper-tungsten composites currently dominate the metal matrix composite market,
commanding about 70 percent of total market share. Frank Polese, vice chair of SEMX Corporation, notes that
the benefits of copper-tungsten include low tooling costs, competitive pricing, ease of fabrication and the ability
to get the highest thermal conductivity at a given rate of expansion. In addition, copper-based composites
generally are more flexible and tougher than other metal matrix composites, giving them the ability to absorb
energy from impacts.
* AlSiC components are strong, possessing a mechanical strength of twice that of copper
AlSiC Components. At least a few companies, SEMX Corporation, Aavid Thermal Products Inc. and Ceramic
Process Systems, have developed various infiltration-manufacturing techniques that now allow SiC to be 'cast'
directly into an overall aluminum net shape of a proposed device. This is important since abrasive products such
as silicon are difficult to handle in a traditional press and sinter manufacturing process. Also, the newer
techniques give a product designer the ability to vary material composition anywhere within the product's
thermal structure - i.e., a more complex pinned or finned heat sink can be designed. In addition, devices are
manufactured with concomitantly lower tooling charges.
Tool life is extended in working with an abrasive material such as silicon and to a very tight location tolerance
where devices will be densely packed, for example, onto a circuit board. Lower tooling charges and
inexpensive base materials such as silicon and aluminum translate into cheap and efficient thermal-management
devices.
AlSiC components are strong, possessing a mechanical strength twice that of copper. Note, though, that AlSiC
components are stiffer and more brittle than copper-based composites (somewhat akin to ceramics),
characteristics that might not be desirable in thin components, for example. AlSiC also has a lower g/cm3
density rating than the copper-based composites- i.e., components are lighter, which may be a factor in certain
applications.
Thus, for example, plate thickness and product weight can be reduced and mounting features designed right into
the product without compromising the thermal performance of the device. This could be of use, for example, in
avionic and defense applications. For these reasons, using AlSiC composites for heat dissipation packages is
another metal matrix market that is growing rapidly.
Markets in Electronics for Metal Matrix Composite Technologies. The end users of metal matrix composite
materials are numerous and expanding rapidly. In the electronics arena, the devices are used as lids and carrier
plates for semiconductor chips, including microprocessors, ASICs, graphics and video chips. The latter includes
anywhere chips are used, such as computers, telecommunications, cars, washing machines and sensors. In
communications, composites are used as heat sinks in bases for wireless components and telecommunications
switches, as well as for the rapidly expanding point-to-point market. Additional significant markets include
satellite communications, motor controllers, microwave housings, avionics and defense industries. Indeed, one
manufacturer even heavily utilizes the technology in sporting equipment.
* Industry Participants
A typical microprocessor application would involve flip chips, a type of semiconductor chip that is directly
connected to the package substrate with the circuitry facing the substrate. That means that the back of the
device (without any circuitry) is exposed upward, and that is where the lid and heat sink are attached to
facilitate removal of unwanted heat. In some configurations and material combinations, it is important that this
microprocessor lid has a CTE that matches either the substrate or the silicon chip.
The substrate and the lid may be matched to prevent any unwanted stress on it from the lid due to CTE
mismatches. Then the potential CTE mismatch between the silicon and the lid will be taken into account with
low stress, die-attach compounds.
The composites may be sold to semiconductors chip makers, wireless equipment manufacturers, original
equipment makers, electronic contract manufacturers, motor control manufacturers from automotive to office
equipment control makers, traction drives, power supplies and electrically powered motion device
manufacturers - indeed, anywhere electrical energy is converted to rotating energy (torque) or where "smart"
electronic devices are employed.
Industry Participants. There are a number companies active in the newest wave of net shape/infiltration
manufacturing techniques for metal matrix composites; SEMX Corporation (SEMX), Aavid Thermal
Technologies Inc. (AATT), Ceramic Process Systems (CPSX), Hitachi Corporation by technology license from
CPSX, and Ametek Corporation (AME).
Ametek Corporation and SEMX Corporation have the broadest material offerings in the low-expansion
substrate business. Aavid Thermal Technologies is active particularly in the motor controller and power supplies
elements of this manufacturing sector. Ceramic Process Systems supplies the defense industries inter alia. And
of course all of the companies support semiconductor and telecommunications applications.
The growth potential in these markets appears significant. SEMX and CPSX for example reported in their most
recent SEC Forms 10Qs that sales from their specialty electronics divisions rose anywhere from 30-86 percent
year-over-year from 1998 to 1999, depending upon whether you utilize the first three or six months of the year
for comparison.
Has the market recognized the apparent potential of this sector? I would say "not yet"; however, one recent
transaction suggests that at least some are beginning to see the economic potential here. Specifically, I refer to
the recent private buyout of Aavid Thermal Technologies Inc. The AATT situation actually involves two
interesting recent events.
First, Willis Stein and Partners, a private capital-management firm with $1.2-billion U.S. equity capital under
management, and which specializes in manufacturing, telecommunications and technology investments, agreed in
August to buy 100 percent of AATT's shares and take AATT private. Interestingly, AATT prior to the Willis
Stein agreement AATT was also itself in the process of acquiring the thermal management division of
Bowthorpe PLC, a British Corporation. Willis Stein's acquisition price for AATT is actually $1 U.S.-per-share
higher of AATT, depending upon AATT's successful completion of the Bowthorpe Thermal Management
Division's acquisition.
Perhaps this indicates the value of this manufacturing sector moving forward.(Recorded 10/99)
Start Audio
¸ Copyright, AudioInvestor.Com, 1999, All Rights Reserved |
