Full opinion text
OPINION ROBINSON, District Judge. I. INTRODUCTION Plaintiffs Biacore, AB and Biacore, Inc. (collectively “Biacore”) filed this suit pursuant to 35 U.S.C. § 271 against defendant Thermo Bioanalysis Corporation (“Ther-mo”) on May 29, 1997, seeking damages (lost profit damages) and an injunction for alleged infringement of a patent that is directed to a matrix coating suitable for use in a biosensor. (D.I.l) Specifically, Biacore charges that Thermo willfully infringed U.S. Patent No. 5,436,161 (the “ T61 patent”) entitled “Matrix Coating for Sensing Surfaces Capable of Selective Biomolecular Interactions, To Be Used in Biosensor Systems,” issued July 25, 1995. (D.I.l) Biacore also alleges that Thermo is inducing infringement of the patent-in-suit. Thermo denies infringement and has counterclaimed for a declaratory judgment of invalidity and noninfringement of the ’161 patent. Thermo challenges the validity of the T61 patent under 35 U.S.C. §§ 102 (“anticipation”), 103 (“obviousness”), and 112 (“written description”). Specifically, Thermo charges that: (1) the patented invention was described in a printed, prior art publication before its development by the patentee (§ 102); (2) the differences between the patented invention and the prior art are such that the claims would have been obvious to one of ordinary skill in the pertinent art (§ 103); and (3) the subject matter of the T61 patent is not disclosed in sufficient detail in the written description of the grandparent application (§ 112). The court has jurisdiction over this matter pursuant to 28 U.S.C. § 1338(a). The parties tried this matter to the court from October 26, 1998 to November 2, 1998. Despite having identified in the pre-trial order claims 1-5, 9-11, and 15 as allegedly infringed by Thermo (D.I.96), “for purposes of trial” Biacore reduced the number of claims, asserting only claims 4 and 5. (D.I. 103 at 4) The following constitutes the court’s findings of fact and conclusions of law pursuant to Fed.R.Civ.P. 52(a). II. FINDINGS OF FACT A. The Parties 1. Biacore, AB is a Swedish corporation with its principal place of business in Uppsala, Sweden. (D.I. 103 at 80; D.I. 1, ¶ 2) Prior to October 1996, Biacore, AB was a subsidiary of the Swedish company Pharmacia AB, operating under the name Pharmacia Biosensor, AB. (D.I. 103 at 78-79) In 1996, Pharmacia AB merged with Upjohn Pharmaceuticals and Biacore, AB was spun off. (D.I. 103 at 78-79) Biacore, AB’s business is totally dedicated to the development, manufacturing, and marketing of affinity biosensors. (D.I. 103 at 80) Since 1990, it has sold its optical biosensor systems in the United States under the trade name BIAcore™. Biacore, AB is the owner of the ’161 patent. (D.I. 96 at 2) 2. Biacore, Inc. is a Delaware corporation with its principal place of business in Piscataway, New Jersey. (D.I.l, ¶ 2) It is the U.S. subsidiary of Biacore, AB and is responsible for the marketing and selling of BIAcore™ optical biosensors in the United States. (D.I. 103 at 79) The BIA-core™ biosensors sold by Biacore, Inc. are manufactured in Biacore, AB’s facilities in Uppsala, Sweden. (D.I. 103 at 80) 3. Thermo is a Delaware corporation with its principal place of business in Santa Fe, New Mexico. (D.I.l, ¶ 3) Since 1994, Thermo has marketed and sold its optical biosensor systems in the United States under the trade name IAsys ™ through its Affinity Sensors division. (D.I. 105 at 413-14) B. The Field of the Invention 4. Biosensors. The subject matter of the T61 patent relates to “the field of biosensors.” (Plaintiffs’ Exhibit (“PX”) 1, col. 1, Ins. 15-16) A biosensor is an analytical device comprising a biological or biologically derived sensing element which is either intimately associated with or integrated within a physical chemical transducer where the transducer may be, for example, optical, electrochemical, piezoelectric, thermoelectric or magnetic. (D.I. 104 at 266) Generally, [t]he usual aim [of a biosensor] is to produce a digital electronic signal which is proportional to the concentration of a specific chemical or set of chemicals. (Defendant’s Exhibit (“DX”) 574 at 3) Bio-sensors are employed in biomolecular interaction analysis, i.e., the study and characterization of the interactions between biologically active molecules. (D.I. 103 at 74-75) For example, in the pharmaceutical industry, biosensors are used to study the binding of a novel drug to the targeted receptor. (D.I. 103 at 75) Biosensors also are employed in the fermentation and bioprocessing, petro- and agrochemical, and pollution industries. (DX 513 at 19-20) 5. A biosensor is composed of two essential elements: (1) a biorecognition system and (2) a transducer. (PX 1, col. 1, Ins. 23-27; DX 513 at 20) In general, biosensors function by first immobilizing on a surface within the instrument ligands or receptors (e.g., whole cells, enzymes, lectins, antibodies, or receptor proteins) that are able to recognize target molecules (analytes) over a host of other biomolec-ules. (D.I. 103 at 183; D.I. 104 at 219-20; DX 513 at 20) The bound ligands then are contacted with a solution or suspension containing analytes having specific recognition sites such that they will bind to the ligands. (D.I. 104 at 218-20) Generally speaking, the binding of an analyte to a ligand (i.e., the biological recognition event) results in a change in one or more parameters associated with the interaction. (DX 513 at 22) The transducer element of the biosensor functions to respond to the products of the biological recognition event, converting the physio-chemical signal into a signal (e.g., an electrical output) that can be either visualized or processed in some fashion, e.g., via a computer. (D.I. 104 at 267; DX 513 at 22) 6. Biosensors employ a number of different types of transduction technologies. These technologies include thermister, electrochemical, potentiometric, optical, piezoelectric crystal, and amperometric transduction. (DX 574 at 3^4; DX 960; D.I. 106 at 759-60) Particularly relevant to the case at bar, optical biosensors employ an optical transducer that “detect[s] the change which is caused in the optical properties of a surface layer due to the interaction of the receptor with the surrounding medium.” (PX 1, col. 1, Ins. 28-31; D.I. 104 at 267) One type of optical biosensor, an evanescent wave optical biosensor, exploits the energy that is propagated beyond a reflecting surface, i.e., the evanescent wave. These biosensors “bring[] about or effect[ ] changes in the reflecting light as a result of interacting with the evanescent field,” i.e., by “taking advantage of the change in refractive index causing differences in the light signal.” (D.I. 104 at 267-68) 7.One type of evanescent wave technology relies on the phenomenon of surface plasmon resonance (“SPR”). SPR “is a quantum optical-electrical phenomenon that arises from the interaction of light with a suitable metal or semiconductor surface.” (D.I. 27, Ex. K at 516) Under certain conditions, the photon’s energy is transferred to plasmons on the surface of the metal or semiconductor. (D.I. 27, Ex. K at 516) The wavelength that excites the plasmons, the resonance wavelength, can be calculated by measuring the amount of light reflected from the surface. (D.I. 27, Ex. K at 516) The resonance wavelength is determined by the interaction between the plasmoris electric field and the matter within the field; thus, any change in the composition of the matter alters the resonance wavelength. (D.I. 27, Ex. K at 516-17) The magnitude of the change in the resonance wavelength is directly proportional to the change in composition of the surface. (D.I. 27, Ex. K at 516-17) As a result, SPR can be “exploited as a direct optical sensing technique that allows the real-time measurement of interfacial refractive index (dielectric) changes ... made at suitable metal or dielectric surfaces ... without the use of labels or probes.” (D.I. 27, Ex. K at 518) SPR optical biosensor technology, therefore, is a method whereby “changes in the refractive index in a layer close to a thin metal film are detected by consequential changes in the intensity of a reflected light beam.” (PX 1, col. 1, Ins. 44-47) Biacore’s bio-sensors employ SPR technology. 