Appendix — Tanabe Seiyaku Co. v. International Trade Commission
Supreme Court brief1997
Ask Donna
What actually matters in this document.
Text
(2) ete ea
“No. F_ 2° 395 SEP - 4 1997,
OFFICE OF THE CLERK”
In The
Supreme Court of the United States
October Term, 1996
’
TANABE SEITYAKU CO., LTD.
AND MARION MERRELL DOW, INC.,
Petitioners,
V.
UNITED STATES INTERNATIONAL
TRADE COMMISSION,
Respondent,
-and-
ORION CORPORATION FERMION, COPLEY
PHARMACEUTICALS, INC., INTERCHEM
CORPORATION AND RHONE-POULENC RORER, INC.,
Respondents.
+
On Petition For A Writ Of Certiorari
To The United States Court Of Appeals
For The Federal Circuit
*
APPENDIX TO PETITION FOR A WRIT OF
CERTIORARI VOLUME II, PAGES 251-462
+
LAWRENCE G. MEYER MicHaet A. Carvin*
James M. SPEARS Davip H. THOMPSON
THomas B. PAHL Cooper & Carvin, PLLC
Gapssy & HANNAH, LLP 2000 K Street, N.W.
1747 Pennsylvania Washington, D.C. 20006
Ave., N.W. (202) 822-8950
Washington, D.C. 20006 e
(202) pe, ei Counsel of Record
WiiuraM J. Sirio, PHD.
PENNIE & EDMONDS
1155 Avenue of the Americas
New York, NY 10036
(212) 790-9090
COCKLE LAW BRIEF PRINTING CO., (800) 225-6964
OR CALL COLLECT (402) 342-2831
INDEX TO APPENDIX
VOLUME i OF II
Opinion, Tanabe Seiyaku Co., Ltd. v. United States
International Trade Comm'n, No. 95-1448 (Fed.
Se ee Fp Re ic 6k de Ranke e eee cee ndebens App.
Order, Tanabe Seiyaku Co., Ltd. v. United States
International Trade Comm'n, No. 95-1448 (Fed.
Cor. Tee FU Ge Bee i no's cece. cht vaneukees App.
Commission Opinion, In the Matter of Certain
Diltiazem Hydrochloride and Preparations, Interna-
tional Trade Comm'n, Investigation No. 337-
TA-349 (filed June 8, 1995 ITC)............. App.
Initial Determination (Public Version), In the Mat-
ter of Certain Diltiazem Hydrochloride and Prepara-
tions, International Trade Comm’n,
Investigation No. 337-TA-349 (filed Feb. 14,
i Bn 8 BRR RAs tag iar ag cn am seg eae rere App.
VOLUME II OF II
Initial Determination (Public Version), In the Mat-
ter of Certain Diltiazem Hydrochloride and Prepara-
tions, International Trade Comm’n,
Investigation No. 337-TA-349 (filed Feb. 14,
BOGS TER Oe ks oe si vec eee a gia seesseannvees App
19 U.S.C. § 1337(a) (relevant portions) ........ App.
Warner-Jenkinson v. Hilton Davis Chem. Co., 117
Se SOE POE Da oe ich ks hos tan oka es es App
Page
20
22
33
ae )
. 435
App. 251
Ill. INFRINGEMENT
A. The Profarmaco Process Does Not Infringe
Claim 1 of the ‘035 Patent
1. The Development of the Profarmaco Pro-
cess
FF CP 1. In approximately late 1982, Profarmaco
began work to synthesize diltiazem. Piselli, Tr. 1966.
FF CP 2. Using the German counterpart to the ‘257
patent, one of the Profarmaco scientists, Dr. Piselli, ran
several experiments involving the N-alkylation step. Pis-
elli, Tr. 1967-1969, 1998.
FF CP 3. In these, he used sodium hydride and
anhydrous (“DMF”) to become more familiar with the
N-alkylation of TZP. Piselli, Tr. 1967-1969, 1998-99.
FF CP 4. Knowing that sodium hydride is unaccept-
able for commercial scale synthesis, Dr. Piselli almost
immediately tried potassium carbonate and DMF. Piselli,
Tr. 1967-69.
