Ex Parte 6440359 et al

Board of Patent Appeals and Interferences

Aug 31, 2007

90006951 (B.P.A.I. Aug. 31, 2007)

The opinion in support of the decision being entered today is not binding
precedent of the Board.
UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte ALCAN INTERNATIONAL LIMITED
____________
Appeal 2007-2359
Reexamination 90/006,9511
Patent 6,440,3592
Technology Center 3900
____________

Decided: August 31, 2007
____________

Before TEDDY S. GRON, CAROL A. SPIEGEL, and MARK NAGUMO,
Administrative Patent Judges.

NAGUMO, Administrative Patent Judge.

1 Request for reexamination filed 24 February 2004. Citations to the
Specification are to the 359 patent as issued.
2 The national stage of PCT/GB98/00849 was entered on 22 September
1999, under 35 U.S.C. § 371; the PCT parent application was filed 20 March
1998, claiming the benefit of EPO 97-301,911, filed 21 March 1997. The
named inventors are Nicholas Charles Parson, Jeffrey David Hankin, and
Kevin Paul Hicklin. The real party-in-interest is listed as Alcan
International Ltd., of Canada (Brief filed 16 August 2006 ("Br."), at 2.)
Appeal 2007-2359
Application 90/006,951

DECISION ON APPEAL
A. Introduction
The patent owner ("Alcan") appeals under 35 U.S.C. § 134 from the
final rejection under 35 U.S.C. § 103 of claims 1–5 and 7–20, which are all
of the pending claims in this reexamination of U.S. Patent 6,440,359. We
have jurisdiction under 35 U.S.C. § 6(b). We REVERSE.
The subject matter on appeal relates to aluminum-magnesium-silicon
("AlMgSi") alloys said to be particularly suitable for "the bottom end" of the
extrusion market because they are strong (i.e., have an Aluminum
Association alloy AA6063-T5 or -T6 rating) and can be extruded at a
relatively high rate. (359 patent at 1:11–14.) (AA6063 alloys are said to
account for 80% of all extruded aluminum goods. (Id. at 1:9–10.)) The
inventors state that the industry has obtained such alloys by reducing the
content of Mg2Si while increasing the amount of "excess silicon" and adding
copper ("Cu"), and manganese ("Mn") to obtain extrudability, tensile
strength, and other desired properties. (359 patent at 1:15–25.) However,
according to the inventors, there has been a bias against reducing the content
of Mg to less than 0.35 wt% in general purpose AlMgSi extrusion alloys.
(359 patent at 1:26-31.) The claimed compositions have an Mg content of at
most 0.34 wt%.
Claim 1 is representative of the issues necessary to resolve this appeal
and reads:
A DC cast extrusion ingot of an alloy of composition consisting
of
0.20–0.34 wt.% Mg,
2
Appeal 2007-2359
Application 90/006,951

0.35–0.60 wt.% Si,
0.02–0.15 wt.% Mn,
up to 0.10 wt.% Cu,
up to 0.35 wt% Fe,
optionally Ti and B as grain refiners,
incidental impurities up to 0.05 wt% each, up to
0.15 wt.% total,
balance Al,
provided that when Mg is at least 0.30 wt% and Cu is at least
0.05 wt.%, then Fe is greater than 0.15 wt%,
wherein excess Si, as defined by the formula
Excess Si = Si - (Mg/1.73) - ((Fe + Mn)/3),
is present in said composition
and wherein Fe is present in the ingot substantially as α-AlFeSi,
by virtue of homogenisation.
(Br. App. at 20.; Examiner's Answer at 4; paragraphing and indentation
added.)
B. Findings of Fact
The following findings of fact and any set out in the Discussion are
supported by a preponderance of the evidence of record. To the extent any
finding of fact is a conclusion of law, it should be treated as such.
1. The patent under reexamination, U.S. 6,440,359 B1 ("359 patent"),
issued 27 August 2002, based on application 09/355,497, the national stage
of PCT/GB98/00849, which names Nicholas Charles Parson, Jeffrey David
Hankin, and Kevin Paul Hicklin as the inventors and Alcan Int'l Ltd, of
Canada, as the assignee.
2. The national stage was entered on 22 September 1999; the PCT
application was filed 20 March 1998.
3
Appeal 2007-2359
Application 90/006,951

3. The benefit of the 21 March 1997, filing date of EP 97-301,911 under
35 U.S.C. § 119 is claimed in the application.
The Examiner's Rejections
4. The Examiner has rejected claims 1–5, 7–13, and 20 under 35 U.S.C.
§ 103(a) over the combined teachings of Reiso3 and either WO95/067594
("WO") or EP 0,716,716 B15 ("EP"), collectively, "EP/WO". (Examiner's
Answer mailed 1 December 2006 ("Answer"), at 5.)
5. The Examiner has rejected claims 1–5, 7–13, and 15–20 under 35
U.S.C. § 103(a) over the combined teachings of Reiso, Timsit6, and either of
EP/WO. (Answer at 8.)
6. The Examiner has rejected claim 14 under 35 U.S.C. § 103(a) over the
combined teachings of Reiso, EP/WO, and Marchive7. (Answer at 10.)