8. Another type of evanescent wave system technology employs “an integrated optical chip called the resonant mirror (RM),” which “comprises a glass prism with the top surface coated with a low refractive index silica spacer layer which is in turn coated with a thinner high refractive index monomode wave-guide of titania, hafnia or silicon nitride. This is then coated with the bioselective layer.” (D.I. 27, Ex. K at 519) In operation, a laser light directed at the prism “is repeatedly swept through an arc of specific angles,” generating, inter alia, an evanescent wave at the waveguide surface that penetrates into the sample. (D.I. 27, Ex. K at 519-20) “This wave detects surface binding events by detecting the changes in the refractive index which in turn change the resonance angle that is tracked by diode arrays.” (D.I. 27, Ex. K at 520) Thermo’s bio-sensors employ a resonant mirror. 9. Hydrogel. The ’161 patent specifically discloses a matrix coating that is comprised of a hydrogel. A gel, of which a hydrogel is a type, is “a solid colloidal dispersion consisting of a network of particles and a solvent that is immobilized in this network.” Dictionary of Biochemistry 192. A hydrogel is a material that imbibes or absorbs a large amount of water, a common example of which is gelatin. (D.I. 106 at 760-61) Because hydrogels are composed mostly of water, thus resembling the environment in which most biom-olecules are found, they have good biocom-patibility, i.e., bound biomolecules are more likely to be stable. (D.I. 106 at 761-62) Polysaccharide hydrogels and water-swellable polymer hydrogels are conventional ligand immobilization reagents. (DX 574 at 2) 10. Ligand Immobilization. In the context of the technology at issue, a ligancl is a molecule that binds to a macromolec-ule. See Dictionary of Biochemistry 273. Ligand immobilization is a method of fixing a biomolecule to a surface in some particular orientation. (D.I. 106 at 752-53) This procedure has long been employed in various types of chromatography. (D.I. 104 at 242-44) Although there are many ways by which to bind ligands to a surface, with respect to the technology at issue, they are immobilized via covalent bonding with reactive groups in the hydrogel matrix. (D.I. 106 at 752-53) In addition to biosensor technology, ligand immobilization is used in a wide variety of fields, including diagnostic assays, enzyme immobilization, and protein purification. (D.I. 106 at 753) 11. Activation. According to the ’161 patent, the hydrogel is activated to contain two types of chemical groups: charged groups capable of concentrating oppositely-charged biomolecules and reactive groups capable of covalently binding the' concentrated biomolecules. In the context of ligand immobilization, “activated” refers to the state of reactivity required to covalently bind another biomolecule under conditions that would not result in alteration of the biomolecule itself, with the exception of that alteration necessary to allow for the covalent binding. (D.I. 106 at 762) In contrast to activated groups, which are able to react with- and bind a biomolecule under the mild conditions necessary for biomolecule immobilization, reactive groups react under reasonable, or “normal,” conditions. (D.I. 106 at 766) Charged groups, as that term is understood in the art of ligand immobilization, are groups containing either a positive or negative charge. (D.I. 106 at 764) They function to concentrate or attract oppositely-charged biomolecules. (D.I. 106 at 764) In general, the term “charged groups” describes the use of an electrostatic concentration. (D.I. 107 at 867) 12. In the context of the ’161 patent, the ligands are concentrated into the hy-drogel matrix via the electrostatic charge created by the presence of oppositely-charged groups incorporated into the hy-drogel. (D.I. 107 at 867-68) The reactive groups in the hydrogel then act to covalently-bind the concentrated ligands to the hydrogel in an orientation that preserves the ligands’ affinity function. (D.I. 107 at 868) As a result, the immobilized ligands are able to attract the analytes from the solution. ' (D J. 107 at 868) C. The Technology Developed by the Biacore Researchers 13. As of 1983, a- number of obstacles faced researchers attempting to develop a functional biosensor. These problems had to do with capacity, activity, and nonspecific binding. (D.I. 103 at 188-90; D.I. 104 at 271-72) With regard to capacity, the two-dimensional (i.e., planar) surfaces employed in the prototypical biosensor limited the amount of available surface area. (D.I. 103 at 188-90; D.I. 104 at 271-72) Even if the ligands were tightly packed on the surface, there was insufficient ligand immobilization to yield a signal that would be of use for biosensor purposes. (D.I. 103 at 188-90; D.I. 104 at 271-72) The capacity problem was exacerbated by the activity problem, which was two-fold. (D.I. 103 at 188-90; D.I. 104 at 271-72) Specifically, the ligands would bind to the surface in an orientation that would prevent them from interacting with the ana-lytes. (D.I. 103 at 188-90; D.I. 104 at 271-72) Moreover, direct adsorption often would cause the ligands to denature, i.e., breakdown, thereby losing their ability to function. (D.I. 103 at 188-90; D.I. 104 at 271-72) Finally, nonspecific binding, i.e., unwanted binding events at the surface, would contribute to the signal coming from the biosensor unit and thereby confound the data. (D.I. 103 at 188-90; D.I. 104 at 271-72) In addition, a problem specific to evanescent wave optical biosensors concerned maximizing the biomolecular interactions throughout the available detection field, which extends a few hundred nanometers above the sensor surface. (D.I. 104 at 271; PX 33 at 101020) 14. In 1984, Pharmacia, AB created a division, Pharmacia Biosensor, AB (“Phar-macia”), solely for the purpose of developing a functional affinity-based biosensor for the study of biomolecular interaction. (D.I. 103 at 71-72; PX 360 at BIA 003080-81) Interest in the field had been stimulated by the publication in 1983 of an article by researchers at Linkoping University in Sweden demonstrating, for the first time, the use of SPR for biosensing applications. (D.I. 104 at 269-70, 275-74) Competition in the field was high. (D.I. 104 at 275) Initially, the Pharmacia researchers employed functioning coupling reagents and methods that were being used at that time for affinity chromatography and enzyme immobilization. (D.I. 104 at 247-48) They worked with two-dimensional silicon surfaces, immobilizing ligands via either silan-ization of the surface or direct adsorption. (D.I. 103 at 185-87) Neither method, however, yielded