Cabot Corp. v. Solution Technology, Inc.

122 F. Supp. 2d 599 (2000)

CABOT CORPORATION, Plaintiff,
v.
SOLUTION TECHNOLOGY, INC., Defendant.

No. 3:96CV505-MCK.

United States District Court, W.D. North Carolina, Charlotte Division.

October 10, 2000.

John R. Wester, Lawrence C. Moore, III, Robinson, Bradshaw & Hinson, P.A., Charlotte, NC, Matthew Weil, Howard, Rice, Nemerovski, Canady, Falk & Rabkin, Irvine, CA, David Stein, McDermott, Will & Emery, Newport Beach, CA, Fay E. Morisseau, McDermott, Will & Emery, Irvine, CA, for Cabot Corporation, a Delaware Corporation, plaintiff.

David Bennet, Worcester, MA, for Saint-Gobain Ind Cer, interested party.

Neil C. Jones, Dority & Manning, Greenville, SC, Patrick J. Flinn, Alston & Bird, Atlanta, GA, Wellington M. Manning, Jr., Dority & Manning, Greenville, SC, Malcolm B. Wittenberg, Crosby Heafey Roach & May, Oakland, CA, Andre G. Bouchard, Bouchard & Friedlander, Wilmington, DE, for Solution Technology Inc., a North Carolina Corporation, defendant.

Fay E. Morisseau, McDermott, Will & Emery, Irvine, CA, Irving M. Brenner, Smith Helms Mulliss & Moore, L.L.P, Charlotte, for Cabot Corporation, a Delaware Corporation, counter-defendant.

Wellington M. Manning, Jr., Malcolm B. Wittenberg, Irving M. Brenner, (See above), for Solution Technology Inc., a North Carolina Corporation, counter-claimant.

ORDER

McKNIGHT, United States Magistrate Judge.

This patent case is before the undersigned United States magistrate judge by consent of the parties pursuant to 28 U.S.C. ž 636(c). At issue is whether this case is exceptional based upon willful infringement and litigation misconduct, as "exceptional" is understood for purposes of applying 35 U.S.C. ž 285, and whether costs and attorneys' fees should be awarded to Cabot.^[1] These matters were exhaustively discovered and tried to the bench. After searching review of the evidence in light of appropriate standards of proof and relevant case authority, it is the conclusion of the undersigned that the exceptional case standard has not been met and, therefore, that attorneys' fees should not be awarded.

FINDINGS OF FACT

In all actions tried upon the facts without a jury ..., the court shall find the facts specially and state separately its conclusions of law thereon, ....

Rule 52(a), Fed.R.Civ.Pro.

Accordingly, I essay to set in order a narrative of these things clearly and convincingly shown and relevant, and not found in the various discovery orders of this Court.^[2]

Preliminary

Chemical mechanical planarization or polishing ("CMP")

CMP is a technique for flattening surfaces by means of abrasion. It has application in building semiconductor wafers. In this process, an eight-inch piece of single-crystal silicon is used as a plate onto which the manufacturer deposits layers of circuitry. The metal layer is insulated by depositing an oxide layer, essentially glass (TEOS). Oxide CMP planarizes the glass to an appropriate level.

In CMP, the chemical part is the oxidizer. The mechanical part is the abrasive. To achieve CMP, it is necessary to have an abrasive and a chemical component which interact with the surface. Abrasives include alumina and silica. The abrasive in the dispersion (for example, alumina or cerium oxide) achieves the polishing.

Roughly, one adds to the abrasive dispersion different oxidizers (such as ferric nitrate, potassium iodate, hydrogen peroxide, ferric sulfate, or potassium ferric cyanide), which are the chemical component. The oxidizers change the surface of the metal being polished on the surface layer in such a way that the abrasive can then more easily remove it.

Aluminas

There are precipitated and fumed aluminas, the terms describing the chemical process of their generation. For the processes at issue, fumed alumina has a more advantageous viscosity than precipitated alumina because the liquid is thinner yet maintains good stability.

Boehmite is an alkoxide-based alumina containing gamma alumina.

Paul J. Yancey's education and work; the founding of Solution Technology, Incorporated ("STI"); STI's approach to testing its slurries

Paul J. Yancey ("Yancey") was educated as a chemical engineer. By 1994, Yancey had been working as a chemical engineer for nearly thirty years, and, at least since 1985, in the field of dispersions (suspensions of particles in liquid media).

Yancey joined Union Carbide Corporation in 1966. He worked in Union Carbide's Linde division in the manufacture of aluminum oxide materials for ceramics for polishing and growing single crystals of aluminum oxide and other materials used in lasers. He grew crystals and learned techniques for polishing them. He developed a slurry for polishing materials consisting of a permanent dispersion of particles, and this product figured importantly in the Linde product line.

Yancey left Union Carbide for Sawyer Research Products. There he worked in the manufacture of synthetic quartz for use in the high frequency circuit business. He did some polishing and grew sapphire crystals.

Thereafter, he worked for Allied Chemical and came to Charlotte, N.C., with that company in 1979. His work for Allied Chemical involved growing single crystals and polishing materials for lasers and bubble memory technology. These materials were polished with abrasives.

Thereafter, he started the United States subsidiary of Baikowski International, which is a supplier of high purity alumina. In the course of this work, he developed material made from aluminas which were very good polishing powders.

In 1985, Yancey left Baikowski to start Solution Technology, Incorporated ("STI"). He was interested in selling alumina compounds in liquid form in permanent suspension. Baikowski was not interested in pursuing such materials, so Yancey, having 19 years of experience with polishing substrates and three or four years of experience with permanent slurries, started STI and began manufacturing liquid compounds which were permanently stable, meaning, in Yancey's understanding, that they had a shelf life of at least one year.

STI's first products were based on aluminas that were in permanent dispersions and were primarily directed to the optics industry. STI sold dry powders of high purity aluminas and several other optical materials. One of STI's first customers was the Airtron division of Litton Industries, which polished gallium arsenite substrates for use in some high end integrated circuits for industries such as communications. STI branched out from aluminas into compounds based upon cerium oxide, zirconium oxide, diamond, and silica.

When STI was founded, Yancey worked alone and mixed the slurries. At that time, STI did not have testing equipment. Even as late as 1996, Yancey would ask potential customers to test most of the slurries.

However, STI did have certain testing equipment. In 1992, after IBM became a customer, STI purchased a MicroTrac particle size analyzer from Leeds and Northrup, a BET surface area analyzer, and a zeta potentiometer, primarily for quality control. The particle size analyzer gave a printout of particle size distribution, a histogram of sizes within each particle size range or a cumulative distribution that essentially added up each of those size ranges until they reached 100 and plotted the curve. The particle size analyzer worked by introducing ultrasonically a pH-adjusted predispersed sample into the circulation system of the analyzer. An inappropriate pH causes agglomeration of the particles upon introduction into the sample well of the analyzer. The agglomeration will show up on the printout as a bump on the right side of the curve in the 10 to 15 micron range, indicating large particles, which is anomalous. A solution with 10-15-micron particles could scratch the surface of what is being polished. Yancey has operated MicroTrac and Horiba particle size analyzers.

The BET surface analyzer measures the surface area of the outside surface of particles. The zeta potentiometer measures the charge on a particle in a solution. The repulsion resulting from higher charges promotes stability when the particles are sufficiently small relative thereto.

A "retain" is a small portion of a quantity manufactured. It is drawn off for analytical and historical uses. Retains of STI's slurries were taken just after they were mixed, before the settling period. They were stored. In the current facility, they are stored in the mezzanine of the building ten to twelve feet above the floor. The current facility is air-conditioned on weekdays.

The abrasives which Yancey used at STI to produce polishing slurries included precipitated aluminas, calcined aluminas, fumed aluminas, cerium oxide, zirconium oxide, fumed and precipitated titanium dioxide, diamond, silicon carbide, and boron carbide.

Rodel

As of September, 1997, STI became a wholly-owned subsidiary of Rodel. Since the early 1970s, Rodel has manufactured products for sale to the semiconductor industry. Rodel manufactures two primary kinds of polishing products, pads and slurries.

As of the time of trial, Rodel was 48% owned by Rohm & Haas (Class D shares), a public company; 32% by the Rodel Foundation, a 501(c)(3) charity; and the remainder by Rodel employees. Rodel's approximate revenues in 1996 were $105 million; in 1997, $150 million; in 1998, $155 million.

Rodel was founded in 1968 by William Budinger, who at the time of trial was Chairman and sole director, who was CEO of Rodel until 1999, and who at the time of trial owned the Class A shares, which have 99% of the vote.

STI's sales of alumina to IBM

As of July, 1994, STI had supplied to IBM precipitated alumina dispersions under the labels 201 A/60, 201 A/140, 201 A/210 (which was a combination of precipitated alumina and silica), and 201 A/280 (which contains gamma alumina). 201 A/60 consists of a boehmite alumina with a particle size in the fifty to seventy nanometer range. (A nanometer ("nm") is one billionth of a meter. There are 1,000 nanometers in a micron.) STI sold 201 A/60 to IBM between 1990 and 1995 or 1996.

By contrast, Met 202, which STI marketed at Semicon/Southwest in October, 1994, was a fumed alumina, not a boehmite alumina.

Sematech

Sematech is a consortium consisting of the government and semiconductor companies. The companies pay a large fee and supply technicians to Sematech for a tour of duty of several years. Sematech members include Texas Instruments, IBM, Motorola, AMD, Intel, LSI Logic, and Hewlett Packard, all manufacturers of integrated circuits. Since STI supplied slurries to the industry, it was ineligible for membership in Sematech.

Sematech periodically conducts trials in which companies test wafers according to a very strict protocol and examine the results to establish a benchmark. STI was selected by Sematech for benchmarking trials for CMP.

Trade shows

The two which figure into this case are Semicon/West and Semicon/Southwest. The Southwest show is somewhat smaller and more concentrated in integrated circuit fabrication technology than West, which displays some peripheral technologies as well.

Sequential

1990

Early in 1990, STI sold alumina slurries to IBM for CMP in polishing integrated circuits. These slurries were of the product Ultra-Sol 201 A, a very small particle boehmite alumina dispersion, having particle size of 60 nm. The boehmite alumina included in the 201 A/60 was produced by Vista Chemical of Houston, Texas, by an organometallic process. It is a dispersable or precipitated alumina. The chemical component of those slurries was nitric acid. Yancey did not sell IBM boehmite alumina and ferric nitrate.

STI bought the small particle, high purity boehmite alumina in dry powder form from Vista Chemical. STI would mix the dry powder with water and other additives. On certain occasions, STI bought the alumina with water already added.

When Yancey shipped lots 1656, 2021, and 2361 to IBM in 1990, he considered them to be stable, permanent, uniform dispersions. Yancey personally shipped these lots.

March 2, 1990. On this date, IBM ran a particle size distribution analysis of an STI slurry, lot 1656 of Ultra-Sol 201 A, on a Coulter LS Particle Size Analysis machine. Yancey has never operated this machine. Mark Jaso, the IBM analyst, sent a copy of the analysis to Yancey. The analysis presented graphically the percentage of particles in each of a series of discrete size ranges (in microns). It revealed a very small percentage in the 10-micron range, the largest entity being 16 or 17 microns, and a majority in the less-than-.4-micron range. In addition to the number distribution graph, the analysis included a volume distribution graph showing a severe agglomeration at between 4 and 18 microns. Jaso also tested lot 1625 on that date, with similar results.

Yancey did not, and does not, know the procedure which IBM used for preparing the sample, how much water was used, the chemical composition of the water, for how long the tests were run, or the skill and experience of the person who conducted the tests. Nonetheless, Yancey interpreted the 10-micron result in the test of lot 1656 as an anomaly resulting from not taking sufficient caution in preparing the sample, because it was his understanding that in testing a highly dispersed colloid, agglomerates result from improper sample preparation. Yancey believed that any reading of greater than one micron was an anomaly. Although he sold 201 A/280, which has particles greater than one micron, for CMP polishing, he believed that the actual presence of particles of a size greater than one micron could produce severe wafer scratching in the CMP polishing process.

IBM did not complain of scratching and continued to buy STI slurries.

In 1990, IBM was in a screening mode as regards this chemistry and this process. IBM was engaged in fundamental research. In the course of this research, IBM screened four or five versions of STI's 201 A product and eventually settled on 201 A/280 (i.e., particle size, 280 nm). IBM's analysis of that product showed particles extending out into the micron regime.

In 1990, IBM was in the process of refining its measurement of slurry properties. IBM researchers had questions about getting true information relative to an absolute standard. In successive runs, they observed some displacement in relative graphic peak location and height. Despite these variations, they obtained reproducible measurements. That is, although there was a certain variability in the data, they did not see peaks appear and disappear. They did not have a secondary measurement of particle size, but they were not off by as much as twenty to forty microns. The researchers did not know whether the testing process caused the alumina particles to agglomerate.

IBM believed its results and believed that there was some correlation between particle size and polishing results. IBM's decisions were driven by polishing results, not particle size. IBM did not rely on the particle size analysis to predict polishing results.

October, 1990. Lot 1656 was sold to IBM.

1991

STI sold alumina slurries to IBM (201 A/140, 201A/210, and 201A/280).

Mid-1991. Yancey, who had been in communication with Mark Jaso at IBM, now began communication with Dr. Frank Kaufman. They spoke first by telephone in reference to IBM's receiving STI's slurry samples for experiments. Yancey and Kaufman spoke up to several times per month from then until Kaufman left IBM in 1995.

December, 1991. Lot 2021 was mixed and sold to IBM.

1992

Yancey met Richard R. Cavagnaro. They met first by telephone, then by personal contact in the spring of 1992.

Cavagnaro worked for DeGussa Corporation from 1988 to 1996. From 1990 to 1994, he was the product manager for fumed silicas, including fumed metallic oxides such as aluminum oxide C.

Yancey had attempted to purchase aluminum oxide C powder from Cavagnaro.

Yancey had evaluated DeGussa's Lepandin 20N aluminum oxide C based dispersion with a small amount of alcohol added. Lepandin 20N is made of aluminum oxide C, methanol, and an organic auxiliary material.

DeGussa's technical bulletin number 56, issued in 1984 and containing a table entitled "Characteristics of Lepandin 20N" listing those characteristics, was handed out to customers in 1988 when Cavagnaro was employed by DeGussa. Cavagnaro gave a copy of this bulletin to Yancey.

The list of characteristics of Lepandin 20N in the bulletin includes aluminum oxide C, water, and methanol. It does not describe the organic auxiliary material, nor does it state that the pH was adjusted with a potassium hydroxide or caustic solution. A phrase following an asterisk below the "Characteristics of Lepandin 20N" chart states, "contains approx. 1% of an organic auxiliary material."

Cavagnaro recommended to Yancey that he make Lepandin 20N ÔÇö type dispersions.

Cavagnaro never told Yancey what the organic auxiliary material was. The material was added to improve the stability of the slurry. It was a surfactant, keeping materials in suspension, bridging the gap between methanol and water.

In November, 1990, DeGussa published an updated version of bulletin number 56. This bulletin describes aluminum oxide C but does not contain a recipe for Lepandin 20N. Cavagnaro gave this bulletin to customers and was doing so when he met Yancey. He also handed out a sheet describing how to mix slurries.

In the spring of 1992, aluminum oxide C was on allocation because of increased market demand relative to production capacity. Cavagnaro knew that Yancey was expecting to use quantities of aluminum oxide C that Cavagnaro would be unable to provide in Yancey's slurry development program for the electronics market.

Cavagnaro knew Yancey had bought only a ten-kilogram bag of aluminum oxide C by that time, but he did not know how much of that supply Yancey had used. Cavagnaro told Yancey that it was not cost effective to ship slurries across the ocean, that Yancey should make his slurries at home with aluminum oxide C powder.

May or June, 1992, until the beginning of 1994. This was the approximately 18-month period of allocation, during which DeGussa would not sell aluminum oxide C to STI.

Early 1992. STI mixed some fumed aluminas.

The source of the alumina from which STI made these dispersions was DeGussa Corporation. Yancey knew only of DeGussa as a source of fumed alumina. STI wanted to make fumed alumina dispersions and sell them commercially in 1992, but did not do so because STI was unable to purchase the raw material, aluminum oxide C, from DeGussa.

February 21, 1992. STI mixed two slurries from aluminum oxide C, the DeGussa product bought by STI. One of these slurries was made with 20% solids and 10% alcohol, the balance water. The other was made the same way but without the alcohol. The slurries were made from the recipe for DeGussa's Lepandin 20N. In mixing the two slurries, STI wanted to determine if it could duplicate the properties of the DeGussa product. Thus, STI mixed up the slurry with alcohol (DeGussa's recipe) and without alcohol (because STI did not need the alcohol for its purposes).

STI's production log has entries for lots STI-2084 and STI-2085 under the date, February 21, 1992. The entry for STI-2084 reflects an aluminum oxide C that has been cold milled, milled at low temperature, in a very high speed mixer at 20% solid alumina oxide material in the liquid dispersion, with a pH of 4, and alcohol added. This is a fumed alumina dispersion. The entry for STI-2085 is the same as for STI-2084 except that it does not list an alcohol component. A note following the description of STI-2085 states, "No change in particle size between milled, milled with alcohol or standard prep."

Yancey had a copy of DeGussa's bulletin number 56, containing the Lepandin 20N recipe, on February 21, 1992, and followed that recipe in mixing lot STI-2084. The formulation for STI-2084 in the production log book indicates the characteristics listed for Lepandin 20N in the DeGussa brochure. However, STI-2084 did not contain any auxiliary organic material, and STI-2085 contained aluminum oxide C (fumed alumina purchased from DeGussa), water, but no alcohol and no organic auxiliary material.

In 1992. Yancey unsuccessfully attempted to obtain commercial quantities of aluminum oxide C. Commercial quantities became available for purchase in 1994. In 1994, Yancey had a small quantity of fumed alumina in stock that he had purchased several years earlier. The fumed alumina slurries which Yancey took to the Semicon/Southwest trade show were made from this fumed alumina, purchased from DeGussa before 1992.

By late 1992. Yancey realized that CMP could be a good business for STI, in part because Kaufman had asked him what STI's production capacity could be. He sought to learn more about CMP, to attend trade shows and examine various products that were being used in the industry.

November 23, 1992. STI assigned lot number 2361 to a slurry mixed three months earlier. Lot 2361 is a combination of lots 2102 and 2237. Lot 2102 was mixed on March 9, 1992. Lot 2361 was mixed on July 16, 1992. The lots were combined in August, 1992. The lot number was assigned in November, 1992.

In 1992. STI had a MicroTrac machine. The MicroTrac tests dry alumina powder, which was the form in which the powder came to STI.

In 1992. STI had three employees: Yancey, a secretary, and a technician.

