|
| |
| -
2004 - |
| Jan. 1, 2004 |
No TeleSeminar--HAPPY NEW YEAR! |
| Jan. 8 |
Host: Gary
Rubloff, University of Maryland
Presentation by: Laurent Henn-Lecordier, Department of
Materials Science and Engineering / Institute for Systems Research, University of Maryland
Topic: "Monitoring
and Control of Binary Gas Mixtures from Solid Phase MOCVD Sources using an
Acoustic Sensor"
Abstract:
The delivery
of reactant species for an increasing variety of new materials is
important to Si ULSI, wide-bandgap semiconductors, and other areas. ESH benefits from improved delivery systems (e.g., for MOCVD
or PECVD) can be anticipated in at least two areas: (1) advanced process
control, where higher yields and equipment effectiveness convey ESH
rewards; and (2) greater flexibility in chemical process design, where a
wider variety of precursors meet manufacturability constraints. We have
developed and demonstrated significant advances in controlling the
delivery of low vapor pressure sources for CVD-based processes.
In-line acoustic sensors have been used for several years in MOCVD source
delivery systems to monitor and control the upstream composition of binary
gas mixtures obtained from temperature- and pressure-controlled
“bubbler” vessels. Since the vapor pressures of some commonly used
MOCVD solid sources is low, extending into the sub-Torr range, it becomes
difficult to maintain a constant – but minute - concentration of reagent
in a high flow of carrier gas. In addition, aging effects in the source
severely affect its stability resulting in low reproducibility of the
process and film properties. These effects have been minimized by
replacing the source at an early stage, typically before half its normal
lifetime.
In this study, an Inficon Composer acoustic sensor was
implemented to measure and control the concentration obtained from two
solid phase sources using H2 as a carrier gas, including (1)
trimethylindium (TMI), which is used to grow GaInAs III-V compound
semiconductors for optoelectronics, and (2) bis(cyclopentadienyl)
magnesium (Cp2Mg) which is used in part as a p-type doping
element in nitride-based compound semiconductors for blue LED’s.
Both sources are crystalline solids with low vapor pressures (2.5
and 0.04 Torr at 25°C
respectively for TMI and Cp2Mg.
Using the acoustic sensor, reagent levels as low as 1 ppm were monitored
and found in close correlation with the expected concentrations over a
broad range of total pressure from 500 down to 60 Torr. Closed loop
process control was implemented to maintain the TMI and Cp2Mg
composition on target in the presence of long term temperature drifts.
Despite induced variations of the precursor vapor pressure up to 50%, the
upstream composition remained stable within ±0.15% for TMI (at 0.5 mol.%
set point) and ±0.3% for Cp2Mg (at 0.04 mol.% set point).
Induced short term disturbances were also significantly minimized and the
effect of downstream condensation was put in evidence. Potential benefits
resulting from the implementation of this sensor-based control methodology
will be discussed, such as the improved long term process reproducibility
required for high performance devices and the increase of the solid source
usable lifetime. [L. Henn-Lecordier, J.N. Kidder, G. W. Rubloff, Department of
Materials Science and Engineering/Institute for Systems Research, University of Maryland] (PDF)
|
| Jan. 15 |
Host: Kim Ogden,
University of Arizona
Presentation by: Kim Ogden, University of Arizona
Topic: "Survival and Adaptation of Bacteria in Ultrapure Water
Systems"
Abstract: Bacteria isolated from ultrapure water (UPW) systems
were examined for their ability to survive in UPW. The ultimate goal is to
elucidate potential carbon and energy sources for the bacteria and
determine how to destroy them. Two strains of Ralstonia
pickettii isolated from different areas within the UPW system
(pretreatment and polishing loop, and referred to as strains 3A1 and
MF254A, respectively) and a strain of Bradyrhizobium
sp. were compared to increase our understanding of the fundamental
behavior of bacteria contaminating UPW. R.
picketti MF254A was capable of cryptic growth if the cells supplied as
nutrients were heat-killed, however, R.
picketti 3A1 and Bradyrhizobium
sp. were not. If the cells supplied as nutrients were killed using UV254
light, cryptic growth was not observed. However, cells treated with UV254
light typically adapt to the UV light and thus are harder to destroy.
Preliminary protein electrophoresis results indicate that bacteria
in UPW “turn on” 2 to 3 proteins. Identification of these proteins is
underway. SEM analysis shows
that the bacteria do encapsulate in Ge surfaces. (PDF)
|
| Jan. 22 |
Host: Karen
Gleason, Massachusetts Institute of Technology
Presentation by: E. Todd Ryan, Advanced Micro Devices, Inc.
Topic: "Integration of Nanoporous Materials into Advanced
Microprocessors"
Abstract:This
talk will provide a brief overview of back-end-of-line processing with
copper interconnects and a survey of integration challenges for nanoporous
materials. The talk will specifically focus on how the plasmas used
for integrated circuit fabrication alter and damage nanoporous
organosilicate glass (OSG) films and how to deposit an effective metal
barrier onto nanoporous films. Effectively monitoring integration
damage and metal interdiffusion, preventing these problems, or repairing
damage are necessary to successfully integrate nanoporous OSG films.
Thus it is critical to understand the properties of nanoporous films
and how they are altered by the integrated circuit fabrication process. (PDF) |
| Jan. 29 |
Host: Anthony Muscat,
University of Arizona
Presentation by: Eniko T. Enikov, Aerospace and Mechanical
Engineering, University of Arizona
Topic: Design, Analysis and Fabrication of MEMS Thermal
Micro-Actuators for Tactile Displays and Switches
Abstract: Modern integrated circuits (ICs) contain more than 300
million transistors per square centimeter. Such Very Large Scale
Integration (VLSI) of electronic elements has resulted in a tremendous
technological progress. The origins of this phenomenal progress can be
traced back to the invention of integrated circuits in 1959 by Jack Kilby
(Texas Instruments) and later improved by Robert Noyce into a planar
technology. The great success story of integrated circuits has sparked
interest in developing other highly integrated systems based on
semiconductor processing technology. The resulting technology, now known
as micro-electromechanical systems (MEMS), includes many useful devices
based on IC processing technology and integrates electrical, mechanical,
optical, chemical, and/or biological processes. While many MEMS devices
have permanently entered the commercial world in the form of automotive
pressure sensors, accelerometers or printer heads, the quest for more
efficient and powerful micro-actuators continues.
In part one of this talk our work on the development of thermal
micro-actuators will be presented Potential applications of these devices
are micro-relays, tunable impedance RF networks, and miniature medical
instrumentation. The specific motivating application in our case is a
tactile display, which utilizes a hybrid actuation a macroscopic vibrating
piezoelectric plate combined with a MEMS array of thermally actuated
micro-mechanical switches. Several fabrication processes will be discussed
including an SU8-based electroforming, SU8/metal composite and purely
metallic devices. Non-traditional substrates such as RF-printed circuit
boards have also been successfully utilized to fabricate the devices. The
PCB-based devices exhibited similar characteristics, thus opening the
possibility of integrating RF MEMS directly on PCBs. The actuators were
benchmarked with respect to power consumption, stroke, and response time.