3 Oddvin Reiso, The Effect of Composition and Homogenization Treatment
of Extrudability of AlMgSi Alloys, 1 Proc. 3d Int'l Aluminum Extrusion
Technology Seminar, 1984 31 (1984).
4 Alcan Int'l Ltd. (Canada), WO 95/06,759, Extrudable Al-Mg-Si Alloys,
based on international application PCT/GB94/01,880, published 9 March
1995.
5 Alcan Int'l Ltd. (Canada), EP 0,716,716 B1, Extrudable Al-Mg-Si Alloys,
published and granted 12 August 1988, based on international application
PCT/GB94/01,880.
6 Roland S. Timsit and Benjamin J. Janeway, Aluminum Brazing Sheet, U.S.
Patent 5,232,788, issued 3 August 1993.
7 D. Marchive, High Extrudability Alloys in the 6000 Series, Light Metal
Age 6 (April 1983)
4
Appeal 2007-2359
Application 90/006,951

7. The Examiner has rejected claim 14 under 35 U.S.C. § 103(a) over the
combined teachings of Reiso, Timsit, EP/WO, and Marchive. (Answer
at 11.)
359 Patent Disclosure
8. According to the 359 patent, the invention relates to AlMgSi alloys
containing a relatively low amount of Mg that meet certain strength
requirements and that can be extruded at "the highest possible rates."
(359 patent at 1:5–14.)
9. The principal additional components of the aluminum alloy are said to
be magnesium, silicon, manganese ("Mn"), copper ("Cu"), and iron ("Fe").
(359 patent at 2:1–10.)
10. The Specification adopts a convention that all amounts are reported in
weight-percent (e.g., 359 patent at 1:66-67); accordingly we shall write
"0.34 wt.%" as "0.34 %" or as "0.34" except when quoting the record.
11. The Specification teaches that "incidental impurities" may be present
at up to 0.05 % each, with a total of 0.15 %. (359 patent at 2: 11–13.]
12. Each component is said to have upper and lower limits, beyond which
adverse consequences for the alloy arise.
13. Extrusion pressure is said to increase with increasing Mg content,
eventually becoming unacceptably high. If the Mg content is too low,
"extrusions" (i.e., extruded products) cannot attain the required strength.
The Mg content of the invention alloys is 0.20–0.34, preferably 0.20-0.30.
(359 patent at 2:59–63.)
5
Appeal 2007-2359
Application 90/006,951

14. According to the Specification, if the Si content is too low, the alloy
strength is too low; and if the Si content is too high, the extrudability of the
alloy is reduced. The Si content of the alloy is 0.35–0.60,
preferably 0.40-0.59. (359 patent at 2:64–3:5.)
15. The presence of iron is said to be unavoidable in practice because it is
too expensive to remove; and, in certain applications at least 0.15 % Fe is
said to be desirable. Iron may be present up to 0.35 %. (359 patent
at 3:6-16.)
16. Most importantly, Fe is said to be present in as-cast alloy ingots as
large plate-like β-AlFeSi particles (359 patent at 3:12–13), which are said to
contribute to a reduction in surface quality at a given temperature and
extrusion speed (Parson's Declaration8 at 6).
17. According to the Specification, β-AlFeSi is preferably converted to
substantially (at least 80%) α-AlFeSi by a heat-treatment referred to as
"homogenisation" (henceforward "homogenization"). (359 patent at 3:13-16
and at 4:19–26.)
18. According to the Specification, the inventors discovered that adding
an appropriate level of Mn, given by the equation
wt% Mn ≥ 0.3 × wt% silicon - 0.12
promotes the β- to α-AlFeSi transformation during homogenization,
decreases the AlFeSi particle size, and increases the "degree of
spheroidization" of the AlFeSi particles in the alloy. (359 patent at 3:24-52.)