In 1992. STI sold 201 A/60, 201 A/140, and 201 A/280 to companies involved in the high end optics industry and the computer peripheral industry.

1993

First part of 1993. By this time, Cabot was selling oxide CMP slurries to the semiconductor industry for CMP on glass interlevel dielectric insulating layers, a fumed silica abrasive removing silicon dioxide.

In 1993 through the first quarter of 1994.

Bruce Zwicker of Cabot saw that there was a market for another type of CMP product, tungsten CMP. Oxide CMP slurries will not perform tungsten CMP because a different substrate is involved. Zwicker submitted a research proposal in Cabot's Cabosol division to develop a CMP slurry for the application of tungsten CMP. In tungsten CMP, circuit manufacturers want to use tungsten as an interconnecting layer between two layers to carry current rather than to insulate current. A rectangular area is etched out of a layer of glass, and tungsten is deposited atop the substrate. CMP then removes between 800 and 2000 angstroms of the tungsten layer down to a level at which the surface as a whole is planar. A goal of the effort was to achieve high selectivity, meaning a high polishing rate for tungsten and a low polishing rate for glass.

A research effort was begun.

Members of the research team were listed inventors of the Cabot patent (patent number 5,527,423). Neville, a PhD physicist, was director of research in the Cabosol division at the time. He is an expert in particle technology and fumed metal oxide technology. Fluck, who is a colloid chemist, formulated the slurry and worked on its stability. Hung, a chemical engineer and Cabosol process engineering specialist, was responsible for the manufacturing process of fumed alumina used to develop the abrasive for the tungsten slurry. Scherber and Dr. Lucarelli are chemists.

Cabot devoted approximately one year to developing the invention patented as the Neville patent (herein referred to as "the Cabot patent"). The research proposal was submitted in the spring of 1993, and work began in the fall of 1993. The first slurry samples based on fumed alumina were made in the period January through March, 1994. Late in the first quarter of 1994, Cabot sent samples to its customers.

Towards the end of 1993. Yancey hired a fourth person, Jim Fox, to take over the marketing and sales operation. It was at this time that STI became more interested in the CMP industry.

1994

January through March, 1994. Cabot made its first slurry samples. Late in the first quarter of 1994, Cabot sent samples of its slurries to its customers pursuant to a nonanalysis and nondisclosure agreement. Samples of the experimental tungsten slurries were sent to Motorola (fumed alumina) and, in May, to Sematech (fumed and precipitated alumina).

April, 1994. STI moved to its current facility in Monroe, N.C.

Spring, 1994. Prior to the Silicon/West trade show, STI started testing fumed silicas. In April or May, STI contacted DeGussa concerning silicon oxide and made inquiries about fumed alumina. STI wanted to get into the CMP business and approached DeGussa about it because DeGussa had products that were capable of producing the particles that STI needed to begin investigating this new business. STI wanted fumed alumina and fumed silica from DeGussa. DeGussa began to work with STI to give STI products and assistance in making these dispersions.

At this time, STI had precipitated alumina products. STI was interested in fumed alumina because it would be another option for customers. Fumed alumina dispersions had particle size distributions that were narrow as compared to other STI products and in a range that STI did not have (100-300 nm).

At this time, Yancey knew that Cabot was selling fumed silica products to the semiconductor industry. He was aware of the interest in the semiconductor industry in using products such as fumed silica.

June, 1994. Rodel began to sell MSW1000 as a commercial product. MSW1000 is a tungsten polishing slurry for CMP. Rodel claimed a 3-month shelf life for this slurry but did not say that it was stable or very stable.

June, 1994. By this time, various companies were using ferric nitrate as an oxidizer for chemical mechanical polishing. IBM used ferric nitrate for metal CMP, and this set the standard. IBM developed CMP in the mid-1980s and was the first company to do so.

Week of July 18, 1994. Yancey attended the Semicon/West trade show. STI did not have a booth at this show.

At Semicon/West, Yancey attended Cabot/Rippey's CMP seminar, "Planarization Technology: Chemical Mechanical Polishing (CMP)," in which Scherber of Rippey and Neville of Cabot presented papers. The seminar was held on July 19th from 8:30 a.m. until noon, and between 300 and 400 people attended.

The presentation discussed the results of a new slurry that Cabot and Rippey had tested, an alumina slurry used for polishing tungsten. The presenters did not go into detail as to the formulation of their slurries, and they did not mention fumed alumina.

There were ten or more twenty-minute presentations at that seminar. There were also presentations from IBM, Applied Chemical Solutions (Ed Ferri), Speedfam, Strasbaugh, and OnTrak.

Yancey did not hear the phrase "fumed alumina" at this trade show. As of the time of Semicon/West, Yancey believed that Rippey was using a precipitated calcined alumina in its polishing slurries. He believed this because for the past two years STI had been selling to Rippey and, later, to Cabot, increasing amounts of an aluminum oxide called M100, a dry gamma alumina powder, a precipitated calcined alumina with a large particle size between one and one and one-half microns. Rippey had told Yancey in many conversations that they were using M100 in their tungsten slurry polish.

As a result of the Cabot/Rippey presentation, Yancey wanted to get samples of the Cabot/Rippey alumina in order to analyze it. He wanted to determine its characteristics. He wanted to know whether the alumina was fumed or precipitated (although whether he could have determined this is uncertain).^[3]

During Semicon/West Yancey met for an hour with Sue Davis of Sematech to gain her assistance in understanding the industry. He wanted her support for STI becoming involved in a Sematech bench-marking that was coming up that autumn. Sematech would supply wafers that STI could use for polishing, to develop data for STI's products.

Davis supplied Yancey with names of industry contacts. She supported the idea of Yancey's going to Semicon/Southwest. She wanted to put Yancey in touch with Sematech personnel and suggested that STI join the Semiconductor Equipment Manufacturing Institute ("SEMI"), sponsor of the Semicon trade shows. (Jim Fox, STI's vice president for sales at that time, contacted the Sematech personnel suggested by Davis.)

At this time, STI sold abrasives but not oxidizers for CMP. At Semicon/West, Yancey discussed oxidizers, including ferric nitrate, with Ed Ferri of Applied Chemical Solutions. This was the first time that Yancey heard about ferric nitrate as an oxidizer for CMP.

July 22, 1994. Yancey wrote to Sue Davis, thanking her for meeting him at Semicon/West. He asked her for a bibliography of CMP papers. Referring to an article by Michael Martinez, he wrote, "I noticed a paragraph which reads, `CMP consumables are supplied primarily by Rippey and Rodel, both of which received development contracts from Sematech.'" He asked if Sematech was going to offer other consumables contracts (thus addressing her in her role as a Sematech employee). Further, he stated that STI was well ahead of other suppliers such as Cabot/Rippey and Rodel in the area of (precipitated) alumina dispersions.

July 26, 1994. Yancey wrote to Sue Davis, again inquiring about the bibliography of CMP papers, but making no reference to fumed alumina.

August 5, 1994. At the request of Jim Fox, Yancey wrote to Raul Jairath of Sematech. Davis was Jairath's supervisor. The letter refers to Yancey's attempts to contact Jairath by telephone. It references STI's 201 A/280 precipitated alumina product but does not mention fumed alumina.

August 18, 1994. Yancey again wrote to Sue Davis, thanking her for the CMP bibliography, and telling her that STI had joined SEMI and decided to exhibit at Semicon/Southwest. His letter refers to developing fumed silica sols and to having an alumina being used for tungsten CMP (here referring to IBM's purchase of precipitated alumina products from STI) but makes no mention of fumed alumina slurries.

Around August 20, 1994. Yancey attended an introductory overview CMP course at Rensslaer Polytechnic Institute ("RPI"), given by Swartz and Kaufman. There he learned that Rippey was using ferric nitrate as an oxidizer.

August 26, 1994. Yancey again wrote to Jairath. This letter speaks of STI's interest in getting more involved in the CMP slurry business including planarization of both oxide and metals. It states that STI had joined SEMI and would exhibit at Semicon/Southwest "to formally introduce our company to the device manufacturing industry." It refers to having attended an introductory CMP seminar given by Geraldine Swartz and Frank Kaufman at RPI. It refers to STI's development of silica sols, and "our nonsettling colloid alumina which we have used for years in ultra high technology markets." It does not make reference to a fumed alumina product.

As a result of this letter, a meeting between Sematech and STI was set for the day after Semicon/Southwest, October 20, 1994.

August 29, 1994. Entry for this date in STI's product log book: Lot number consecutively assigned the designation STI-3090, labelled "Experimental Silica."

August 30, 1994. Entry for this date in STI's product log book: Lot number STI-3091, labelled "Experimental 201 A/280," which is precipitated alumina.

October 6, 1994. Cabot patent application filed. Yancey did not know of the filing at this time.

October 12, 1994. Entry in STI's product log book: STI-3100 "assigned to Exp. Alum. Oxide C." "Alum" means alumina. Alumina Oxide C is a fumed alumina. Following this entry is the phrase, "See Paul's notes."^[4] As of October 12th, Yancey knew that Rippey was using ferric nitrate as an oxidizer for the chemical part but did not know that Rippey was using fumed alumina for the mechanical part.

Between July, 1994 (Semicon/West) and October 12, 1994. During this time period there is no log book entry in STI's product log book indicating that STI mixed up a fumed alumina. Between September 29, 1994, and October 6, 1994, STI was working with silica and precipitated alumina. October 12th is the first date in the STI log book showing that STI mixed up a fumed alumina after Semicon/West.

Between January 1, 1993, and October 12, 1994, STI mixed up no fumed alumina. Any fumed alumina slurry mixed up by STI would have been at least 21 months old when Yancey in his letter to Kaufman on October 25, 1994, stated that the slurry could easily be expected to have a shelf life of 6 months.

Between October 12, 1994, and October 18, 1994. Yancey did not attempt to polish any semiconductor wafers during this time period. At this time, Yancey did not have any facilities in-house for using the slurries to polish wafers. He still relied on customers to test his products.

October 18, 1994. Yancey attended the smaller Semicon/Southwest trade show held in Austin, Texas. STI had a booth at this show and tried to introduce its products to the semiconductor industry.

Yancey decided to take products which he had not manufactured or sold. He had not sold a fumed silica or fumed alumina slurry, but he had sold a precipitated alumina slurry and a cerium oxide. He decided to take fumed silica because of industry interest in using it to polish oxide layers. (He bought the silica dioxide from DeGussa.)

Yancey decided to include a fumed alumina slurry among the products he displayed at Semicon/Southwest. He did not manufacture the fumed alumina, but rather bought fumed alumina powder from DeGussa. He was influenced in his decision by the renewed availability of aluminum oxide C from DeGussa. The half-gallon sample he took (in a transparent container) was from the lot that he mixed on October 12th.

STI also displayed its traditional colloid aluminas, including the Ultra-Sol 200 and the 201 A/280. STI exhibited a cross-section of its product line, including some fumed silicas, precipitated aluminas, the fumed alumina, and a cerium oxide.

Yancey decided to take no oxidizers to the show because STI had never sold an oxidizer. STI did not want to sell oxidizers but did so because it was demanded by the industry in order to sell CMP slurries requiring abrasive and oxidizer components.

Cabot and Rippey had a booth at the show. They gave an invitation-only presentation, which was their commercial introduction of their new tungsten CMP slurry, named Semi-Sperse W-A355, which Cabot had developed in 1993 and 1994. Yancey was not invited to this presentation and did not attend. At the presentation, Cabot disclosed information as to the characteristics of their new product: The weight percent of fumed alumina in the slurry was 3; the viscosity of the solids content of the slurry was less than 5 centipoise (essentially, it flowed like water); it had a mildly acidic pH of 4; the particle size of the fumed alumina abrasive was approximately 100 nm (Cabot disclosed a particle size distribution of fumed alumina with a number mean of 94.3 nm, or approximately. 1 micron, and a distribution between approximately 0 and approximately 225 or 230 nm); and its shelf life, during which its properties would not change, was 3 months. (Cabot had conducted shelf life studies and had a formalized shelf life study under way as part of the Sematech project. It was not typical to quote a 3-month shelf life for a CMP slurry. One year was typical.) Semi-Sperse Fe 10, ferric nitrate 10% in the solution, was the oxidizer component of the slurry. Its viscosity was less than 5; its pH, 1.7; and its shelf life an atypical 3 months.

Yancey did not learn at Semicon/Southwest that this new product had a solid content of 3%, a viscosity of 5 centipoise, a pH of 4, a mean particle size distribution of around 94 nm, and a shelf life of 3 months.

Three Cabot or Rippey employees came to STI's booth and examined the dispersions. Yancey showed them STI's fumed alumina. They told him that Cabot was also coming out with a fumed alumina. This was Yancey's first awareness that Cabot was coming out with a fumed alumina.

October 20, 1994. Yancey and Fox from STI met with Jairath of Sematech. Yancey took the STI fumed alumina product to the meeting. STI's participation in Sematech's benchmarking trial was discussed.

October 21, 1994. STI mixed up lot 3113, "Alum. Oxide C Exp," a fumed alumina.

October 24, 1994. Yancey by letter thanked Jairath for meeting with him and for the opportunity to participate in benchmarking, and confirmed that STI would supply Jairath with some of STI's slurries.

Attached to the letter was a particle size distribution of the two materials which Yancey sent to Sematech. One is for XMet202 (the same product as Met202, "X" meaning experimental). The particle size distribution is centered around .1-.2 micron with virtually nothing about .5 micron, and with a small tail just over I micron (which Yancey believed to be anomalous). Yancey believed that there were no particles greater than I micron in size in this development sample.

October 25, 1994. Yancey wrote to Kaufman at IBM to give him information about STI's fumed alumina dispersion, XMet202 (Met202). Yancey hoped to persuade IBM to buy more. In the letter, Yancey says that STI had been studying the stability of its new fumed alumina product "for some time" but had only recently made a "testing grade" batch, which was lot 3100, mixed thirteen days previously, on October 12th. He says that he had not received the chemical data for lot 3100 but includes data from 201 A/280 (precipitated alumina), lot 1090, for comparison, noting that the difference is "dramatic," meaning between 201 A/280 and lot 3113.

The letter goes on to say that within 10 days Yancey would send to Kaufman "a full set of data on lot 3100" and lot 3113. The new information would contain chemical purity information on the alumina particles and TEM pictures.^[5]

Yancey observes that his fumed alumina product seem "very stable." "We can easily state a three month shelf life at this point, and I expect that it will easily pass six months." Yancey did not mix any fumed alumina slurry in 1994 before October 12th. When he says that he had been studying the stability of the alumina product for some time, he had in mind his 1992 slurries, which he had allowed to sit on the shelf to determine if they had a shelf life of a year or more.^[6]

Yancey attached to the October 25th letter two charts of particle size, one for XMet202 and one for 201A-280. The chart indicates that as to the mean particle size distribution for Met202, 50% of the particles have a particle size of .15-.17 micron.^[7] The particle size distribution for XMet202 indicates some greater-than-1-micron particles, which Yancey believes to be anomalous.^[8] The particle size range for 201A/280 included 3 or 4 major peaks going as high as 6 or 7 microns and as low as .2 micron.^[9]

October 28, 1994. STI sent to Jairath at Sematech at no charge 5 gallons of Ultra-Sol XMet202 fumed alumina dispersion, development sample let STI-3113.

In the weeks following October 25, 1994. In a series of voice mails to Kaufman, Yancey discussed ferric nitrate and the Sematech project. At that time, Kaufman was the IBM representative to the technical advisory board for the benchmarking project.

A voice mail said that STI was interested in taking part in the benchmarking study and for that needed the tungsten slurry; that STI had the abrasive and was looking for a list of possible oxidizers which could be used with an abrasive. Yancey asked Kaufman if he had any suggestions. Kaufman was surprised that Yancey would ask for advice on the identity of an oxidizer, which IBM considered confidential.

Several weeks later, Kaufman received another voice mail from Yancey to the effect that STI was proceeding and had determined from Sematech that ferric nitrate was a suitable oxidizer.

Kaufman never discussed with Yancey what materials IBM polished with STI's slurries, only that IBM was polishing a metal of some sort. Nor did Kaufman disclose anything about any additives or chemistries that IBM might or might not have added to what IBM received from STI. Nor did he discuss with Yancey IBM's method of preparing a slurry sample for particle size testing, nor how IBM actually made up a slurry, nor how IBM conducted the particle size analysis, including sonication (which tends to reduce particle size). Nor did IBM disclose to Yancey any information about any other slurries from any other company, including particle size analysis or polishing results.

By October, 1994, IBM had published that ferric nitrate could be used as an oxidizer in a CMP polishing system. It had done so in U.S. Patent 5,262,354, an IBM patent dated November 26, 1993: "metallization can be removed from the surface of a dielectric by chemical mechanical polishing with an alumina slurry in dilute nitric acid or by using other acidic slurries, or example, ferric nitrate." The patent does not refer to tungsten. There is no evidence that Yancey knew of this patent. Yet use of ferric nitrate was in the public domain, and Yancey testified that he heard that IBM was using ferric nitrate.

In the autumn of 1994, Yancey knew that Rippey, IBM and AMD were using ferric nitrate as an oxidizer. Ferric nitrate was the first oxidizer he selected for STI's product line, and he selected ferric nitrate in the period after he learned that these other companies were using it.

November, 1994. Cabot first offered to sell for profit slurries falling within the claims of its patent.

December 5, 1994. Yancey wrote to Jairath acknowledging the invitation to participate in the benchmarking test. The second paragraph refers to oxidizer additives:

We have decided to participate with one of our developmental products. Ultra-Sol Met202 Fumed Alumina CMP Dispersion, which we discussed during our meeting and followed with a five gallon sample. In discussions with other companies in the CMP business, Applied Chemical Solutions for one, it has become apparent that the oxidizer additives should be made a part of the slurry package. With this in mind, we are also submitting a 20% solution of ferric nitrate nonahydrate for use with the Met202 Alumina. If you prefer another type of oxidizer system we would be happy to package your preference in a ready use, volumetrically measurable solution.

Yancey did supply the ferric nitrate oxidizer. Jairath did not request a different oxidizer system. At this time, STI was not selling oxidizers for CMP.

Along with this letter, Yancey sent a specification sheet for Met 202 alumina CMP dispersion, an x-ray diffraction trace of the aluminum oxide C in the material, and an application note showing how to dilute various abrasive and oxidizer concentrations to form desired solids and oxidizer concentrations in a slurry (a dilution table). Yancey wrote the application notes around the time he sent this letter to Jairath in December, 1994. It was Yancey's belief at the time that ferric nitrate and hydrogen peroxide oxidants were commonly used in the industry.

December 14, 1994. The first sale occurred of a Cabot slurry that fell within the claims of the patent.

In 1994, IBM did not order any Met 202 from STI.