The fabricated nickel actuators are shown to be robust with displacements
in the range of 76 micrometers using 80 mW of power. Actual cooling
transients were captured using a two-step constant-current excitation
method. It is further demonstrated through analytical models that the
thermal cooling times limit the bandwidth of these devices below 1KHz.
(PDF) |
| Feb. 5 |
Host: Krishna
Saraswat, Stanford University
Presentation by: Chi
On Chui,
Department of Electrical Engineering,
Stanford University
Topic: "Novel
Germanium Technology and Devices for High Performance MOSFETs"
Abstract: The
saturation of Si MOSFET drain current upon dimension shrinkage may limit
the prospect of future scaling. The lower effective mass and lower valley
degeneracy of Ge could alleviate the problem by providing a higher source
injection velocity. However, surface passivation for gate insulator and
field isolation, and n-type dopant incorporation are the two classic
problems that obstruct CMOS device realization in Ge for four decades. In
this talk, we will present various novel Ge technologies on surface
cleaning, gate dielectric, and dopant incorporation. In addition, we
will disclose an innovative self-aligned gate-last fabrication process
not only to demonstrate functional Ge MOSFETs, but also to provide a
vehicle to characterize many novel material integration schemes.
(PDF)
|
| Feb. 12 |
Host: Anthony Muscat,
University of Arizona
Presentation by: James Watkins, Associate Professor, Department
of Chemical Engineering, University of Massachusetts
Topic: Supercritical Fluid Technology for Semiconductor Device
Fabrication Deposition of Metals and Mesoporous Silicates from Carbon
Dioxide
Abstract: Supercritical fluids (SCFs) offer a unique technology
platform for semiconductor devices. While the environmental advantages of
SCFs such as carbon dioxide are attractive, it is their physicochemical
properties that are enabling and may ultimately drive their use in device
fabrication. In particular the absence of surface tension, favorable
transport properties and densities that approach those of liquids provide
a means for solution-based processes in an environment that behaves much
like a gas. These attributes are ideally suited for executing materials
chemistries within the smallest device features. This talk will describe
applications of SCFs developed in our laboratories relevant to the
fabrication of next-generation interconnect structures including the
deposition of Cu, Co and other pure and doped metal films and the
preparation of well-ordered, mesoporous, ultra-low dielectric constant
films that are sufficiently robust to survive CMP.
As device dimensions decrease below 45 nm, the deposition of high purity
Cu and alternative metal barrier systems within high aspect ratio features
becomes a significant challenge. We have found that the reduction of Cu(II)
or Cu(I) precursors with H2 or alcohol yields remarkably pure, conformal
films with resistivities as low as 2.0 microohm-cm. Excellent adhesion was
achieved through the use of surface pre-treatments. The approach is
suitable for the single step fill of narrow features or the preparation of
seed layers for subsequent plating. Deposition in CO2 is not limited to
Cu, but can be extended to other pure and doped thin metal films,
including cobalt, for Cu line capping applications. In all cases precursor
and reagent volatility constraints, which are often limiting in CVD, are
eliminated through the use of SCFs.
Reduced device dimensions will also place greater demands on ILDs,
requiring the development of robust, mesoporous films that can be extended
to dielectric constants well below 2.5. Much of this discussion will focus
on a new approach to mesoporous silicates that involves the infusion and
selective condensation of metal oxide precursors within one phase domain
of a highly ordered, preformed block copolymer template dilated with
supercritical carbon dioxide. The template is then removed to produce the
mesoporous oxide. To date we have replicated ordered spherical and
cylindrical morphologies to yield silica, organosilicate and mixed silica/organosilicate
mesostructures in films over 1 micron thick while maintaining all the
structural details of the sacrificial copolymer template. Varying
precursor loading and selection produced a first-generation family of
films with dielectric constants as low as 1.8. A film with k = 2.2 was
selected for further evaluation and found to survive CMP in a planar test
stack. One advantage of the process is the elimination of excess alcohol
from the reaction media, which provides a pathway for rapid and high
degrees of network condensation. Ultimately, structure on both the local
and device levels can be achieved in three dimensions wholly in the
polymer template using established techniques prior to infusion of the
inorganic phase.
(PDF)
|
| Feb. 19 |
No
TeleSeminar
|
| Feb. 26 |
No TeleSeminar -- 8th
Annual ERC Site Review Meeting (February 25-26-27, 2004 in Tucson, AZ) |
| March 4 |
No TeleSeminar |
| March 11 |
Host: Paul
McIntyre, Stanford University
Presentation by: Piero Pianetta, EE & SSRL,
Stanford University
Topic: "Spectroscopic Determination of Work Functions: Possible
Application to Silicon Gate Technologies"
Abstract: The work function is a
fundamental property of a material which determines the details of electron
emission from surfaces, the characteristics of electrochemical reactions as
well as determining the band offsets between different semiconductors. In
this talk, methods of measuring work functions of surfaces such as
photoemission and Kelvin probe techniques will be discussed as well as their
applicability to interfaces.
(PDF) |
| March 18 |
Host: Charles
Musgrave, Stanford University
Presentation by: Charles Musgrave, Stanford University
Topic:
The Chemistry
of Atomic Layer Deposition of High-K Dielectrics: Selection of Precursors,
and Substrates
Abstract: Atomic
layer deposition is a deposition process capable of depositing uniform and
conformal ultra thin films over large areas and has been proposed as a
possible technology for future integrated circuit processing. We have used
quantum chemistry to predict the atomistic mechanisms of the atomic layer
deposition of ZrO2, and HfO2, with various precursors and substrates. The
resulting energetics are used to evaluate the viability of processes
involving different ALD precursors, surface functional groups, substrate
choices, area-selective processes and to provide a fundamental basis for the
design of new ALD chemical processes that are environmentally benign.
