Expert witnesses proven to be pivotal in Dow Chemical patent infringement suit

Expert witnesses proven to be pivotal in Dow Chemical patent infringement suit

Dow Chemical Co v NOVA Chemicals Corp, 2014 FC 844

The Federal Court found that NOVA Chemicals Corporation (“NOVA”) infringed Canadian Patent No. 2,160,705 (the ‘705 Patent), owned by The Dow Chemical Company (“Dow”) and subsidiaries, by NOVA’s use of its “SURPASS” polyethylene product. [172] The ’705 Patent relates to a particular polymer composition for Dow’s “ELITE” polyethylene product. [3, 252]

The defendant NOVA argued invalidity of the ‘705 Patent for lack of utility, claims broader than any invention made or disclosed, anticipation, obviousness, double patenting, and insufficiency of the specification. [173] NOVA was unsuccessful on all counts.

The case involved lengthy disputes over technical terms in claim construction which had significant impact on the infringement and invalidity analyses. The claims at issue were claims 11, 29, 30, 33, 35, 36, 41, and 42. [119] Claim 11 and claim 10 from which it depends are representative of the technical terms that are in dispute:

10.      An ethylene polymer composition, comprising:

(A)      from 10 to 95 percent (by weight of the total composition) of at least one homogeneously branched linear ethylene/α-olefin interpolymer having:

(i)         a density from 0.88 grams/cubic centimetre (g/cm3) to 0.935 g/cm3,

(ii)        a molecular weight distribution (Mw/Mn) from 1.8 to 2.8,

(iii)      a melt index (I2) from 0.001 grams/10 minutes (g/10 min) to 10 g/10 min,

(iv)      no linear polymer fraction,

(v)        a single melting peak as measured using differential scanning calorimetry; and

(vi)      a short chain branching distribution index (SCBDI) of greater than 50 percent; and

(B)      from 5 to 90 percent (by weight of the total composition) of at least one heterogeneously branched ethylene polymer having a density from 0.91 g/cm3 to 0.965 g/cm3.

11.      The composition of claim 10, wherein the homogeneously branched linear ethylene/α-olefin interpolymer has a slope of strain hardening coefficient of from 1.3 to 10.

Claim construction

The issues before the parties regarding claim construction were as follows:

1. Who is the person skilled in the art?

2. What is the meaning of “ethylene polymer composition” in the asserted claims of the ‘705 Patent?

3. What is the meaning of the term “comprising” in the asserted claims of the ‘705 Patent?

4. What is the meaning of the terms “homogeneously branched” and “heterogeneously branched” in the asserted claims of the ‘705 Patent?

5. What is the meaning of “linear” and “substantially linear” in the asserted claims of the ‘705 Patent?

6. What is the meaning of “slope of strain hardening coefficient” in the asserted claims of the ‘705 Patent?

7. What is the meaning of “linear polymer fraction” in the asserted claims of the ‘705 Patent? [13]

The parties did not disagree that the person skilled in the art would be a scientist with at least a Bachelor’s degree in chemical engineering, chemistry, materials science or polymer science, and at least two years’ experience in the characterization of polyolefins, including an understanding of polymer compositions and the materials science of polymers. [36]

Each of the disputes regarding claim construction were won by Dow.

NOVA argued that the term “ethylene polymer composition” could include an arbitrary selection of polymer molecules taken from a single polymer,[39] but the Court favoured the position of Dow that an ethylene polymer composition would be understood as a blend of distinct polymer components.  [42]

NOVA argued that the term “comprising”, used in each of the claims, should carry the restrictive definition of “consisting of”, but that in claim 41 “comprising” should be given the conventional meaning of “including but not limited to”. [43] However, the Court held that the term “comprising” in all cases should carry the conventional meaning in patent law of “including but not limited to”. [45, 54]

NOVA argued that the terms “homogeneously branched” and “heterogeneously branched” should be distinguished by a bright line test, where homogeneously branched means having a Composition Distribution Branch Index (CDBI) of greater than either 30% or 50%, and heterogeneously branched means having a CDBI of less than either 30% or 50%. [55] The expert witnesses disagreed on whether the bright line should be held at 30% or 50%. [58] Finding NOVA’s preferred definitions to be problematic, [63] the Court favoured the definition provided in the ‘705 Patent, which is also the definition favoured by Dow. [64-65] Under this construction, firstly, a homogeneously branched polymer is a polymer in which the comonomer is randomly distributed within a given interpolymer molecule and wherein substantially all the interpolymer molecules have the same ethylene/comonomer ratio within that interpolymer. Secondly, a heterogeneously branched polymer has a distribution of branching different from and broader than the homogeneously branched ethylene/α-olefin interpolymer at similar molecular weight and Short Chain Branching Distribution Index (SCB) averages. [65]

