Overview:
Perhaps the most effective platform for
knowledge transfer and intellectual exchange is through sharing of facilities.
Special facilities built to support STC programs are listed in Table
V. Approximately $4MM in NSF funds have been invested in unique equipment and
laboratories, including funds provided by NSF, the participating universities;
the Kenan Institute for Engineering, Science, & Technology; and facilities
shared with the Kenan Center. We are making make every effort to publicize the
availability of these facilities for collaboration (as well as contract work,
charged on an out-of-pocket basis). We consider these facilities to be a
national resource and encourage their use as such.
 |
The desired collaborative working environment for STC and associated personnel |
|
State-of-the-art equipment/facilities needed by STC personnel to achieve their goals |
|
Unique facilities to the scientific community at large primarily for collaborative research |
In addition to the $4MM in facilities funded by the NSF approximately $3MM in capital and operating funds have been contributed by the University of NC for the Dry Fab of the Future (DF2DF). This facility will become fully operational during the first quarter of 2004.
Original Shared Experimental FacilitiesIn support of initial goals, in 1999 CERSP funded five Shared Experimental Facilities:
•
NMR
Laboratory for the Study of Compressible Media
•
Colloid,
Interfacial Science and Light Scattering Laboratory
•
Reaction
Kinetics Laboratory
•
Computational
Facilities
•
Distributed
Collaboratories
These
facilities, operational for several years, and their uses are described in the
following section.
1.
NMR Laboratory for the Study of Compressible Media
NMR
is the premier analytical tool for the study of supramolecular systems: for
identification of
molecules,
determination of molecular structure and transport properties of selected
species in mixtures,
and
characterization of inter-molecular interactions. We concluded that the most
effective use of Center
funds
was to acquire a high field spectrometer (600 MHz - standard bore) with an array
of special purpose
probes.
A committee decided upon a three-channel spectrometer having four probes having
various
function-alities.
Special large diameter sample vessels of PEEK resin have been developed for
routine use in order to achieve the necessary sensitivity in high-pressure CO2
solutions. The NMR laboratory, located at UNC-CH, has two permanent staff
members plus part time assistants to insure that all nine spectrometers
including the 600 MHz system are operational and are available. The labora-tory
manager has trained graduate students to use the 600 MHz spectrometer as well as
lower field spectrometers. Hourly
use charges are the same for all machines and are based upon pro-rated operating
costs for the entire laboratory. Equipment has been available for routine use
for two years and is available to outside users. The 600 MHz NMR facility is
heavily used for polymer studies, primarily by members of the STC. To date more
than a dozen scientists and students within the STC and several from outside the
STC have used this equipment. This facility has served its purpose and has been
instrumental in numerous papers as well as contributing to at least 10 PhD
theses.
2.
Colloid, Interfacial Science and Light Scattering Laboratory
Five
major pieces of equipment comprise this facility, located primarily at
UT-Austin: (1) High-pressure
Pendant
Drop Interfacial Tensiometer, (2) Static and Dynamic Fluorescence Spectrometer
(3) Low Angle
Dynamic
Light Scattering System. (4) Small Angle X-Ray Scattering (SAXS) Facility and
(5) High Pressure
In
Situ Ellipsometer.
In addition, a 6000 psig static and dynamic light scattering facility was built
at UNC-CH.
Facilities
are available to researchers outside the Center, though to
date
most use has been internal. At least 15 STC faculty, post-docs and students have
used these facilities
extensively,
plus several visiting scientists. At least 20 papers, presentations and theses
have resulted
from
work done using these facilities.
3.
The Reaction Kinetics Laboratory (RKL) At
NC State we have designed and built continuous and batch reactors including
ancillary support
equipment,
feed and recovery systems and on-line monitors. In addition, RKL includes
specialized off-line
equipment
at NCA&T. Over a dozen faculty members, students and post-docs have used
this equipment.
The
continuous reactor at NCSU has been used to study polymerization of vinylidene
fluoride and acrylic
acid.
Five students and post-docs have utilized the facility, including one receiving
industrial support. This
reactor
has been used not only to study kinetics but also to produce polymer in great
enough quantities to
allow
determination of molecular weight distributions. The batch reactor is
essentially a reaction
calori-meter.
Three students have used it to study kinetics of emulsion polymerization of
methyl
meth-acrylate
and acrylic acid. The off-line equipment installed at NCA&T includes: (1) A
high-pressure
vapor/liquid
equilibrium (VLE) apparatus, (2) an isothermal micro-calorimeter, (3) a
differential scanning calorimeter and (4) a mass spectrometer/gas chromatograph.
All equipment is available for CERSP investigators with pre-scheduling and
provision of personnel, necessary chemicals and supplies for measurements.