(PDF) |
| March 25 |
Host: Farhang Shadman, University of Arizona
Presentation by: Terrence McManus, Intel Fellow, Director of
EHS Technologies, Intel Corporation
Topic: "Design for the Environment at Intel" (PDF) |
| April 1 |
Host: Ara
Philipossian, University of Arizona
Presentation by: Professor
Ed Paul, Stockton College, Pomona NJ
Topic: "A Model of Chemical Mechanical Polishing"
Abstract: In chemical mechanical polishing, the material
removal rate depends on many variables. In this talk, a first principles
model will be discussed which explains how the removal rate varies with
changes in oxidizer and inhibitor concentrations, abrasive diameter and
abrasive loading, polishing pressures and speeds, and pad stiffness. The
results will be applied to data for tungsten CMP, explaining why the removal
rate initially increases and then approaches a maximum value as
concentrations of oxidizer or abrasive increase, and as mechanical pressures
and speeds increase. The model has been found useful in academic
descriptions and industrial applications of CMP processes. (PDF) |
| April 8 |
Host: Paul
Blowers, University of Arizona
Presentation by: Paul Blowers and
Monica Titus, Department of Chemical and Environmental Engineering,
The University
of Arizona
Topic: "The Use of Life
Cycle Assessment as a Screening Tool for Environmental Performance:
Supercritical Carbon Dioxide Use in Wafer Rinsing"
Abstract: Supercritical carbon dioxide (scCO2) is
attractive as a replacement for many solvents because it is cheap,
non-toxic, and easy to use. While scCO2 has been used in food
processing and for rinsing machined parts, researchers have recently
developed the technology to implement scCO2 as a solvent in
semiconductor manufacturing to remove photoresist from wafers during
processing. This use replaces an ultrapure water (UPW) rinsing step that
can use millions of gallons of water a year and will have a direct impact on
water usage throughout semiconductor manufacturing communities.
We use Life Cycle Inventory (LCI) to quantify the environmental benefits and
costs of implementing the new technology through material and energy
balances. This data can be used to justify additional improvements in
processing technology associated with the change. We find that the new
technology has a smaller energetic footprint and consumes less mass of raw
materials compared to UPW water rinsing. The purification steps necessary
before and after using water lead to larger energy costs. On the other
hand, scCO2 usage allows recycling of both scCO2 and,
possibly, the necessary cosolvent without incurring large separation costs.
Overall, energy usage for compressing and heating the carbon dioxide into
the supercritical region cause the largest environmental impacts for this
process. Pilot testing of the technology is ongoing and should be
accelerated in order to achieve a large environmental savings in more
facilities. (PDF) |
| April 15 |
Host: Anthony Muscat,
University of Arizona
Presentation by: Professor Neal R. Armstrong, Department of
Chemistry, University of Arizona
Topic: "Disks to Rods:
Nanometer-Scale Characterization and Control of Self-Organizing Molecular
Systems of Interest for Emerging (Hopefully) Electronic and Energy
Conversion Technologies"
Abstract: There
is substantial interest in new molecular electronic materials which would
(inexpensively) provide for logic circuits, displays and solar cells, on
flexible substrates -- "made by the mile" and "sold by the inch" (Thanks to
3M for the quote). There are SIGNIFICANT challenges, however, to creating
electronic quality materials from organic assemblies, not least of which is
that charge mobilities are strongly affected by the microscopic
(nanometer-scale) organization of the molecules within these materials.
This talk will focus on some of our recent efforts to create a new class of
discotic mesophase (liquid crystalline) materials, which might be solution processable
into transistor and/or solar cell structures. The individual molecules are
disks, which self-organize into rod-like aggregates, with coherence lengths
of up to ca. 300 nm. I'll show some of our very preliminary FET studies of
thin films of these materials, and prognosticate on what it will take to
turn materials like these into viable technologies. (PDF) |
| April 22 |
Host: Steve Beaudoin,
Purdue University
Presentation by: Gautam Kumar, School of Chemical
Engineering, Purdue University
Topic: "Validation of a Model for Undercut Etch Cleaning"
Abstract: Undercut removal of
micron-scale particles was modeled on the basis of a force balance on the
adhering particles. The adhesion between particles and surfaces was
predicted based on the particle and substrate morphology, composition, and
mechanical properties, as well as the composition of the cleaning medium.
The adhesive force was computed by summing the van der Waals and
electrostatic forces over the area of interaction between the particle and
the substrate. Estimated adhesive forces were in agreement with experimental
data presented in the literature. As the surface was etched during the
undercut cleaning, the contact area between the particle and substrate
changed, creating a corresponding change in the magnitude of the
electrostatic and van der Waals forces that control the particle adhesion.
Prior to the onset of etching, the net van der Waals and electrostatic
forces between the particles and surface were attractive. However, as the
etch proceeded, attractive van der Waals forces were seen to be reduced
while repulsive electrostatic forces increased, until the particles became
dislodged from the surface.
Removal of 7 and 15µm polystyrene latex (PSL) spheres from a TEOS-sourced
silicon dioxide surface was examined. 20:1 buffered hydrofluoric acid (BHF)
was used to etch the silicon dioxide surface. Etching was performed under
non-flow conditions. Pre- and post-etch wafer scans were obtained using a
Tencor SP1 Surfscan system. The fraction of particles adhering to the
surface for different etch times was studied. It was observed that particles
with nominally the same size and shape were not all removed at the same etch
time owing to variations in the van der Waals force of adhesion. These
variations were attributed to varying particle and surface roughness. The
adhesion model was used to predict the percentage of particles that remain
on the wafer surface after a given etch. The experimental data was found to
lie within 10% of the predictions when equilibrium contact between the
particles and the substrate was assumed. (PDF) |
| April 29 |
Host: Christopher Ober,
Materials Science & Engineering,
Cornell University
Presentation by: Dario Goldfarb, IBM and Chris Ober, Cornell
University
Topics: Goldfarb: "The Demands of Advanced Lithography-ESH
Issues in the Lithographic Process"; Ober:
"scCO2 Processing Methods for ESH Friendly Lithography"
Abstract:
When
considering new processes for the semiconductor industry, ESH and
performance issues need to be considered in concert. These two presentations
will discuss the present and future performance aspects of lithography and
where ESH opportunities may exist. Also covered will be the capabilities of
supercritical (sc) CO2 in photoresist processing and how this may provide
opportunities for future lithographic process design. SC CO2 offers a
non-VOC, clean solvent with no surface tension and low viscosity. A possible
route to an all-dry lithographic process will be discussed.
Goldfarb (PDF);
Ober (PDF) |
| May 6 |
Host: Yoshio Nishi,
Stanford University
Presentation by: Dr.
Hyoungsub Kim, Stanford University
Topic: "Nanoscale zirconia and
hafnia dielectric grown by atomic layer deposition: crystallinity, interface
structures and electrical
properties"
Abstract: Presently, metal
oxides having higher dielectric constants than SiO2 are being
investigated to reduce the leakage current by increasing the physical
thickness of the dielectric. Among many possible deposition techniques,
atomic layer deposition (ALD) has drawn attention as a method for preparing
ultra-thin metal oxide layers with excellent electrical characteristics and
near-perfect film conformality due to the layer-by-layer nature of the
deposition mechanism.