FF CP 5. The potassium carbonate/DMF combina-
tion - which was the first one that Dr. Piselli tried - was
selected based on a 1978 article by Professor Makosza, an
organic chemist known as the “inventor of phase trans-
fer,” which specifically disclosed the use of potassium
carbonate and DMF in similar reactions. Piselli, Tr.
1967-69; RX 3025.
FF CP 6. The Makosza article described the possi-
bility of replacing the reagents described in the ‘257 pat-
ent with potassium carbonate and DMF. Piselli, Tr. 1967;
RX 3025.
App. 252
FF CP 7. The article specifically described the
advantages of potassium carbonate/DMF over sodium
hydride, including the eliminations of potentially dan-
gerous reactions caused by anhydrous organic solvents.
Piselli, Tr. 1968; RX 3025.
FF CP 8. Dr. Piselli had previously used such a
system at Profarmaco and he therefore followed
Makosza’s suggestions and tried potassium carbonate/
DMF in his first experiments. Piselli, Tr. 1969.
FF CP 9. This potassium carbonate/DMF process -
the first one that Dr. Piselli tried — was successful. Piselli,
Tr. 1969.
FF CP 10. Within two months, Dr. Piselli had devel-
oped an industrial process using potassium carbonate/
DMF. Piselli, lr. 1969.
FF CP 11. Profarmaco used this process for produc-
ing bulk diltiazem from approximately mid-1983 to July
15, 1986. Piselli, Tr. 1970; RPX 4026.
FF CP 12. In order to increase the consistency of the
yield, Profarmaco experimented with the addition of
water to the reaction and found that [C] % by volume of
water caused more consistent yields. During the summer
of 1986, Profarmaco therefore modified its process to
include the addition of [C] % water to its potassium
carbonate/DMF process. Piselli, Tr. 1969-1970, 2001-2002;
RPX 4026. :
FF CP 13. Shortly after December 30, 1986, Profar-
maco first learned from a French pharmaceutical firm,
App. 253
Sanofi, of the Tanabe European Patent Application corre-
sponding to the ‘035 patent. Russolo, Tr. 1924; Piselli, Tr.
2001; RX 3930C.
FF CP 14. This was the first time anyone at Profar-
maco became aware of the existence of the ‘035 patent or
any of its counterparts. Russolo, Tr. 1924; Piselli, Tr. 1970,
2001.
FF CP 15. Profarmaco reviewed this patent applica-
tion and after such review concluded that its potassium
carbonate/DMF process did not infringe. Russolo, Tr.
1934.
FF CP 16. Profarmaco therefore continued using
this process for five additional years. Russolo, Tr. 1934.
FF CP 17. In April 1989, after expiration of the ’257
patent and with the end of MMD’s Waxman-Hatch exclu-
Sivity on the horizon, Gyma, Profarmaco’s exclusive
agent in the United States, wrote to MMD requesting
disclosure of any process patents which MMD contended
might cover processes for the manufacture of diltiazem.
RX 3947C at 214-215; Russolo, Tr. 1924-25.
FF CP 18. MMD responded shortly thereafter by
identifying four patents, including the ‘035 patent. RX
3947C at 216; Russolo, Tr. 1924-25.
FF CP 19. Gyma forwarded MMD’s process patent
disclosure letter to Profarmaco for review. Russolo, Tr.
1924-25.
FF CP 20. Profarmaco, after reviewing the ‘035 pat-
ent, continued to use its potassium carbonate/DMF pro-
cess. Russolo, Tr. 1934.
App. 254
Receipt by Profarmaco of the
EPO Comparative Test Report
FF CP 21. On June 13, 1991, Profarmaco received
from its Italian patent attorneys (in connection with an
inquiry from Profarmaco on a different matter) Tanabe’s
October 1, 1984 submission to the European Patent Office,
including the Comparative Test Report. Russolo, Tr.
1925-27; RX 4032-C.
FF CP 22. Dr. Russolo, Profarmaco’s Managing
Director and General Manager, testified that, Profarmaco
is a conservative company, and immediately decided to
ascertain whether it could develop a process using a base
not specified in the ‘035 patent claims and, particularly,
bases and solvents that Tanabe had expressly represented
to the EPO not to be the subject of its invention. Russolo,
Tr. 1916-1917, 1925-1927.