a workable or usable bio-sensor as the aforementioned salient problems persisted. (D.I. 103 at 187) 15. By September 1985, the Pharmacia researchers had introduced hydrogels to the surface in an attempt, to decrease, or obviate altogether, the incidence of nonspecific binding. (D.I. 103 at 190; D.I. 104 at 274) At that time, it was known that hydrogels, because they are highly water-solvated, form a biocompatible surface. (D.I. 104 at 274; D.I. 106 at 761-62) The Pharmacia researchers believed that the attachment of a hydrogel would hamper the ability of undesired biomolecules (i.e., biomolecules other than the analytes) to contact the surface, thereby minimizing nonspecific binding, while at the same time displaying the required ligand. (D.I. 103 at 190-91; D.I. 104 at 275) In addition, the introduction of a hydrogel would create a three-dimensional matrix thus increasing capacity and exploiting the evanescent wave phenomenon to the greatest extent. (D.I. 103 at 191; D.I. 104 at 267-68, 273-74) Finally, the researchers postulated that attachment of ligands to a fluid hydro-gel structure, rather than a planar surface, would not only increase accessibility, resulting in a commensurate increase in activity, but would also decrease the incidence of ligand denaturation. (D.I. 103 at 191; D.I. 104 at 274) 16. The first hydrogel employed by the Pharmacia researchers was dextran. (D.I. 103 at 192) At that time dextran, a naturally occurring polysaccharide, was being used in chromatography procedures as a matrix for the binding of biomolecules. (D.I. 104 at 244; PX 1 col. 6, In. 6) The researchers selected dextran because it was a biocompatible and, ostensibly, inert material. (D.I. 103 at 192; D.I. 104 at 233; D.I. 106 at 762) Moreover, it was readily available in different grades from Pharma-cia. (D.I. 103 at 192; D.I. 104 at 232-33) The fact that dextran was thought to be inert was important to the Pharmacia researchers, who wished to avoid any nonspecific binding caused by charged interactions between the nonanalyte biomolecules in the solution and charged groups on the biosensor surface. (D.I. 103 at 192) 17. Although dextran’s inert nature was beneficial with respect to reducing nonspecific binding, it was a disadvantage with respect to immobilizing ligands. (D.I. 103 at 192; PX 489 at 289) Therefore, the Pharmacia researchers sought to modify the dextran by introducing reactive groups that could covalently bind ligands into the dextran. (D.I. 103 at 192-94) The scientists experimented with a variety of conventional reagents, including allyglyci-dylether, cyanodimethypyridin, carbonyldi-imidazole, and tresyl chloride. (D.I. 103 at 192-94) In each instance, however, the modified dextran failed to yield a matrix suitable for use in a biosensor, i.e., the signal produced was not “good enough” because an insufficient amount of active ligand was bound. (D.I. 103 at 193) 18. In the summer of 1986, having failed to produce a workable matrix, the Pharmacia researchers began exploring the possibility of employing SPR technology techniques that had been developed elsewhere in the company for use in non-biosensor applications. (D.I. 103 at 195) Specifically, the researchers began to experiment with charged hydrogel surfaces, conjecturing that these surfaces would interact with ligands by electrostatic attraction. (D.I. 103 at 195) The results of the experiments demonstrated that, by using a dextran hydrogel matrix possessing both charged and reactive groups, a dramatic increase in capacity (over 1000%) could be achieved, even at reduced concentrations of ligands. (D.I. 103 at 195-96) The activated hydrogel matrix employed in these experiments was attached to a silicon oxide surface via well-known silanization procedures. (D.I. 103 at 196-97) The resulting sensing element was suitable for use in a biosensor. D. The ’161 Patent Application 19. The PCT Application. On November 9, 1989, three researchers from Pharmacia, Jan Bergstrom, Stefan Lofás, and Bo Johnsson, filed Patent Cooperation Treaty application PCT/SE89/00642 (“the PCT”) entitled “Sensing Surfaces Capable of Selective Biomolecular Interactions, To Be Used in Biosensor Systems.” (Joint Exhibit (“JX”) 1) The application claimed priority from Swedish patent application 8804073 filed on. November 10, 1988. (JX 1 at BIA 001545) The PCT was published on May 17, 1990. (JX 1 at BIA 001545) 20. The PCT is directed to [mjethods for the production, on metal surfaces, of surface layers which are capable of selective biomolecular interactions; sensing surfaces produced by means of these methods; and the use thereof in biosensors, especially in surface plasmon resonance systems. (JX 1 at BIA 001545) The invention also discloses activated surfaces for coupling a desired ligand; surfaces containing bound ligand; and the use of such surfaces in biosensors. (JX 1 at BIA 001547) The PCT teaches a barrier monolayer of an “organic molecule X-R-Y” between the metallic surface of an SPR system and the desired ligands in order to bind the ligands and protect the metal surface. (JX 1) In addition, an optional embodiment discloses a matrix comprised of a hydrogel coupled to the X-R-Y monolayer by which ligands suitable for the target analytes can be immobilized. (JX 1 at BIA 001553-54) Although acknowledging that there exist methods for attaching a hydrogel directly to a surface, the specification of the PCT contends that these methods “ha[ve] a number of obvious drawbacks;” it also recognizes that these “problems” can be “solved at least in part” by known procedures. (JX 1 at BIA 001549-50) 21. The PCT contains 14 claims. Claim 1 of the PCT is a generic claim drawn to a “sensing surface” for use in a biosensor. (JX 1 at BIA 001569) Claim 1 discloses a [s]ensing surface to be used in bio-sensors, characterized by consisting of a film of a free electron metal selected from the group consisting of copper, silver, aluminum and gold and having one of its faces coated with a densely packed monolayer of an organic molecule X-RY ... (JX 1 at BIA 001569) Claim 1 is the only independent claim; all the other claims, which are drawn to specific variations of the sensing surface described in claim 1, contain the limitations found in claim 1. 22. Claim 2 of the PCT discloses an optional embodiment: [A sjensing surface according to claim 1, characterized by containing a biocompa-tible porous matrix which is bound to the monolayer X-R-Y and via which a desired ligand can be bound. (JX 1 at BIA 001569) Claims 3-13, which are drawn to specific variations of the sensing surface described in claim 2, contain the limitations set forth in claim 2. 23. The U.S. Patent Applications. The ’828 patent. The same inventors who filed the PCT application filed the United States counterpart application, Serial No. 681,531, and a Preliminary Amendment with the PTO on May 10,1991. (D.I. 96 at 2; PX 5) As with the PCT, the inventors claimed a priority filing date of November 10, 1988 based upon the Swedish patent application. 