For this research, an ALD system using ZrCl4/HfCl4 and
H2O was built and optimized. The microstructural and electrical
properties of ALD-ZrO2 and HfO2 grown on SiO2/Si
substrates were investigated and compared using various characterization
tools. In particular, the crystallization kinetics of amorphous ALD-HfO2
films were studied using in-situ annealing experiments in a TEM. The
effect of crystallization on the electrical properties of ALD-HfO2
was also investigated using various in-situ and ex-situ
post-deposition anneals. Additionally, a promising new approach for
engineering the thickness of the SiO2-based interface layer
between the metal oxide and silicon substrate after deposition of the
metal oxide layer was suggested.
As one of several possible applications, ALD-ZrO2 and HfO2
gate dielectric films were deposited on Ge (001) substrates with different
surface passivations. After extensive characterization using various
microstructural, electrical, and chemical analyses, excellent MOS electrical
properties of high-k
gate dielectrics on Ge were successfully demonstrated with optimized surface
nitridation of the Ge substrates. (PDF) |
| May 13 |
Host: David
Graves, University of California-Berkeley
Presentation by: Jerry (Cheng-Che) Hsu, University of California-Berkeley
Topic: "Etching Ruthenium with Inductively Coupled Plasma
Studies of Etch Products Downstream and Interactions with Chamber
Walls."
Abstract: An
important potential EHS problem associated with the investigation of many
possible new materials is the creation of unknown by-products during
plasma etching. Among the set of new materials are metals that are under
investigation as gate electrodes. Some of these materials are difficult to
etch and have received relatively little attention to date. The present
study aims to develop systematic methods to evaluate and quantify the ESH
impact of process, chemicals, and process equipment. The ITRS notes that
early identification of ESH impacts is crucial for simultaneous selection
of processes and chemicals and minimizing ESH impacts. Plasma etching is
especially challenging in this regard since the plasma often scrambles the
incoming gases and etched materials to form completely new compounds. The
first step in assessing the potential ESH impacts of these new compounds
is to develop a methodology to identify their nature, source and
transport. .
The ICP was characterized by in-situ
ion and neutral mass spectrometers, a chamber wall-mounted quartz crystal
microbalance, optical emission spectroscopy, a wall-mounted ion flux
probe, and an FTIR spectrometer in the turbomolecular pump foreline. Ru
films were etched from 150 mm diameter wafers placed on a rf-biased
substrate. Ru can be etched
readily by Ar and O2-containing plasma.
Cl2 addition results in significant changes in etch
rate, wall deposition behavior, and downstream etch product composition.
With 10 sccm Ar and 10 sccm O2 at 3*1010 cm-3 plasma density, 10mT pressure and 100V bias voltage,
a 60 Ĺ/min etching rate was observed. In addition, without Cl2 addition, no RuO4 was observed in the
foreline, and almost all etch by-products were deposited on the chamber
wall. With Cl2
addition (Ar/O2/Cl2 plasma), the etching rate
increased by a factor of 5, RuO4 was observed downstream by
FTIR, and virtually zero wall deposition rate was observed.
One interpretation of the observations is that chlorine addition to
the Ar/O2 plasma results in a more volatile Ru-oxychloride etch
product, increasing both film etch rate and chamber wall re-etch rate. A
major conclusion from this work is that etch by-products and chamber wall
interactions are sensitive functions of plasma chemistry, tool type and
operating conditions. This has important implications for simultaneous
process and ESH optimization.
Presentation: (PDF) |
| May 20 |
Host: Pierre
Khuri-Yakub, Stanford University
Presentation by: Utkan Demirci, Stanford University
Topic: "Environmentally Benign Deposition of Photoresist,
Low-k and High-k Dielectrics"
Abstract: We present the theory of operation,
fabrication and the experimental results obtained with a novel
acoustically actuated 2D micromachined microdroplet ejector array. The use
of micro-droplet ejector arrays is proposed for environmentally benign
deposition of photoresist and other spin-on materials, such as low-k and
high-k dielectrics used in IC manufacturing. Direct deposition of these
chemicals will reduce waste and production cost. These ejectors are
chemically compatible with the materials used in IC manufacturing. They do
not harm fluids that are heat or pressure sensitive. Moreover, this makes
them appear as an attractive droplet generation device in biomedicine and
biotechnology applications, such as printing of DNA or protein assays and
drug testing. The radiation pressure associated with the acoustic beam
overcomes the surface tension force and releases droplets into air in
every actuation cycle. The ejectors operated most efficiently at 34 MHz
and generated 28 µm diameter droplets in drop-on-demand and continuous
modes of operation as predicted by the finite element analysis. Various
fluids such as photoresist isopropanol, ethyl alcohol, and acetone were
ejected from all the elements of a 3x3 2D micromachined ejector array
simultaneously. The acoustic actuators are combined with silicon
micromachined fluid pools which control the fluid height and the location
of the focus, and provide ejection in all angles. (PDF) |
| May 27 |
Host: Rafael
Reif, Massachusetts Institute of Technology
Presentation by:
Dr. Cynthia
Folsom Murphy, Research Scientist, Center for Energy and Environmental
Resources, University of Texas at Austin
Topic: "Development
of Parametric Inventories for Semiconductor Wafer Fabrication"
Abstract: Based on
currently available data and information, most of the energy and material
consumption used in the production of semiconductor devices (microchips)
occurs during the wafer fabrication process. However, despite the growing
prevalence of these components in consumer electronics, appliances, and
automobiles, very few life cycle studies have been completed for these
products. This is in large part because the data required for the mass and
energy balances are difficult to obtain and are collected against a
background of rapidly changing product designs and short product life-times.
In order to address the difficulties associated with inventory development
in the semiconductor industry, a methodology has been developed that
accounts for mass and energy use at the unit operation level as function of
basic process and design parameters as well as equipment selection and
capacity within the wafer manufacturing setting. Illustrations of the
methodology are presented for the furnace and wafer clean operations.
Energy consumption is compared over all unit operations for a 6 and 8-layer
metal device. (PDF) |
| June 3 |
Host: Srini
Raghavan, University of Arizona
Presentation by: Professor
Pierre Deymier, Materials Science and Engineering, University of Arizona
Topic: "Biological
Templates for Nanointerconnects"
Abstract: The current state of knowledge suggests that proteins
and assemblies of proteins offer the level of control necessary for
inexpensive and reliable bottom-up fabrication of nano-scale interconnects.
