Board of Patent Appeals and Interferences
Patent and Trademark Office (P.T.O.)
*1 EX PARTE SATOSHI HIYAMIZU AND TOSHIO FUJII
Appeal No. 650-06
February 8, 1988
Application for Patent filed July 27, 1981, Serial No. 286,863. High Electron Mobility Single Heterojunction Semiconductor Devices and Methods for Production Thereof.
John P. Moran for appellants
Primary Examiner--Martin H. Edlow
Before Henon, Rubinson and Hairston
This appeal is from the refusal of the examiner to allow claims 1 and 5 through 13, which constitute all the claims remaining in the application.
The invention relates to improved structure for high electron mobility heterojunction semiconductor devices. Claim 1 is considered to be representative and reads as follows:
1. A high electron mobility heterojunction semiconductor device comprising:
a semi-insulating GaAs substrate;
an undoped A1GaAs layer grown on the semi-insulating GaAs substrate;
a GaAs layer grown on said undoped A1GaAs layer, said GaAs layer having a low N-type impurity concentration;
a N-type A1GaAs layer grown on said GaAs layer, and defining a heterojunction between said GaAs layer and said N-type A1GaAs layer;
at least one gate electrode formed on said N-type A1GaAs layer; and
a pair of source and drain electrodes arranged flanking said at least one gate electrode.
The references relied on by the examiner are:
Mills et al. (Mills) 4,155,784 May 22, 1979
Japan (Mitsubishi) 54-12261 Jan. 29, 1979
Mimura et al. (Mimura), Japanese Journal of Applied Physics, Vol. 19, No. 5 (May 1980), pp. 1225-1227.
All the claims in issue stand rejected as being directed to obvious subject matter within the meaning of 35 U.S.C. 103. As evidence of obviousness, the examiner relies on the teachings of Mimura considered with Dingle, Mills and Mitsubishi. As stated in the final rejection of December 31, 1984 (paper no. 23) the apparent position of the examiner is that if the person of ordinary skill in the pertinent art were confronted with a chromium diffusion problem in a conventional high electron mobility transistor of the nature disclosed by Mimura, such an artisan would look to the teachings of Mills or Mitsubishi for a solution to the problem and would have found it obvious to include a "buffer layer" between the Mimura substrate and the undoped GaAs layer and that the artisan would have further found it obvious that the "buffer layer" may comprise both undoped GaAs or GaA1As layers. As to the additional requirements found in dependent claims 6 through 11, the examiner relies on the teachings of Dingle.
*2 Preliminarily we note that although the Figueroa patent (4,346,394) has been cited and discussed in both the final rejection and answer, it has not been included in the statement of the rejection. Accordingly, we have not considered the teachings of Figueroa in forming our decision. Note In re Hoch, 428 F.2d 1341, 166 USPQ 406 (CCPA 1970).
As an additional preliminary matter, we disagree with the examiner's assertion (answer page 5) that by definition "one of normal (sic, ordinary) skill in the art" is an engineer or scientist of the doctorate level working at least 40 hours per week in semiconductor research or development, as evidenced by the publication attached to appellant's principal brief. The statutory "person having ordinary skill in the art" created by Congress to provide a standard of patentability is a hypothetical person presumedly ossessing knowledge in the field to which the claimed "subject matter pertains," wherein the level of said knowledge resides somewhere between that possessed by the layman and that possessed by the expert. Note Standard Oil Company v. American Cyanamid Co., 774 F.2d 448, 227 USPQ 293 (Fed.Cir.1985); and Kimberly-Clark Co. v. Johnson and Johnson, 745 F.2d 1437, 223 USPQ 603 (Fed.Cir.1984). It is our view that such a hypothetical person is no more definable by way of credentials than is the hypothetical "reasonably prudent man" standard found in laws pertaining to negligence. As to the publication relied on by the examiner as evidence in support of his definition, we find that the publication fails to support said definition. That is to say, the publication fails to indicate how many hours per week the authors work, and although each of the authors has apparently earned a doctorate degree, such would merely indicate to us that the authors are highly skilled and, perhaps, experts. In summary, although the hypothetical "person having ordinary skill in the art" to which the claimed subject matter pertains would, of necessity have the capability of understanding the scientific and engineering principles applicable to the pertinent art, we disagree that the evidence in this application supports the conclusion that such a hypothetical person would require a doctorate or equivalent knowledge in science or engineering.