24. The claims of the U.S. counterpart patent initially were rejected, inter alia, as obvious over the prior art. In their response to the rejection, the applicants stated that [t]he basic concept of the present invention resides in providing a biosensor sensing surface, in the form of a free electron metal film, with a barrier layer comprising monomeric organic molecules which, through self-association, form a well-defined, dense and stable monolayer. (PX 5) The applicants distinguished the prior art based on the presence in the invention of the barrier (i.e., the densely packed monolayer of organic monomeric molecules) alone or in combination with “a porous matrix”: No such barrier layer, nor its combination with a porous matrix such as a hydrogel, is disclosed or suggested by the cited references, either individually or in combination. (PX 5) According to the applicants, the prior art references relied upon by the patent examiner disclose polymeric coatings. The applicants argued that the polymers of these coatings are either not as efficiently densely-packed as is the mono-layer disclosed in the invention, thus providing less protection against corrosion and nonspecific binding, or are not bound to the surface in a manner so as to provide stability, uniformity, and durability. (PX 5) The applicants requested withdrawal of the rejection, concluding that the cited prior art does not disclose or suggest a sensing surface comprising a metal film coated with a densely packed monolayer of organic molecules X-R-Y as defined in claim 1, nor does it disclose or suggest such a barrier layer supporting a three-dimensional porous matrix, preferably a hydrogel, onto which ligands and analytes may be bound. (PX 5) 25.On September 7, 1993, the U.S. counterpart application was issued as U.S. Patent 5,242,828 (the “ ’828 patent”). (PX 4) The specification of the ’828 patent is essentially the same as that of the PCT. The claimed invention relates to the field of biosensors and is more specifically concerned with methods for providing metal surfaces with surface layers capable of selective biomoleeular interactions. The invention also comprises activated surfaces for coupling a desired ligand; surfaces containing bound ligand and the use of such surfaces in bio-sensors. (PX 4, col. 1, Ins. 8-14; see also col. 8, Ins. 16-21: “The invention relating to (i) the aforesaid methods for providing metal surfaces with surface layers capable of selective biomoleeular interactions, to be used in biosensor systems ....”) The examples set forth in the specification of the ’828 patent are drawn to SPR technology and demonstrate a hydrogel attached to a metal surface via an X-R-Y monolayer. (PX 4; D.I. 104 at 316) The specification does not describe a hydrogel attached to a nonmetal surface other than by reference to further patent applications. (D.I. 104 at 232, 316) 26. Whereas the PCT contains 14 claims, the ’828 patent contains 27 claims. (PX 4) Besides claiming the sensing surface described and claimed in the PCT, the ’828 patent also claims a “sensing element suitable for use in a biosensor” comprising a substrate, “a film of free electron metal ... having a first and second major surface, said first major surface being in contact with the substrate,” and “a densely packed monolayer of an organic molecule X-R-Y coated on said second major surface of said film.” (PX 4, col. 14, Ins. 38-63) 27. The sensing surface claimed and described in the ’828 patent essentially is the same as that claimed in the PCT. Claim 1 of the ’828 patent discloses 1. A sensing surface suitable for use in a biosensor, comprising: a film, having two faces, of a free electron metal selected from the group consisting of copper, silver, aluminum and gold; and a densely packed monolayer of an organic molecule X-R-Y coated on one of the faces of said film where X is a group selected from the group consisting of (PX 4, col. 13, Ins. 6-27) Whereas claims 2-16 of the ’828 patent depend in part from claim 1, claims 17-21 depend solely from claim l. 28. As in the PCT, claim 2 of the ’828 patent describes an optional embodiment: 2. The sensing surface according to claim 1, which contains a biocompatible porous ■ matrix which is bound to the densely packed monolayer X-R-Y and via which a desired ligand can be bound. (PX 4, col. 13, Ins. 28-31) Claims 3-16 are drawn to variations of the sensing surface disclosed in claim 2 and contain all the limitations found in claim 2. 29. The ’161 patent. On May 10,1993, the inventors of the ’828 patent filed a continuation application, Serial No. 058,265 (the “ ’265 application”) and a Preliminary Amendment. (PX 3) Claim 1, as set forth in the Preliminary Amendment, disclosed [a] sensing surface suitable for use in a biosensor, comprising a hydrogel which is bound to a surface and via which a desired ligand can be bound, which hy-drogel is activated to contain (i) charged groups for bringing about a concentration of biomolecules carrying an opposite charge to that of said charged groups, and (ii) reactive groups for covalently binding said biomolecules concentrated to said sensing surface. (PX 3 at 68) In contrast to the claims of the ’828 patent, the claims of the ’265 application did not recite a densely packed monolayer of organic monomeric molecules forming a barrier on a metal surface. 30. On November 8, 1993, during the prosecution of the ’265 application, the claims were rejected by the patent examiner for obviousness-type double patenting over the ’828 patent. (PX 3, Tab 6) In response to this rejection, on April 20, 1994, the inventors filed a terminal disclaimer, disclaiming the ’265 application beyond the expiration date of the ’828 patent. (PX 3, Tab 8) 31. The claims were also rejected as obvious in light of the prior art, specifically European Patent Application Publication No. 0 226 470 (the “ ’470 patent”) in combination with other prior art references. (PX 3, Tab 6) In this regard, the patent examiner opined that the prior art already disclosed the use of an activated hydrogel on a surface and that such “an ‘activated’ hydrogel would inherently provide for the claimed features of charged and reactive groups.” (PX 3, Tab 6 at BIA 000116) In response to this rejection, the applicants distinguished the prior art, individually and in combination, asserting that none of the cited references (1) discloses or suggests the concept of combining charged and reactive groups; (2) contains any example where the hydrogel has been provided with both charged and reactive groups; or (3) discloses or suggests that an activated hydrogel would have a concentrating effect on biomolecules. (PX 3, Tab 7 at BIA 000153) 32. Although the claims of the ’265 application ultimately were allowed, the application subsequently was abandoned, and on July 22, 1994, the inventors filed another continuation application, Serial No. 279,-089. (PX 3, Tabs 11, 12, and 13) The claims of this application were initially rejected by the patent examiner, inter alia, “as being indefinite for failing to particularly point and distinctly claim the subject matter which applicant regards as the invention.” (PX 3, Tab 15 at BIA 000333) The examiner noted, however, that the subject matter of the application was allowable: The prior art discloses the use of charged species to concentrate biomolec-ules to an area and the use of charged species to improve binding capabilities, however fails to disclose the use of a hydrogel, bound to a substrate, that has charged groups for concentrating biomo-lecules and uncharged groups (“reactive groups”) for binding an analyte. (PX 3, Tab 15 at BIA 000334) Following revision of the application in accordance with the examiner’s comments, this application issued as the ’161 patent on July 25, 1995. (PX 3, Tab 17) E. The ’161 Patent 33. The abstract and specification. The abstract describes the invention claimed in the ’161 