In Re Joan Stevens, Peter P. Van Daele and Roeland G. Baets, 26 F.3d 138 (Fed. Cir. 1994)

Applicants (Stevens) appeal the rejection of claims 1 through 14 of United States Patent Application No. 07/166,026 ('026). The Board of Patent Appeals and Interferences of the United States Patent and Trademark Office rejected the claims under 35 U.S.C. § 103 (1988). Because the claims are obvious in view of the prior art, this court affirms the Board's decision.

BACKGROUND

The microelectronics industry constantly seeks more advanced technology to place more circuits on a semiconductor chip. Within this semiconductor chip technology, the claimed invention addresses the problem of combining circuits that have different thicknesses. The invention describes circuits which incorporate relatively thick light-emitting diodes (LEDs) into the same gallium arsenide (GaAs) substrate as the substantially thinner field-effect transistors (FETs). These FETs control the LEDs.

Claim 1 states:

--a semi-insulating GaAs substrate layer,

--at least one channel etched into said substrate layer,

--a contact layer of a first conductivity type grown in and restricted to said channel(s),

--double heterostructure light-emitting diodes (LED) comprising a first layer of a ternary compound Alx Ga1-x As of a first conductivity type, a second light-emitting GaAs layer or a doped Aly Ga1-y As layer of said first conductivity type, a third layer of a ternary compound Alx Ga1-x As of a second conductivity type and a fourth doped GaAs layer of said second conductivity type grown on said third LED layer forming ohmic contact, each of said Leds being restricted to a channel and the individual LEDs being mutually insulated,

--a GaAs layer in which mutually insulated active FET area are formed,

--insulation means mutually insulating LEDs and FETs,

--a LED short circuit preventing layer at the edge of said channel(s),

--n-type and p-type metal depositions forming ohmic FET-source and FET-drain contacts, LED-electrodes and LED and FET interconnections,

--Schottky-type metal FET-gate electrodes,

--a LED and FET series resistance reducing metallization layer,

--LED windows etched in said fourth n-type GaAs LED layer [s].

Claims 1 through 13 define the integrated circuit device. Claim 14 describes a recording head incorporating that device. The examiner rejected claims 1 through 13 under 35 U.S.C. §§ 102(e) and 103 as being anticipated by, or obvious over, Wada et al., U.S. Patent No. 4,719,498 (Wada). The examiner rejected claim 14 as obvious over Wada in view of Iwata et al., Japanese Kokai No. 61-199681 (Iwata).

Wada teaches an integrated circuit device. Wada differs from the claimed invention in four areas: (1) the use of a laser diode (LD)¹  instead of an LED, (2) the use of only one light emitting element, (3) the lack of resistance-reducing metallization layer, and (4) the lack of a window etched into the LED.

According to undisputed facts on this record, one of ordinary skill in the semiconductor chip art would know that, under low current densities, an LD operates as an LED. Therefore, the examiner determined that the knowledge of a skilled artisan would eliminate the first difference between Wada and the claimed invention. Additionally, one of skill in the art would fabricate more than one light emitting element at a time. Thus, the examiner determined that the second difference between Wada and the claimed invention was inherently eliminated during high volume production. The examiner did not address the third and fourth differences.

The Board reversed the examiner. The Board detected no record evidence that an artisan of ordinary skill would normally operate the Wada invention at the lower current densities necessary to make the diode function as an LED. The Board also rejected the examiner's inherency argument for the second difference. The Board reasoned that, after production, multiple light emitting elements would ordinarily be separated for use.

The Board, however, rejected the claims on other grounds. The Board rejected the first 13 claims as obvious in view of Wada in conjunction with Pankove,²  Itoh,³  Sze,⁴  and Levine.⁵  The fourteenth claim fell when the Board added Iwata to this same combination.

OPINION

Obviousness is a question of law, but this determination occurs in the context of a factual inquiry regarding the scope and content of the prior art. Graham v. John Deere Co., 383 U.S. 1, 17 (1966). This factual inquiry examines what a reference would have taught or suggested to one of ordinary skill in the art at the time of the invention. See Northern Telecom Inc. v. Datapoint Corp., 908 F.2d 931, 934, 15 USPQ2d 1321, 1323 (Fed. Cir.), cert. denied, 498 U.S. 920 (1990). Thus, although reviewing an obviousness determination by the Board de novo, this court reviews the underlying factual determination of the meaning of references to one of ordinary skill for clear error. In re Beattie, 974 F.2d 1309, 1311, 24 USPQ2d 1040, 1041 (Fed. Cir. 1992).