The dispute over the terms “linear” and “substantially linear” essentially reduced to one over the meaning of “long-chain branching” (LCB). [66] NOVA argued that LCB includes the existence of octane comonomer short chain branching. However, the common meaning of the term does not include this kind of branching,[66] and on the evidence the Court found that the term LCB as used in the patent could not include this kind of branching. [72] The Court thus favoured the common meaning, which is the meaning used in the ‘705 Patent, which is also Dow’s preferred definition. [73, 74]

The parties disagreed on the meaning of “slope of strain hardening coefficient” in the ‘705 Patent. [75] Central to this issue was that NOVA argued that an engineering stress/strain curve is to be used to calculate this coefficient, but Dow contended that a person skilled in the art would infer that a load/elongation curve should be used. [92] The Court concluded that since information about gauge length, which is necessary for using an engineering stress/strain curve, [90] is not included in the patent, then Dow’s interpretation that a load/elongation curve should be used should prevail. [93] The Court further concluded that a person skilled in the art would know how to calculate the required coefficient using the load/elongation curve. [102]

NOVA disputed DOW’s characterization of the term “linear polymer fraction” but did not put forward any clear construction of its own. [105] Relying on witness testimony that the term “linear polymer fraction” means a high density fraction, [108] the Court accepted Dow’s construction that the person skilled in the art would understand “linear polymer fraction” to mean a higher density fraction that may have some light branching. [110]

Infringement

Having decided each issue of claim construction in favour of Dow, the Court subsequently found that NOVA infringed each of the claims in dispute. The claims at issue were claims 11, 29, 30, 33, 35, 36, 41, and 42. [119] The infringement analyses reduced to four sub-issues, outlined below.

Are NOVA’s Reactor 1 polymers linear or substantially linear as those terms are defined?

Having decided under claim construction that the term “long chain branching” does not include octane comonomer short chain branching, the Court found that NOVA’s SURPASS product was indeed linear or substantially linear as defined in the ’705 Patent. [125, 127]

Do NOVA’s Reactor 1 polymers have the required SCH value?

Seeking to prove this fact, Dow had an expert witness reproduce 14 polymers that represent various polymers produced by NOVA’s Reactor 1 in a mini-plant. [128] NOVA disputed the expert witness’s reproduction of the polymers, [130] but the Court decided on the evidence that the reproductions were accurate. [135] Also in dispute was whether an engineering stress/strain curve or a load/elongation curve should be used to calculate the SCH value of these reproductions. Having decided during claim construction that a person skilled in the art would know to use a load/elongation curve, and that the skilled person could calculate the SCH value accordingly, the Court then found on the evidence that the 14 polymers that were reproduced had an SCH value that infringed upon the claims of the ‘705 Patent. [152]

Does the whole polymer made in Reactor 2 satisfy part (B) of Claim 11?

The Court found that the NOVA’s product from Reactor 2 infringed claim 11 based on tests performed by Dow’s expert witnesses [160] and testimony from NOVA’s expert witnesses [155, 158] which together demonstrated that the physical requirements in claim 11 were met.

Does the “higher density” component of Reactor 2 satisfy part (B) of Claims 11, 29, 30, 33, 35, 36, 41 and 42?

At issue was whether a high density (HD) version of SURPASS, made in NOVA’s Reactor 2, was heterogeneously branched so as to satisfy component B of the claims at issue. [162, 166] Relying again on tests performed by an expert witness of Dow, [169] and having decided the meaning of the term “heterogeneously branched” under the claim construction analysis, the Court found that NOVA’s HD SURPASS polymer infringes each of the claims at issue. [172]

Validity

NOVA argued invalidity of the ‘705 Patent for lack of utility, claims broader than any invention made or disclosed, anticipation, obviousness, double patenting, and insufficiency of the specification. [173] NOVA was unsuccessful on all counts.