8 Declaration of Nicholas Charles Parson, one of the inventors, under
37 C.F.R. § 1.132, filed in the reexamination on 3 January 2006.
6
Appeal 2007-2359
Application 90/006,951

19. The Specification teaches that Mn is not generally useful for
improving the toughness of alloys of the present invention. (359 patent
at 3:17-20.)
20. To avoid increased levels of dispersoid formation upon quenching,
Mn levels are kept below 0.15%, preferably below 0.10%. (359 patent
at 3:20-23.)
21. Cu is said to improve tensile strength without a comparable increase
in extrusion breakthrough pressure, but to increase corrosion problems.
(359 patent at 3:53–58.)
22. The Specification discloses that Si in excess of the amount required to
combine with all the Mg as Mg2Si and with all the Fe and Mn as
Al(Fe,Mn)Si (a nonstoichiometric formula) is desirable to achieve the
desired tensile strength; but that too much excess Si adversely affects the
extrusion surface quality and "formability" of the alloy. (359 patent
at 3:59-4:11.)
23. According to the Specification, as-cast extrusion ingots are
homogenized to bring the soluble secondary magnesium-silicon phases into
suitable form, to dissolve the silicon, and partly to convert β-AlFeSi
particles into substantially α-AlFeSi particles. (359 patent at 4:19-24.)
24. Homogenization is accomplished by heating the ingot to a
temperature higher than 530°C for 30 minutes to 24 hours, depending on the
temperature. (359 patent at 4:24-26.)
25. Following homogenization, the ingot is cooled to a
temperature ≤ 425°C, preferably at a rate of at least 150°C per hour.
(359 patent at 4:27–32.)
7
Appeal 2007-2359
Application 90/006,951

26. The homogenized ingot is then extruded and aged at elevated
temperatures to attain peak strength. (359 patent at 4:40-67.)
Reiso
27. Reiso indicates that, in 1984, it was generally accepted that Mg and Si
content, as well as the homogenization process used, affected the
extrudability of AlMgSi alloys. (Reiso at 31, col. 1.)
28. According to Reiso, the literature was "somewhat contradictory," and
his goal was to clarify the influence of Mg and Si and the cooling rate after
homogenization on the extrudability of these alloys. (Reiso at 31, col. 1.)
29. Reiso presents studies of three series of alloys, I, II, and III.
30. In Series I, the Mg content was held constant at about 0.40, while the
Si content was varied between 0.29 and 0.98 (alloys 1–5); and then the Si
content was held constant at about 0.43 while the Mg content was varied
from 0.30 to 0.70 (alloys 6–10). (Reiso at 33, Table 1.)
31. Alloy 6 in Series I has Mg of 0.30, Si of 0.43, and Fe of 0.21. (Reiso
at 33, Table 1.)
32. In Series II, the Mg content was held constant at about 0.50 while the
Si content was varied between 0.31 and 0.98; then the Si content was held
constant at about 0.50 while the Mg content was varied between 0.28
and 0.78. (Reiso at 33, Table 1.)
33. Alloy 16 in Series II has Mg of 0.28, Si of 0.48, and Fe of 0.19.
(Reiso at 33, Table 1.)
34. All other alloys reported by Reiso have an Mg content of 0.39 or
greater. (Reiso at 33, Table 1.)
8
Appeal 2007-2359
Application 90/006,951

35. In Series III, three alloys having 0.45, 0.66, and 0.45 Mg were
prepared for studies of the effect of cooling rate after homogenization.
(Reiso at 33, col. 2.)
36. Samples were extruded and examined by scanning electron
microscopy and energy dispersive X-ray analysis. (Reiso at 34, col. 1.)
37. According to Reiso, Mg-rich and Si-rich alloys contained, "in addition
to the primary AlFeSi constituents, coarse phases containing Mg and/or Si
with varying Mg:Si ratios." (Reiso at 34, col. 1.)
38. However, Reiso reports that example 16 "had a negligible amount of
such phases." (Reiso at 34, col. 1.)
39. Reiso also reports that only extrusions of alloy 5 (Si-rich) and alloy 10
(Mg-rich) of Series I showed the coarse Mg-Si phases. (Reiso at 34, col. 1.)
40. Hence, it is reasonable to infer that alloy 6 of Series I showed a
"negligible amount" of coarse Mg-Si phases.
41. Reiso developed a model for damage-limited extrusion of AlMgSi
alloys. (Reiso at 38, col. 2ff.)
42. According to Reiso:
[t]he temperature in the surface of the extrusion is thought to be
the major factor in limiting the extrusion speed for a given
section. For a given extrusion speed (V1) the temperature of the
extrusion will increase with increasing deformation resistance,
i.e., increasing amounts of Mg and Si in solid solution or
increasing cooling rate after homogenization for a given alloy."
(Reiso at 38, col. 2.)
43. Reiso explains that when an alloy contains coarse Mg-Si phases that
are too big to go into solution in a given extrusion process (too low
9
Appeal 2007-2359
Application 90/006,951