In determining the obviousness of a claimed invention that combines known elements, the Board must determine "whether there is something in the prior art as a whole to suggest the desirability, and thus the obviousness, of making the combination." Lindemann Maschinenfabrik GMBH v. American Hoist & Derrick Co., 730 F.2d 1452, 1462, 221 USPQ 481, 488 (Fed. Cir. 1984). The required suggestion need not be explicit, but may arise from "knowledge generally available to one of ordinary skill in the relevant art, which would have led one skilled in the art to combine the relevant teachings." Ashland Oil, Inc. v. Delta Resins & Refractories, Inc., 776 F.2d 281, 297 n. 24, 227 USPQ 657, 667 n. 24 (Fed. Cir. 1985), cert. denied, 475 U.S. 1017 (1986).

According to the Board, Wada teaches a double heterojunction diode that satisfies the structural requirements of claim 1 but operates the diode as a laser. Furthermore, the Board noted that Sze, a textbook, teaches various advantages and disadvantages to using LEDs as optical sources rather than LDs. For instance, Sze teaches that LEDs work better than LDs at higher temperatures and in environments subject to temperature fluctuations. According to Sze, as well as Pankove, double-heterojunction diodes can function as LEDs.

Although Sze teaches both advantages and disadvantages in the use of LEDs, this court detects no clear error in the Board's determination that an artisan desiring higher temperature operation would make the substitution. Indeed, the reference explicitly states the proposition that LEDs offer superior performance under these conditions. Moreover, the reference also indicates that the output signal is more uniform during temperature fluctuations. Similarly, this court detects no clear error in the Board's finding that Pankove teaches the ability to use the same circuit design as an LD or an LED by varying the current density.

The second difference between the claimed invention and the prior art Wada device is simply the use of multiple light emitting elements. The Board determined that one of ordinary skill would know to integrate and insulate a pair of double heterojunction diodes, as taught by Itoh. The Board further reasoned that more than one LED on the integrated device satisfied the multiplicity requirement of the claim. Moreover, the Board found that the need for a second optical coupler in an integrated circuit would have motivated one of ordinary skill in the art to provide two such diodes on the integrated structure of Wada. This court detects no clear error in these findings. In light of Wada's express suggestion to create high integration densities using the Wada circuit, one of ordinary skill in the art would have been motivated to provide multiple diodes.

The third difference between the claimed invention and the prior art is the use of a resistance reducing metallization layer in the connection of the LED and the FET. The Board found that depositing thicker layers of a metal was a common method of reducing resistance as taught by Levine. The Board's determination that Levine teaches that the resistance of a thin metal film is a function of its thickness is not clearly erroneous. Moreover, the Board's analysis that the skilled artisan would deposit additional layers of metal until the resistance was reduced enough to permit the devices to function is not erroneous.

The fourth difference is the window etched into the LED. The Board found that Sze taught it was known in the art to etch a window in the outer surface in order to use surface emissions from the LED. Sze teaches the use of LEDs as surface emitters as well as edge emitters. The Board found that an artisan preferring surface emission to edge emission would know to etch a window in the outer layer. Indeed, Sze teaches that such windows facilitate transmission of a signal through an optical fiber. This court detects no clear error in the Board's findings on this point.

Finally, the Board rejected claim 14 on the references applied to claims 1 through 13 taken in view of Iwata. Iwata discloses an optical print head comprising an array of LEDs and associated drivers. The Board found that the capability of high integration expressly taught by Wada as a benefit would have induced one of ordinary skill in the art to fabricate the Iwata print head using a modified Wada device. Wada teaches operation of an optoelectric device at high integration densities. These high densities are necessary for the Iwata print head and provide ample motivation to incorporate the claimed invention into an Iwata print head.

CONCLUSION

This court finds that Wada, in conjunction with secondary references, renders the claimed invention obvious. Each of the references relied upon by the Board fall within the inventor's field of endeavor. Indeed, many of the references are general textbooks on the subject. These teachings clearly evince "knowledge generally available to one of ordinary skill in the relevant art." Ashland Oil, 776 F.2d at 297 n. 24. The Board set forth the objective bases for its conclusion that each of the references would have been combined. Accordingly, the Board's rejection of those claims pursuant to 35 U.S.C. § 103 is affirmed.

MICHEL, Circuit Judge, dissenting.

I respectfully dissent. For two reasons, I believe the Board erred in rejecting claims 1-13 for obviousness over Wada in view of Pankove, Itoh, Sze and Levine, and claim 14 on the same ground over the references applied against claims 1-13 in view of Iwata. First, even assuming the references are combinable, they do not teach all features of the claimed device. Second, the evidence of motivation or suggestion to combine is insufficient, and indeed was mostly assumed by the Board and speculated by the panel.