Turning to a consideration of the rejection of the claims in issue as being directed to obvious subject matter under 35 U.S.C. 103, we find from a careful review of the record in this application that the rejection cannot be sustained. We reach such a conclusion substantially for the reasons presented by appellants in their briefs and our additional comments below.
Under 35 U.S.C. 103 where the examiner has relied on the teachings of several references, the test is whether or not the references viewed individually and collectively would have suggested the claimed invention to the person possessing ordinary skill in the art. Note In re Kaslow, 707 F.2d 1366, 217 USPQ 1089 (Fed.Cir.1983). It is to be noted, however, that citing references which merely indicate that isolated elements and/or features recited in the claims are known is not a sufficient basis for concluding that the combination of claimed elements would have been obvious. That is to say, there should be something in the prior art or a convincing line of reasoning in the answer suggesting the desirability of combining the references in such a manner as to arrive at the claimed invention. Note In re Deminski, 796 F.2d 436, 230 USPQ 313 (Fed.Cir.1986). Furthermore, it is well settled that where the claimed invention solves a problem, the discovery of the source of the problem and its solution are considered to be part of the "invention as a whole" under 35 U.S.C. 103. Note In re Kaslow, supra; In re Nomiya, 509 F.2d 566, 184 USPQ 607 (CCPA 1975); and In re Sponnoble, 405 F.2d 578, 160 USPQ 237 (CCPA 1979).
*3 Reviewing the prior art relied on by the examiner, we initially note that none of the references teaches or suggests the problem and its possible causes as identified by appellants beginning at the last paragraph of page 4 of the specification. However, even if we presume for the sake of argument that the admissions found in the paragraph bridging pages 2 and 3 of appellants' disclosure coupled with the discussion found in Mills and Dingle pertaining to the effects of impurity migration would have suggested the problem noted by the examiner at page 3 of his final rejection and would have served as motivation to look to the prior art for a solution to the problem, we are not convinced that the collective teachings of the references would have suggested appellants' claimed solution. For example, Mimura merely teaches a structure similar to that which is admitted by appellants to be known in their specification and the reference includes no teachings or suggestions pertaining to the problem or its solution as identified by appellants.
Mills teaches the inclusion of a gallium arsenide buffer layer between the substrate ?? into the active layer so as to prevent migration of the substrate ?? into the active layer. However, both Mimura and the admitted ?? illegible text?? already include such a buffer layer. As to the teachings of Mitsubishi, we do not consider the reference to teach the interchangeability of the claimed ternary material for the binary material taught by Mills, as seemingly urged by the examiner at page 7 of the ??. The functions for the respective ?? taught by the references are different. That is to say, Mills teaches the inclusion of the binary buffer layer for the purpose of preventing the migration of chromium to the active layer, whereas, Mitsubishi teaches the inclusion of the ternary layer for the purpose of, inter alia, forming the heterojunction at the boundary between the operating layer and the buffer layer. Moreover, the claim requires defining the heterojunction between the upper GaAs layer and the N-type A1GaAs layer. In light of the above, it would appear to us that it would not have been obvious to modify Mimura in the manner urged by the examiner without using appellants' claims as a guide. It is to be noted that simplicity and hindsight are not proper criteria for resolving the issue of obviousness; note In re Horn, 203 USPQ 969 (CCPA 1979). As to the teachings of Dingle which, in some respects, are far more pertinent to the problem and its solution as disclosed by appellants, we note that the solution taught by the reference, namely modulating the doping and bandgap of the multilayered structure (column 3, second full paragraph), is significantly different from the structure specified in the claims in issue. In light of the above, we are convinced that the evidence adduced by the examiner fails to support his conclusion that the claims at bar are directed to obvious subject matter within the meaning of 35 U.S.C. 103.
*4 The decision of the examiner rejecting claims 1 and 5 through 13 under 35 U.S.C. 103 is reversed.
Paul J. Henon
Gene Z. Rubinson
Kenneth W. Hairston