We focus on microtubules (MT),
self-assembling, dynamic nano-structures composed of polymerized
a
and b
tubulin protein monomers, as a model biomolecular template for nano-scale
interconnects. We report on progress in controlling the directed
self-assembly of MTs from metal electrodes on silicon wafers and their
subsequent metallization. (PPT) |
| June 10 |
Host: Krishna
Saraswat, Stanford University
Presentation by:
Werner Pamler,
Infineon Technologies, Corporate Research, Munich, Germany
Topic: "Nano Interconnect Technology - Looking
at the End of the Roadmap"
Abstract:
Moving along
the International Roadmap of Semiconductors we will be facing a series of
severe challenges: keeping the resistivities small in spite of size effects,
lowering the dielectric constant of insulators, and reducing the thickness
of diffusion barriers. The nano-interconnects group at Infineon Corporate
Research has the task to assess these challenges. This presentation will
give an overview of these activities. Emphasis will be placed upon air gap
technology where selective ozone/TEOS deposition may overcome some of the
drawbacks of other approaches. (PDF) |
| June 17 |
Host: David
Mathine, University of Arizona
Presentation by: David Mathine, Optical Sciences Center,
University of Arizona
Topic: "The CMOS Biochip for
Toxicity Testing"
Abstract: Traditional means of
determining chemical toxicity typically involve expensive and laborious
animal studies. These methods cannot keep pace with the demand for
evaluating the toxicity of new chemicals introduced by industry. The advent
of biochip technology promises to yield a high-throughput means of screening
even complex mixtures of chemicals for toxicity. By monitoring the genetic
activity of exposed cells, investigators can identify signature genetic
responses that indicate toxic insult. Specifically, relative levels of gene
expression between treated and control in vitro models are used to
assay whether the treated model exhibits a genetic response characteristic
of physiological stress.
The presentation will report on the design and production of a CMOS chip
capable of performing the requisite electrochemical and optical
measurements, the design and production of a driver device to orchestrate
the chip’s functions, and the development of adequate in vitro models
for toxicity testing. (PDF-part 1)
(PDF-part
2) |
| June 24 |
Host: Kimberly
Ogden, University of Arizona
Presentation by: Kim Ogden & Sally Clement, University
of Arizona
Topic: "Assessment of Industrial Continuing Education Needs"
(including pre-TeleSeminar survey)
Abstract:
This teleconference will serve as one method to assess the future continuing
education needs of our industrial partners. A pre-teleconference survey was
sent out and the results were reviewed. These results will be discussed and
a dialog initiated to determine timing of courses and topic priorities. A
few slides are available below for review to initiate discussion. The goals
for this teleconference include
-Determining when to offer a short course
-Discussing best method to advertise web based university courses
-Determining the feasibility of hands-on training in university test-beds
for employees at all levels
-Discussing water education needs (PDF) |
| July 1 |
No TeleSeminar-Independence
Day |
| July 8 |
Host: Duane
Boning, Massachusetts Institute of Technology
Presentation by: Yun Zhuang, PostDoc, University of Arizona
Topic:
Part 1--
"Experimental and Numerical Analysis of An Inhibitor-Containing Slurry for
Copper CMP";
Part 2
-- "Novel
Method for Direct Measurement of Substrate Temperature during Copper CMP"
Abstract: Part 1
--
A novel slurry containing benzotriazole (BTA)
as the inhibitor was analyzed in terms of its frictional, thermal and
kinetic attributes for copper CMP applications. The frictional analysis
indicated that
‘boundary lubrication’ was the dominant
tribological mechanism. Due to the presence of the inhibitor in the slurry,
copper removal rate exhibited a highly non-Prestonian behavior. Based on the
measured coefficient of friction (COF)
and pad temperature data, a
proven thermal model was used to predict wafer temperature.
The Preston Equation was used to describe the polishing rate when
p´V
was lower than 11,000 Pa-m/s; while a modified Langmuir-Hinshelwood kinetic
model was used to simulate the copper
removal when p´V
was higher than 11,500 Pa-m/s.
Assuming that the adsorbed inhibitor
layer was abraded off instantly from the copper surface
when
p´V
was higher than 11,500 Pa-m/s, the modified Langmuir-Hinshelwood kinetic
model indicated that copper polishing was chemically limited
in this polishing region.
(PDF)
Abstract: Part 2 --
It has been
shown that temperature has a significant effect on removal rate during
copper CMP process. While there has been tremendous effort spent on
developing thermal models to simulate copper wafer temperature based on
measured pad temperature, there has been no report of direct measurement of
copper wafer temperature during polishing. In this study, a novel method was
created to directly measure the substrate temperature during copper CMP.
Using specially designed wafer carriers, real-time copper disc and copper
wafer temperatures were obtained by an infrared camera. Results show that
copper disc and copper wafer temperatures are higher than those of the
leading and trailing edges of the pad. Results also show that there is a
temperature distribution on the copper wafer, which is believed to be
closely related to the slurry flow during polishing. A 3-D thermal model was
developed and used to simulate the pad and wafer temperatures. Simulation
data are shown to agree well with the experimental results. (PDF) |
| July 15 |
Host: David Dornfeld,
University of California-Berkeley
Presentation by:
Part 1--Sarah Boyd, UC-Berkeley;
Part 2--Nikhil Krishnan,
UC-Berkeley
Topic: Part
1--"Applicability of Quantitative Structure Activity
Relationship (QSAR) Methods in Semiconductor Manufacturing";
Part 2--"Industry Needs in
Semiconductor EHS Assessment - Summary of Survey Results"
Abstract: This talk is divided into two parts. In
Part 1, we will discuss recent
work exploring the applicability of toxicological data sources and
quantitative structure activity relationship (QSAR) methods for use in
establishing human health impact metrics in semiconductor process life cycle
assessment. In Part 2, we
will present results of the survey on EHS assessment needs for the
semiconductor industry, conducted by researchers MIT and U C Berkeley
earlier this year. We welcome additional inputs from the member companies
and listeners and feedback on summary points. (Part 1:
PDF) (Part 2:
PDF) |
| July 22 |
Host: Anthony
Muscat, University of Arizona
Presentation by: Keith P. Johnston, Department of Chemical
Engineering, University of Texas-Austin
Topic: "Chemical/mechanical process
for supercritical carbon dioxide cleaning of porous methylsilsesquioxane (pMSQ)
dielectric films with surfactants"
Abstract: In order to fulfill the International Technology
Roadmap for Semiconductors (ITRS) requirements for the performance of
interconnects, new cleaning processes must be made available. Traditional
wet chemistries tend to damage the sensitive interlayer dielectric (ILD)
materials designated for the ITRS 100 nm node. Supercritical carbon dioxide
(scCO2)
cleaning is emerging as a potential low cost, environmentally benign
alternative to conventional cleaning processes. With the inclusion of
hexamethyldisilazane (HMDS) repair in scCO2, a fast and efficient
way to clean and repair the low-k dielectric is within reach. scCO2
solutions exhibit minimal surface tension and thus penetrate and wet
surfaces with small geometries. Here we investigate cleaning and dielectric
repair of pMSQ interlayer dielectric materials and examine the properties of
these films with spectroscopic ellipsometry, scanning electron microscopy (SEM)
and Fourier transform infrared spectroscopy (FTIR),.