BACKGROUND

The Board reversed the examiner's rejection of claims 1-13 under 35 U.S.C. § 102 over Wada because the Board found that Wada does not teach the use of a light-emitting diode as does claim 1 in the application patent. Instead, it teaches the use of a laser diode, but does not teach that a laser diode may function as a light-emitting diode during normal operation. App. at 4-5.

The Board also acknowledged that Wada does not teach the use of multiple, mutually insulated diodes, as does the application patent. App. at 5.

Finally, the Board found that Wada expressly disclosed neither the additional resistance-reducing metallization layer nor the window etched in the fourth layer of the optical semiconductor device. App. at 5. Accordingly, the Board reversed the examiner's rejection based on Wada alone.

The Board, however, entered a new rejection of claims 1-13 on the new ground that the Wada patent in combination with the publications of Pankove, Itoh, Sze and Levine made appellants' claimed device obvious under 35 U.S.C. § 103.

According to the Board, Sze shows that a double heterojunction diode structure, which is disclosed in Wada, can be operated as a light-emitting diode (LED) as well as a laser diode (LD). App. at 108, 6, 12. The Pankove reference adds that whether an optical semiconductor device capable of laser operations operates instead as a light-emitting diode is principally a function of current density. App. at 101. Sze lists numerous reasons why one skilled in the art might prefer a light-emitting diode, e.g., simpler construction, simpler drive circuitry and less temperature sensitivity. App. at 108, 7. Thus, based on the teachings of Pankove and Sze, the Board found that a person skilled in the art of semiconductor devices would have been motivated to so reduce the current so as to operate the laser diode of Wada as a light-emitting diode. App. at 6-7.

Further, the Board found that the Itoh reference shows that a pair of double heterojunction optical semiconductor devices may be integrated and insulated on the same substrate. App. at 102, 7.

Although the Wada patent does not provide a motive for integrating more than one optical/electronic device on a substrate,¹  the Board found that the Sze text suggests opto-isolator and optical-fiber communications applications that would motivate one skilled in the art to integrate multiple optical/electronic devices on a single substrate. App. at 108-09, 13. Sze also shows that it was known in the art to etch a window in the top layer of a surface emitting diode. App. at 109, 7.

Finally, although the Board found that Wada does not teach a resistance-reducing metallization layer on the metal conductor between the optical diode and the driver electronics, Levine teaches that conductivity varies as a function of the thickness of the metal film. App. at 121, 8. Therefore, the Board reasoned that a person of ordinary skill in the art would know that resistance could be reduced by adding another layer of metal conductor on the connecting structure already in place in both the Wada device and the Stevens invention and would be motivated to do so. App. at 8.

With respect to claim 14, the Board rejected it under 35 U.S.C. § 103 as obvious in view of the references applied against claim 1 in view of Iwata. App. at 9. Iwata discloses an optical print head comprising an array of LEDs and associated drivers. In Iwata, the print head is made from a linear array of thirty-two units, where each unit has a linear array of sixty-four LEDs. App. at 87, 90. According to the Board, it would have been obvious to a person skilled in the art to use a device like that disclosed in the Wada patent, but modified to operate as a light-emitting diode, as part of a linear array on a recording head, the subject matter of the Stevens application. App. at 9.

DISCUSSION

Reversal of the Board's rejection of claims 1-13 is appropriate for two main reasons.

First, even if the teachings are combined, the references cited by the Board--Wada, Pankove, Itoh, Sze and Levine--do not fully teach the claimed invention.

The Board relies on the Sze reference to support its assertion that one skilled in the art would have known to operate the LD disclosed in Wada as an LED. The Sze reference, however, states that there are advantages and disadvantages to the use of LEDs as optical sources compared to the use of LDs. App. at 108. The advantages include higher temperature operation, smaller temperature dependence, simpler device construction and simpler drive circuit, while the disadvantages include lower brightness, lower modulation frequency and wider spectral line width. Based on this reference, there is no reason to conclude that one of ordinary skill in the art, confronting the problems faced by Stevens in developing a recording head, would ignore the disadvantages articulated by Sze and operate the LD of Wada as an LED. Although the majority correctly states that an artisan desiring higher temperature operation would be motivated to substitute an LED for an LD, there is no suggestion in the record that appellants in this case were particularly concerned about operating their device in high temperature environments and not concerned (or less concerned) about lowered brightness, lowered modulation frequency or wider spectral line width. Therefore, the majority can only be engaging in speculation when it suggests otherwise.