1995

January or February, 1995, through Spring, 1996. STI participated in Sematech's benchmarking trials. For these tests, two of the three STI products were fumed alumina (Ultra-Sol 200A and Ultra-Sol Met 202), used to polish aluminum. The third, Met 201, was a precipitated alumina, used to polish tungsten and copper. Sematech provided the test wafers. Yancey was the lead engineer in the trial.

Early 1995. STI began active marketing of Met 202. Between 1996 and 1999, STI sold approximately $130,000 of Met 202. Met 202 was marketed by sales agents in Japan, the West Coast, Texas, Florida, and Europe. STI used the results from the Sematech benchmarking trial in its marketing.

June or July, 1995. Yancey learned that IBM was unhappy with STI's boehmite products. IBM did not order any Met 202 from STI in 1995.

Mid-1995. Customers began demanding an oxidizer with the slurries, and STI started selling an oxidizer. The Met 202 and the oxidizer were marked separately.

Late 1995. Yancey decided to apply for patent protection as part of his overall plan to obtain investors in STI. (Yancey had consulted with patent attorneys, applied for and obtained patents between 1970 and 1982.) In December, 1995, Yancey met with his attorneys at Bell, Selzer, Park & Gibson ("BSPG") (Drew Meunier and Ron Linker) in connection with filing a patent application.

1996

Around March 11, 1996. STI took on some investors for the first time (Aurora Ventures). They invested around $700,000 initially. STI hoped that this investment would allow STI to play a dominant role in the CMP slurry business.

Yancey found them to be "greedy." They took over three of the six seats on STI's Board of Directors. Yancey was unpleasantly surprised when, shortly thereafter, he found that they wanted him to sell the company. Yancey himself did not want to sell the company. The investors were not happy about STI's being sued by Cabot.

Once in March and once in April, 1996, Yancey contacted Linker and BSPG about a Cabot patent or patent application. Linker thus knew that Cabot had a patent or an application. BSPG searched the databases for the Cabot patent or application.

March 11, 1996. Yancey contacted Meunier, and they conferred regarding accusations made by Cabot to customers of STI that STI was using Cabot technology.

At this time, Yancey was trying to increase STI's sales of Met 202.

In response to that conversation, in March and April, 1996, Meunier conducted at least two database searches to see if he could locate a Cabot patent or application. He found nothing relevant.

On or before March 18, 1996, Yancey told BSPG about his sales to IBM and his use of ferric nitrate as an oxidizer.

March 18, 1996. Yancey sent a fax memo to Dan Sullivan in IBM's legal department regarding IBM's possibly having intellectual property rights `in a slurry system that STI wanted to make for CMP planarization.' Yancey sent a copy to Meunier, and Linker confirms that BSPG received this copy. The point of Yancey's memo to Sullivan was to check with IBM to make sure that STI did not transgress IBM's intellectual property rights. In the memo, Yancey stated, "[O]ur main concern at this time is clearing the use of the two components, alumina and ferric nitrate."

In the second paragraph, the memo states:

... In about 1989, IBM contacted me regarding using our products in some undefined use in the IC fabrication area. We have been selling an alumina product to one of IBM's groups ever since that `initial contact.... Since ferric nitrate was the one used in the industry, and had no known patent constraints, we developed our process around this combination ÔÇö colloidal alumina combined with ferric nitrate oxidizer.

Thus, sometime around March 18, 1996, Yancey provided BPSG with information about his sales of an alumina product to IBM. He also informed BPSG that he believed that the had developed a process that was a combination colloidal alumina combined with ferric nitrate oxidizer.

Meunier received a copy of this letter in March, 1996. This letter did not lead Meunier to conclude that STI had been selling boehmite alumina to IBM since 1990. The letter does not reference boehmite alumina.

By March 18, 1996, Yancey had known for two years that IBM was using ferric nitrate as an oxidizer. He knew that IBM was buying boehmite alumina from STI and that they were using it to polish metal. He knew that IBM was using many other aluminas. He did not know specifically that IBM was using ferric nitrate with STI's alumina.

April, 1996. At a trade show in Europe, Cabot employees approached one of STI's board members. They told him that STI was selling a product as to which Cabot had patent protection and that Cabot intended to act upon it to put STI out of business.

Hearing this, Yancey again consulted BSPG.

April 2, 1996. At Yancey's request, Meunier conferred with STI Board of Directors member Gerry Cutini regarding Cabot's claims that STI was using Cabot's technology. Cutini had told Yancey of Cabot's claims. This conversation triggered Meunier's second database search for patents that were assigned to Cabot. Nothing relevant was found.

April 10, 1996. In a telephone conference, Yancey and Meunier discussed a new patent disclosure for a copper polish made up of boehmite alumina and ferric nitrate. The boehmite solution was that which STI had sold to IBM since 1990. The invention was the combination of boehmite alumina and ferric nitrate.

In this conversation, Yancey told Meunier that he had been selling to IBM a copper polishing compound without the ferric nitrate oxidizer. As of April 10, 1996, Yancey had known for several years that IBM was using the alumina that they bought from STI for CMP and for polishing metal.

Yancey recalls having told Meunier in this or a later related conversation prior to June 17, 1996, that he had sold boehmite alumina slurries to IBM on a continuous basis from 1990 until 1996. He did not send documentation of these sales to Meunier. Meunier does not recall it being his understanding as of April 10, 1996, that he knew that STI had sold boehmite alumina to IBM since 1990.

In Meunier's April 11th memorandum to the file in reference to this conversation, Meunier states, "Paul Yancey would like to go ahead and pursue patent protection for this particular copper polishing compound in combination with the ferric nitrate oxidizer in the process of polishing copper." This is a reference to the specific copper polishing compound without the ferric nitrate oxidizer that Yancey had been selling to IBM.

After the April 10th conversation with Yancey, at Yancey's instruction, Meunier started doing work as to patents he had already found and looking for additional patents to address the issue of whether the copper polishing compound was something that could be patentable. Yancey told Meunier on April 10th that the copper polishing compound is what Yancey had sold to IBM without the ferric nitrate oxidizer.

April 19, 1996. On this date there was an office conference between Meunier and Yancey to discuss issues regarding preparation of the patentability opinion that was issued on April 30th regarding Yancey's copper polishing compound in combination with a ferric nitrate oxidizer. Yancey told Meunier that his invention utilized an alkoxide based alumina, and they discussed its use in combination with an oxidizer for polishing copper.

April 30, 1996. Meunier wrote to Yancey regarding the patentability investigation regarding chemical mechanical polishing compositions for copper and aluminum, and giving the opinion that Yancey's CMP compound for copper should qualify for patent protection. The invention was a combination of a ferric ion oxidizer and an alkoxide-based alumina, the combination of the two. This combination reduces the extremely corrosive effect of the oxidizer when polishing copper. It reduces the static etch rate. The patent application filed and ultimately abandoned by Yancey was directed to the invention described in this letter. The subject of Claim 1 of this application was not Yancey's invention (combination of boehmite alumina and ferric nitrate) in that it did not include the combination of boehmite alumina and ferric nitrate.

Yancey believed that the patent application he was filing was for the combination of boehmite alumina and ferric nitrate. Yet the April 30th patentability opinion is in one place so worded as to suggest that the writer thought of the invention as simply the use of an alkoxide-based alumina for the removal of copper or aluminum via planarization. Under "Evaluation of Prior Art References," Meunier states:

Based on the patentability search, it is our opinion that none of the above noted references, taken either individually or in combination, discloses or suggests the components in your CMP polishing slurry for polishing copper or aluminum. In particular, none of the references disclose or suggest the use of an alkoxide-based alumina for the removal of copper or aluminum via CMP planarization.

However, earlier on, in "The Invention" section, Meunier writes:

The slurry composition of the invention contains a ferric ion oxidizer and an alkoxide-based alumina. The addition of ferric nitrate or other ferric salts provides the ferric ion which acts as the oxidizing agent in the slurry. The alkoxide-based alumina contains reactive sites which are believed to tie up the nitrate ions allowing ferric ions to freely oxidize the metal being polished. As a result, corrosion of the metal on the wafer is greatly reduced.

In the final paragraph of the letter, Meunier asks that Yancey "provide us with more detailed information on the alumina used in your CMP composition."

As of April 30, 1996, Meunier had not understood that STI had been selling boehmite for several years.

May 13, 1996. Yancey faxed to Meunier the Vista Chemical brochure for DISPAL alumina powders and alumina sols, describing the alumina that STI had been using in its slurries. The DISPAL alumina powders are made by Vista Condea. Boehmite alumina is described in this brochure. At the bottom of the first page that was faxed to Meunier, yancey wrote, "We use a 14N 4-25 (140 m2/gm, nitric, pH 4, 25%)," pointing out to Meunier what Yancey used in the formulation for the copper slurry. In the printed section, left-hand column, of the first page, there is the following sentence: "The alumina in all DISPAL powders and sols consists of boehmite crystallites produced via Vista's alkoxide chemistry." Before sending the brochure to Meunier, Yancey drew a circle around this sentence to highlight it. Yancey wrote at the bottom and drew the circle to inform BSPG of the boehmite alumina sales.

A table on the third page of the Vista brochure listed average particle size for DISPAL alumina powders.

Yancey believed that he had provided full details of his prior sales of the boehmite alumina slurries and the use of ferric nitrate as an oxidizer to BSPG before they prepared the patent application. He relied on BSPG to prepare the patent application directed to his invention of the combination of the boehmite alumina and the ferric nitrate oxidizer.

When Meunier received this document, he was working on the patent application for the copper polishing compound.

As of May 13th, Meunier thought that the invention was directed to the use of a specific type of alumina, and that was boehmite alumina; and further, that the invention did not include the ferric nitrate oxidizer, and that was a preferred embodiment. Yancey told him that the invention included the ferric nitrate oxidizer and was a combination of ferric nitrate oxidizer and alkoxide-based alumina, as in the patentability opinion.

May 14, 1996. Yancey wrote to Meunier describing experiments run by STI on copper wafers, for inclusion in the patent application. In the five experiments, the alkoxide-based alumina is combined with ferric nitrate in various concentrations. Yancey understood his invention to be a combination of ferric nitrate and his 201A/ 280 product. In his description of the second experiment, Yancey states, "[T]he 201A/280 offers a significant advantage over standard alumina slurries made from gamma aluminas using ferric nitrate as the oxidizer."

June 17, 1996. Yancey's attorneys found a published European patent application (EPO 708160A2) on metal oxide slurries including fumed silica and fumed alumina. The patent application was for a chemical mechanical polishing slurry for metal layers. They found no U.S. counterpart but assumed one was pending.

June 17, 1996. BSPG attorney Raymond O. Linker filed for Yancey and STI a patent application regarding a formulation for a CMP slurry. Meunier drafted the application. Yancey reviewed the application and signed under oath the "Declaration and Power of Attorney for Patent Application," declaring that he was the first and sole inventor of the invention.

The application claims an invention broader than ferric nitrate oxidizer in combination with an alkoxide-based alumina, in that in Claims 1 and 12 boehmite alumina by itself is also claimed. (Yet Meunier refers in his April 11th memorandum to a statement by Yancey on April 10th that his invention is ferric nitrate oxidizer in combination with a particular copper polishing compound.)

Meunier did not understand Yancey to have told him prior to June 17th that STI had been selling boehmite alumina abrasives to IBM for polishing metal conductive layers since 1990 or that STI had been selling boehmite alumina abrasives before the critical date, June 17, 1995. Meunier remembers Yancey telling him about the boehmite alumina sales after June 17th.

In connection with filing this patent information, Linker relied upon Yancey to provide accurate information regarding his prior art products as they related to the patent application. By July 2d, Linker knew that STI had sold boehmite alumina for polishing metal layers. When he found out that STI had sold boehmite alumina abrasive in a slurry to polish metal conductive layers, Linker concluded that that claim in the application should not be pursued and that the sales should be disclosed to the Patent Office.

Claim 1 claims a process for polishing the metal surface of a semiconductor device. Yancey did not invent the Claim 1 process.

Claim 12 reads, "A chemical mechanical polishing slurry for polishing metal conductive layers comprising a boehmite alumina abrasive." The claim makes no reference to an oxidizing component or gamma alumina. As of June 17, BSPG had conducted a search to determine whether the invention described in Claim 12 was a new invention and did not see it described in any of the patents they located. The review is set forth in Meunier's April 30th letter to Yancey. When the application was filed, Meunier believed Claim 12 to be a patentable invention.

When the application was filed, Yancey knew that IBM was using gamma alumina powder with ferric nitrate to polish metal, but he did not know if IBM was using ferric nitrate with STI's boehmite product to polish metal. Yancey knew as of 1994 that Rippey and others were using ferric nitrate.

The description and every example in the application describe a combination of alkoxide based slurry with ferric nitrate.

Late June, 1996. Yancey displayed his products and data from the Sematech benchmarking at the San Jose, California, trade show. Shortly after Yancey returned, he learned that the Cabot patent had issued. The very next day, a letter arrived from Cabot attorney Faye Morriseau saying that Cabot had sued STI in California.^[10]

Upon receiving notice of the suit, Yancey consulted Linker and Meunier. He faxed the complaint to Linker and met with them the next day. At this meeting, they reviewed the complaint and the patent and discussed invalidating the patent with prior art and obtaining a license, as well as the possibility of joint testing.

June 18, 1996. The Cabot patent issued. The claims in this patent are slightly different from those in the European application.

At Column 4, line 21 and following, the patent reads:

In order for the polishing slurry of the present invention to be an effective alternative to conventional slurries, it is important that the aggregates of the metal oxide particles are uniformly dispersed in a stable aqueous medium. By uniformly dispersed is meant that the aggregates are isolated and well distributed throughout the medium. By stable is typically meant that the aggregates will not re-agglomerate and settle out (e.g., form a hard, dense sediment). In a preferred embodiment, the aggregates will remain stable for at least a three month period of time. Critical to achieving slurry stability, it has been further discovered that the metal oxides particles of the present invention, in addition to having an aggregate size distribution less that [sic] 1.0 micron, have an average or mean aggregate diameter of less than about 0.4 micron and that the particles of the present invention have a force sufficient to repel and overcome the van der Waals attractive forces between the particles. The mean aggregate diameter refers to average equivalent spherical diameter when using TEM image analysis, i.e. based on the cross-sectional area of the aggregate. By force is meant that either the surface potential or the hydration force of the metal oxide particles must be sufficient to repel and overcome the van der Waals attractive forces between the particles.^[11]

Within a few days, or at most a few weeks, of the issuance of the Cabot patent, Budinger discussed STI's retains with Yancey. Knowing that Yancey had worked with IBM, Budinger called Yancey to see if STI had any retains that could invalidate the patent. Budinger knew that Cabot had sued STI and was very concerned that should Cabot take over STI, Rodel would lose access to the retains.

Budinger and Yancey also discussed Rodel's desire to obtain an invalidity opinion regarding the Neville patent. (They did not discuss the possibility of an invalidity opinion being prepared by anyone else.) Budinger told Yancey that if Yancey would give Rodel the retains, Rodel would give Yancey a copy of the opinion. Yancey agreed to give Rodel, and Rodel's attorneys at the firm of Foley & Lardner, access to the retains. Yancey did not restrict Foley & Lardner's ability to test the retains. The retains provided by STI were used as part of the basis for the conclusions reached by Foley & Lardner in its opinion.

June 19, 1996. Meunier and Linker provided Yancey with copies of the abstract European patent application and the application itself.

June 20, 1996. Yancey read the application and understood it. He understood that Cabot was claiming particles in a size range of 1 micron or less, with a mean aggregate diameter of less than .4 microns, and that the application discussed the importance of a stable slurry, which it claimed, one in which the aggregates would not reagglomerate, settle out, and form a hard, dense sediment.

June 21, 1996. At BSPG's offices, Yancey went through the European patent application's description and claims with his attorneys, Linker and Meunier, compared it to STI's products, and decided that although Met 202 fell within the scope of the claims of the European patent application, other STI products might invalidate it. At that meeting, they decided that they would need an opinion regarding Met 202 and invalidity to address the European and the presumed U.S. applications. They did not know at that time that Cabot's U.S. patent had issued on June 18th.

At this meeting, Yancey and the BSPG attorneys reviewed the claims in the European patent application one by one. A number of claims at the end of Cabot's European patent application set out ranges for particle size, among other things. At the June 21st meeting, Yancey provided his attorneys with numerous product descriptions, including Met 201, having aggregate particle distributions and mean diameters which would not fall within the ranges claimed in Cabot's European patent application but had been sold for a long period of time. Yancey told his attorneys about a product that had a bimodal size distribution, some of the particles of which would be within the range and others would be outside the range. Yancey did not provide them with prior art from his sales which would fall within the ranges of the Cabot European application. As a result of their discussion, Yancey believed that Met 202 would infringe claims of the European patent if it issued in the U.S., but he felt that these claims would be invalid in light of STI's Met 201 family of products (201A, 201A/60, 140, 210, 280), which he was selling to IBM at that time.

Also at this meeting, Yancey and his attorneys reviewed the prior art patent search conducted in connection with Yancey's copper polishing slurry compound to determine whether any of the patents that BSPG had discovered could serve as a basis for invalidating what Cabot was claiming to be a new invention. This included the Cadien and Rosstoker patents. It was discussed that the prior art patents that BSPG had already discovered and reviewed were not specific enough to serve as a basis for invalidating the Cabot patent. They discussed the fact that these patents may not describe the abrasive particles with as much specificity as the Cabot application and that, therefore, it would be difficult to use these references to show invalidity.

At some point during this time period, Budinger enlisted the assistance of investment banker and Rodel director Gary Hultquist in connection with evaluating STI as a potential subject of acquisition. Budinger sought Hultquist's expertise as an investment banker to evaluate STI, and for Hultquist to negotiate with STI, which at that time was a potential competitor of Rodel. STI was willing to talk with Hultquist. Hultquist is not a patent litigation specialist and was not engaged by Rodel for that purpose in dealing with STI.

June 23 or 24, 1996. Yancey learned of the Cabot patent-in-suit, U.S.Patent No. 5,527,423 ("the Cabot patent" or "the Neville patent").

Prior to filing the instant suit against STI, Cabot never informed STI of the existence of the patent or of the patent application that matured into the patent. Nor did Cabot request STI to cease and desist making, using, or selling Met 202. The suit was Yancey's first notice that Cabot had been issued a patent in this area.

June 25, 1996. Yancey received a copy of Cabot's complaint in the instant suit. Upon obtaining this notice, Yancey called BSPG, faxed them a copy of Cabot attorney Fay Morriseau's correspondence notifying him of the suit along with a copy of the patent, which he read and understood, and scheduled a meeting. In the meantime, at his attorneys' direction, Yancey collected documents and retains. He collected documents having to do with his business with IBM in the early 1990s.