patent as follows: A matrix coating suitable for use in a biosensor is provided. This matrix coating comprises a hydrogel bound to a surface and via which a desired ligand can be bound. This hydrogel is activated to contain charged groups for bringing about the concentration of biomolec-ules carrying an opposite charge to that of said charged groups, and reactive groups for covalently binding the biomo-lecules concentrated to the matrix coating. (PX 1, Abstract) 34. The specification of the ’161 patent, which is essentially the same as that of the ’828 patent, includes the following field of the invention: The present invention relates to the field of biosensors and is more specifically concerned with methods for providing metal surfaces with surface layers capable of selective biomolecular interactions. The present invention also comprises activated surfaces for coupling a desired ligand; surfaces containing bound ligand; and the use of such surfaces in biosensors. (PX 1, col. 1, Ins. 15-21) The invention claimed is described as [a] generally useful sensing surface for biosensor systems, especially SPR, ... fulfilling] the following desiderata: ... [C]hemically resistant to the media employed. ... [Compatible with proteins and other biomolecules and ... not interacting] with any molecules other than those desired. ... [C]apable of providing for covalent binding of such a large number of ligands as is required for a general applicability of this technique to a variety of analytical problems. ... [P]rovid[ing] a tridimensional matrix for the sample solution for binding the target molecules therein. In this manner a greater part of the volume influencing the resonance effect, by way of its refractive index, will be utilized as compared to cases where a two-dimensional surface would be used. (PX 1, col. 3, Ins. 19-39) 35.The specification indicates that the scope of applicability of the claimed invention extends beyond the field of biosensor technology. Specifically, it is noted that • [further scope of the applicability of the present invention will become apparent from the detailed description and drawings provided [herein], (col.3, lns.43-45) • [t]his type of surface modification can be utilized also in other fields of technology where a specific, or alternatively, a low non-specific, interaction is required between a surface on one hand and proteins or other biomolec-ules on the other hand. Examples that may be mentioned are parts of chromatographic systems for biomo-lecule separations .... It would also be possible to construct capillary-type chromatographic columns in conformity with these principles. Furthermore, it is evident that a surface structure' may be modified so as to acquire biocompatibility, for use in environments of the “in vivo” type. Depending on the particular field of use contemplated, the actual choice of, for example, the hydrogel, can be made such that undesired interactions are minimized. To those skilled in the art, a number of additional fields of use will be readily obvious, along the lines of the aforesaid examples, (col.6, lns.20-38) • [i]t will be readily evident that ion exchanging groups, metal chelating groups and various types of receptors for biological molecules — such as are known from conventional liquid chromatographic procedures — may be employed for the construction of systems which are suitable for selection purposes even in complex measuring systems. (eol.7, lns.40 — 48) 36. The claims. The ’161 patent contains 15 claims. Claim 1 is a generic claim, drawn to a “matrix coating suitable for use in a biosensor.” Claims 2-4 are drawn specifically to particular modifications of the matrix described in claim 1. Claim 15 is drawn to a “sensing element suitable for use in a biosensor.” Claims 1 and 15 are the only independent claims; claims 2-14 depend, at least in part, on claim 1. Biacore alleges that Thermo has infringed claims 4 and 5 of the ’161 patent. Thermo seeks a declaratory judgment of invalidity with respect to claims 1-5, 9-11, and 15. 37. Claim 1 reads: 1. A matrix coating suitable for use in a biosensor, comprising a hydrogel which is bound to a surface and via which a desired ligand can be bound, which hydrogel is activated to contain (i) charged groups for bringing about a concentration of biomolecules carrying an opposite charge to that of said charged groups, and (ii) reactive groups for covalently binding said biomolecules concentrated to said matrix coating. (PX 1, col. 12, In. 63 — col. 13, In. 2) 38. Claim 1 is directed to a matrix coating “suitable for use in a biosensor.” As defined in the patent, a biosensor is a unique combination of a receptor for molecular recognition, for example a selective layer with immobilized antibodies, and a transducer for transmitting the interaction information to processable signals. (PX 1, col. 1, Ins. 23-27) This broad definition comports with the definitions found in the literature relevant to biosensor technology. 39. The disclosed matrix coating is further described as comprising a “hydrogel which is bound to the surface.” The patent defines hydrogel by reference to Merrill et al., Hydrogels for Blood Contact (1986). (PX 1, col. 5, Ins. 49-52) According to Merrill, a hydrogel presents a surface layer of bound molecules which by reason of their chemical nature hold a large fraction of water, in which the molecules are predominantly in an amorphous, water-solvated state, and in which the thickness of the layer is of the order of 30 Á minimum up to any indefinitely higher limit. (JX 3 at 2; PX 1, col. 5, Ins. 49-52; D.I. 103 at 199; D.I. 106 760-61) Dr. William H. Scouten, Thermo’s expert witness, opined that this definition “does not differ substantially from what a person of ordinary skill in the art would understand the plain meaning of the word ‘hydrogel’ ” to be. (D.I. 107 at 858-59) The means by which the hydrogel is bound to the surface is not limited in the patent to any specific binding chemistry; thus, any form of contact, covalent, physical, or adhesive, is sufficient. (D.I. 104 at 231-32, 294; D.I. 107 at 859-60) Nor is the type of surface to which the hydrogel is bound limited despite the fact that the examples set forth in the specification refer only to metallic surfaces. (D.I. 103 at 205-07; D.I. 104 at 246-47, 251-52, 287-92; D.I. 106 at 766; D.I. 107 at 858) 40. The hydrogel disclosed in claim 1 must be able to bind the desired ligands. (see also PX 1, col. 1, Ins. 48^9: “[A] sensing surface composed of ... ‘ligands.’”) According to the patent, ligands, or receptors, are “molecules or molecular structures which interact selectively with one or more biomolecules.” (PX 1, col. 1, Ins. 48-51) Within the context of claim 1, ligand is used in the same manner as that term is employed in the field of affinity chromatography. (D.I. 107 at 863) The patent does not specifically limit the means by which the ligands are bound to the hydrogel. (D.I. 104 at 253-54) 41. Said hydrogel is “activated to contain” two types of chemical groups. These groups are defined by their function. (D.I. 104 at 235) Specifically, the groups are: (1) charged groups for concentrating oppositely-charged biomolecules and (2) reactive groups for covalently binding the concentrated biomolecules. (PX 1, col. 12, In. 66 — col. 13, In. 2) The patent does not specify the degree or amount of charge required with respect to the charged groups or the degree of ligand concentration required by the reactive groups. (D.I. 106 at 780, 791) Nor does the patent state the relative ratio of the two chemical groups or the chemical nature of the groups, except as that is limited by the function to be performed. (D.I. 104 at 234-35) In fact, the patent allows for the two groups to be one and the same, i.e., the same chemical group on the hydrogel could serve both functions. (D.I. 104 at 234; D.I. 106 at 777) No mention is made in the patent as to the process whereby the charged and reactive groups are put onto the hydrogel. (D.I. 104 at 244-45) 42.Dependent claims 2-5 and 9-11 provide as follows: 2. The matrix coating according to claim 1, wherein said hydrogel is a poly-saccharide or a swellable organic polymer. 