temperature or not enough time), "these coarse phases are then thought to
represent regions of local weakness in the structure." (Reiso at 39, col. 1.)
44. These local weaknesses are thought to induce a change in the
initiation of tearing from type I (tearing of the Al matrix) to "type II"
(tearing when the temperature exceeds the melting temperature of the
eutectic composition of Mg-Si phases, AlFeSi phases, and the Al matrix).
(Reiso at 39, col. 1.)
45. According to Reiso, the effects of cooling rates on different
compositions can be analyzed in a similar fashion: "at some alloy
content . . . the given cooling rate is too slow to prevent precipitation of
coarse Mg-Si phases surviving the extrusion process and a shift in the
mechanism of tearing initiation results." (Reiso at 39, col. 1.)
46. Reiso also explained the effect of cooling rate after homogenization
on alloy structure with his model. (Reiso at 39, col. 1.)
47. According to Reiso:
[a]t slow cooling rates the Mg-Si phases are allowed to grow so
big that they will not go into solid solution before the material
reaches the extrusion die, and a shift in the mechanism of
tearing initiation results in lower extrusion speeds. The 'critical
cooling rate' will vary with Mg and Si content as the particle
size will depend upon factors such as supersaturation,
nucleation rates and growth rates.
(Reiso at 39, col. 1.)
48. Reiso concludes that "[c]oarse Mg-Si phases in the structure should be
avoided," although some precipitation is beneficial because it reduces
deformation resistance, provided that the precipitates are "of a subcritical
size to avoid local weakening of the structure." (Reiso at 39, col. 2.)
10
Appeal 2007-2359
Application 90/006,951

49. Reiso advises that this goal may be achieved by increasing the
nucleation rate by "choosing alloys with a Si content in excess of the amount
necessary to form Mg2Si or by adding small amounts of dispersoids (Mn, Cr,
Zr) to the alloy." (Reiso at 39, col. 2, to 40, col. 1, citing Bichsel9.)
50. Bichsel, which Reiso cites as reporting a "positive effect by addition
of only 0.06% Mn" (Reiso at 40, col. 1), describes alloys having
about 0.54% Si and 0.55% Mg. (Bichsel at 58, table at col. 1; translation
at 1.)
51. Reiso does not exemplify alloys that contain Mn or Cu.
WO 95/06,759 and EP 0,716,716
52. Patent publications WO and EP claim the benefit of priority of the
same Great Britain patent application and have the same inventors, who are
entirely distinct from the inventors named for the 359 patent undergoing
reexamination.
53. WO is the published application.
54. EP is the European patent Specification.
55. The Examiner has cited the two patent publications in parallel and the
disclosures are substantially the same, although the EP disclosure is
somewhat more extensive than the WO disclosure. Neither the Examiner
nor Alcan have argued either disclosure separately.

9 H. Bichsel et al., Metallographic Investigations on the Influence of Small
Additions of Manganese to AlMgSi 0.5 Alloys, Proc. 7th Int'l Light Metals
Congress, 68 (1981) (English translation provided by Alcan during
prosecution).
11
Appeal 2007-2359
Application 90/006,951

56. Accordingly, we shall treat WO and EP as cumulative and cite the EP
disclosure exclusively.
57. EP relates to "intermediate strength extrudable AlMgSi alloys" having
the following compositions in weight %:
Mg 0.25 – 0.40
Si 0.60 – 0.90
Mn 0.10 to 0.35; and up to 0.35
Fe up to 0.35
others up to 0.05 each, 0.15 total
Al balance.
(EP at 2:15–21.)
58. According to EP, the "nominal composition" of its inventive
composition is:
Mg 0.35 Si 0.70 Fe 0.2.
(EP at 2:39-50.)
59. EP explains that a "balanced alloy" is "one in which just enough Si is
present to combine with all the Mg, Fe, Mn as Mg2Si and Al(Fe,Mn)Si."
(EP at 2:41–42.)
60. Thus, EP's inventive alloys are said to be "high excess Si alloys." (EP
at 2:39–40.)
61. Preferred compositions are described as having excess Si of at
least 0.3%. (EP at 3:27–29.)
62. According to EP, if the Mg content is too low, it is hard to achieve the
required strength in aged extrusions, while if the Mg content is too high, the
extrusion pressure becomes unacceptably high. (EP at 3:9–11.)
12
Appeal 2007-2359
Application 90/006,951