During the renewal we will enhance our capabilities in this shared Experimental
Facility through the design and purchase of the equipment necessary to make
“emul-sion polymers” using the new CCSSFMP being developed by DeSimone
and George Roberts.
4.
Computational Facilities The
Beowulf Super-Computer Cluster at UNC-CH provides a computer platform to perform
simulations
of
large systems over long length scales, both of which are necessary to simulate
micellar and polymer melt
systems.
The Cluster at UNC-CH consists of 17 Compaq DS10 Alpha server workstations, each
powered
by
the Alpha 21264 processor and of 33 Dell Precision 330 based on Pentium-4
processor. All computers
are
connected into a local area network via 100Mb/s ethernet switches. The Cluster
became operational in
June
2000 and was further extended in September 2001. The Cluster is used is to carry
out simulations of
micelles,
charged polymers and dense polymer systems, which require an exceptional
compu-tational
power.
It is connected to the Internet via dedicated PC, acting as a gateway, making it
possible to submit jobs and acquire data from any networked PC. To date it has
served about ten researchers. At NCSU a 12-node Beowulf cluster was installed,
consisting of six separate machines, each supporting two 667 Mhz
alpha
processors with 4MB of cache on each processor. Each of these six machines has
512MB of RAM, a
14.4GB
hard disk drive, and 100MBit ethernet for data communication. A separate RAID
array (108GB
capacity)
is used for data storage, supplemented by a Sony DDS-4 DAT drive for data
backup. The Beowulf
system
can be used for both parallel and serial calculations. The primary function of
this facility is to carry
out
large-scale molecular simulations of CO2 solvent systems, including
micellar solutions. It has supported
ten
researchers.
5.
Distributed Collaboratories All
five universities provide videoconference (VC) centers easily accessible to
CERSP personnel.
Rather
than create new VC facilities, we have been collaborating with University staff
to support our weekly
CERSP
video meetings using existing facilities. We have upgraded these facilities to
better meet general
educational
needs, e.g., both our meetings and distance education classes at UNC-CH benefit
from larger
video
displays that we purchased. SmartBoards enable us to broadcast seminars outside
CERSP via the
Internet
when appropriate. To facilitate small group videoconference meetings between
universities, we
purchased
VCON Vigo Executive teleconferencing units. These work in conjunction with a
laptop,
SmartTech
Smart-Board Overlay, NEC 42. plasma display and a conference phone (which we
purchased) to
provide
an interactive environment in which students and faculty can share ideas through
audio, video and
an
interactive white board. UNC and UT-A mounted the plasma display and Smartboard
overlay on the wall of a conference room, with a swivel arm for ease of use. NCSU
and NCA&T mounted the equipment on a cart for use in multiple rooms. Staff
from each university operates specialized facilities, coordinated by an STC
technician. We are beginning to
investigate the utility and effectiveness of wireless mobile collaboration
technology in support of scientific work. Using supplemental funds we are
purchasing state-of-the-art PDA communication devices and services that provide
mobile access to e-mail, documents, scientific data and the Internet. We believe
it is important to investigate how new information and communications technology
may benefit scientific collaboration. These will be initially provided to the
management team and will later be made more generally available if they prove
effective.
Summary
of Key Capital Equipment Funded by NSF
Additional
NSF STCERSP Facilities
Several significant capital investments in facilities other than those
officially designated as “Shared Experimental Facilities”.
These, of course, are also available for collaboration both within and
outside the CERSP:·
Three laboratories (NCA&T, NCSU, and UT-A) for
evaluating membrane separations of CO2 are described in detail in the
“Accomplishments” section of Thrust Area C
·
An isothermal microcalorimeter with environmental
chamber, on-lone GC system and VLE cell with magnetic stirrer system has been
ordered for installation at NCA&T.
·
A polarization microscope at UNC-CH, suitable for
liquid studies in CO2
·
An x-band microwave system for enhanced EPR studies
Dry
FAB of the Future Demonstration Facility (DF2DF)
During
the second five-year funding cycle, the University of North Carolina—both
UNC-CH and NCSU—have funded an additional Shared Experimental Facility called
the "Dry FAB of the Future" Demonstration Facility (DF2DF).
CO2-based
processes are now listed on the International Technology Roadmap for
Semi-conductors (ITRS) as a stated process need of end-user customers for BEOL
cleaning operations. Of the approximate 500 steps required to manufacture a
state-of-the-art IC, about 25% of these are cleaning steps and a subset of these
are being targeted for use with CO2. This is in line with the
industry-accepted method for introduction of a new technology one to two nodes
into the future (implementation into production in 2004 - 2006). Discussions
with leading OEMs for the micro-electronics industry continue to reaffirm the
potential for broad acceptance of CO2 for cleaning in the FAB. Our
External Advisory Board members credit the CERSP for bringing CO2 to
the forefront of realistic consideration in the microelectronics industry. We
believe that the ITRS targeted imple-mentation of CO2-based cleaning
processes in 2004 - 2006 will be the foundation upon which CO2 will
be seriously considered for use in additional unit operations necessary for
manufacture of devices, such as in lithography, materials deposition and
chemical mechanical planarization. CERSP plans to play a leading role to make
that happen.