FTIR studies and SEM images corroborate to show repair of the interlayer
dielectric after HMDS treatment in scCO2. After
depressurization, ellipsometric studies indicate that the films contract to
a value close to their original thickness.
The via cleaning experiments took place on a dual damascene structure that
was comprised of TEOS ILDs
(interlayer dielectrics) and SiCN etch stop layers at the trench level
and at the via open level. Pure CO2 did not clean the vias. With various
processes utilizing water, CO2 and surfactant, no post etch residue is left
in the via and smooth side
walls in the trench are also seen. The cleaning time was
reduced by utilizing a chemical/mechanical method compared with our previous
studies utilizing purely chemical dissolution. These experiments provide
proof of concept for cleaning with CO2 and surfactants,
which has
the potential to remove via veil residues, water, etch gases and other
contaminates from pMSQ films.
We acknowledge support from the etch and ESH groups at International
SEMATECH.
[Keith Johnston2,
John Keagy2, Xiaogang Zhang2,
Joseph Pham2,
Peter F. Green2,
Todd Rhoad 1 Josh Wolf1
1International SEMATECH
Assignee from Intel, Portland OR, USA;
2The
University of Texas at Austin TX, USA] (PDF) |
| July 29 |
Host: Stacey
Bent, Stanford University
Presentation by: Stacey Bent, Stanford University
Topic: "Area-Selective ALD of HfO2"
Abstract: Compared to other
deposition techniques, atomic layer deposition (ALD) is promising for
preparing a variety of materials because it can
produce high quality films with excellent conformality and precise film
thickness control. Typically, the process permits nano-scale control of
materials in the vertical direction. We are investigating an area-selective
ALD technique which will enable micro- and ultimately nano-scale definition
of the lateral structure. Our approach is to chemically modify the
substrate surface in order to impart spatial selectivity to ALD. Our focus
is on the high k
material HfO2. Using a variety of analytical techniques, we show
that functionalizing the surface with organic monolayers can block the ALD
chemistry in the growth of HfO2. Both solution- and gas-phase
delivery of the blocking layers have been explored. The efficacy of
blocking is found to depend strongly on the quality and the chain length of
the attached monolayers. The process has been found to be effective at ALD
temperatures up to 300 oC and for different ALD precursors [e.g.
HfCl4 and Hf(N(CH3)2)4].
The potential of the area selective process for defining lateral structure
has also been examined using several different patterning methods, including
microcontact printing, and selective functionalization of patterned SiO2/Si.
The propagation of the two-dimensional patterns generated by these methods
into three-dimensional structures using area-selective ALD has been
explored. Results on the area selectivity of these processes will be
presented. (PDF) |
| Aug. 5 |
No TeleSeminar -- RETREAT
PREPARATION |
| Aug. 12 |
No TeleSeminar |
| Aug. 19 |
No TeleSeminar --RETREAT
MEETING-STANFORD |
| Aug. 26 |
No TeleSeminar |
| Sept. 2 |
Host: Karen
Gleason, Massachusetts Institute of Technology
Presentation by: April Ross, MIT
Topic: "Chemical Vapor Deposition of Organosilicon Composite
Films for Porous Low-k Dielectrics"
Abstract: There is an
increasing need for low-k solutions in integrated circuits that can be
extended to future generations. Two different pathways are explored to
lower the dielectric constant of these materials relative to silicon dioxide
(k~4). By incorporating atoms and bonds that have a lower polarizability,
such as alkyl groups, or by lowering the density of the material, either
sterically or through the integration of air, the value of the dielectric
constant is reduced.
Creating porous films using an organosilicate glass (OSG) matrix is an
avenue for introducing void space and thereby decreasing density and
lowering the dielectric constant. Since the introduction of porosity
severely diminishes the mechanical integrity of the material, understanding
the relationship between film structure and mechanical properties becomes
extremely important.
Depositing matrix and porogen precursors simultaneously using CVD creates
many obstacles as the deposition requirements of the two species are
significantly different. A novel approach to alleviate this problem is to
use pre-formed porogens, such as polystyrene micro-spheres or cyclodextrin.
This allows use of the deposition conditions that give the optimal matrix
material structure, with disregard to the effect on the porogen deposition.
An ultrasonic atomizer can be used to introduce these porogen materials into
the reactor for deposition. Another advantage gained from this technique
is that the pore size is controlled and determined by size of the porogen.
Cyclodextrin is a prime porogen candidate due to its low decomposition
temperature and small molecular size. (PDF) |
| Sept. 9 |
Host: Farhang Shadman,
University of Arizona
Presentation by: Dr. Mansour Moinpour, Engineering
Manager-Materials Technology, Intel Corporation
Topic: “CMP
Consumables: Meeting Technology Challenges with Quality Incident Free
Performance”
Abstract: Semiconductor
interconnects are requiring an increase in the number of metal layers and
decreasing dielectric constant. The chemical mechanical planarization (CMP)
has been one of the key enabling modules in the past 10 years for sub 0.35
micron Back End Of the Line (BEOL) processing. The CMP process is using more
complex processing steps, and greater number of materials to meet the
topography and defect challenges. A quick overview of CMP trends and
technology challenges is provided in this presentation. The overall incoming
quality of fabrication materials have improved, but the quality improvement
has stalled in the past couple of years. To take the next leap in quality,
suppliers need to move to a quality incident free mindset. This mindset
comes from developing PCS culture, starting with a solid foundation in the
data collected and identification of relevant process parameters. These key
and control parameters can come from many points in the process, raw
materials, etc. A practical example of key parameter characterization and
identification process control strategies is presented. (Paper
presented at the CAMP CMP Conf., Lake Placid, NY Aug. 2004) (PDF) |
| Sept. 16 |
Host: Ara Philipossian, University of Arizona
Presentation by: Dr. Robert P.
Meagley, Researcher in Residence, MAP Program Manager Components Research,
Intel Corporation
Topic: "Fundamental
Challenges for Lithographic Roadmap"
Abstract:
Photolithography has met the challenge of the 90 nm node through application
of higher NA as well as RET in 193nm and 248nm technologies. Printing even
smaller CDs has pushed the size of the resist polymers to within a factor of
100 of molecular sizes, a direct outcome of the ITRS. As a consequence,
performance characteristics less significant for previous generations have
begun to take on new priorities, for example collapse margin and line width
roughness (LWR). These aspects are key to the successful extension of 193nm
technology and as well to the successful introduction of EUV, a key enabling
technology. This is especially a consequence of importance to Intel's
lithographic roadmap, for through accelerated development of advanced
technologies, we expect to operate in the molecular regimes first.
To improve performance of patterning, resists have been designed to optimize
performance through engineering various properties (Tg, profile, surface
properties, etc.). However new materials designs may be applied to push the
limits of performance even further into the molecular scale. Indeed, new
materials are anticipated to enhance performance across photolithographic
technologies. Upstream engagement with suppliers, universities and consortia
to facilitate early learning for the parties engaged will be highlighted.