63. EP instructs that if the Si content is too low, the alloy strength
becomes too low, while if the Si content is too high, the extrudability
becomes too low. (EP at 3:12-14.)
64. According to EP, the Fe is present in an as-cast alloy ingot as large
plate-like β-AlFeSi particles, which are preferably homogenized to convert
the β-form to the α-form. (EP at 3:15–18.)
65. EP states that it was known that excess Si (as defined supra) stabilizes
the β-AlFeSi form, "which has a detrimental effect on extrudability and in
particular on extrusion surface quality." (EP at 3:18–20.)
66. EP states that "Mn is included in the alloys in order to improve
extrusion surface quality. Mn acts to accelerate the β to α-AlFeSi
transformation during homogenisation, so that the resulting homogenised
ingot has improved extrudability, that is to say improved extrusion surface
quality." (EP at 3:22–24.)
67. EP teaches that following homogenization at 550–600°C, the hot
homogenized ingot is extruded under conventional hot extrusion conditions
and quenched, preferably in still air. (EP at 3:34–40.)
68. According to EP, aging at 150–200°C for 1–48 hours, yielding
ultimate tensile strengths "of at least 240 MPa, often greater than 250 MPa,
with acceptable toughness." (EP at 3:42–50.)
69. Comparison of the EP alloy composition ranges with the 359 patent
composition ranges on appeal shows that there is substantial overlap for all
components except for Si:
EP 359 Patent
Mg 0.25–0.40 0.20–0.34
13
Appeal 2007-2359
Application 90/006,951

Si 0.60–0.90 0.35–0.60
Mn up to 0.35 0.02–0.15
Fe up to 0.35 up to 0.35
Cu up to 0.05 up to 0.10
70. The extent of "overlap" between the EP/WO disclosure and
the 359 patent claimed subject matter can also be seen in Figure 1 of
the 359 patent.
{Figure 1 of the 359 patent is shown below:}10

{Figure 1 is said to compare the composition of the invented alloy with the
prior art}
71. Region A is said to be the broad range of alloy compositions, while
region B is said to be the narrow (preferred) range of alloy compositions.
(359 patent at 2:1–6.)

10 The text in curly braces preceding and following the Figures are provided
to ensure compliance with section 508 of the U.S. Rehabilitation Act for
publication of this Decision on the USPTO website pursuant to the Freedom
of Information Act. They are not part of the Decision.
14
Appeal 2007-2359
Application 90/006,951

Timsit
72. According to the face of the Timsit patent, the assignee is Alcan Int'l
Ltd., the same entity as the patent-owner requestor in this reexamination
proceeding.
73. Timsit describes an aluminum alloy composite sheet said to be useful
in brazing.
74. Timsit explains that aluminum components are commonly joined by
disposing an aluminum brazing alloy between the surfaces to be joined and
subjecting the assembly to a brazing temperature, i.e., a temperature at
which the brazing alloy melts while the components remain unmelted.
(Timsit at 1:7–13.)
75. Timsit explains further that when the assembly is cooled, the brazing
alloy forms a fillit or joint that bonds the components. (Timsit at 1:13–15.)
76. According to Timsit, the conventional brazing sheet is prepared by
cladding a core aluminum alloy sheet with a sheet of high-silicon aluminum
alloy that is relatively expensive to produce and difficult to process in scrap-
recovery operations. (Timsit at 1:22–31.)
77. Timsit provides a brazing sheet in which the core alloy and cladding
alloys may be highly variable. (Timsit at 1:53–2:30.)
78. The key, according to Timsit, is to coat the cladding with a mixture of
a brazing flux material and metal particles capable of forming in situ a
eutectic alloy with aluminum. (Timsit at 1:45–52.)
15
Appeal 2007-2359
Application 90/006,951

79. Timsit teaches that other metal particles may be added to the eutectic
alloy forming metal to yield joined parts with improved properties. (Timsit
at 3:12–45.)
80. In particular, Timsit teaches that:
[a]dditions of Mn and excess Si may yield superior
extrudability of the brazed components. All the superior
properties described above would be achieved by diffusion of
brazing-mixture additives into the brazed components and their
subsequent reaction with elements or precipitates in the core
alloys. Diffusion may be accomplished either during brazing or
by an appropriate heat treatment following brazing.
(Timsit at 3:38–45.)
Marchive
81. Marchive teaches that homogenizing is a two-step treatment
comprising a high temperature stage and a cooling stage. (Marchive at 8,
col. 1.)
82. According to Marchive, if the cooling rate is in excess of 500°C per
hour, a solid solution supersaturated in Mg2Si will be obtained, having high
resistance to deformation and lower extrusion velocity. (Marchive at 8,
col. 1.)
83. On the other hand, Marchive teaches that if the cooling rate is slow,
e.g., 20°C per hour, the solid solution will contain little Mg or Si, because
Mg2Si will have fallen out of the solution as coarse precipitates. (Marchive
at 8, col. 1.)
84. In Marchive's words, "[t]he cooling rate therefore has to be adjusted to
yield finely divided Mg2Si precipitates which can redissolve during
extrusion." (Marchive at 8, col. 1.)
16
Appeal 2007-2359
Application 90/006,951