3. The matrix coating according to claim 2, wherein said hydrogel is a poly-saccharide selected from the group consisting of agarose, dextran, carrageenan, alginic acid, starch, and cellulose, and a derivative of any of the foregoing. 4. The matrix coating according to claim 3, wherein said hydrogel consists of dextran. 5. The matrix coating according to claim 4, wherein said charged groups and said reactive groups of said dextran are carboxyl groups, part of which are in the form of reactive esters, hydrazides, thiols, or reactive disulfide-containing derivatives. :¡< ‡ ‡ $ 9. The matrix coating according to claim 2, wherein said charged groups and said reactive groups of said hydro-gel are carboxyl groups, part of which are in the form of reactive esters, hydra-zides, thiols, or reactive disulfide-con-taining derivatives. 10. The matrix coating according to claim 1, wherein said charged groups and said reactive groups of said hydro-gel are selected from the group consisting of hydroxyl groups, carboxyl groups, amino groups, aldehyde groups, carbonyl groups, epoxy groups, and vinyl groups for immobilizing a desired ligand, and, optionally, a biospecific ligand bound via said groups. 11.The matrix coating according to claim 1, wherein said charged groups are carboxyl groups. (PX 1, col. 13, Ins. 3-19; col. 13, In. 31-col. 14, In. 11) 43. Independent claim 15 is drawn to “[a] sensing element for use in a bio-sensor,” said sensing element comprising: a substrate; and a matrix coating comprising a hydro-gel supported on said substrate via which a desired ligand can be bound, which hydrogel has been activated to contain (i) charged groups for bringing about a concentration of biomolecules carrying an opposite charge to that of said charged groups, and (ii) reactive groups for covalently binding said biom-olecules concentrated on said matrix coating. (PX 1, col. 14, Ins. 23-32) The patent does not limit the term “sensing element” to any particular type or types of element capable of detecting an analyte. In addition, the patent does not specifically limit the type of surface to be employed or the means for supporting the hydrogel on the substrate. F. The Prior Art 44. The publications characterized by Thermo as prior art include: (1) the ’470 patent published June 24, 1987; (2) an article entitled Polysaccharide Derivatives as Coats for Nylon Tube Urease authored by Francis N. Onyezili and AMntunde C. Onitiri and published in Analytical Biochemistry, Vol. 117, in 1981 (the “Onyezili reference”); (3) an article authored by Carl Fredrik Mandenius et al. entitled Reversible and Specific Interaction of Deh-ydrogenases with a Coenzyme-Coated Surface Continuously Monitored with a Reflectometer that was published in Analytical Biochemistry, Vol. 157, in 1986 (the “Mandenius reference”); (4) a paper authored by Dr. Scouten et al. entitled Immobilizing Fluorescently-Labeled Albumin for Use in a Fiberoptic Bilirubin Monitor that was presented at the Chemically Modified Surfaces symposium in June 1987 (the “Scouten paper”); (5) a survey article authored by Dr. Scouten entitled A Survey of Enzyme Coupling Techniques that was published in Vol. 135 of Methods in Enzymology, Immobilized Enzymes and Cells Part B in 1987 (the “Scouten survey article”); (6) an article entitled Simple Hydrazidation Method for Carboxymethyl Groups on Cross-Linked Dextran authored by Hiroshi Akanuma and Makoto Yamasaki and published in the Journal of Biochemistry, Vol. 84, in 1984 (the “Akanuma reference”); (7) an article authored by Russell G. Frost et al. entitled Covalent Immobilization of Proteins to N-Hydroxysuccinimide Ester Derivatives of Agarose — Effect of Protein Charge on Immobilization that was published in Biochimica et Biophysica Acta, Vol. 670, in 1981 (the “Frost reference”); (8) an article authored by Suresh B. Shuk-la entitled Preparation of an Active Ester Agarose Derivative Having a Positively Charged Spacer Arm; Enhanced Coupling to Acidic Proteins that was published in Affinity Chromatography and Biological Recognition in 1983 (the “Shukla reference”); (9) an article entitled Covalent Immobilization of Enzymes on Ionogenic Carriers authored by V.P. Torchilin et al. and published in the Journal of Solid-Phase Biochemistry, Vol. 2, in 1977 (the “Torchilin reference”); (10) U.S. Patent No. 3,619,371 entitled “Production of a Polymeric Matrix Having a Biologically Active Substance Bound Thereto” issued on November 9, 1971 with a priority date of July 3, 1967 (the “Crook patent”); and (11) a 1986 brochure for Activated Affinity Supports Affi-Gel 10 and 15 by Bio-Rad Laboratories (the “Bio-Rad brochure”). All of these references are within the field of ligand immobilization. It is undisputed that these references were publicly available more than one year prior to the priority date at issue and, thus, constitute prior art. Thermo contends that four of these references, the ’470 patent, the Onyezili article, the Mandenius reference, and the Scouten paper, each standing alone, anticipate the asserted claims of the ’161 patent with the exception of claim 5. Alternatively, Thermo contends that the claimed invention is obvious in light of the identified prior art. 45. The teachings of the ’470 patent. The ’470 patent discloses a microchemical analytic apparatus comprising a solid substrate having a surface that carries a hy-drogel formed thereon and covalently bound thereto. (DX 541) The patent considers the invention’s use in an electrochemical biosensor as well as its suitability for use in other types of biosensors, such as thermistors and optical biosensors. (DX 541 at Col. 3, In. 55 — col. 4, In. 3; col. 7, Ins. 2-20; D.1.107 at 813) 46. Example 5 of the ’490 patent describes a method for preparing a matrix coating comprising an acrylate hydrogel bound to a glass slide. According to the patent, the carboxyl groups contained within the polymer matrix may be activated by treatment with, for example, an aqueous solution Woodward’s Reagent K .... The activated copolymer may then be reacted with functional groups such as antibody protein molecules, antigens, or haptens. (DX 541, col. 8, Ins. 43-49) According to Dr. Scouten, the Woodward’s Reagent K, which possesses an overall neutral charge with a positive amine and a negative sulfate group, reacts with the carboxyl groups on the polymer to form a “charged and activated reactive ester.” (D.I. 106 at 792) Thus, the activated hydrogel disclosed in example 5 of the ’470 patent contains charged groups, the Woodward’s Reagent K sulfate groups as well as any of the original carboxyl groups that did not react, and reactive groups, the reactive esters, that “happen to be the same thing.” (D.I. 106 at 793; D.I. 107 at 816-17) 47. The surface