New lithographic materials will be discussed in terms of their performance
challenges and a vision of creative and dynamic materials development as
well as enhanced execution to the roadmap will be communicated in this
presentation. (PDF) |
| Sept. 23 |
No TeleSem -- end of 2003-2004
season |
| Sept. 30 |
Host: Reyes
Sierra, University of Arizona
Presentation by: Reyes Sierra,
Department of
Chemical and Environmental Engineering, University of Arizona
Topic: "Biological Removal of
Copper in CMP Effluents”
Abstract: The rapid expansion of copper chemical mechanical
planarization (CMP) in semiconductor fabrication facilities has generated a
significant increase in the quantities of wastewater requiring treatment.
CMP effluents contain high concentrations of soluble copper and a complex
mixture of organic constituents. Physico-chemical methods currently applied
for the removal of copper in CMP effluents, (eg. ion exchange, filtration
and microfiltration, coagulation/electrocoagulation, etc.), are often
expensive and energy intensive. Environmental biotechnologies have been
shown to offer interesting potentials for metal removal and recovery.
Biological treatment could also provide an attractive approach to meet
regulatory challenges associated with copper in CMP effluents. In addition,
biological treatment offers the possibility remove organic contaminants
along with copper.
The aim of this research is to evaluate the feasibility of an innovative
system configuration that combines a crystallization reactor and a
sulfate-reducing anaerobic bioreactor for the simultaneous removal of copper
and organic contaminants from CMP effluents. Copper biomineralization, a
process stimulated by biogenic sulfides produced by sulfate-reducing
bacteria, will occur in the crystallization reactor. Degradation of organic
contaminants and sulfate conversion will be attained in the expanded
granular sludge bed (EGSB) bioreactor. The crystallization reactor is a
fluidized bed containing fine sand. The sand surface offers centers for CuS
nucleation and subsequent crystal growth. Potential advantages of a
bioreactor/crystallization system as compared to a stand-alone anaerobic
bioreactor include: recovery of valuable metals in purified form; reduced
dilution of active biomass with minerals; decreased inhibitory and toxicity
effects of heavy metals on microorganisms; and operation under optimal
conditions for sulfidogenesis and mineral deposition, facilitating process
control. (PDF) |
| Oct. 7 |
Host: IAB >>
Rescheduled to future date |
| Oct. 14 |
Host: Duane Boning,
MIT
Presentation by: Julie Snook, Application Engineer,
CON-TACT Program, Brewer Science
Topic: "A Novel
Planarization Technique to be Used in the Manufacture of Semiconductor
Devices"
Abstract: The increasingly
complex nature of integrated circuits has exposed a need for interlayer
planarization in device manufacture. The method of choice to provide the
required planarization is Chemical Mechanical Planarization (CMP). A novel
method of planarization has been developed, CON-TACT, which offers
significant performance benefits. This method is flexible and has been
successfully applied to a variety of applications. This presentation will
discuss the technical performance benefits, such as pattern-density
independent planarization, as well as the environmental benefits of this new
method. (PDF) |
| Oct. 21 |
Host: Kimberly
Ogden, University of Arizona
Presentation by: Mark A.
Burns, Department of Chemical Engineering and Department of Biomedical
Engineering, The University of Michigan
Topic: "Microfabrication
Biochemical Analysis:
Bulk Properties to
Single Molecules"
Abstract:
Microfabrication holds great promise for the construction of inexpensive
biochemical analysis systems. Multiple analysis systems can be constructed
on an individual die resulting in systems with higher throughput, lower
biochemical costs, and minimal user intervention relative to other
technologies. Such combined biochemical analysis systems can, in essence,
function as micron-scale intelligent sensors.
We are constructing such devices on silicon, glass, and polymer substrates
for the analysis of DNA and other biochemical samples. The devices consist
of micron-scale reaction, separation, and/or detection systems connected by
a series of micromachined channels. Liquid samples and reagents are
injected into these devices and then moved between components by
hydrophobic/hydrophilic patterning, pressure sources, and/or phase-change
valves. Temperature controlled chambers can be used for a variety of
reactions including selective amplification and digestion of DNA. The
products can then be separated on micron-scale electrophoresis units with
on-chip fluorescence detection. Individual analysis components as well as
complete integrated systems will be described. (PDF) |
| Oct. 28 |
Host: David
Graves, UC-Berkeley
Presentation by:
Yassine
Kabouzi, Department of Chemical Engineering,
University of California-Berkeley
Topic:
"Abatement of greenhouse gases using surface-wave microwave discharges
sustained at atmospheric pressure"
Abstract:
Reducing perfluorinated compound (PFC) emission from chamber cleaning and
dielectric etch tools continues to be an important challenge for the
semiconductor industry. In some cases, there are advantages to the use of
abatement techniques using plasmas at atmospheric pressure. To date, this
approach has received less attention than plasma abatement at low pressure.
In this talk, a promising atmospheric pressure microwave-based plasma
abatement method is described.
Microwave surface-wave plasmas can be operated at atmospheric pressure and
have been efficiently used to eliminate PFC-containing gases. As an example,
we present results from the abatement of SF6 diluted in N2
(0.5-2.4%). O2 is added to the gas mixture to ensure oxidation of
PFC fragments and to provide scrubbable byproducts (acid-like byproducts).
The destruction and removal efficiency (DRE) of SF6 is
investigated as a function of microwave power, discharge tube radius,
nitrogen flow rate and PFC inlet concentration. The DRE increases with
increasing microwave power as a result of increasing electron density and
residence time. Reformation of PFC molecules from fragments is the main
mechanism limiting abatement efficiency. Reformation increases with
decreasing gas temperature and increasing PFC inlet concentration. DRE
values of virtually 100% have been achieved with surface wave plasmas,
provided microwave power is high enough (3 kW) and the discharge tube radius
is small (4 mm). (Work done at the University of Montreal)
(PDF) |
| Nov. 4 |
Host: Srini Raghavan,
University of Arizona
Presentation by: Thomas J.
Wagener, Applications Manager, FSI Surface Conditioning Division, Chaska,
Minnesota
Topic: "Particle
Removal Using Cryogenic Aerosols"
Abstract:
Removing particles in both the FEOL and BEOL is becoming more
challenging. For FEOL cleans, the 2003 ITRS roadmap calls for less than 0.5
Angstrom material loss for the 65nm technology node and beyond. To achieve
this material loss goal, more dilute and reduced temperature chemistries are
being implemented. To minimize the drop in particle removal efficiency
(PRE), energetic particle removal techniques such as megasonics and/or
atomized spray may be utilized. However, care must be taken to prevent
structure damage. Shrinking features in the BEOL, along with the
incorporation of low-k materials, provide additional challenges. One is
not only concerned with structure damage, but also moisture absorption,
chemical reactions, and the removal of particles from phobic surfaces.