The Rejections
85. The Examiner finds that Reiso teaches cast aluminum ingots having
all limitations of claims 1–5, 7–13, and 15-20 but for:
(1) the use of Mn in the claimed amount together with excess Si;
(2) homogenizing the ingot to convert β-AlFeSi to α-AlFeSi; and
(3) thermally aging the ingot at a heating rate of 10 to 100°C per hour
(required by claims 9 and 15).
(Answer at 6 and 8.)
86. In particular, the Examiner finds that alloys 6 and 16 are compositions
meeting the Mg, Si, and Fe limitations of Alcan's claims, including the
Mg2Si and excess Si amounts (required by claim 3). (Answer at 5.)
87. The Examiner finds that Reiso teaches that coarse Mg-Si phases may
be avoided and the extrudability improved in the AlMgSi alloys by adding
small amounts of dispersoids such as 0.06% Mn. (Answer at 6.)
88. The Examiner reasons that it would have been obvious to include 0.06
wt% Mn in alloys 6 and 16 to improve the extrudability, as taught by Reiso.
(Answer at 6.)
89. The Examiner finds that EP and WO teach homogenization to convert
β-AlFeSi to α-AlFeSi and thermal aging at the required heating rate to
develop the required strength properties. (Answer at 7.)
90. The Examiner also finds that EP and WO teach that Mn is added to
improve extrusion surface quality, and that it accelerates the β- to α-AlFeSi
transformation. (Answer at 7.)
17
Appeal 2007-2359
Application 90/006,951

91. The Examiner reasons that it would have been obvious to apply the
thermal treatments taught by EP and WO to Reiso alloys 6 and 16 because
they would have had a reasonable expectation of successfully obtaining
ingots having the improved properties taught by the references. (Answer
at 7–8.)
92. The Examiner finds that Timsit teaches that it was known to improve
the extrudability of aluminum alloys by adding Mn and excess Si in
combination. (Answer at 8; (Timsit at 3:38–45.)
93. The Examiner reasons that one of ordinary skill in the art would have
applied the teachings of Timsit regarding the improved properties arising
from the addition of Mn and excess Si to aluminum alloys, to the alloys
taught by Reiso, in particular, alloys 6 and 16. The Examiner applies the
same reasoning applied to the teachings of the EP/WO and concludes that
the claims would have been obvious over the combined teachings of Reiso,
Timsit, and the EP/WO references. (Answer at 9–10.)
94. With regard to the limitations of claim 14, namely, that the
homogenized ingot be cooled to 425°C or less at a rate of at least 150°C per
hour, the Examiner finds that Marchive teaches quenching at rates between
20 and 500°C per hour at page 8. (Answer at 10 and at 11.)
95. The Examiner reasons that Marchive teaches a broad range that
encompasses or overlaps the range recited in claim 13, and that in the
absence of a new or unexpected result, modification to select a value within
the claimed range would have been obvious. (Answer at 10 and 11.)
18
Appeal 2007-2359
Application 90/006,951

The Counterarguments
96. Alcan does not appear to argue for the separate patentability of any
claims relative to a given rejection.
97. Rather, regarding the combined teachings of Reiso and the WO/EP
references, Alcan argues that the teachings of Reiso to add Mn to aluminum
alloys do not apply to aluminum alloys 6 and 16. (Br. at 10.)
98. Alcan supports its arguments with citations to Reiso, Bichsel, and the
Declaration of Parson. (Br. at 10 ff.)
99. Alcan argues against the rejection based on the alleged obviousness of
adding Mn to enhance the β- to α-AlFeSi transformation in a similar manner,
urging that the record does not support the equivalence of the high-Si, high-
Mg alloys of the EP/WO references to the relatively low-Si and low-Mg
alloys 6 and 16 of Reiso. (Reply Brief filed 29 January 2007 ("Reply Br."),
at 3.)
100. Alcan finds that EP/WO require a Si-content of 0.60% or greater and
that there is no teaching regarding inventive alloys having less than
0.40 % Mg. (Reply Br. at 2.)
101. More particularly, Alcan finds that the teachings of the EP/WO
publications regarding the addition of Mn "is limited to a narrow range of Si
content that excludes Reiso alloys 6 and 16." (Reply Br. at 3.)
102. Moreover, according to Alcan, it is in the context of a teaching that
excess Si stabilizes the β-AlFeSi phase that has a detrimental effect on
extrudability and extrusion surface quality. (Reply Br. at 3, citing EP
at 3:18–21.)
19
Appeal 2007-2359
Application 90/006,951