Within a few days of receiving notice of the suit, it appeared to Yancey that Met 202 infringed claims of the Cabot patent.

June 26, 1996. Yancey met with Linker and Meunier. They discussed Cabot's suit against STI. Yancey took with him to the meeting the invoices from selling STI aluminas from the early years when STI was selling to IBM. He took several retains. He brought whatever he could find on the old sales and materials showing that years before he was selling a product that was similar to that described in Cabot's patent. He discussed these material with Linker and Meunier. Yancey said that he thought that he had prior art that would invalidate the Cabot patent. Meunier did not advise him on that date that the Cabot patent was invalid.

Also at this meeting they discussed Yancey's marketing at the upcoming Semicon/West trade show.

June, 1996. Rodel contacted STI about merging with Rodel. These merger discussions only lasted a few days. They resumed shortly thereafter. The discussions were on again, off again. (STI was sold to Rodel on September 24, 1997.)

By mid-1996, STI had twenty employees. STI's 1996 annual sales were around $900,000. In 1996, Cabot's sales were around $1.85 billion.

July 1, 1996. Following the June 26th meeting with his attorneys, Yancey found information on prior sales of products, primarily IBM data on particle size distributions, and ran some particle size distribution and zeta potential tests on available STI product. On July 1st, Yancey took this information to BSPG. He discussed this information with Linker and Meunier.

As to particle size distributions, Yancey brought test results from an Horiba LA-910 laser scattering analyzer. The particle size distribution tests were run on lots 2021, 2022, 3133, and 3563d.

The 201A/60 lot 3653 displays a course tail at around one micron. Yancey pointed this out to Linker and Meunier and told them his conclusion that the course tail was an anomaly of the testing procedure.

Yancey also brought the October 10, 1990, IBM Coulter distributions for lot 1656 and told his attorneys why he thought that the course tail was anomalous, arising from the testing procedure, although the particle size analysis from a volume basis for this lot indicated that almost all of it was greater than 1 micron.

Lots 2021 and 2022 were old lots sold to IBM in 1990 or 1991. Lots 3133 and 3563d were more recent. Yancey brought them to show that the materials were essentially the same then as five or six years earlier. Yancey did not tell his attorneys that lots 3133 and 3563d were slurries that could be used in invalidating the Cabot patent.

Yancey took product samples with him to this meeting to show that STI had been selling these small particle slurries in 1993 or before. He also brought sales records which indicated sales of the 201A family, including of lots 2021 and 2022, to IBM on January 8, 1992.

Yancey discussed the course tail phenomenon, which he deemed anomalous, with his attorneys. He told them that it is caused by improper mixing procedures. Had the sample been prepared properly, he told them, there would have been no course tail.

Linker and Meunier told Yancey to provide additional information about lots 1560 and 1625.

In this meeting, Yancey and his attorneys went through the claims of the Cabot patent-in-suit.

July 2, 1996. At around four o'clock, Yancey met with Linker and Meunier, and this time also with Larry Jones. As of this date, the attorneys did not have a file history of the Cabot patent. They had the information about alumina particles in Yancey's prior sales, including the source of the particles (Vista) and Yancey's statement that the alumina was of high purity.

Linker relied upon Yancey to give accurate and complete information.

It was Linker's belief that the retains tested by Yancey on July 1st had neither partially gelled nor partially settled.

Yancey told his attorneys at this meeting that the retains were stored in 8- to 32-ounce plastic containers sealed under a screw sealing cap and kept in an attic; that some of the retains were still in good shape; and that some of the retains thus stored may have settled or partially settled, and may have gelled or partially gelled.

At this meeting Yancey gave his attorneys particle size analyses for six retains and told them that they were stable. However, he drew the retains for lots 1625 and 1656 in 1990, recalls that they were stable at that time, but had not looked at the retains in preparation for this meeting. He drew the retains for lots 2021 and 2022 in December, 1991, but did not examine them again before the meeting. The retains were stored close to his office.

Linker did not look at the retains. Yancey said that the prior slurries were colloidal dispersions, from which Linker inferred that they were uniformly dispersed. Yancey told Linker (without written documentation) that his prior art slurries had a force sufficient to repel and overcome the van der Waals forces between the particles, rendering the slurries stable. Linker relied upon this statement.

They discussed use of STI's slurries to polish metal. Nothing in writing was produced in reference to this discussion, e.g., to record the use of STI's slurries to polish a tungsten layer which further comprised at least one underlayer selected from the group consisting of titanium, titanium nitride, titanium tungsten, or mixtures thereof.

As of July 2d, Linker understood that the prior art slurries Yancey was relying on were precipitated aluminas. Linker further understood that oxidizing components and surfactants were commonly used in slurries of this type.

Yancey brought a list of additional customers and documentation. This list was derived by going through all the invoice files from October of 1993 (the cutoff date, before which products could be used to invalidate the patent-in-suit; thus, lots 3133 and 3563 are not on the list because they were produced after October, 1993) and compiling all records of 201 product sales. The list contained dates of sale, the surface area of the alumina in the product, the particle size, the zeta potential in milli-volts, the percent abrasive in the slurry, the lot number, and whether the use was for optics or metal film. On this list beside the entries for lot numbers 1625 and 1656 there is the comment "M. Jaso, see IBM PSD's," PSD standing for particle size distribution.

On the Plaintiff's Exhibit version of this computer printout (# 77), but not on the Defendants' Exhibit version (# 134), written notations appear under the column entitled "Comments," and one notation appears to the right of that column beside the entry for lot 1656. These notations were made by Yancey and are in his handwriting. Check marks were placed beside lots whose retains were still available in 1996, after the retains for these lots were located. These marks were placed during the first two weeks of July, 1996 (1560, 1982, 2021, 2047, 2361 (3 entries), and 2557).

Yancey added the remainder of his notations to the list just before the Ricerca testing in late October, 1996, and after he had received BSPG's September 11, 1996, opinion. Beside lot 1656 of 201A, Yancey wrote "partial gel." (Lot 1656 is referenced in the BSPG opinion.) Beside lot 2021 of 201A/210, Yancey wrote "partial settle." (Lot 2021 is referenced in the BSPG opinion.) Yancey wrote "OK" beside the three entries for lots 2361 (201A/240) and for 2557 (201A/60). By "OK" Yancey meant that the retains showed no degradation. Beside the lot 2361 entries is noted "OK." (As to lot 1560, despite the production log note that the lot had been destroyed, there was a bottle that was not destroyed.)

At the July 2d meeting, Yancey and his attorneys discussed the details of all of the materials that Yancey had collected with a view to whether the Cabot patent could be shown to be invalid. They discussed whether STI's prior sales could be used to invalidate the patent. Four or five lots were selected for use in invalidating the patent (including lots 1625, 1656, 2021, and 2361), on the basis that they represented different products and products sent to at least one company other than IBM. Yancey was instructed to see if he could find the retains for these lots for further testing. Yancey told his attorneys that retains were taken and stored from every batch of material made by STI. He told them how this was done.

At the July 2d meeting, Yancey and his attorneys worked through the claims of the Cabot patent.

At the July 2d (or, perhaps, the July 1st) meeting, Yancey asked his attorneys whether STI could continue selling Met 202. Yancey's primary concern was whether he could market Met 202 at the upcoming Semicon/West trade show. His attorneys told him to stop selling Met 202 for a brief period, including July 5th. They told him that he could attend the trade show and continue to market his products.

At the July 2d meeting, for the first time, Yancey was informed, verbally, by his attorneys that they had an opinion that the Cabot patent was invalid or that there was considerable doubt as to its validity. They told him this after discussing the information Yancey had provided and reviewing the claims, and after Yancey had told them that STI's prior products had all of the features of what was described in the Cabot patent. Jones and Linker verbally told Yancey that it was their opinion, having gone through the 39 claims of the patent, that each claim was invalid. Even before they gave him this opinion, Yancey believed that the Cabot patent was invalid.

As of July 2, 1996, Yancey believed that Met 202 fell within claims 17, 18, 26, 27, 29, 31, 32 (when sold with an oxidizing metal salt, ferric nitrate), and 39, of the Cabot patent.

As of July 2, 1996, Yancey knew that retains from lot 2021, 2022, 3133, and 3563 had been tested on an Horiba machine at STI.

Yancey's attorneys told him at this meeting that the information he had presented having to do with prior sales appeared quite relevant to demonstrating the invalidity of the Cabot patent. They discussed what could be done to test the assertion that the retains satisfied the parameters of the claims, to build to the conclusion that the patent was invalid.

Between June 18, 1996, and July 2, 1996, STI did not sell Met 202. Having met with his attorneys on July 1st and 2d, Yancey understood that he needed to obtain an opinion on the validity of the Cabot patent. He understood that he would obtain such an opinion from BSPG.

July 2, 1996. STI shipped a quart sample of Met 202 without charge to the Materials Analysis Group of Phillips Semiconductors.

July, 1996. During this month, Yancey was in California for three weeks.

Between July 2, 1996, and September 11, 1996, Yancey did not look at the retains for lots 1656, 2021, 2022, and 2361. They were stored approximately 100 feet from his office.

Between July 2, 1996, and October, 1996, Jones first saw the retains in storage at STI.

July 5, 1996. Larry Jones wrote to Fay Morisseau, telecopy to Yancey, claiming to be able to invalidate the Cabot patent and offering to test STI's products jointly:

We have already determined, however, that Solution Technology made and sold, on multiple occasions prior to the patent's critical date, polishing slurries which included particles having the characteristics described in the claims of the patent.

And later in the letter, he stated:

Our client's December 1991 production and sale of Lot No. 2021 is one of several instances which, if an Answer to the Complaint is required, would support an allegation that the claims of the '423 patent are invalid inasmuch as the claimed invention was on sale in this country more than one year prior to the application's filing date (i.e., before October 6, 1993). Thus, we suggest that your client dismiss its Complaint before it becomes necessary for us to file such an Answer.

If, prior to dismissing its Complaint, your client requires further corroboration of Solution Technology's pre-critical date activities, we offer the following proposal. Solution Technology would provide, for each of several pre-critical date batches of the claimed slurry, copies of the purchase orders and invoices for the relevant products (which invoices reflect thereon Solution Technology's Lot Numbers) and an opportunity for analysis of the retains from those Lot Numbers by an independent third-party laboratory to be selected by agreement of our clients. Solution Technology would bring to the laboratory the retains in the containers in which they have been maintained and from which samples may be withdrawn for testing by the laboratory....

This letter was Yancey's first written notice from his BSPG lawyers that they had formed an opinion that the Cabot patent was invalid.

At this time, Yancey was willing to share STI's retains with Cabot. He was only unwilling to share the sources of his raw materials and some proprietary business information. The only other restriction was that he wanted to share the retains through a third-party professional analytical laboratory which would conduct the tests via a mutually agreed protocol. Yancey's purpose was to show Cabot invalidating prior art with a view to settling the lawsuit.

July 9, 1996. STI sold five gallons of Met 202 to OnTrak. This was STI's first sale of Met 202 after receiving notice of the Cabot patent.

July 10, 1996. Yancey told Larry Jones by telephone that he had found the retains for lot numbers 1625, 1656, 2361, and 2021. He found them by directing his employees to go through dozens of boxes looking for the lot numbers on the retains. The retains were not stored in numerical order. The lot numbers were chosen by Yancey in his discussions with his attorneys because they represented different STI products and customers.

After telling Jones that he had the retains, Yancey had them analyzed in STI's laboratory as to surface area, particle size, and zeta potential. He sent this information to BSPG in the form of particle size distribution curves.

July 11, 1996. Fay Morisseau responded to Larry Jones' letter of July 5th, rejecting the offer of joint testing. In his response, he stated:

As you know, a patent is presumed to be valid when it is issued. This presumption of validity applies to the Cabot patent which Cabot contends Solution Technology has infringed. To overcome this presumption of validity, Solution Technology must prove by clear and convincing evidence that the Cabot patent is invalid. Your letter of July 5, 1996 falls far short of this standard.

Cabot cannot evaluate Solution Technology's contention of invalidity without reviewing all evidence relevant to Solution Technology's allegations. To this end, you proposed to provide Cabot additional information and an opportunity to test product samples retained from these sales. Cabot wishes to proceed along the lines of your proposal and to accomplish this evaluation as quickly as possible. Accordingly, Cabot proposes that Solution Technology agree to provide the following information and documents to Cabot by Friday, July 26:

1. A list showing all product sales which are relevant to Solution Technology's allegation of invalidity....

2. All documents concerning the sale and delivery of the product which are relevant to Solution Technology's allegation of invalidity....

3. Copies of all advertisements, product bulletins, specifications, ..., for the products which are relevant to Solution Technology's allegation of invalidity.

4. All test reports and test information for these products, as well as all information which would relate to aggregate size distribution, mean aggregate diameter, zeta potential, and surface area of the particles.

Cabot can agree to proceed in this informal manner, as you have suggested, only if Solution Technology is able to produce the requested information and documents within the time frame indicated above. Hopefully, this will give Cabot time to review the information and the documents Solution Technology produces and, if necessary, take the testimony of representative of Solution Technology knowledgeable concerning the conception, reduction to practice, testing, manufacturing, shipping, retention of samples, selling, delivery, invoicing, and/or use of the product as well as other relevant areas requested by Cabot in order to learn the complete history of the development, marketing and sale of this product.

Additionally, Cabot may wish to take the testimony of representatives of buyers such as IBM to verify the information provided by Solution Technology. We would expect Solution Technology to co-operate with Cabot in securing the availability of such representatives.

Finally, Cabot may need to test the "retains" Solution Technology identifies, .... We hope that Solution Technology will agree to the foregoing proposal.... In any event, Cabot will be in a position to respond to your request that it dismiss the Complaint only after it has had a chance to review the information, documents and testimony relating to invalidity.

July 15, 1996. Jones sent to Yancey a draft letter to Morisseau, a draft disclosure agreement, and a draft stipulation and protective order. The draft disclosure agreement included the sentence, "Counsel and technical personnel from both Cabot and STI will be permitted to attend and observe all such tests and related activities." This sentence was included in the disclosure agreement that Jones sent to Morisseau.

July 16, 1996. In a telephone conversation between Jones and Yancey, Yancey told Jones that "a couple of the retains have gelled." The context of this remark was that Yancey and Jones were anticipating a joint testing program whereby some of the retains from STI's collection of pre-critical date retains would be subjected to independent testing. Yancey was telling Jones that some of those retains had gelled. Yancey told Jones that retains that have gelled can be ungelled. Jones' understanding was that some retains had gelled by this time in 1996, but they did not need to rely on those since they needed only to show one instance in which there was an invalidating manufacture or usage, and that being the case, they would use retains that even in 1996 were still in an ungelled state.

Jones was concerned that a slurry that had gelled would be challenged by Cabot as being unstable, and STI need not confront this because there were other retains available which did not raise that question. Use of a slurry that had gelled and then was ungelled would raise other issues that Jones did not want to pursue: Whether and when it gelled would have to be determined, but whether that determination could be made retroactively was uncertain. An example would be for a slurry that was made in 1990, and in 1996 was found to have gelled in the interim, to determine when it had gelled and thus whether it was stable for a sufficient period of time to permit usage by customers in the ordinary manner as intended. Jones did not believe that gelling by itself established instability. Gelling could frustrate use of that lot to corroborate the precritical date activities, but gelling in and of itself did not mean, to Jones, that the lot was not an invalidating manufacture or that its use would not have been invalidating.

July 17, 1996. Larry Jones responded to Morisseau's letter of July 11th, copy to Yancey, stating:

In reviewing the proposed documents [draft "Disclosure Agreement" and "Stipulation and Protective Order"] you will note that we have identified certain specific sales which are relevant to Solution Technology's allegation of invalidity. Please understand, however, that notwithstanding the breadth of your request, we have not identified all such sales. As you know, it is necessary for Solution Technology to show only a single instance in which the claimed solution was sold in the United States prior to the patent's critical date. Nevertheless, Solution Technology has selected the products from more than one sale, and such evidence of multiple sales should be more than adequate to satisfy any reasonable concerns that your client may have.

[Regarding taking testimony] ... We recognize that such discovery may be appropriate during the ordinary course of a lawsuit. Please be advised, however, that if Cabot does attempt to proceed with such extensive formal discovery at this preliminary stage, Solution Technology almost certainly will proceed with the filing of a counterclaim seeking a declaratory judgment of the patent's invalidity. By doing so, Solution Technology will then be able to use the fruits of Cabot's discovery to support Solution Technology's motion for summary judgment on the issue of invalidity. As I indicate in my letter of July 5, the disclosure and independent analysis proposed by Solution Technology are to be undertaken, if at all, to avoid the time and expense of formal discovery and the necessity of attacking your client's patent in a public forum. If Cabot insists on proceeding with formal discovery, however, it should expect Solution Technology to similarly proceed in its efforts to obtain an adjudication that Cabot's patent is invalid and may not be rightfully asserted against Solution Technology or anyone else.

(Emphasis in original.)

July 17, 1996. Jones first reviewed the file history of the Cabot patent.

July 31, 1996. Horiba LA-910 laser scattering particle size distribution analyses were performed by STI on retains from STI's lots 1656, 2021, and 2361. The results were given to STI's counsel.

Lot 1656 used Vista alumina of type 11 N 4-25. Lot 2021 used Catapal Vista alumina of type 11 N 4-20. Lot 2361 also used 11 N 4-20. The "11" is the surface area divided by 10; thus, the surface are of the particles for each of these lots is 110 meters squared per gram. Yancey provided this information to his attorneys.

During July, 1996, Yancey learned that Rodel was obtaining an opinion on the invalidity of the Cabot patent.

At some point during this general time period, BSPG reviewed test data for the retains. This was done at STI's facility. At that time, Jones had no reason to doubt that the retains that were studied to produce that test data had neither gelled nor settled. Jones expressed to Yancey his expectation that what he would receive would be test data of retains that were clear or clean. There were communications with Yancey which over a period of time led Jones to the firm conviction that BSPG would be relying on the testing of retains which were clear or clean, as opposed to any the use of which might interject other issues unnecessarily.

August 6, 1996. Rodel Executive Vice-President and Chief Operating Officer Wayne C. Wilson wrote to Yancey regarding Rodel's possibly purchasing STI, stating:

Following our recent meeting with you, we remain highly interested in working with you to make every effort to arrive as soon as possible at a structure and terms to meet the goals of all concerned. To this end, we've engaged Gary Hultquist of Hultquist Capital, LLC to assist us.

We have been advised that we will need the information on the attached exhibit in order to proceed in making a valuation analysis....

Time is critical for us, as I'm sure it is for you.