Cryogenic aerosols have been used to increase manufacturing yield of
integrated circuits for nearly a decade. The benign characteristics of
this all-dry process makes for straightforward insertion into any process
flow requiring the removal of particle defects. These characteristics
include no charging, no water marks, no scratching, no chemicals and no
material loss, addressing many of the challenges addressed above. In
addition, recent studies show that cryogenic aerosols have superior PRE
from phobic surfaces when compared to conventional wet cleans. The above
mentioned attributes of cryogenic aerosols will be presented. (PDF) |
| Nov. 11 |
No TeleSeminar -- VETERANS DAY |
| Nov. 18 |
Host: Pierre
Khuri-Yakub, Stanford University
Presentation by: Utkan
Demirci, Stanford University Electrical Engineering Department, and E.L.
Ginzton Lab
Topic: "Acoustically
Generated Picoliter Droplets Using 2-D Micromachined Microdroplet Ejector
Arrays"
Abstract:
There is growing demand in the fields of semiconductor manufacturing and
biotechnology to reliably generate repeatable, uniform, picoliter-size fluid
droplets. Such droplets can be generated using MEMS
(Micro-Electro-Mechanical Systems) technology.
We propose 2-D micromachined
microdroplet ejector arrays for environmentally benign deposition of
photoresist and other spin-on materials, such as low-k and
high-k dielectrics used in
integrated circuit (IC) manufacturing. Direct deposition of these
chemicals will reduce waste and production cost. These ejectors are
chemically compatible with the materials used in IC manufacturing, and do
not harm fluids that are heat or pressure sensitive.
Moreover, these ejectors are attractive to biomedicine and
biotechnology for droplet generation in applications such as printing of DNA
or protein assays and drug testing.
Two novel methods for generating
millions of droplets per second using acoustically actuated 2-D
micromachined microdroplet ejector arrays will be presented. First, membrane
based 2-D micromachined ejector arrays will be introduced. Each element of a
membrane based 2-D ejector array consists of a flexurally vibrating circular
membrane on one face of a cylindrical fluid reservoir. The membrane has an
orifice at the center. A piezoelectric transducer generating ultrasonic
waves, located at the open face of the reservoir, actuates the membrane and
droplets are ejected through the membrane orifice. The ejectors
operated most efficiently at 1.2 MHz and
generated 3-7 µm diameter droplets. Second, acoustic focus based 2-D
micromachined ejector arrays will be demonstrated. The radiation pressure
associated with the acoustic beam overcomes the surface tension force, and
releases droplets into air in every actuation cycle. The ejectors operated
most efficiently at 34.7 MHz, and generated 28 µm diameter droplets in both
drop-on-demand and continuous modes of operation, as predicted by the finite
element analysis. Photoresist, water, isopropanol, ethyl alcohol, and
acetone were ejected from a 4x4 2-D micromachined ejector array. The theory
of operation, fabrication and the experimental results obtained with novel
acoustically actuated 2-D micromachined microdroplet ejector arrays will be
presented. (PDF) |
| Nov. 25 |
No TeleSeminar - THANKSGIVING
HOLIDAY |
| Dec. 2 |
Host: Greg McRae, MIT
Presentation by: Greg McRae, MIT
Topic:
"Integrating
Environmental Considerations into Technology Selections Under Uncertainty"
Abstract: A critical
challenge facing the chemical and microelectronics industries is how to
achieve improvements in both economic and environmental performance.
Deciding which process chemistry or technology is the bestis not easy
especially when there are multiple and often conflicting objectives. This
paper describes new decision analytic methods, life cycle assessment tools
and fast algorithms for uncertainty analysis that can help structure the
decision processes in such a way to reduce uncertainties in outcomes. In
particular, value-of-information concepts and hierarchical modeling are used
as a way to decide whether there is sufficient knowledge to make a decision
or if more resources need to be allocated to reduce the risk of bad
outcomes. Within a uniform model environment, process models combined with
uncertainty analyses of the effects of parametric and structural
uncertainties are used to identify critical elements of the decision process
and to determine if more refined modeling or data collection must be carried
out. The new framework will be illustrated with a case study of the issues
involved in choosing between two gases (NF3 or F2) for cleaning a reactor
used in manufacturing microelectronics. The case study will show how the
choice of process boundaries, levels of modeling detail and data quality can
influence the decision outcomes. (PDF) |
| Dec. 9 |
Host: Ara
Philipossian, University of Arizona
Presentation by: Dr.
Leonard Borucki, Intelligent Planar
Topic: Flash
Heating in Chemical-Mechanical Polishing
Abstract: Silicon
dioxide removal rates can be modeled within an RMS error of a few percent
using a two-step Langmuir-Hinshelwood model that abstractly includes both a
mechanical removal step and a rate-limiting chemical step. The model also
describes copper polishing when the chemistry conforms to the two-step
framework. The chemical step in the model requires a reaction temperature.
Early in the development of the model, the temperature formula
 ,
(1)
was discovered to be a key part of the explanation for why very different
rates are sometimes observed at fixed pV but for different
combinations of p and V. In Eq. (1), Ta is
the platen temperature, and
b
and the exponent a are fitting parameters. Extracted values of the
exponent a for oxide polishing typically fall in the range 1.6-1.8.
However, it was not clear why the temperature rise in Eq. (1) should be a
power law in V or why a should have the values that it does.
This talk will discuss the resolution of these questions. A physical
analysis of asperity tip heating suggests that as an asperity passes over a
point on the wafer, the temperature of the wafer surface on the average
momentarily flashes to
 .
(2)
For polishing on a k-groove pad, the factors on the right side of Eq. (2)
that do not explicitly depend on p or V are independent of
V except for the pad heat partition factor
gp.
Results of finite element heat transfer calculations for nano-lubricated
sliding of an asperity tip over a silicon wafer with a thin SiO2
layer show that the heat partition factor is well-approximated by a power
law in V. The magnitude of the exponent in the power law furthermore
is in natural agreement with the empirically extracted values of the
exponent a. The analysis not only provides explicit formulas for the
empirical parameters
b
and a, but also relates them to wafer and pad thermal, mechanical and
surface properties. Thus, it is possible when using (1) as a compact model
to relate the extracted parameters to independently verifiable physical
properties. Alternatively, if physical properties are well known, it is
possible to use (2) as the compact flash heating model and a reduced set of
fitting parameters in the Langmuir-Hinshelwood model. (PDF) |
| Dec. 16 |
No TeleSeminar - CHRISTMAS
HOLIDAY |
| Dec. 23 |
No TeleSeminar - CHRISTMAS
HOLIDAY |
| Dec. 30 |
No TeleSeminar - CHRISTMAS
BREAK AT U. OF A. |
|