described in example 5 was never placed in a biosensor. (D.I. 107 at 934) Dr. Anthony P.F. Turner, Biacore’s expert witness, testified that, although one skilled in the art would know that under certain conditions the matrix disclosed in example 5 might contain charged groups that would attract oppositely-charged biomolecules, the resulting concentration, if any, would be insufficient to produce a useful signal. (D.I. 104 at 306-07) According to Dr. Turner, the concentration of ligands necessary to produce a useful signal would vary depending on a number of factors, including the use of the biosensor and the activity of the biological receptor being used. (D.I. 104 at 307) 48. With respect to claims 1 and 15, the ’470 patent does teach a hydrogel that is both bound to a surface and activated to contain charged and reactive groups. Relevant to claims 9-11, the charged and reactive groups disclosed are carboxyl groups, some of which are in the form of reactive esters. The ’470 patent, however, does not teach the use of charged groups for concentrating oppositely-charged biom-olecules. Nor does the ’470 patent teach that the ionic concentration should be such that electrostatic concentration can be achieved. (D.I. 104 at 256-57, 305; D.I. 107 at 928-29) 49. With respect to claim 2, the ’470 patent teaches the use of a hydrogel that is a swellable organic polymer. (D.I. 107 at 817) With respect to claims 3-4, the ’470 patent does not teach the use of a polysac-charide, or more specifically, the use of dextran. (D.I. 107 at 816-17) 50. The teachings of the Onyezili reference. The Onyezili reference addresses the immobilization of the enzyme urease inside nylon tubes, a procedure used in medical biochemistry. (DX 533) More particularly, the reference teaches the use of polysaccharide derivatives, specifically a dextran derivative, in order to provide a more hydrophilic coat inside the nylon tubes and to eliminate the nonspecific binding of urease to the tubes. (DX 533 at 121) In the experiment, amino “arms” or “coats” were incorporated into an alkylat-ed nylon tube by filling the tube with the polyamine derivative of dialdehyde dextran (“DPA”). (DX 533 at 121-23) The tube was activated by filling it with glutaraldeh-yde in a borate buffer. (DX 533 at 121-23) Subsequently, the tube was filled with a solution of urease, an enzyme that converts urea into ammonia. (DX 533 at 121-23) The activity of the immobilized urease was determined by measuring the enzyme-catalyzed hydrolysis of urea in EDTA buffer, i.e., by assaying the effluent for ammonia. (DX 533 at 121-23) 51. With respect to claims 1 and 15, the Onyezili reference teaches a dextran matrix covalently bound to a nylon surface. Said matrix is activated to contain reactive groups (the carbonyl and aldehyde groups of gluteraldehyde). (D.I. 107 at 824-26) According to Dr. Scouten, the matrix also contains charged groups, the amines of the amino DPA “arms” or “coats” that are incorporated into the tube. (D.I. 107 at 927) Dr. Scouten indicated that these groups, although they interact with the gluteraldehyde, retain their positive charge even in the presence of excess glu-teraldehyde. (D.I. 107 at 823-24) The reference itself, however, states that [m]ore significantly, O-alkylated nylon tubes modified with DPA bound virtually no urease without activation with glu-taraldehyde. This observation would suggest that, in these tubes, urease would not be immobilized by nonspecific bonds but would be bound by the covalent linkages between the carbonyl groups (from gluteraldehyde) on the tube and amino groups in the enzyme. (DX 533 at 124) Dr. Scouten felt that this statement applied “after a washing procedure necessary for use of the material,” although there is no indication of such in the reference. (D.I. 107 at 926) Therefore, according to Dr. Scouten, the statement does not suggest that there was no concentration by charge of the urease prior to activation by gluteraldehyde. (D.I. 107 at 926) He conceded, however, that nothing in the reference indicated that concentration by charge occurred. (D.I. 107 at 926) 52. Dr. Scouten also conceded, assuming arguendo the presence of charged groups, that the reference does not disclose explicitly the use of charged groups for bringing about a concentration of oppositely-charged biomolecules. (D.I. 107 at 923) He also admitted that, if what is required first is concentration of biomolec-ules by charged groups, then the reference also does not teach reactive groups that function to covalently bind biomolecules having been so concentrated. (D.I. 107 at 923-24) He testified, however, that the reference does report reactive groups for covalently binding biomolecules. (D.I. 107 at 924) 53. Dr. Scouten also admitted that the Onyezili reference does not disclose a bio-sensor as that term is defined in the ’161 patent. (D.I. 107 at 922-23) Accordingly, he conceded that the reference does not teach the use of a matrix in a biosensor. (D.I. 107 at 922-23) The reference does not describe the element used to monitor the binding event. (D.I. 107 at 922-23) Specifically, the article does not indicate whether the method for detecting ammonia in the effluent involved manual assay or the use of a transducer. (D.I. 107 at 922-23) According to Dr. Scouten, had the reference described employment of the latter, then it would have disclosed the use of a biosensor as defined in the T61 patent. (D.I. 107 at 922-23) It was Dr. Scouten’s opinion, however, that the surface disclosed in the Onyezili reference is suitable for use in a variety of types of biosensors. (D.I. 107 at 823) 54. With respect to claims 2-4, the reference teaches the use of a polysaccharide hydrogel, specifically dextran. With respect to claim 10, the reference discloses a matrix coating wherein the charged groups are amines and the reactive groups are carbonyl and aldehyde groups. 55. The teachings of the Mandenius reference. The Mandenius reference reports the findings of an affinity-based study in which the reversible affinity binding of NAD -dependent dehydrogenase to an NAD-eoated silicon surface was monitored using a reflectometer. (DX 580) As part of the experiment, after silanization, silicon chips were coated with a layer of T500 dextran in order to “bypass” possible nonspecific binding. (DX 530 at 283-85; D.I. 107 at 906-08) The hydroxyl groups of the dextran were activated using tresyl chloride after which NAD analogs were covalently fixed to the surface and the time course of affinity binding measured. (DX 530 at 283-85) 56. With respect to claims 1, 11, and 15, the reference teaches the use in a biosensor (reflectometer) of a dextran matrix coating bound to a silicon surface. (D.I. 107 at 827-30) The hydrogel is activated to contain reactive groups in the form of tresyl groups, tresyl being a kind of sulfonyl ester which acts like “a sticky molecular gluing agent.” (D.I. 107 at 827-30) With regard to the presence of charged groups, Dr. Scouten opined that dextran possesses an inherent negative charge due to the presence of carboxyl groups in the polysaccharide. (D.I. 107 at 827-30; see discussion infra at Part II.G) It is undisputed that activation of the dextran with tresyl chloride would not result in the incorporation within the hydrogel of charged carboxyl groups. 