Attached to the letter was the referenced list, the second item of which was the following:

Copy of any correspondence with Cabot, or others, indicating applicability or lack thereof of Cabot, and/or other, patents to Solution Technology, Inc. FeNO3 product(s); copy of any other documents which indicate that FeNO3 product is noninfringing or existing patents in the field.

Wilson reported to Budinger's brother, who reported to Budinger. At that time, Wilson was more interested than was Budinger in acquiring STI.

August 7, 1996. Yancey sent a fax memorandum to Larry Jones regarding Rodel's request for information about the Cabot patent. Attached to the memorandum was a copy of Wilson's August 6th letter to Yancey. In his memorandum Yancey wrote:

Attached is a letter which we received today from Rodel, Inc., a company interested in purchasing us. We are moderately interested in proceeding. They have engaged an investment banker and the due diligence is about to begin.

I am concerned about what we want to say, and how we want to say it, (regarding the item 2) of the Exhibit, attached. Could you offer some ideas on how to deal with that issue. Is the ferric nitrate really an issue in the patent? If we were to get the alumina claims thrown out, is the ferric nitrate a problem. I thought there was other prior art regarding that. Our prior art in the alumina area is an asset we hold with respect to Rodel ÔÇö it should be of some value to them. How do the ferric nitrate claims affect this.

Before August 8, 1996. Larry Jones had decided to rely upon the STI slurries which had been sold to IBM in order to invalidate the Cabot patent. As early as the July 2d meeting, their focus was on the IBM transactions.

August 8, 1996. Yancey prepared handwritten notes regarding a telephone conversation he had that date with his attorneys, Jones and Meunier, regarding "Rodel request," stating:

Larry started out conversation by defining "willful infringement vs. unwillful infringement." Willful infringement means that we are continuing to use the patent without a written legal opinion; [therefore] BSP & G will give me a formal written opinion to present to Rodel (under NDA) so that I can't be shown to have "willful infringement" in mind.

At this time, Yancey was aware that Rodel was obtaining its own opinion.

In the second paragraph, Yancey wrote:

I will call Mark Jaso and discuss the situation with him to let him know that his slurries are being used as the basis for invalidating the patent.

In the third paragraph, Yancey wrote:

Drew Meunier will contact IBM counsel and let them know of the situation.

BSPG never contemplated writing an opinion for the purpose of presenting it to Rodel. Jones never participated in any discussion which in any way concerned BSPG providing a written opinion to Rodel or for the benefit of Rodel. Jones never indicated or implied to Yancey that a purpose of an opinion that was forthcoming was to satisfy Rodel or for the purpose of being communicated to Rodel.

August 8, 1996. Meunier did his first work toward preparing the September 11th written opinion. On August 20th, Meunier devoted 2.75 hours to working on the opinion; on August 21, 7.75 hours; on August 22, 7.75 hours; on August 23, 3 hours; on August 29, 0.25 hours; on August 30, 6.5 hours; on September 3, 0.5 hours; on September 8, 2 hours; on September 10, 0.25 hours; and on September 11, 1.5 hours.

August 13, 1996. Yancey sent a memorandum by fax to Jones, the second paragraph of which states:

I would also like to remind you that you promised an opinion on the Cabot matter. We will need this soon, possibly as early as August 23. Be sure to include an opinion regarding the ferric nitrate prior art that Drew has produced lately.

August 21, 1996. Yancey had a brief conversation with Jones. Jones' notes of that conversation include the following: "Rodel interested in the opinion letter." In testimony, Jones stated that this could refer to the BSPG letter or to the one Rodel had commissioned from Foley & Lardner because he could not remember when he learned that Rodel was obtaining the Foley & Lardner opinion. Tr. Vol. 8 (Sept. 1, 1999), p. 15 at lines 16 - 19, p. 16 at lines 3 - 8.

At that time, STI and Rodel were having very limited discussions about merger. In the course of these negotiations, Rodel wanted to see the BSPG opinion letter. Sometime in September or October, 1996, Gary Hultquist was allowed to read the opinion letter but was not given a copy of it.

As of August 23, 1996. Jones was aware of no significant time pressure to complete the opinion letter. Completion was not expedited. The date would be determined by when BSPG's review of the available evidence had been concluded along with the review of prior art patents and other references upon which the opinion would be based, but not in the sense of a calendar date. From July 2d, it was understood and discussed by Linker, Meunier, Jones, and Yancey that there would be an opinion prepared and that Meunier would be the principal author of that opinion.

The Rodel acquisition talks were on again ÔÇö off again during this time period.

August 26, 1996. Jones first reviewed and edited Meunier's draft of the opinion letter.

September 10, 1996. Jones and Yancey talked by telephone for ten minutes. Yancey told Jones that Rodel had retained investment banker Gary Hultquist.

Between September 11, 1996, and late October, 1996, retain samples were drawn and sent to Ricerca for testing. Three retains were provided to Ricerca, from lots 1560, 2022, and 2361, at the STI facility. Yancey gave them to Rodel's counsel from the law firm Foley & Lardner, who gave them to Ricerca. An independent engineering laboratory drew the samples.^[12] The attorneys from Foley & Lardner were present when the samples were brought out from storage, sampled, and sealed. Larry Jones was also present. They were shown the storage area. The storage area and sample collection process were videotaped by the attorneys from Foley & Lardner. The Foley & Lardner attorneys chose the retains on behalf of Ricerca. They were given a choice of available retains, and they chose the ones which they wanted to examine. Each of the retains was sitting on a table visible to all in attendance. The retains were extracted by pouring or with a straw and placed into 2- or 4- ounce plastic bottles.

Prior to turning the samples over to Foley & Lardner, Yancey had reached an agreement with that firm as to their use of the retains. The agreement was that they would take the retains under a confidentiality agreement and have them analyzed by a 3d-party laboratory that Yancey had approved (Ricerca). The results would be sent to Foley & Lardner for their purposes.

Yancey received a copy of the Foley & Lardner opinion containing Ricerca's results.

September 11, 1996. BSPG's validity opinion as to the Cabot '423 patent was issued. It concluded:

As set forth above, it is our opinion that all of the claims of the Neville '423 patent are invalid based on the prior art references and the product sales made by Solution Technology between 1990 and 1993. Therefore, the Neville '423 patent is unenforceable and Solution Technology should be free to continue to produce chemical mechanical polishing slurries.

Validity Opinion, Defendants' Exhibit 1, at 18.

As to which:

(1) Meunier had a major role in analysis and drafting of the opinion. He relied on Yancey for information about the retains and their testing. Meunier believed that none of the retains referenced in Table Two had gelled or settled. Meunier believed that the retains were representative of the material in the lots which were sold to customers.

(2) Linker spent roughly five and one-half hours in its preparation. Before he started working with Yancey and STI, Linker had no experience in chemical mechanical polishing of semiconductors.

(a) He was aware of some of the patents cited in the opinion on June 21st when he reviewed Cabot's European application.

(b) By September 11, 1996, Linker had concluded that the Cabot patent was invalid based upon STI's prior sales to IBM of boehmite alumina. (Yet as of September 11th, STI was still prosecuting its patent application as it filed it on June 17th. It had been filed and was pending at that time. Although preparations were made to do so, STI never actually informed the patent office that it had sold boehmite alumina slurries as early as 1990. Once it became evident to Linker that these prior sales of boehmite alumina existed and that they had relevance to Linker's opinion on the pending patent application, Linker had a supplemental disclosure statement prepared for purposes of bringing that information to the attention of the patent office. The supplemental disclosure statement was never filed.)

(c) In the opinion, Linker states, "In addition, because the alumina solutions were colloidal dispersions, it can be assumed that the particles possessed a force sufficient to repel and overcome the van der Waals forces between the particles and are therefore stable." Validity Opinion, Defendants' Exhibit 1, at 13.

In making this assumption, Linker relied on his discussion with Yancey and his own understanding of physical chemistry from university courses. He consulted a physical chemistry treatise. This book's discussion of the charges on particles in a colloidal dispersion was consistent with what Yancey said (although it did not say that if a solution is described as a colloidal dispersion it can be assumed that the particles possess a force sufficient to repel and overcome the van der Waals forces between the particles and, therefore, the solution would be stable).

(d) Linker had written documentation on zeta potential for lots 2021, 2022, and 3133.

(e) In the opinion, Linker states, "[B]ased on the other characteristics of the particles and knowledge in the art, in may be inferred that may [sic; many] of these references describe alumina particles having surface areas within the claimed range." Validity Opinion, Defendant's Exhibit 1, at 14. The references do not specifically describe the surface areas of the alumina particles within the claimed range. Nor did Linker perform any calculations of other particle parameters to determine what the surface area was. Linker's inference is based on what he was told by Yancey.

(f) In the opinion, Linker states, "The surface areas of STI's pre-1993 slurries are about 70 m2/g and fall into the claimed surface area range." Validity Opinion, Defendants' Exhibit 1, at 15. This was based on what Yancey told Linker.

(g) In the opinion, Linker states, "Alternatively, the zeta potential of the slurries described in Rosstoker '194 may be assumed to be greater than +/- 10 mV because the slurries are in an acidic medium." Validity Opinion, Defendants' Exhibit 1, at 15. This was based on what Yancey told Linker.

(h) Table Two, page 12 of the Validity Opinion, lists three lot numbers, 1656, 2021, and 2361. At the time of writing this opinion, Linker believed that none of the retains for these listed lots had gelled. Linker relied on Yancey for this information. Jones' understanding, based on what Yancey had told him about the retains, also was that none of these retains had experienced gelling or settling. The BSPG lawyers intended to rely only on those retains for which there could not be a question raised as to their stability.

(3) Mark Jaso's particle size distribution graph for lot 1656 appears in the opinion at Exhibit 3. The Horiba particle size analysis of lot 2021 appears at Exhibit 4, and for lot 2361 appears at Exhibit 5. Sales documentation for each of these lots follows the particle size analysis in the exhibits.

(4) The opinion relies on lots 1656, 2021, and 2361, and evidence of prior sales, including to IBM, for its conclusion of invalidity. The relevant documents related to the sales of these lots and the test results from the retains are attached as exhibits.^[13] Yancey did not tell his attorneys about lot 2361's mixing history on July 2, 1996, but he showed them the log book before September 11, 1996.

(5) The particle size tests relied on in this opinion were conducted on July 31, 1996. (Of the six particle size analyses which Yancey gave to his attorneys on or before July 2, 1996, three (3563d, 1625, and 1656; 3563d was for 201A/60, the smallest-particle product STI had made) showed entities greater than 1 micron. Yancey told his attorneys that these were anomalous. As of July 2, 1996, Yancey had not shared any specific particle size information with his attorneys regarding lot 2361.)

(6) Linker told Yancey that he could rely on the opinion. Yancey read it and relied upon it to continue marketing Met 202. He shared it with his Board of Directors (Geatz, Cutini, Hutaff, Cinnamea, Albert) and officers (Heid, Roberts, Fox).

As of September 11, 1996, Yancey had not conducted any aggregate size analyses of his prior art patents or slurries based upon transmission electron microscopy (TEM), the method of determining the size of particles or aggregates used in the Cabot patent.

September 11, 1996. Yancey sent a fax memorandum to Jones, stating:

Regarding the protective order competitive issue, Cabot's arguments regarding the competitive situation and confidential information can be turned around and used against them. We have no problem showing them our supposed infringing product, which as you state, is fumed alumina. And I don't care if I see their "patented" product because I have seen it quite enough.

The evidence does not establish that Yancey saw Cabot's patented product before September 11th. He saw information about it.

September 19, 1996. Meunier prepared a Supplemental Information Disclosure Statement for the copper polishing compound patent application of June 17th, to disclose to the Patent Office that there had been boehmite alumina sales. Between June 17th and September 16th, Meunier had been informed by Yancey for the first time of the details of the boehmite alumina sales. This Supplemental Information Disclosure Statement was never filed with the Patent Office; the application was abandoned.

September 19, 1996. Yancey sent a fax memorandum to Budinger regarding the Cabot patent. By this time, Budinger had read the Cabot patent, which claimed particles of 1 micron or less in size.

The second page of Yancey's memorandum included some of the information Yancey had received in March, 1990, from IBM's Mark Jaso regarding the particle size of STI's lot 1656, indicating that the particles are generally less than 1 micron in size. Yancey did not include volume information showing a number of particles greater than 1 micron in size.

The third page of Yancey's memorandum was a copy of the computer printout list of sales that Yancey had given to BSPG with some customer names and surface area information redacted and containing check marks but no other handwriting. Yancey had not yet written in the notes that appeared on a later version of this document (for lot 1656, "partial gel"; for lot 2021, "partial settle"). (These were added when Yancey was looking for the retains for Ricerca.) There is no reference in the memorandum to settling or gelling of the retains.

October 1, 1996. Jones sent a letter to Morisseau including the following stipulation:

With respect to the definition of "alumina product," I can stipulate on behalf of STI that the only prior art activities of STI (i.e., STI's manufacturing, usage and sales) upon which STI will seek to rely to invalidate the claims of the patent, pertain to STI's "alumina slurries or dispersions sold for use and/or intended for use in polishing metal layers on substrates for integrated circuits." STI does not intend to rely on any alumina slurries or dispersions sold by STI for use other than in polishing metal layers on substrates of integrated circuits.

October, 1996. Yancey prepared the retains for Ricerca and made the "partial get" (lot 1656) and "partial settle" (lot 2021) notations on the computer printout of STI alumina sales. By "partial get" he meant that the lot had become very thick. By "partial settle" he meant that some of the solid material was in the lower portion of the bottle.

Yancey gave Ricerca retains from lots 1560, 2022, 2361, and 2557.

Lot 1560 was mixed on June 4, 1990, using lot 1558, water, and acid. The alumina in lot 1558 was of type 21 N 4, another Vista Chemical product. A note to the side of the production log entry for 1560 says "Gelled 8/12/92 Discarded." A note to the side of lot 1558 says, "GELLED Solid 6/4/90" (capitals in original). "Gelled solid" means it does not flow. The June 4, 1990, note is in Yancey's handwriting. Yancey does not recall if he told his attorneys about the gelling when he met with them on July 2, 1996.

November 4, 1996. Yancey sent a fax memorandum to Budinger regarding Rodel's possible acquisition of STI. The second and third pages contain a list of points that Yancey wanted to make to Budinger to interest him in acquiring STI, one of which is the following:

We have enough data to beat the suit and invalidate the patent. This has been documented by our counsel's opinion letter which was reviewed by Gary Hultquist.

Yancey's final point was the following:

If for no other reason, Rodel should invest in STI to aid in the defense against Cabot's patent. By assuming control in STI Rodel will assume control of our patent counsel's legal opinion and assure Rodel a position in current and future nanometer slurry product development for advanced CMP applications without the threat of excessive damages should any litigation ensue.

Prior to Rodel's purchase of STI, Yancey did not give a copy of the BSPG opinion to Rodel or to anyone affiliated with Rodel. Gary Hultquist was given access to the opinion but was not given a copy.

November 22, 1996. The Patent Office rejected Yancey's patent application.

December, 1996. Robert Heid, STI's Vice President of Sales, discussed settlement of the instant suit with Cabot's Executive Vice-President for Manufacturing, William Noglows.^[14] They met at Cabot's development facility in Aurora, Illinois.^[15]

Heid took with him to the meeting a set of overhead slides which he prepared in December, 1996. The overheads were prepared to aid in discussing settlement with Cabot. The fourth slide contained the following language, emphasis, and capitalization:

4. Summary

We Will Not Lose: we will quit first ÔÇö not validating patent and leaving Cabot to fight other Goliaths with no precedence but the ghost of our data and retains remain The Industry Will Not Accept A Monopoly: Cabot will be forced to license or technology will evolve in a different direction

Our Desire is NOT To Win: everyone else wins for free & we forfeit a lucrative co-licensing opportunity

As to the "Summary" overhead slide, Yancey testified that it was a summary of STI's "strategy," but denied that STI had a strategy to quit the lawsuit before losing, this being an option but not a strategy. Yancey denied that STI decided to quit this lawsuit before it lost to avoid validating Cabot's patent.

As to the meaning of the terms or phrases in the "Summary" slide, Yancey testified that "We" means STI; "We will not lose" refers to the instant lawsuit; "We will quit first" signals that STI will quit the lawsuit before losing it; "Not validating patent" refers to the Cabot patent; "Not validating the Cabot patent and leave Cabot to fight other Goliaths" means that if STI quits this lawsuit and there is no finding in this lawsuit that the patent is not invalid, then Cabot will have to sue others in the future, other "Goliaths" such as Rodel, Moore, IBM, and AMD; "precedent" refers to this Court finding the Cabot patent not to be invalid; "But the ghost of our data and retains remain" refers to the sales which STI made to IBM and others between 1990 and 1993 and which could invalidate the Cabot patent.

Noglows does not recall seeing the summary document at the December meeting. Heid discussed no feature of that document except that the industry would not accept a monopoly.

In 1996, STI sold approximately $7,000 worth of Met 202 domestically. In 1997, STI sold approximately $5,000 or $6,000 worth of Met 202 domestically. STI first sold Met 202 in mid-1995. Of the $130,000 worth of Met 202 sold through June, 1998, $18,000 was domestic, the balance overseas.

1997

January 7, 1997. Yancey sent a fax memorandum to Noglows on the subject of "settlement meeting followup," and attached a copy of "data from IBM going back to 1990 showing particle size distributions sent from them to us identifying the products by name" (emphasis in original). This was the number analysis for lot 1656. The IBM volume analysis, showing particles greater than 1 micron, was not included. The memorandum does not mention that there had been settling in lot 2021 or gelling in lot 1656.

January 21, 1997. Yancey sent a fax memorandum to Neville regarding settlement, supplying a list of contract laboratories which could test the retains and a proposal for testing the retains. In this memorandum, Yancey states:

As we discussed on the phone several weeks ago, we would like to come to a solution to this lawsuit issue.... We are going into this testing mode with the understanding that we will reach an agreement on some sort of licensing arrangement or compromise and that we will not continue the lawsuit to its conclusion. We do not want to spend the time and money providing Cabot with this information, only to find the lawsuit continuing.

Yancey testified at trial that he did not intend at that time by this language to make dropping the suit a precondition for going forward with the joint testing.

On or about January 27, 1997. The Foley & Lardner opinion was sent to Budinger.

In early 1997, Yancey found the Coca-Cola bottle containing Cabot slurry on his desk,

February 3, 1997. Matthew Neville sent a letter to Yancey responding to Yanceys' January 21st memorandum, and discussing the timing and efficacy of joint testing, sharing of prior art, and conditions under which Cabot would drop the lawsuit.

February 10, 1997. Yancey sent a fax memorandum to Neville regarding settlement, stating:

If you will agree to my original understanding that some conclusion to this controversy is guaranteed, other than the lawsuit, I will be much more willing to further discuss the nature of our particles.