103. Alcan finds that EP/WO further emphasizes the high-Si-content by the
preference for alloys containing at least 0.3% excess Si. (Reply Br. at 3; EP
at 3:27–29.)
104. With regard to the rejections based further on Timsit, Alcan argues
that the teachings of Timsit make no sense in the extrusion art. (Br. at 17.)
105. With respect to the rejections of claim 14 based further on Marchive,
Alcan argues only that Marchive does not supply what is lacking in the other
references, and that those rejections should be reversed for that reason.
(Br. at 17–18.)
C. Discussion
At the outset, we commend the respective briefs filed by Alcan and
the Examiner. Findings of Fact are generally supported with specific
citations. Side issues have generally been kept to a minimum and the focus
has remained on the arguments for and against patentability. The reasons for
and against combining various teachings are generally set out clearly. In
short, the briefs have assisted our study of the case and enhanced our
understanding of the issues.
The central and dispositive issue in this appeal is whether a person
having ordinary skill in the art would have applied the teachings of Reiso
and EP/WO regarding the addition of Mn to AlMgSi alloys to alloys 6 or 16
of Reiso. For the reasons that follow, we conclude that the preponderance of
the evidence indicates that an ordinary worker would not have added Mn to
alloys 6 or 16 of Reiso because they would not have recognized those alloys
as having a problem that Mn addition would have solved.
20
Appeal 2007-2359
Application 90/006,951

The record shows that it is the Mg- and Si-rich aluminum alloys that
have significant coarse phases or precipitates of Mg2Si. (FF 37.) Alloys
having low Mg content, such as alloys 6 and 16, showed "negligible
amounts of such phases." (FF 38, 40.) Reiso teaches adding Mn to AlMgSi
alloys as a way to disperse or minimize the size of Mg2Si precipitates.
(FF 48–49.) Consistently, Bichsel, which Reiso cites as reporting a "positive
effect by addition of only 0.06% Mn" describes the treatment of alloys
having about 0.54 Si and 0.55 Mg. (Bichsel at 58, table at col. 1; translation
at 1; FF 50.) Thus, we find that Reiso does not generally teach that addition
of Mn to AlMgSi alloys of all compositions would have been expected to
provide beneficial extrusion properties.
The Examiner's argument that the combination of compositions that
are useful for the same purpose (Answer at 12, citing In re Kerkhoven, 626
F.2d 846, 850, 205 USPQ 1069, 1072 (CCPA 1980)) is not persuasive. The
"Kerkhoven rationale" is inapposite here because the question of adding two
different substances to a composition for a common purpose does not arise.
The Examiner has argued that it would have been obvious to add one
substance, Mn, to two different types aluminum alloys, namely a relatively
high Mg-alloy, and a relatively low Mg alloy. The Examiner has assumed
that the function of Mn is the same in all relevant aluminum alloys. Alcan,
however, has come forward with persuasive evidence from Reiso itself
showing that the alloys differ in the critical problem, namely the need to
minimize the size of Mg2Si precipitates. Moreover, the application of the
"Kerkhoven rationale" depends on subsidiary findings that the art of the
invention is relatively predictable and well understood. In particular, the
idea that additives known to be useful by themselves would have been
21
Appeal 2007-2359
Application 90/006,951

obvious to combine because they would have been expected to perform the
same function is premised on an often unexamined assumption that the
additives act linearly and independently (or in some other well understood
way) of one another. The record here, however, indicates that the various
additives to aluminum alloys interact with one another in complicated,
nonlinear and unpredictable ways. Thus, the Examiner's arguments for
obviousness based on Reiso's teachings are not persuasive.
In the Answer, the Examiner offers, apparently for the first time, a
second rationale for obviousness based on the teachings of the EP/WO
publications that Mn improves extrusion surface quality by accelerating the
β- to α-AlFeSi transformation (EP at 3:22–26; FF 66) and thus counteracts
the stabilization of the β-phase by excess Si (EP at 3:18–20; FF 65).
(Answer at 7 and at 12–13.) The Examiner argues that "[a]ny Mn addition is
beneficial in this way." (Answer at 13.) The Examiner does not, however,
address whether alloys 6 and 16 are sufficiently similar to the higher Mg-
content alloys that EP/WO exemplifies such that both would have been
expected, reasonably, to have similar β-α transformation issues that would
be relieved by the addition of manganese. Rather, the Examiner implicitly
finds that the teachings are generally applicable to aluminum alloys, and
concludes that the addition of Mn to alloys 6 and 16 would have yielded
improved extrudability (id. at 14).
To the extent that the Examiner was justified in relying on the
accuracy of this interpretation of a nonpatent publication, the burden was
shifted to the Applicant to come forward with evidence and reasoning to
doubt the Examiner's interpretation of the record. Alcan responded that the
teachings of the EP/WO publications regarding the addition of Mn "is
22
Appeal 2007-2359
Application 90/006,951