Lack of utility

NOVA argued that Dow promised that the invention would have synergistically enhanced physical properties. [179] After examining the patent, [184-185] the Court determined that no specified level of improvement was promised, [192] and that the standard for utility is a “mere scintilla” of utility. [195] This requirement for utility was met. [196]

Claims broader than any invention made or disclosed

NOVA argued that the ranges for density, melt indexes, weight percentages, and range of comonomers used to describe the claimed compositions were broader than the invention made. [200, 203] However, on the evidence, the Court found that sufficient researched was performed by Dow to support these claims. [202, 203]

NOVA also argued that the claims were missing an essential element relating to synergistically enhanced properties that made the claims broader than the invention disclosed. [206, 207] Having already decided that the ‘705 Patent does not promise synergistically enhanced properties, and considering Dow expert testimony that such properties are not essential elements to the claim, this argument failed. [208-209]

Anticipation

NOVA argued that two patents anticipated the ‘705 Patent: Canadian Patent Application No. 2,416,003 (the ‘003 Application) and U.S. Patent No. 4,629,525 (the ‘525 Patent). Considering that the ‘003 Application allows the reader to obtain a composition that was outside the claims of the ‘705 Patent, [215] the Court concluded that the ‘003 Application did not anticipate the ‘705 Patent since performing the ‘003 Patent would not necessarily result in infringement of that patent. [217] Considering that the ‘525 Patent teachers you to blend the whole of one polymer with the whole of another polymer, [222] whereas the ‘705 Patent teaches blending two components each of which have a polymer, [228] the Court concluded that the ‘525 Patent does not anticipate the ‘705 Patent.

Obviousness

NOVA argued that the ‘705 Patent was obvious. [231] However, the Court noted that in 2002 NOVA appeared to have problems with developing a polymer that was of the standard as that described in the ‘705 Patent. [248, 249] As such, the invention could not have been obvious as of the time of filing the ‘705 Patent, in 1994. [249]

The Court was of the view that the “obvious to try” test was not applicable in this case but applied the test anyway, [257] only to find that the invention not obvious to try. [260-263]

Double patenting

NOVA argued that Canadian Patent No. 2,120,766 covers the same subject matter of claims 41 and 42 of the ‘705 Patent, and that Canadian Patent No. 2,153,978 cover the same subject matter as all asserted claims of the ‘705 Patent. [266] Relying expert testimony of a Dow witness that the patents at issue differ, the Court rejected these arguments. [269, 270]

Insufficiency of the specification

NOVA argued that the specification of the ‘705 Patent was insufficient for teaching an insufficient method for calculating the SHC. [271] Having resolved the issue that the patent correctly teaches how to calculate the SCH with a load/elongation curve under the claim construction analysis, and expert evidence that further explains the method for calculating the SHC, [274] the Court rejected this argument. [275, 277]

NOVA also argued that the term “comprising” permits an unidentified third polymer. [271] However, having decided the meaning of the term “comprising” under the claim construction analysis, the Court decided that the person skilled in the art would understand that the composition can include a third polymer. [278]

NOVA also argued that the ‘705 is insufficient for not disclosing the best mode, but the argument failed as such a requirement only applies to inventions relating to machines. [279]

NOVA also argued that the terms “homogeneously branched” and “heterogeneously branched” would not be meaningful to a person skilled in the art, but this argument was rejected relying on expert testimony of a Dow witness to the contrary. [281]

Evidentiary Issues

Dow argued that a number of Dow’s expert reports contained portions that should be inadmissible for improperly splitting Dow’s case. [Sched. A, para 4] The principal argument was that these portions improperly contain evidence on reply that should have been included in the examinations in chief. The Court rejected the majority of the arguments, but several paragraphs of the reports were deemed inadmissible. [Sched. A, para 67] Considering that Dow was successful on every substantive issue before the Court, it would seem that these redactions were inconsequential.

Appeal

The decision is being appealed.

Commentary

This case was heavily driven by expert evidence. The Court drew on the testimony given by the expert witnesses on both sides of the dispute to decide the issues of claim construction that led to the findings of infringement and validity.

This case also demonstrates the importance of the behaviour of expert witnesses at trial. The Court repeatedly sided with the Dow’s expert witnesses over NOVA’s expert witnesses when there was a conflict. In some instances the Court commented that Dow’s expert witnesses were more forthcoming (in different words):

“Dr. Young [a DOW witness] was a responsive witness while Dr. Cakmak [a NOVA witness] sometimes was reluctant to answer questions directly where the answers had an impact on his opinions” [150]

“Dr. Scott [a DOW witness] addressed the patent as a whole while Dr. Speed [a NOVA witness], when dealing with the first reference in the ‘705 Patent, ignored the parenthesized words in the Patent defining ‘destructive synergism’”. [194]

“I accept Dr. Young’s [a DOW witness] testimony as to where to measure the SSH. I prefer Dr. Young’s testimony over that of Dr. Cakmak [a NOVA witness] who in other areas was reluctant to admit the obvious”. [276]