(emphasis in original) By this Yancey meant that he would be willing to share the secrets of STI's material, including manufacturer's information, such as that Vista had assigned a surface area of 110 m2/g (which Yancey had already offered to give Cabot in testing the retains).

February 11, 1997. Yancey sent a letter to Neville, stating:

If you will agree to my original understanding that some conclusion to this controversy is guaranteed, other than the lawsuit, I will be much more willing to further discuss the nature of our particles under some form of mutual technology exchange which we have suggested in the past.

(emphasis in original) Yancey believed that Neville was asking for much more than what was required for determining the nature of the particles with respect to the claims of the patent; that he was asking for all of STI's technology, including manufacturing data; and that, if Yancey gave him the information he requested, Cabot could make every STI slurry.

February 24, 1997. Neville sent a letter to Yancey regarding joint testing and discovery.

On or around March 5, 1997. Cabot served the following request for admission:

As stated in the October 1, 1996 letter from Larry Jones (defendants' counsel) to Fay Morisseau (plaintiff's counsel), the only prior art activities of STI (i.e., STI's manufacturing, usage and sales) upon which STI will seek to rely to invalidate the claims of the Patent pertain to STI's "alumina slurries or dispersions sold for use and/or intended for use in polishing metal layers on substrates for integrated circuits."

STI's answer to this request for admission was as follows:

To the extent this request is understood, denied. As reflected in such documents as the previously served "Defendant's Supplemental Initial Disclosures," the slurry products upon which STI intends to rely to invalidate the claims of the '423 patent are the "alumina slurries or dispersions sold for use and/or intended for use in polishing metal layers on substrates for integrated circuits," but STI knows that those same slurry products were also used by some of the customers identified in "Defendant's Supplemental Initial Disclosures" for polishing metal layers in other types of products, including optical products.

(emphasis in original)

September 22, 1997. Ruling on Cabot's motion in limine to exclude evidence related to purchasers and users of STI's pre-critical date slurries not involved in the production of integrated circuits, and denying Cabot a protective order to prevent a deposition of Seagate Technology, Inc., this Court ordered that STI "may not introduce any evidence at the trial of this matter of the use of a pre-critical date STI slurry unless that evidence relates to the use of the slurry for polishing metal layers on substrates for integrated circuits." This Court further ordered that the scope of the Rule 30(b)(6) deposition of Seagate Technology, Inc., a Minnesota hard-drive manufacturer, be "limited to questions regarding Seagate Technology's purchase and use of STI slurry for polishing metal layers on substrates for integrated circuits."

September 22, 1997. Morisseau telephoned Jones to suggest that, before incurring the expenses of this deposition, they consult Seagate's counsel as to whether Seagate used STI's slurry on integrated circuits. In its website, which Jones had visited and from which Jones had downloaded a hard copy of Seagate's product line, Seagate held itself out as a manufacturer of integrated circuits. Morisseau warned that, should the deposition prove fruitless and this could have been ascertained in advance of incurring its expenses, Cabot would use it as evidence supporting a finding of exceptional case to seek an award of attorneys' fees.

Jones declined to telephone Seagate's attorney. On September 23d, Jones went to the deposition in Minnesota and collected documents before Seagate's counsel was informed of this Court's September 22d ruling.

September 24, 1997. Rodel bought STI for $7,990,000. Rodel bought STI because of STI's slurry technology, including alumina slurries and the Met 202 fumed alumina product. As a result of the sale, Yancey personally received $1 million cash, $1.7 million in stock, and a three-year employment contract with compensation of $185,000 per year and stock options spread over ten years.

After Rodel acquired STI, Budinger, in his capacity as a director of STI, made the decisions concerning how the instant suit would be handled. From the time of the acquisition of STI by Rodel, Yancey made no decisions concerning the conduct of this litigation.

At the time of the acquisition of STI by Rodel, Budinger believed that the Cabot patent was invalid. His opinion was informed by the Foley & Lardner opinion.

As of the time of the acquisition of STI by Rodel, it was not Budinger's intention to protract the litigation. He wanted to end the litigation. He called Cabot chief Sam Bodman and told him that Rodel wanted to end the litigation. Bodman asked Budinger for evidence as to invalidity of the Cabot patent.

In the ensuing period, Budinger initiate other settlement contacts.

At the time of acquisition, Rodel attributed virtually a zero value to Met 202.

October 3, 1997. Cabot withdrew its claims for damages.

November 18, 1997. Budinger sent a letter to Bodman, to which Bodman did not personally respond, expressing at the outset "my hope that we can meet in the near future to discuss in a serious and meaningful way a settlement of this case that is in the best interest of all concerned." Attached to the letter are records of what STI maintained was prior art invalidating the Cabot patent. The penultimate paragraph states:

The information contained in this letter and the enclosures constitute a communication in furtherance of a settlement between Cabot and Solution Technology subject to the Federal Rules of Evidence. In the event you have no interest in pursuing settlement discussions promptly, Solution Technology will disclose this information in the course of discovery in the case. For the time being, this letter will serve as supplementation of Solution Technology's previous discovery responses.

However, Budinger did not include the Foley & Lardner opinion which he had received in January, 1997, or any reference to that opinion, nor any Ricerca analytical data. The final draft of the letter was written by Conrad Kaeding, Rodel's in-house counsel.

November 24, 1997. Cabot general counsel Robert Rothberg sent a letter to Rodel counsel Andre Bouchard, responding to Budinger's November 18th letter to Bodman, and stating that Cabot will review the information Budinger sent to Bodman. Cabot declined to accept as such "a supplemental discovery response that purports to be inadmissible and is not signed by an attorney of record." The letter further stated:

Once we have had a chance to review and analyze your official supplemental discovery response, and have obtained any additional information necessary to understand fully the issues raised in Mr. Budinger's letter, we will let you know of any change in our position.

1998

Budinger decided that STI would stop manufacturing and selling Met 202 because its business prospects were not encouraging. Since being placed on the market, Met 202 had sold less than $200,000.

Budinger decided to have his attorneys apply to this Court to end the case by having an injunction entered against STI. Budinger had obtained an estimate from his attorneys that trying the case would cost almost $1 million. Budinger did not think that that amount of money should be spent on a product that did not have a significant commercial future. Settlement talks with Cabot had not succeeded.

Budinger did not intend not to win the lawsuit.

September 17, 1998. The Court entered a consent temporary injunction enjoining STI from making, using, or selling Met 202. The injunction was voluntary and was not entered based on admissions of either party or on the merits of the case.

November 2, 1998. STI sent a letter to the Court requesting a conference to discuss whether the trial set for January 6, 1999, should be cancelled in light of STI's willingness to agree to the entry of a final injunction. The letter reiterated STI's position that the Cabot patent is invalid and unenforceable. However, STI would consent to entry of a final injunction

in view of a significant change in circumstances that has occurred in recent months. Specifically, Solution Technology has discontinued manufacturing and selling Met 202 and has no intention of manufacturing or selling Met 202 in the future. Accordingly, no useful purpose would be served by further litigation or trial of this action.

November 3, 1998. Cabot sent a responding letter to the Court, stating that Cabot did not consider the case to be moot, that the proposed injunction failed to address STI's infringement of the Cabot patent and its validity, and that issues of exceptional case and fees remain. Cabot wanted Rodel to agree to be bound as a party to the injunction.

November 3, 1998. The Court denied STI's November 2d request for a conference.

In November and December, 1998, STI filed three motions to quit the instant law-suit:

November 5, 1998. STI filed a motion for entry of partial judgment (1) granting an injunction on Cabot's claim, (2) dismissing STI's counterclaim as moot, and (3) reserving jurisdiction to consider any application by Cabot for attorneys' fees. STI would stipulate to infringement but not to non-invalidity. STI's motion was denied on the ground that Cabot's injunction would be unenforceable in contempt proceedings without a determination as to validity.

November 12, 1998. At a hearing on STI's motion for entry of partial judgment, STI stated its willingness to stipulate to infringement.

November 20, 1998. STI moved to reconsider. STI would agree to stipulate in the injunction that it would not challenge the validity or enforceability of the Cabot patent in any contempt proceeding to enforce the proposed injunction.

December 1, 1998. The undersigned denied STI's motion to reconsider, concluding that "a contempt proceeding based upon a judgment arising out of this case, in which validity has been exhaustively litigated, would be inappropriate unless validity were determined or non-invalidity stipulated as to these challenges."

December 16, 1998. STI filed a renewed motion for entry of partial judgment stipulating that

for the purpose of this case only (and not for the purpose of any other action involving products not subject to the injunction in this action) that Met 202 infringes the '423 Patent and that the '423 Patent is not invalid and not unenforceable (including over the prior art and conduct cited in this case, e.g., STI's notice pursuant to 35 U.S.C. ž 282). The validity and enforceability of the '423 Patent shall be the law of this case, but this case only.

Proposed Order accompanying STI's Renewed Motion to Enter Partial Judgment, at 2.

STI's Order enjoined making, selling, and using Met 202, and in light of that injunction dismissed STI's counterclaim as moot, leaving for trial the question of exceptional case and award of attorneys' fees and costs.

December 16, 1998. Cabot responded to STI's renewed motion, stating:

Cabot does not consent to the entry of the proposed order and injunction. Nonetheless, based on Cabot's conclusion that the injunction STI has moved the Court to enter is valid and enforceable and would provide substantially the relief Cabot has sought in this litigation, Cabot will not appeal from the entry of the order and injunction.

December 17, 1998. The Court granted STI's renewed motion for entry of partial judgment and entered STI's proposed Order.

CONCLUSIONS OF LAW

I. Willfulness of infringement

A. Standard as to willfulness of infringement

Willful infringement may require a lawsuit that would otherwise be unnecessary, thus rendering the case so exceptional as to warrant attorney fees under ž 285. Kloster Speedsteel AB v. Crucible, Inc., 793 F.2d 1565, 1580 (Fed.Cir.1986). A finding of willful infringement is legally sufficient to establish an "exceptional" case under ž 285. Del Mar Avionics, Inc. v. Quinton Instrument Co., 836 F.2d 1320, 1329 (Fed.Cir.1987). The Federal Circuit has consistently upheld awards of attorney fees when willfulness of infringement is established. S.C. Johnson & Son, Inc. v. Carter-Wallace, Inc., 781 F.2d 198, 200 (Fed.Cir.1986).

Willful infringement requires a finding by clear and convincing evidence in view of the totality of the circumstances that the alleged infringer acted in disregard of the patent and lacked a reasonable basis for believing it had a right to act as it did. Amsted Industries v. Buckeye Steel Castings, 24 F.3d 178, 181 (Fed.Cir.1994); Ortho Pharmaceutical Corp. v. Smith, 959 F.2d 936, 944 (Fed.Cir.1992); Kalman v. Berlyn Corp., 914 F.2d 1473, 1483 (Fed. Cir.1990); see also, Hall v. Aqua Queen Manufacturing, Inc., 93 F.3d 1548, 1555 (Fed.Cir.1996); State Industries, Inc. v. Mor-Flo Industries, Inc., 883 F.2d 1573, 1581 (Fed.Cir.1989). A finding of willful infringement is a finding of fact, not a conclusion of law. Electro Med. Sys. S.A. v. Cooper Life Sciences, 34 F.3d 1048, 1056 (Fed.Cir.1994); accord, National Presto Industries, Inc. v. West Bend Co., 76 F.3d 1185, 1193 (Fed.Cir.1996).

Willfulness is a question of the alleged infringer's intent. Ortho Pharmaceutical, 959 F.2d at 944 (Fed.Cir.1992) ("Ortho's intent and reasonable beliefs are the primary focus of a willful infringement inquiry."); Gustafson, Inc. v. Intersystems Indus. Products, 897 F.2d 508, 510-11 (Fed.Cir.1990) ("Whether an act is "willful" is by definition a question of the actor's intent, the answer to which must be inferred from all the circumstances.") Read Corp. v. Portec, Inc., 970 F.2d 816, 828-29 (Fed.Cir.1992) ("Willfulness is a determination as to a state of mind.").

In the instance of reliance on an opinion letter, "counsel's opinion must be thorough enough, as combined with other factors, to instill a belief in the infringer that a court might reasonably hold the patent is invalid, not infringed, or unenforceable." Id.; see also, Westvaco Corp. v. International Paper Co., 991 F.2d 735, 743-44 (Fed.Cir.1993) ("Opinion letters should be reviewed to determine whether they evidence an adequate foundation based on a review of all necessary facts or whether they are conclusory on their face.... In considering the reasonableness of the accused infringer's reliance on an opinion of counsel, the opinion letter should be reviewed for its overall tone, its discussion of case law, its analysis of particular facts, and its reference to inequitable conduct." (citations and internal quotation marks omitted)). "Objective evidence must be considered to determine whether a defendant was justified in relying on patent counsel's advice, i.e., whether the patent counsel's opinion was competent." Id.

When the infringement is alleged to be willful, and the defense of reasonable reliance on the opinion of counsel is raised, the trier of fact must ask what the competence of the opinion and the circumstances of its being obtained have to say about intent. The Federal Circuit's recent opinion in Comark instructs the trier of fact to look to intentional withholding of information from counsel, relating objective reasonableness to intent as follows:

As a general matter, a potential infringer with actual notice of another's patent has an affirmative duty of care that usually requires the potential infringer to obtain competent legal advice before engaging in any activity that could infringe another's patent rights.... In determining whether willfulness has been shown, we look to the totality of the circumstances, understanding that willfulness, as in life, is not an all-or-nothing trait, but one of degree. It recognizes that infringement may range from unknowing, or accidental, to deliberate, or reckless, disregard of the patentee's legal rights.... We must look at exculpatory evidence as well as evidence tending to show deliberate disregard of Comark's rights in determining whether substantial evidence supports the jury's verdict. The correct legal standard, therefore, is whether, in light of all the evidence, there is substantial evidence to support the jury's finding of willfulness by clear and convincing evidence.

It is well settled that an important factor in determining whether willful infringement has been shown is whether or not the infringer obtained the opinion of counsel.... However, the legal opinion must be "competent" or it is of little value in showing the good faith belief of the infringer. Those cases where willful infringement is found despite the presence of an opinion of counsel generally involve situations where opinion of counsel was either ignored or found to be incompetent.... Thus, before we may consider the exculpatory value of an opinion of counsel, the legal advice contained therein must be found on the totality of the circumstances to be competent such that the client was reasonable in relying upon it.

The reason a potential defendant obtains an opinion from counsel is to ensure that it acts with due diligence in avoiding activities which infringe the patent rights of others. Obtaining an objective opinion letter from counsel also provides the basis for a defense against willful infringement. In order to provide such a prophylactic defense, however, counsel's opinion must be premised upon the best information known to the defendant. Otherwise, the opinion is likely to be inaccurate and will be ineffective to indicate the defendant's good faith intent. Whenever material information is intentionally withheld, or the best information is intentionally not made available to counsel during the preparation of the opinion, the opinion can no longer serve its prophylactic purpose of negating a finding of willful infringement.

Comark Communications, Inc. v. Harris Corp., 156 F.3d 1182, 1190-91 (Fed.Cir. 1998) (citations omitted) (internal quotations omitted).

To justify reasonable reliance, the opinion of counsel must be authoritative. It should include a thorough review of prior art and the prosecution history. Review of the file history has been held to be a normal and necessary preliminary to a validity opinion. Underwater Devices Inc. v. Morrison-Knudsen Co., 717 F.2d 1380, 1390 (Fed.Cir.1983). Review of the file history is a factor in determining willfulness. It is not, however, by itself apart from the totality of the circumstances, a sufficient predicate to such a determination. Radio Steel & Mfg. Co. v. MTD Products, Inc., 788 F.2d 1554, 1558 (Fed. Cir.1986).

Oral opinions carry less weight, partly because of problems of proof of their content and basis. Minnesota Min. and Mfg. v. Johnson & Johnson, 976 F.2d 1559, 1580 (Fed.Cir.1992).

B. Analysis as to willfulness of infringement

According to Cabot, Yancey willfully infringed upon Cabot's patent rights: He copied Cabot's invention, misled his attorneys, and sought an opinion letter for the improper purpose of selling STI to Rodel. We are to infer copying from Yancey's exposure to information about Cabot's product, his interest in competing with Cabot in selling products to Sematech, and his treatment of and claims about lot 3100. We are to infer intentional misleading of his attorneys from Yancey's July 2d interpretation of the IBM test data and use of gelled or settled retains for STI's post-July 2d confirmatory testing. We are to infer improper purpose from his having decided independently on June 26th that Cabot's patent was invalid and from Yancey's three communications on August 6 - 8 about obtaining a formal written opinion to present to Rodel, set in the context of pressure to sell from recently-acquired investors.

Applying the "clear and convincing" evidentiary standard to the totality of circumstances as established at trial, crediting all reasonable inferences as to intent that the evidence will afford, I find that Cabot has not proven willfulness of infringement.

1. As to the allegation of improper purpose in obtaining the September 11th opinion

Cabot forcefully contends that the September 11th opinion was prepared, not to ascertain the propriety of continuing to make or sell Met 202, but "to present to Rodel." By August 6th, time was critical in the negotiations between Rodel and STI for the purchase of STI. Rodel asked STI for information STI had on the Cabot patent. In his August 7th letter to his attorneys, Yancey references STI's prior art in response to Rodel's request (but makes no mention of an opinion letter). Then on August 8th, we have in Jones' notes that BSPG would give Yancey "a formal written opinion to present to Rodel ... so that I can't be shown to have willful infringement in mind." Five days later Yancey is writing to Jones again, reminding him of the promised opinion letter, and telling him the letter would be needed soon, possibly as early as August 23d. In a November letter to Rodel CEO Budinger he urges Rodel to buy STI, if for no other reason, to assume control of the STI's opinion letter to aid in the defense against Cabot's patent.

Taken alone, this evidence would demonstrate that Yancey sought the formal written opinion to assist in his negotiations with Rodel, which were not completed until September, 1997. However, despite the alleged pressure from investors, to find that this was Yancey's sole, original, or controlling purpose requires an August date for the decision to get an opinion, after Rodel was actively in the picture, rather than a July date prior to the intensification of those negotiations. While the testimony of Jones, Linker, and Meunier does not surface Yancey's undeclared intentions, it establishes the July date for agreeing to give the opinion. This is fatal to the argument that Yancey's original or sole or overriding intent in obtaining the opinion was to sell STI to Rodel.

It is clear that in August Yancey wanted the written opinion to present to Rodel so that he could not be shown to have willful infringement in mind. But Jones did not collaborate with him in an improper purpose. BSPG agreed to issue the opinion, and agreed to do so for a proper purpose, to advise Yancey whether he could continue to make and sell Met 202, and BSPG agreed to do so in July. Yancey's desire to show the opinion to Rodel (which in any event was obtaining its own opinion), a use ancillary to that for which it was originally sought, was not then improper.