limited to a narrow range of Si content that excludes Reiso alloys 6 and 16."
(Reply Br. at 3.) In particular, Alcan argues that EP/WO require a
Si-content of 0.60% or greater and that there is no teaching regarding
inventive alloys having less than 0.40 % Mg. (Id. at 2.) Moreover,
according to Alcan, it is in the context of a teaching that excess Si stabilizes
the β-AlFeSi phase that has a detrimental effect on extrudability and
extrusion surface quality. (Id. at 3, citing EP at 3:18–21.) Alcan finds that
the high-Si-content is further emphasized by the preference alloys containing
at least 0.3% excess Si. (Id.; EP at 3:27–29; FF 61.)
All the direct teachings of adding Mn to aluminum alloys in the prior
art of record relate to high Si- and high Mg-content alloys. In this context, it
is appropriate to consider Timsit's statement that "[a]dditions of Mn and
excess Si may yield superior extrudability of the brazed components."
(Timsit at 3:38–39.) This statement is not limited to any particular class of
aluminum alloy. Nevertheless, we consider its face value in weighing the
evidence as a whole. Amgen Inc. v. Hoechst Marion Roussel, Inc., 314 F.3d
1313, 1355, 65 USPQ2d 1385, 1416 (Fed. Cir. 2003) (reh'g and reh'g en
banc denied) ("[W]e hold a presumption arises that both the claimed and
unclaimed disclosures in a prior art patent are enabled.") It is not clear
whether the origin of the caution is due to the diffusive mechanism proposed
by Timsit to introduce the Mn and excess Si to the bulk of the aluminum
alloy components being brazed, or to uncertainties as to which alloys will
show improvements in extrudability. In an art in which "other impurities"
are limited to "up to 0.05% each, 0.15% total" (EP at 2:20), it is not
plausible that the ordinary worker would consent to "using up" a significant
fraction for impurities by adding a potentially active ingredient without
23
Appeal 2007-2359
Application 90/006,951

reason. We do not accord significant weight to Timsit's vague and general
teaching compared to the more specific disclosures of Reiso, Bichsel, and
the EP/WO publications.11
Based on the foregoing, the preponderance of the evidence, favors
patentability of the claimed subject matter over the applied prior art.
Accordingly, we conclude that, on the present record, the Examiner has not
shown that it would have been obvious to modify Reiso alloys 6 or 16 by
adding Mn in the amounts required by the Alcan claims.
As none of the other references relied on by the Examiner as evidence
of obviousness address the obviousness of adding Mn to AlMgSi alloys, we
need not consider them further.
D. Summary
In view of the record and the foregoing considerations, it is
ORDERED that the Examiner's rejection of claims 1–5, 7–13,
and 20 under 35 U.S.C. § 103(a) over the combined teachings of Reiso and
either WO95/06759 or EP 0,716,716 B1 is REVERSED;
FURTHER ORDERED that the Examiner's rejection of
claims 1–5, 7–13, and 15–20 under 35 U.S.C. § 103(a) over the combined
teachings of Reiso, Timsit, and either WO95/06759 or EP 0,716,716 B1 is
REVERSED;

11 Alcan's objection that Timsit, which is concerned with brazing, makes no
sense in the extrusion art is understandable (why would one extrude an
article comprised of two parts brazed together?) but misdirected. We
understand the Examiner to rely on Timsit's statements as evidence that the
effect of Mn on AlMgSi alloys was known and considered to be general.
For the reasons given, we find that Timsit is not persuasive.
24
Appeal 2007-2359
Application 90/006,951

FURTHER ORDERED that the Examiner's rejection of
claim 14 under 35 U.S.C. § 103(a) over the combined teachings of Reiso,
either WO95/06759 or EP 0,716,716 B1, and Marchive is REVERSED;
FURTHER ORDERED that the Examiner's rejection of claim 14
under 35 U.S.C. § 103(a) over the combined teachings of Reiso, Timsit,
Marchive, and either WO95/06759 or EP 0,716,716 B1 is REVERSED; and
FURTHER ORDERED that the reexamination application is
returned to the Examiner for action not inconsistent with this Decision.

REVERSED

Christopher C. Dunham
Cooper & Dunham LLP
1185 Avenue of the Americas
New York NY 10036

mg
25