57. The reference does not disclose explicitly the use of a matrix carrying a charge. (D.I. 107 at 911) Accordingly, it does not teach the use of charged groups for concentrating oppositely-charged biom-olecules. (D.I. 107 at 918-19) In fact, dex-tran was selected in order to avoid nonspecific binding: To bypass possible nonspecific binding we decided first to coat the silicon chip used with dextran as this has previously been shown to allow fibrinogen to be desorbed conveniently from a silicon surface by buffer solutions which otherwise would not have been possible .... (DX 530 at 283) Nor does the Mandenius reference teach that the alleged inherent charge of the dextran matrix, or the incorporation of charged groups in a dextran matrix, will be beneficial in bringing about a concentration of oppositely-charged biomolecules. (D.I. 107 at 911-12, 918-19) Dr. Scouten averred, however, that one of ordinary skill in the art would know that inherent in dextran are charged carboxyl groups. (D.I. 107 at 911) He further opined that one of ordinary skill would have recognized that charged groups would be advantageous because they would facilitate concentration of the enzyme into the gel. (D.I. 107 at 910) Dr. Scouten pointed to the fact that in the reference the dextran-coated chips were activated by tresyl chloride dissolved in pyridine, a base. (D.I. 107 at 913-14) According to Dr. Scouten, under those conditions, the carboxyl groups inherently present in the dextran would have been negatively charged. (D.I. 107 at 913-14) 58.The structure disclosed in the Man-denius reference that Dr. Scouten contended met the conditions of claim 1 was not used in a biosensor. Instead, that structure was further treated with a relatively high concentration of NAD in a sodium phosphate buffer (0.1 M, pH 7.5, no longer basic conditions) in order to effect ligand immobilization before its use in a bio-sensor. (D.I. 107 at 914, 917-18) According to Dr. Scouten, “those conditions may or may not cause concentration of the NAD,” which still would have been positively charged at that pH. (D.I. 107 at 915, 917-18) Dr. Scouten opined, although he had “not looked up the binding of NAD,” that there are conditions [under which] this material that Mandenius describes would be very useful in making a bio-sensor and that the actual use of that would both have the charged groups concentrating the biomolecules and the reactive groups binding to the biomolec-ules that were concentrated. (D.I. 107 at 918) 59. With respect to claims 2-4, the hy-drogel taught in the reference is dextran, a polysaccharide. (D.I. 107 at 830) 60. The teachings of the Scouten paper. The Scouten paper teaches methods for immobilizing fluorescent-labeled bovine serum albumin (“BSA”) on cellulose membranes. (DX 524) These membranes are incorporated into a fiber optic probe used to monitor bilirubin concentrations directly in the bloodstream. (DX 524) One method taught in the paper involves treating dialysis membranes with polyethylenimine and then reacting those membranes with a gluteraldehyde solution. (DX 524 at 120-21) Fluorescent-labeled BSA then is added to the membranes and allowed to react. (DX 524 at 120-21) 61. With respect to claims 1 and 15, the Scouten paper teaches the use in a fiber optic biosensor of a polyethylenimine hy-drogel bound to a dialysis membrane. The hydrogel is activated to contain reactive groups in the form of aldehyde and vinyl groups of glutaraldehyde. (D.I. 107 at 832-34) According to Dr. Scouten, the hy-drogel also has incorporated into it positively-charged amine groups — these groups being part of the polyethyleni-mine’s backbone. (D.I. 107 at 832-34, 939-10) 62. Dr. Scouten conceded that this paper does not disclose charged groups that are functioning to bring about a concentration of biomolecules carrying an opposite charge. (D.I. 107 at 939-40) He also admitted that the article does not teach that one should employ conditions that would allow the charged groups to electrostatically attract biomolecules into the matrix. (D.I. 107 at 939-40) In fact, the conditions under which Dr. Scouten employed the structure are not set forth in the reference except to state the use of a phosphate buffer, the pH of which is unspecified. (D.I. 107 at 941-43) Dr. Scouten never performed any tests to determine whether or not electrostatic concentration occurred under the experimental conditions he employed in developing the disclosed procedure. (D.I. 107 at 943) 63. With respect to claims 2 — 4, polye-thylenimine is a swellable organic polymer, but it is not dextran. With respect to claim 10, the disclosed matrix coating is activated to contain charged groups that are amines and reactive groups that are aldehyde and vinyl groups. 64. Dr. Scouten opined generally that it would have been apparent to one of skill in the art possessing knowledge of organic chemistry that incorporated in the matrix coatings disclosed in the aforementioned references are charged groups that would act, under the proper conditions, to attract and concentrate ligands. (D.I. 107 at 835-36) Moreover, Dr. Scouten opined that one of ordinary skill in the art would have known from, for example, ion exchange chromatography literature, of the conditions, i.e., the pH, necessary to take advantage of the charged groups to concentrate the desired biomolecules prior to covalent binding. (D.I. 106 at 781; D.I. 107 at 964) 65. The teachings of the Scouten survey article. The 1987 Scouten survey article lists a number of methods for covalently coupling enzymes to a variety of matrices. (DX 518 at 38-41) Specifically, the article mentions the use of carbodiim-ide as a coupling agent with, inter alia, agarose and cellulose matrices. (DX 518 at 38-41) In addition, it discloses a number of activation reagents that are used for hydrogels, including hydrazine and N-hy-droxysuccinimide (“NHS”), both of which can be employed to provide negatively charged groups on a carboxymethyl (“CM”)-dextran hydrogel matrix. (DX 518 at 54-55) 66. The teachings of the Akanuma reference. The Akanuma reference discloses in the context of affinity chromatography a method for the conversion of CM-Sephadex (cross-linked dextran) into its hydrazide derivative. (DX 519) Specifically, the reference discloses a procedure whereby CM-Sephadex is treated with a carbodiimide, resulting in the formation of ester linkages, i.e., the formation of lactone rings on the dextran derivative. (DX 519 at 1358-60) The resultant beads are treated with hydrazine to form hydrazinocarbo-nylmethyl-Sephadex, a hydrazide derivative of CM-Sephadex. (DX 519 at 1358, 1360-61) Analysis of the product so formed revealed that more than 90% of the car-boxyl groups were converted to hydrazide groups. (DX 519 at 1360) The reference “propose[s]” the use of this procedure “as a general and effective method for the conversion of carboxy-methylpolysacchar-ides into their hydrazide derivatives.” (DX 519 at 1360-61) 67. With respect to claim 5, the Akanu-ma reference teaches the use of activation reagents to produce an activated dextran hydrogel matrix having carboxyl groups at least 90% of which are in the form of reactive hydrazides. 68. The “Charged Concentration References.” The remaining prior art references, the “Charged Concentration References,” are indicative of the knowledge as of November 1988 of the concept of charged attraction, i.e., Coulomb’s Law. (D.I. 107 at 836) In general, these references teach the combined use of charged and reactive groups in order to enhance ligand immobilization. 69. The teachings of the Frost reference. The Frost reference address