Much turns on this. Even though Rodel was obtaining its own opinion, the BSPG opinion might be valuable to Rodel if the opinion concluded that the Cabot patent was invalid. The opinion would be a selling point for STI if it were valuable to Rodel. Hence, Yancey, pressured and eager to sell, would seek to manipulate the outcome of the opinion by misrepresenting the significance of the IBM test results and, perhaps, by failing to follow his attorneys' instructions as to the condition of the retains. But if Yancey sought the opinion in order to decide if he could continue to make and sell Met 202, the motivation to manipulate the opinion's conclusions by misrepresentation and failing to follow instructions is simply not there, even if we assume that Yancey were so inclined.

The evidence for Yancey's having sought the opinion in July is persuasive. Yancey, of course, testified that he sought the opinion to determine if STI could continue to market Met 202, that his attorneys permitted him to sell STI products at the July trade show, and that at that time he anticipated a written opinion. Each of the three BSPG attorneys who took part in the writing of the opinion testified that Yancey was promised a formal opinion from around July 2d. Jones' testimony was emphatic that the opinion was not prepared for Rodel:

We never contemplated writing an opinion for the purpose of presenting it to Rodel. We did ÔÇö we were asked to cooperate with Rodel in arranging independent testing of retains by Rodel, but Rodel retained its own counsel who wrote an opinion about the Cabot patent. I never participated in any discussion which in any way concerned us providing a written opinion to be provided to Rodel or for the benefit of Rodel.

Jones, Aug. 31 at 142; also Jones, Sept. 1 at 10.

Linker told Yancey on July 2 that BSPG believed the patent to be invalid and, further, that as of the beginning of July he intended to contribute to a written opinion. Meunier discussed preparing a written opinion with Yancey when BSPG obtained a copy of the European patent application in June, 1996. Yancey authorized BSPG to proceed with the opinion at that time.

Yancey's communications of August 6th - 8th are not inconsistent with this evidence. The Rodel negotiations did not intensify until well after the July 2d meeting when a written opinion was authorized. When they did take up in earnest, Yancey wanted to use the opinion as a selling point. Since time was then critical in the negotiations, Yancey was eager to get the opinion. Rodel had not accused Yancey of willful infringement, and Yancey would not have wanted Rodel to suspect him or his company of that. None of this is inconsistent with a written opinion having earlier been authorized to determine if Met 202 could continue to be marketed.

2. As to the allegation of intentional misleading by Yancey of his attorneys

Cabot argues that Yancey prevented his attorneys at BSPG from preparing a competent opinion by misleading them. First, lacking precise knowledge of how the solution was prepared and the test conducted, Yancey unreasonably told his attorneys that course tails in the IBM test record were anomalies. Second, contrary to their instructions that the retains be clear or clean, two of the three retains which Yancey tested for his attorneys had either gelled or settled, in violation of his duty of due care.

As to the first, Yancey did not know, and did not tell his attorneys he did not know, how IBM prepared its sample, the water composition of the sample, or the credentials of the person who prepared the sample and ran the test. The September 11th opinion depends in part on lot 1656 having no particles greater than .339 microns. Yet IBM's October, 1990, test shows particles up to 18 microns.

As to the second, Jones expected to receive test data on retains that had neither gelled nor settled, that were clear or clean, in order to avoid interjecting other issues unnecessarily, and he told this to Yancey. Jones' communications with Yancey led him to believe that he would receive such data. Although the retains were within 100 feet of Yancey's office, he did not check on their condition. Yet lots 1656 and 2021 had settled or gelled. In July, 1996, Yancey tested a sample from lot 1656. He stated that its partially gelled state did not cause him concern about testing for particle size.

Comark does not read intent out of willfulness. That opinion looks to the intentional withholding of information, in which case an opinion cannot serve its prophylactic purpose. In light of the totality of these circumstances, Cabot does not prove by clear and convincing evidence that Yancey intentionally withheld material information or that he did not have a good faith basis for believing that the alumina particles were less than one micron in size. When Yancey told his attorneys that the course tail was anomalous, he was not fabricating a convenient interpretation out of thin air to obtain a favorable opinion, or simply speculating. He drew from his extensive experience with testing slurries and STI's tests of its UltraSol slurries. STI's testing of retains later in the summer indicated no large particles in the slurries. In addition, Cabot offers no evidence that the interpretation of the course tail as anomalous was inaccurate or unreasonable, or that the July Horiba tests were inaccurate.

As to the retains, the testing of which for the written opinion was requested in early July, Cabot shows that Yancey did not take adequate measures to predetermine the condition of the retains, but not that he intended to deceive his attorneys: First, he gave them a clear retain (lot 2361), which does not prove Yancey's intent as to the other two but which speaks to his overall purpose. BSPG would have found the Cabot patent invalid on that basis alone. Second, on July 5th and 17th Yancey offered to Cabot the opportunity to test the retains, including the lots used in the written opinion. (Contrary to Cabot's interpretation, a reading of all relevant documents in context indicates that Yancey desired but did not require dismissal before testing.) Yancey would not have offered to Cabot what he intended to conceal from his own attorneys.

3. As to what Yancey told his BSPG attorneys, particularly Meunier, about STI's pre-critical-date sales to IBM, in their discussions leading to Yancey's patent application

Cabot contends that Yancey did not disclose to BSPG STI's prior sales to IBM, dating back to 1990, before BSPG filed STI's patent application on June 17, 1996. Cabot assigns STI the task of reconciling a failure to treat the IBM sales as prior art in connection with its patent application and its conclusion that the IBM sales were invalidating prior art for the Cabot patent. Cabot discovers in this another instance of Yancey's withholding of crucial information from BSPG to obtain that to which he was not entitled, and in any event of not acting reasonably. STI counters by arguing that Meunier, an attorney at that time new to practice, failed to appreciate the significance of STI's prior sales, information as to which was in fact provided to BSPG, and that Vista had been commercially selling four types of boehmite alumina slurry. As to which, I conclude the following:

First, the evidence does not convincingly demonstrate that Yancey did not tell BSPG about the prior sales to IBM prior to the drafting and filing of his patent application. If anything, it demonstrates just the opposite. Linker testified that Yancey both provided BSPG with information about his sales of an alumina product to IBM and told BSPG that ferric nitrate had been used as an oxidizer in the industry before or during March, 1996. According to Linker, Yancey informed BSPG that he believed that he had developed a process that was a combination of colloidal alumina and ferric nitrate oxidizer. Linker stated in his testimony that Yancey told BSPG about sales of alumina to IBM, boehmite alumina, during April, 1996.^[16] Moreover, Yancey's March 18, 1996, letter to IBM stating that STI had been selling an alumina product to one of IBM's groups since around 1989, was given to Meunier. In his record of a conversation with Yancey on April 10, 1996, Meunier notes that Yancey stated that he had been selling a copper polishing compound without ferric nitrate oxidizer to IBM. In his notes of an April 19, 1996, meeting with Yancey, Meunier makes reference to a slurry STI had sold to IBM since 1989.

Second, the evidence of Yancey's involvement in preparing the patent application does not prove that he intentionally misled his attorneys, nor does it prove a pattern of deceptive behavior. First, as previously noted, whether he told his attorneys about the IBM sales before the application was drafted and filed is not established by the evidence. Second, the description of the invention and the examples in the application accurately reflect the combination of alumina slurry and ferric nitrate.

Against this, of course, must be weighed the evidence of Yancey's opportunities to review the application and attestation of inventorship. Claims 1, 4, 5, 7, 10-14, and 20-22 of the patent application cover only boehmite alumina for use in CMP, and other claims cover boehmite alumina in combination with ferric nitrate. Meunier reviewed his work with Yancey several times. By letter to Yancey on April 30th, Meunier reported that in his search of prior art patents he found nothing suggesting the use of boehmite alumina for a CMP slurry for polishing copper, and therefore, that the use of boehmite alumina in a copper CMP would be patentable. Meunier then prepared a draft of the patent application and sent it to Yancey on June 4th. Yancey reviewed the draft on June 10th and provided Meunier with comments. Meunier prepared a second draft, sent it to Yancey on June 11th, and received comments from Yancey upon his review. Meunier then prepared a final draft of the application.

Certainly, with years of patent experience, Yancey knew that he could not file a patent application on a product that had been sold more than one year earlier. There are two options: Either he misunderstood the application, or understood it incompletely, relying on the description and the examples in light of his direction to his attorneys to prepare an application for a patent on the combination of alumina slurry and ferric nitrate, or he understood the application, remaining silent as to the objectionable claims.

If, as he testified, Yancey had told his attorneys about the IBM sales and had told his attorneys that the patent he sought was on the combination of alumina slurry and ferric nitrate, then the best interpretation of this evidence is that he relied on the description and examples in attesting to inventorship. He failed to use due care in reviewing the application, but this is a far cry from intentional deception. Without clear and convincing proof of when and under what circumstances Yancey told his attorneys about the IBM sales, it is not justified to go further and attribute to him deceptive intent.

Finally, in light of the totality of the evidence, the events surrounding the filing of Yancey's patent application do not establish that Yancey did not believe that the slurries he sold to IBM in the early 1990s were invalidating prior art. Nor do they provide sufficient ground for inferring a pattern of deliberate deceit in the patent application and the September 11th opinion. Nor do they provide an inferential basis for concluding that Yancey failed to use due care in obtaining the September 11th opinion.

4. As to whether STI could reasonably rely upon the September 11th opinion because of assumptions contained therein

In short, yes. The September 11th opinion is thorough and well reasoned. It bears the BSPG pedigree. It was not unreasonable for Yancey to rely upon it.

Nothing in the record undermines the validity or reasonableness of the assumptions. Specifically, it was not unreasonable for Linker, a veteran patent attorney, to consult a physical chemistry textbook in reference to what Yancey had told him, so that he felt comfortable with the assumption that the particles possessed a force sufficient to repel and overcome the van der Waals forces and were therefore stable. Further, the Rosstoker patent does not assume, but calls its suspension colloidal.

5. As to what we may infer from Yancey's having decided on June 26th that the Cabot patent was invalid.

Cabot argues that Yancey concluded independently of his attorneys, without relying on their advice, that the Cabot patent was invalid. He reached this conclusion on June 26th after receiving notice of the lawsuit on June 25th and despite the advice of his attorneys at their June 21st meeting. According to Cabot, he then did what he could to get an opinion which would say that the Cabot patent was invalid. He limited the type of information that went to his attorneys. When he got the opinion, he shared it with Rodel in order to persuade Rodel to buy STI, which would satisfy STI's "greedy" investors. The July 2d and September 11th opinions are incompetent because they are not based upon the best information known to STI. According to Cabot, the failure to supply that information was part of Yancey's deliberate, willful strategy, conceived and implemented because of and in spite of the conclusion of his attorneys on June 21st that a basis had not been shown them for finding the patent invalid.

Much, then, turns upon what happened at that June 21st meeting. What can be legitimately inferred from the evidence about the independence of Yancey's June 26th conclusion that the Cabot patent was invalid depends in part upon what can be determined to have happened at his June 21st meeting with his attorneys.

At the June 21st meeting, Yancey, Meunier, and Linker discussed the European patent application. Yancey brought product descriptions having aggregate particle distributions and mean diameters which would not fall within the ranges in the Cabot patent claims but which had been sold for several years. These would not invalidate the Cabot patent. Upon discussion it was determined that Ultra-Sol Met 202 may fall within the claim range of the Cabot patent and that it was essentially the same material that was combined with the gamma alumina to form a slurry having a bimodal distribution. Finally, it was determined that the prior art patents may not describe the abrasive particles with as much specificity as the Cabot patent application, rendering it difficult to premise invalidity upon them.

Cabot concludes that a reasonable person would leave the June 21st meeting thinking that his product infringes, his prior art products would not invalidate the patent, and the prior art patents would not invalidate the patent. All that happened between June 21st and June 26th is that Cabot sued STI. Thus, Yancey's conclusion that the Cabot patent was invalid must have been reached independently of the advice of his attorneys. He could not have relied on their advice because they told him he had not given them the basis for a conclusion that the patent was invalid.

Cabot's position is unpersuasive. First, it is offered in partial explanation of why Yancey tried to manipulate the outcome of the July 2d and September 11th opinions by limiting the type of information that went to his attorneys. But Cabot fails to establish that he attempted this manipulation. The significance to Cabot's account of his independent conclusion is the link it affords in this chain of motivation and action. Second, the evidence does not establish that at the June 21st meeting Yancey was told that the prior art would not establish invalidity, only that it may not do so. In the memorandum of the June 21st meeting there are references to there being prior art and to the patent being invalid. The BSPG attorneys had had a full copy of the European application for only two or three days prior to the meeting. (By September they had decided to the contrary, that the prior sales of the STI slurries, along with the prior art or combinations of the prior art, would invalidate the 39 claims.) Third, Yancey had a basis for an independent conclusion that the patent was invalid, namely, his own experience of some twenty years, as well as familiarity with his products, manufacturers' particle size descriptions, and the tests that had been run on his products. For these reasons I do not find the record to have established that Yancey's June 26th conclusion was reached contrary to that of his attorneys or that notice of suit on June 25th was the motivating factor of his conclusion.

6. As to copying

a. Standard as to copying

Intentionally copying a patented invention is a basis for finding willful infringement. Jurgens, 80 F.3d at 1571; In re Hayes Microcomputer Products Patent Lit., 982 F.2d 1527 (Fed.Cir.1992), abrogated on other grounds by Bott v. Four Star Corp., 807 F.2d 1567 (Fed.Cir.1986). "The inquiry is whether the infringer intentionally copied the ideas of another, ..., or deliberately copied the ideas or design of another." Stryker Corp. v. Intermedics Orthopedics, Inc., 96 F.3d 1409, 1414 (Fed.Cir.1996) (citations and internal quotation marks omitted).

At least one patent opinion has embraced the application of an access and similarity test to prove copying. Pacific Furniture Manufacturing Co. v. Preview Furniture Corporation, 626 F. Supp. 667, 676-78 (M.D.N.C., 1985); see, Keeler Brass Co. v. Continental Brass Co., 862 F.2d 1063, 1065 (4th Cir.1988).

b. Analysis as to the allegation of copying

Cabot fails to prove copying primarily because STI established that on February 21, 1992, over one year before Cabot initiated its efforts to develop a slurry for polishing tungsten, Yancey mixed two slurries using Aluminum Oxide C, a fumed alumina made by DeGussa. In mixing these slurries he was guided by a DeGussa recipe, Lepandin 20N, in a product brochure he had received in the late 1980s. One of the slurries (lot 2084) was made with alcohol, the other (lot 2085) without alcohol. Lot 2085 was the basis for Met 202. The evidence does not show that Yancey had access to information about Cabot's slurry before Semicon Southwest, when he displayed Met 202.

Moreover, Cabot fails to show that STI copied Cabot in the use of ferric nitrate as an oxidizer for metal CMP. Evidence adduced at trial establishes that during the relevant time period ferric nitrate was an industry standard, used by companies such as IBM and Advanced Micro Devices.

Indeed, Yancey, desirous of competing with Cabot in selling product to Sematech, mixed lot 3100 six days after Cabot, having filed its patent application, began informing certain of its customers of its use of fumed alumina, and ten months after DeGussa ended its allocation, and like Cabot, claimed a three-month shelf life, but unlike Cabot, before testing it. One would be inclined to ask how Yancey would feel comfortable telling Kaufman three months unless he knew from some tested source that three months was accurate (assuming, as the record suggests, he did not look back at his 1992 formulations).

However, Cabot does not show that Yancey had access to Cabot's product or the information about it. Other sources, such as Rodel, claimed the three-month shelf life, and Kaufman admitted that three months was reasonable. As against the inferences Cabot would draw from the temporal proximity between when Cabot began telling some of its customers about its product and the mixing of lot 3100, there remains the evidence of what Yancey made in February, 1992, and where he could have learned of the use of ferric nitrate: in light of which, copying is not clearly and convincingly proved.

7. Conclusions as to willfulness of infringement

It is a high standard: Clear and convincing evidence, having regard to the totality of the circumstances. By the application of which I conclude that willfulness of infringement has not been established, either by copying, deliberate manipulation of information supplied to counsel, failure to follow counsel's instructions, or failure to act as would a reasonable person with any confidence that a court might hold the patent valid or not infringed. Consequently, that ground for determining the case exceptional under ž 285 is not available to Cabot.

II. Litigation Misconduct

Cabot argues that STI crafted and pursued a vexatious litigation strategy, and that this is an independent reason for awarding attorneys' fees. This vexatious strategy consisted of carrying on unnecessarily protracted discovery but quitting before a decision could validate Cabot's patent.

A. Standard as to litigation misconduct

Litigation misconduct and unprofessional behavior may suffice to render a case exceptional. Multiform Desiccants, Inc. v. Medzam, Ltd., 133 F.3d 1473, 1481-82 (Fed.Cir.1998); Serrano v. Telular Corp., 111 F.3d 1578, 1585 (Fed.Cir.1997) (award of attorneys' fees upheld where based in part on "litigation conduct that resulted in needless expense," including relitigation of issues already decided and continuing to sell products embodying only minor differences from infringing products as to the sale of which it had been enjoined); Sensonics, 81 F.3d at 1574; Jurgens v. CBK, Ltd., 80 F.3d 1566, 1570 (Fed.Cir.1996) (Bad faith includes "inequitable conduct during patent prosecution, bringing vexatious or unjustified suits, attorney or client misconduct during litigation, or unnecessarily prolonging litigation."); Read Corp. v. Portec, Inc., 970 F.2d 816, 831 (Fed.Cir.1992); Badalamenti v. Dunham's, Inc., 896 F.2d 1359, 1364 (Fed.Cir.1990); Spectra-Physics, Inc. v. Coherent, Inc., 827 F.2d 1524, 1537 (Fed. Cir.1987); Kloster Speedsteel AB v. Crucible Inc., 793 F.2d 1565, 1580 (Fed.Cir. 1986); Rohm & Haas Co. v. Crystal Chemical Co., 736 F.2d 688, 691-92 (Fed.Cir.), cert. denied, 469 U.S. 851, 105 S. Ct. 172, 83 L. Ed. 2d 107 (1984); Stevenson v. Sears, Roebuck & Co., 713 F.2d 705, 713 (Fed.Cir. 1983).^[17]

"It is the judicial duty to refuse to condone behavior that exceeds reasonable litigation tactics." Sensonics, 81 F.3d at 1575. However, "[e]ven an exceptional case does not require in all circumstances the award of attorneys' fees." S.C. Johnson, 781 F.2d at 201. The purpose of ž 285 is "to provide discretion where it would be grossly unjust that the winner be left to bear the burden of his own counsel which prevailing litigants normally bear." J.P. Stevens, 822 F.2d at 1052 (emphasis in original); accord, Sun-Tek Industries v. Kennedy Sky Lites, 929 F.2d 676, 679 (Fed.Cir.1991) ("Section 285 may be used to award attorney fees to prevent "gross injustice" when a party has litigated vexatiously."); Machinery Corp., 774 F.2d at 472; Rohm & Haas Co., 736 F.2d at 691 ("The exercise of discretion in favor of such an allowance [of fees] should be bottomed upon a finding of unfairness or bad faith in the conduct of the losing party, or some other equitable consideration of similar force, which makes it grossly unjust that the winner of the particular lawsuit should be left to bear the burden of his own counsel fees which prevailing litigants normally bear." Quoting Park-in Theatres, Inc. v. Perkins, 190 F.2d 137, 142 (9th Cir.1951)). The court "may consider the litigation actions of both sides in connection with ž 285." Sensonics, 81 F.3d at 1575. The court's decision as to the award of attorneys' fees "should be in furtherance of the policies of the laws that are being enforced, as informed by the court's familiarity with the matter in litigation and the interest of justice." S.C. Johnson, 781 F.2d at 201.