In
order to further the implementation of CO2-based techniques in
microelectronics and more importantly to observe first hand the technical and
scientific challenges associated with implementation, we will establish the
"Dry FAB of the Future" as a demonstration facility. Both liquid and
supercritical CO2 have significant potential for replacing current
aqueous and organic solvent based processes in use in FABs today. We consider CO2-based
processes as “dry” processes, since at standard temperature and pressure CO2
is a gas. Therefore CO2 is amenable for use in tools/chambers that
can be easily evacuated and interfaced with vacuum cluster tools. To achieve our
stated technical and economic goals, CERSP will seek additional new partnerships
with the private sector, through the assembly of a critical mass of integrated
CO2-based wafer research and development facilities. DF2DF
will present an excellent opportunity for industry, providing access to
state-of-the-art, "quick-turn-around" equipment and prototyping
infrastructure. It will enable industrial partners, in close collaboration with
STC researchers, to perform increasingly difficult "proof-of-concept"
and first stage fabrication runs using industry-compliant wafers and processes.
This capability ensures that the "dry" CO2 technologies
developed within our Center are manufacturing-worthy and cost-effective, thus
assuring seamless transition to the high-volume manufacturing stage after the
necessary and difficult demonstration at the feasibility stage.
DF2DF
will initially consist of lithography, cleaning/drying/stripping, chemical
mechanical planarization (CMP), and surface analysis tools uniquely designed to
operate using highly formulated CO2–based processes.
1.
Lithography UNC-CH
and NCSU have recently purchased an ASML 5000/900 series 193 nm scanner with
0.63 numerical aperture (NA) capable of generating 130 nm high resolution images
over a 2 cm x 2 cm area with precise alignment. This tool will be housed in the
Plasticization
of photoresists with CO2 can be used to densify resists materials, an
alternative to soft baking as a method to control acid diffusivity and, hence,
line edge roughness (LER). In addition, we can take advantage of the excellent
wetting properties and very low viscosity of liquid and supercritical CO2
to deposit, develop and strip photoresists—what we call as totally “dry”
lithography. For example, in lithographic applications, novel 193-nm resists
have been designed with novel photo-acid generators that can be used in totally
“dry” lithography where deposition, development and stripping steps are all
done using CO2. Even polymers like poly-styrene that are not soluble
in liquid or supercritical CO2 undergo swelling. Decreased glass
transition temperature and viscosity result. We will investigate the effect of
this plasticization with CO2 on line edge roughness. Analogous to
soft baking, this mild method can result in a densifi-cation of the resist. By
this, acid diffusivity can be controlled and optimized in order to minimize LER.
With different pressure/temperature let-down schemes, one can also decrease the
density of a resist, which will also influence resist performance.
2.
Cleaning/drying/stripping and SurfaceAnalysis Cleaning
with supercritical CO2 will
happen
commercially in 2004/2005. A number
of
key
companies are spending millions to bring the
super-critical
fluid technology into commercial reality,
including
TEL, MICELL, BOC, Air Products, DNS,
Kobe
Steel,
3.
Chemical Mechanical Planarization (CMP) CMP
is a rapidly growing process used in the microelectronics industry to make metal
and dielectric layers on silicon substrates smooth and defect free with vertical
dimension control. Most current processes use water as the solvent for CMP
slurry, leading to both technical and environmental difficulties. One
particular technical draw-back of the aqueous-based CMP processes is the
incompatibility of porous low-k inorganic and organic interlayer dielectric
materials with water. There exists a
great demand for new CMP tech-nologies that circumvent the technical and
environmental drawbacks of the current aqueous and chlorinated organic solvents.
In this regard we are pioneering the
effort to establish a “dry” CMP process based on condensed CO2.
The development of a “dry” CO2-based process for CMP will require
CO2-based etchants, abrasives, oxidants and an integrated tool and
process capable of handling liquid and supercritical CO2. As we study
the fundamental chemical processes in CO2, we plan to design and
build the world’s first CO2-based CMP tool as part of the DF2DF
suite of capabilities.
These facilites are contained within the Triangle
National Lithography Center at NC State University.
To access the DF2DF
for experimental work, contact
Prof. Carl Osburn,
NCSU.