The Board also refers to the Itoh reference to demonstrate that it was known how to integrate and insulate a pair of heterojunction diodes. Relying upon Itoh, the Board concludes that integrating many light-emitting diodes with driver electronics on a single substrate would be within the ordinary skill in the art. Just because Itoh teaches how to integrate and insulate two diodes on a single miniature substrate, however, does not mean that one skilled in the art would know how to integrate multiple light-emitting diodes on a single such substrate along with their accompanying driver circuitry since "multiple" clearly means both two and more than two. Such an increase on so tiny a substrate is not routine. In fact, the specification in appellants' patent application describes the difficulties of integrating a large number of LED's on a single substrate. The specification also notes that, prior to the claimed invention, the state of the art in optoelectronic integrated circuits was limited "mainly to the integration of individual light sources with driver electronics," rather than "an array of light sources." App. at 23 (emphasis added).

Further, the Board relies on the Pankove reference to support its finding that current density determines whether a semiconductor device functions as an LD or LED (and, since Sze teaches certain advantages of using LEDs over LDs, one skilled in the art would arguably be motivated to and know how to operate the LD disclosed in Wada as an LED by varying the current density). The Board, however, acknowledged in its decision for reconsideration that Pankove does not say that only current density determines whether an optical semiconductor device functions as an LD or LED. App. at 12. Assuming, arguendo, that this is the only difference between an LD and an LED, the prior art still does not teach why appellants would be motivated to use an LED in place of the LD taught by the Wada reference. As noted above, the disadvantages of using an LED instead of an LD are just as prominent from Sze as are the advantages. Moreover, in a recording head used for linewise recording information on a photoreceptor, these disadvantages might be serious. Therefore, without more, there is insufficient basis for the Board's determination that Wada in combination with Pankove and Sze would have motivated one in appellants' position to operate the LD as an LED.

Finally, the Board also acknowledges that Wada does not disclose the series resistance reducing metallization layer, and looks to the Levine prior art reference for the teaching that the resistance of thin metal film is a function of thickness. Based on the teaching in Levine, the Board concludes it would have been obvious to deposit similar metals on the connecting structure in Wada until the resistance thereof is sufficiently reduced for the device to function in the manner of the claimed invention. This analysis is flawed. The Board itself acknowledged in its decision for reconsideration that, while the resistance of a thin metal can be a function of its thickness as taught by Levine, there are other factors that affect resistance. App. at 13. The Board and the solicitor fail to cite prior art references which explain why one skilled in the art when faced with Stevens' problem would have selected to vary this particular property (i.e., thickness) as opposed to other properties not disclosed in order to obtain the desired resistance.

As to claim 14, the Board rejected it based on the references applied to claims 1-13 taken in view of Iwata. I believe the Board erred in rejecting claim 14 since Iwata does not overcome the deficiencies of Wada and the remaining references. Iwata only discloses an optical print head comprising an array of devices containing LEDs and associated drivers. Although claim 14 also claims a linewise array of devices described in claim 1 to form a recording head, since the five prior art references cited against claims 1-13 do not render claims 1-13 obvious, neither do these references in combination with Iwata render claim 14 obvious.

Second, even if the Wada patent in combination with the Pankove, Itoh, Sze and Levine publications fully teach all features of the claimed invention, there is no motivation in the prior art to combine these isolated teachings. We stated in In re Fine, 837 F.2d 1071, 1075, 5 USPQ2d 1596, 1599-1600 (Fed. Cir. 1988), that

Obviousness is tested by "what the combined teachings of the references would have suggested to those of ordinary skill in the art." In re Keller, 642 F.2d 413, 425, 208 USPQ 871, 881 (CCPA 1981). But it "cannot be established by combining the teachings of the prior art to produce the claimed invention, absent some teaching or suggestion supporting the combination." ACS Hosp. Sys., 732 F.2d at 1577, 221 USPQ at 933.... One cannot use hindsight reconstruction to pick and choose among isolated disclosures in the prior art to deprecate the claimed invention.

In the present case, there is nothing in Wada or in any of the other references which would motivate one skilled in the art to combine the separate teachings of the cited prior art to come up with all the features of the claimed device. The Board appears to have succumbed to "hindsight reconstruction" by starting with the features of the claimed invention, combing through the prior art to find teachings on the various aspects of the claimed invention, and then declaring the claimed invention obvious in light of those references. In In re Fine, we specifically warned against such "pick [ing] and choos [ing] among isolated disclosures in the prior art" and "fall [ing] victim to the insidious effect of a hindsight syndrome." 837 F.2d at 1075, 5 USPQ2d at 1600 (citation omitted).

CONCLUSION

Given the gaps in the prior art teachings compared to the features of the claimed invention and the absence of sufficient suggestion to combine, I conclude the Board erred in its determination that Wada in combination with Pankove, Itoh, Sze and Levine fully teach the claimed invention. Accordingly, reversal of the Board's rejection of claims 1-13 for obviousness under 35 U.S.C. § 103 is required.

In my view, the Board also erred in rejecting claim 14 since Iwata does not overcome the deficiencies of Wada and the other references.