Whether litigation misconduct renders the case exceptional for purposes of ž 285 is determined by the trier of fact in light of the totality of the circumstances. L.A. Gear, Inc. v. Thom McAn Shoe Co., 988 F.2d 1117, 1128 (Fed.Cir.1993).

B. Analysis as to litigation misconduct

1. As to the inferences which may be drawn from STI's "summary document" and STI's attempts to end the lawsuit

Cabot contends that the document summarized STI's strategy from at least December, 1996, to assure that the Cabot patent was not validated, a strategy that dictated decisions such as STI's attempt to quit the lawsuit. Cabot does not argue that quitting the lawsuit was vexatious, but that the strategy of quitting the lawsuit so as not to risk a judicial determination that Cabot's patent was valid, was vexatious. "We will not lose: we will quit first.... Our desire is NOT to win ..." proclaims the document, and, according to Cabot, Yancey's testimony that this was STI's strategy is correctly interpreted as referring to STI's litigation strategy. As referenced in that document, STI wanted a royalty-free license from Cabot to practice the claimed invention while excluding others from doing so; despite STI's asserted belief in the patent's invalidity. Cabot finds motivation in pressure from STI's investors to sell STI. Yancey wrote that the suit was "more an asset than a liability" to Rodel. Yancey also viewed the suit as an asset in selling STI to Rodel. While those negotiations continued, he had an incentive to keep the suit going.

Yancey pursued his strategy by offering STI's retains to Cabot for inspection if Cabot agreed not to proceed with the suit. Thus, Jones' July 5th letter to Morisseau offered lot 2021 but by Cabot's reading required dismissal as a precondition to proceeding with the tests. Twenty-three months later, STI initiated the series of communications with the court attempting to end the lawsuit: STI's November 3d letter to the court, its November 4th "Motion to Enter Partial Judgment," its November 23d motion to reconsider the denial of that motion, and its November 16th "Renewed Motion to Enter Partial Judgment." Cabot reasons that if the court accepted one of these offers of injunctions, then STI would not have lost, because the preclusive effect of the injunction could be limited. STI would then have achieved the aim of its strategy. Cabot argues that this is proof positive of the strategy announced in the document.

By Cabot's account, Rodel CEO Budinger's conduct after taking the reins of the litigation is further evidence of the strategy. Budinger obtained Foley & Lardner's opinion, yet specified that that opinion not address Claim 39 (slurries including, among other things, fumed alumina comprising at least 50% gamma phase), which is admittedly infringed and for which STI has no tenable claim of prior art. As to Claim 39, Budinger, who did not read the September 11th opinion, had no reason to believe the patent was invalid. As an inventor and expert in slurry chemistry, he knew that Met 202 had little commercial value. He testified to watching every dollar. He testified that he conducted litigation with the intent to win. Yet he continued this litigation for 14 months before the motions were filed to end the suit.

STI responds that there is no evidence that the "settlement document" was anything other than one page out of the set of overhead slides that was to be taken to Chicago for the settlement meeting. It was nothing more than a summary of points for settlement discussions, a strategy of settlement, not of litigation. The document simply says that STI would rather enter into a license with Cabot than have the Neville patent declared invalid; and that STI was prepared to walk away from the case rather than endure a trial. Even if STI had to resign the contest because of the expense of trial, there were larger companies that could present the same evidence. In his testimony, Yancey did not mean that it expressed STI's litigation strategy, but STI's Chicago settlement discussion strategy.

STI argues that Cabot never explains why STI would drag Cabot through 30 months of expensive litigation never intending to win, STI being an extremely small company in comparison to Cabot, and Met 202 generating meager revenues. According to STI, Budinger made numerous attempts to settle the case. When Cabot refused to settle, STI litigated to win. But with trial approaching Budinger was told that trial would cost almost one million dollars. He concluded that spending a million dollars to vindicate a product that promised little revenue was not justified.

By STI's account, Budinger decided not to get an opinion from Foley & Lardner on Claim 39 because it was not relevant to Rodel's business.

According to STI, no evidence shows that Budinger was informed of STI's quit-before-losing strategy, so if he pursued this strategy he arrived at it himself. Cabot's theory of prolonging the litigation and quitting before losing would therefore require the conspiratorial efforts of Yancey, BSPG attorneys, Budinger, Dority & Manning attorneys, Bouchard, and Smith Helms attorneys.

I resolve this question as follows:

The phrases said to express a litigation strategy should be read in context:

We Will Not Lose: we will quit first ÔÇö not validating patent and leaving Cabot to fight other Goliaths with no precedence but the ghost of our data and retains remain The Industry Will Not Accept A Monopoly: Cabot will be forced to license or technology will evolve in a different direction Our Desire is NOT To Win: everyone else wins for free & we forfeit a lucrative co-licensing opportunity

STI says it does not want to win because it wants a license. The industry will force Cabot to license anyway. If Cabot persists, STI will quit before Cabot can validate its patent. Cabot will have to litigate against larger competitors without a determination of validity, but STI's data will remain. So the best course for Cabot is to license STI. In context, the threat of quitting before losing is marshalled to persuade Cabot to give STI a license. STI says it is pointless for Cabot to seek judicial validation of its patent through litigation because STI will not give Cabot that opportunity.

"[W]e will quit first" is a statement about what STI will do rather than lose. It is difficult to read it as anything other than a statement about how STI in December, 1996, intended to defend the suit, and it is unreasonable to suppose that Yancey was not at least aware of it. I do not conclude otherwise. In December, 1996, Yancey and STI intended to quit before losing.

"[W]e will quit first" does not confess that STI will protract the suit through expensive, unnecessary and burdensome discovery. That is a separate tactic, perhaps, for bringing Cabot to heel, but a separate defensive measure from what is admitted by saying "we will quit first."

At issue is whether this statement of intent is representative of a grand strategy actually carried out, as evidenced by the motions to end the suit on the eve of trial. Can we conclude that these motions are the smoking gun, as it were, establishing the declared intent to "quit first" as the overarching strategy actually carried out'? I conclude not, because Budinger took the reins, there is no evidence that Budinger knew of or independently derived the strategy, and there is the independent and sufficient reason given at trial for ending the suit, namely, the projected million-dollar cost of trial.

Having reason to believe the patent was not invalid, and thus continuing the litigation with at least some apprehension that he would not prevail, Budinger nonetheless did not have his attorneys file a motion to end the suit for fourteen months. This may suggest a protract-the-litigation strategy, but it does not prove a quit-before-losing strategy. Apart from the fact that for fourteen months Budinger did not direct his attorneys to file motions to end the suit there is no evidence that Budinger adopted a prolonging strategy or directed his new attorneys to engage in delaying tactics. There is no evidence that his new attorneys did so. However peculiar his letter to Bodman, Budinger in fact made settlement overtures during this period.

STI's end game was played by new attorneys and directed by Budinger. Because of their untainted intervention, I conclude that Cabot has not established that the improper tactic suggested by the overhead slide in December, 1996, was a grand strategy actually carried out.

2. As to STI's conduct during discovery

A perduring ground of contention in this litigation has been the parties' conduct in discovery and representations to the court. This remains a theme of Cabot's case for assigning to STI's litigation behavior the appellation "vexatious." In pursuit of that assignment, Cabot draws attention to a series of instances said to signify intent. Connecting these dots, inferring from specific to general, Cabot discovers a pattern, each instance an exemplar and furtherance of a plan to draw out the litigation and make it more expensive, in short, a strategy, vexing Cabot and the court, protracting what tiny STI knew to be the inevitable demise of its fortunes on the field, to coax a weary but well-heeled Cabot to the licensing table.

Each of the instances has been described, argued over, and in some manner or other addressed by this court. Stipulations have been enforced, discovery compelled, the necessity and scope of depositions reviewed, orders entered (I counted five, more or less).

In assessing the contribution of this conduct to Cabot's proof of vexatious litigation, I have before me my duty, rehearsed in rule and case, motion and argument, keeping me immediately, poignantly, even painfully, aware of this and any court's solemn duty to control the behavior of parties in this stylized combat, to punish its excesses, to vindicate truth and the search for truth within the Rules. This is a duty that is shirked on peril of defeating the search for truth, substituting gamesmanship for its ordered discovery, rendering justice inaccessible to parties lacking means, and enthroning falsehood. That is intolerable. Mr. Wester's arguments to this effect were eloquent, powerful, and persuasive.

Then is the pattern of STI's behavior so in excess of the normal give and take of discovery in patent litigation and accepted bounds of its conduct, as to render the case exceptional, to require this court in the interest of substantial justice, fulfilling its regulatory duty under the rubric of ž 285, to tax STI the financial consequences of its alleged perfidy?

We must attend to the totality, "quit before losing" along with other acts said to have been taken in service of the grand strategy. "Quit before losing" is discussed above. What remains is whether the other instances submitted by Cabot render STI's conduct in totality exceptional. They do not.

First, both sides were aggressive, both sides fought hard, both sides required the intervention of this court. Cabot accuses STI, but Cabot was itself compelled to disclose the characteristics of its SEMISPERSE and CAB-O-SPERSE silica slurry products prior to the Cabot patent, to produce unredacted documents, to answer deposition questions, and to declare its claim construction of the Cabot patent. Cabot's behavior is not decisive in assigning degree of fault to STI, but it is a consideration in this totality, big-picture analysis, where the actions of STI were partly taken in response to those of its accuser.

Second, Yancey did not require dismissal before handing over his retains. Jones' letter does not require dismissal. The fact of the matter is that from mid-July, 1996, those retains were available for Cabot's inspection and for testing by a third-party laboratory. Cabot protests that STI wanted to dictate the terms of the inspection and that disclosure would have been insufficiently complete for Cabot to safeguard its interests. I do not read the exchange of letters as foreclosing negotiation of the terms of disclosure, at least not on STI's part. I conclude from the evidence that at least during the crucial half-year between July, 1996, and December, 1996, Yancey's desire to settle was sufficiently sincere that the settlement overtures could not be reasonably interpreted as part of the strategem, and if not, if Yancey was in some measure sincere about settling on genuinely negotiated terms, then the extent to which a stratagem, if indeed it were, of drawing out the litigation and making it more expensive was in place, or in any event controlling of decisions, is substantially brought into question.

Third, there is simply no evidence that Jones' forays across the nation, from deposition to deposition to deposition, and his procedural maneuverings, were guided by any consideration beyond his judgment as to what was prudent to defend the case. I looked at these instances very carefully in considering Rule 37 cost shifting and concluded that it was not warranted under the circumstances. In that review I considered the arguments for and against his purportedly offending behavior. That does not decide the matter for ž 285 purposes. But even if the law did, this evidence does not support an analogical respondeat. On these facts, I conclude that to visit upon STI the alleged sins of its former and erstwhile defender is not justified.

For these reasons I conclude that STI's litigation conduct is not shown to be exceptional as that term is understood for purposes of ž 285. This is so if the acts adduced by Cabot are considered individually, together, and as part of a grand design to protract the litigation and finally quit before losing.

3. Conclusions as to litigation misconduct

STI dulls the sharp edges of Cabot's attack. The evidence in its totality does not prove a discovery strategy aimed at prolonging the process of obtaining relevant information. Nor, in the final analysis, does it establish that STI filed its end-game motions to effect a quit-before-losing stratagem.

III. Conclusions as to exceptional case and award of attorneys' fees

A. Standard as to exceptional case and award of attorneys' fees

The court in exceptional cases may award reasonable attorney fees to the prevailing party.

35 U.S.C. ž 285.

When deciding whether to award attorneys' fees under 35 U.S.C. ž 285, the trial court asks whether there is clear and convincing evidence that the case is "exceptional", and if so, whether it is appropriate to award attorneys' fees to the prevailing party. Interspiro U.S.A., Inc. v. Figgie Intern., Inc., 18 F.3d 927, 933 (Fed.Cir. 1994); accord, J.P. Stevens Co., Inc. v. Lex Tex Ltd., Inc., 822 F.2d 1047, 1050 (Fed. Cir.1987); compare, Machinery Corp. of America v. Gullfiber AB, 774 F.2d 467, 470-71 (Fed.Cir.1985) (a four-part standard: "(1) the case is exceptional; (2) the district court may exercise its discretion; (3) the fees must be reasonable; (4) the fees may be awarded only to the prevailing party.")

Cases have been deemed "exceptional" for reasons such as willful or intentional infringement, vexatious or unjustified litigation, or inequitable conduct before the Patent and Trademark Office. When bad faith and willful infringement are found, the standard "is more readily met." Sensonics, Inc. v. Aerosonic Corp., 81 F.3d 1566, 1574 (Fed.Cir.1996). "[W]e are aware of few cases in which a patent owner has been granted attorney fees solely on the basis of litigation misconduct, without a concurrent finding of willful infringement." Beckman Instruments, Inc. v. LKB Produkter AB, 892 F.2d 1547, 1552 (Fed.Cir.1989). "When infringement was found not willful, the district court's denial of attorney fees or increased damages has not been disturbed." S.C. Johnson & Son, Inc. v. Carter-Wallace, Inc., 781 F.2d 198, 200 (Fed.Cir.1986).

B. Conclusions as to exceptional case and award of attorneys' fees

Willful infringement and litigation misconduct are grounds for finding a case to be exceptional under ž 285. Willful infringement is a finding of fact; whether it is exceptional is an interpretation driven by case analogies and the judgment of the trier of fact in light of all the circumstances clearly and convincingly shown. By that standard, willful infringement is not established in this case and cannot serve to render the case exceptional.

The cases teach that litigation misconduct without willful infringement can serve as a basis for finding a case to be exceptional, but it is unusual, and it is not the situation before us. Cabot's strongest argument, the alleged quit-before-losing stratagem, does not enjoy the necessary evidentiary support linking the overhead slide with the motions filed in late December, 1998. Moreover, although discovery was prolonged, expensive, and contested by both Cabot and STI, there was no egregious abuse of the judicial process. Even if the litigation misconduct as alleged had been shown to be such, it would not by itself be sufficient on these facts absent a finding of willful infringement.

All things considered, it is not grossly unjust that Cabot and STI should bear the burden of their respective attorneys' fees.

Accordingly, I conclude that this case is not exceptional under ž 285 and costs and attorneys' fees should not be awarded.

NOTES

[1] Cabot withdrew its prayer for damages, including lost profits and treble damages for infringement.

[2] STI's motion strike portions of Cabot's affidavit [doc. 190] is denied.

[3] Yancey got a sample on two occasions. The second occasion was a litigation-related swap in 1996 or 1997. As far as may be discerned from trial, on the first occasion, Yancey walked into his office one morning, date uncertain, and found a Coke bottle full of Cabot slurry on his desk. Yancey testified under oath that he does not know and never inquired where the Coke bottle came from. As it stands, this strange fact is not tied to other evidence or itself sufficiently developed to ground an inference of misconduct.

[4] Yancey testified that, although he is a careful record keeper, he cannot find "Paul's notes."

[5] In Kaufman's communications with Yancey from 1991 through 1995, Yancey referred to particle size determination by laser light scattering, a process which at that time was conventional in the industry in conducting particle size analysis of components of slurries, and which is quite different from TEM. Yancey never referred to TEM measurements until his October 25th letter.

[6] Kaufman believes that a very stable slurry would have meant at least one year's stability. Anything that had only a 3-month stability would be only barely stable.

[7] The particle size in the patented Cabot invention was less than approximately .4 micron.

[8] Yancey gave a copy of this letter and its accompanying particle size charts to his attorneys at BSPG in the materials which he supplied on July 1, 1996.

[9] After initial screening in the early 1990s, IBM obtained satisfactory results with the 201A/280 product and continued to experiment with it. IBM tested STI's 201/60, 201/140 and 201/180 slurries. The 201/180 became a slurry for an IBM development project.

[10] Neither Cabot nor STI has offices in California. Cabot's corporate offices are in Boston, and its CMP division is in Aurora, Illinois.

[11] Compare Yancey's October 25, 1994, letter to Kaufman, which references a shelf life of three months and TEM analysis. In Kaufman's communications with Yancey from 1991 through 1995, Yancey referenced particle size determination by laser light scattering, which is different from TEM. Yancey never referred to TEM measurements until the October 25th letter.

[12] Ed Hardin of Southern Engineering in Charlotte supervised the drawing of the samples.

[13] Jones believed that prior sales of lot 2361 to IBM by itself invalidated the Cabot patent. Trans., Vol. 8, Sept. 1, 1999, at p. 62, lines 17-21, p. 64, lines 2-9, 19-21. According to Jones, BSPG would have issued the same opinion based upon information as to what went into STI's precritical date products even if the retains had not existed. The existence of the retains was fortuitous. Jones believes that the patent claims would have been invalid based on the precritical date activities of STI even if there had not been any such retains. Id. at p. 59, lines 6-14. Jones believed that the purpose of the opinion was to provide STI with information on which it could rely in deciding whether to continue or discontinue certain business operations, particularly the manufacture of the accused products. Id. at p. 59, line 23 ÔÇö p. 60, line 1.

[14] Noglows had held that position since early summer, 1998. Prior to assuming this position, Noglows since November, 1992, had been General Manager of Cabot's Cabosol Division, and for the two years prior was Manager of Cabot Australia Asia. As General Manager, his responsibility had been to find opportunities to leverage Cabot's core technology (combustion and plant technology). One of the applications was CMP and dispersions.

[15] This was a 40,000-square-foot building which Cabot purchased for CMP research.

[16] As to which, see Transcript, August 27th at 160-63.

[17] At least one court has concluded that discovery misconduct should be dealt with under Rule 37 rather than ž 285. Arbrook, Inc. v. American Hospital Supply Corporation, 645 F.2d 273, 279 (5th Cir.1981).