FF CP 23. On June 27, 1991, two weeks after receipt
of the Comparative Test Report, Profarmaco held an R&D
meeting attended by, among others, Drs. Russolo and
Piselli. Russolo, Tr. 1927-29; Piselli, Tr. 1970-71; RX 3928-
aa
FF CP 24. At that meeting, Dr. Piselli was directed
to try to develop an N-alkylation process using sodium
carbonate as the base. Russolo, Tr. 1927-29; Piselli, Tr.
1970-71.
FF CP 25. Specifically, the meeting minutes state:
“try the attachment of the chlorobase [i.e., 2-dimet-
hylaminoethyl-chloride (“DMC”)] with sodium carbo-
nate/DMF with different percentages of water.” RX 3928-
C; Russolo, Tr. 1928-29; Piselli, Tr. 1970-71.
App. 255
FF CP 26. Sodium carbonate was chosen as a target
base because Tanabe had identified the base in the Com-
parative Test Report as being outside the scope of its
invention. Russolo, Tr. 1928-29.
FF CP 27. Profarmaco therefore viewed the use of
sodium carbonate as a “zero-risk situation by using what
inventors were saying not to be part of the invention.”
Russolo, Tr. 1926-1927.
FF CP 28. DMF was identified because that was the
solvent that Profarmaco was then using in its current
potassium carbonate/DMF process which was “a very
good process.” Russolo, Tr. 1928-29; Piselli, Tr. 1973.
FF CP 29. Four days after this R&D meeting, on July
1, 1991, Dr. Piselli conducted the first experiment using
sodium carbonate in the period following receipt of the
Comparative Test Report. Piselli, Tr. 1972; RX 3926C; RX
3932C.
FF CP 30. Dr. Piselli used sodium carbonate and
DMF with C percent water. RX 3926 at Prof 436; Russolo,
Tr. 1929-30; Piselli, Tr. 1972.
FF CP 31. During approximately the next eight
months, Profarmaco was able to develop a new process
for N-alkylating TZP using sodium carbonate as the base.
Piselli, Tr. 1971-1972; RX 3926C; RX 3932C.
re Cr 32. in developing this process, Profarmaco
conducted approximately 100 experiments. Piselli, Tr.
1972; RX 3932C.
FF CP 33. Profarmaco experimented by including
and not including a phase transfer catalyst, by conduct-
ing the reaction at various temperatures, by varying
App. 256
times, by using various solvents, and by adding or
removing water. Piselli, Tr. 1973-74; RX 3932-C.
FF CP 34. Experimental evidence provided by com-
plainants in this investigation shows that water removal
is not critical in order to get a low yield from the Profar-
maco process. Taber Tr. 2094-95.
FF CP 35. Profarmaco discovered during the course
of these experiments that only by removing water
(through azeotropic distillation) could Profarmaco
achieve an industrially valid high-yield process. Piselli,
Tr. 1975.
FF CP 36. Dr. Piselli characterized the removal of
water as very important and essential. Piselli, Tr. 1975.
FF CP 37. If water is not removed from the current
Profarmaco process, the reaction is “never complete” and
there are by-products and impurities. Piselli, Tr. at 1975.
FF CP 38. Profarmaco also discovered that if the
temperature of the reaction is less than [C] °[C], then the
N-alkylation reaction cannot be completed. Piselli, Tr. at
1976.
FF CP 39. During the next eight months, Profar-
maco scientists conducted approximately 100 experi-
ments with different base/solvent combinations, and by
February 1992, determined to use sodium carbonate and
toluene, a base and a solvent, both of which were
expressly identified by Tanabe in the Comparative Test
Report as not included within its invention. Russolo, Tr.
1930; RX 3928C at 312, 315.
App. 257
FF CP 40. On March 6, 1992, the sodium carbonate
and toluene process went to the Pilot Plant. RX 3016C at
Prof 166.
FF CP 41. By June 4, 1992, that process had been
prepared for production and was ready for use. Piselli, Tr.
1978; RX 3928C at 316.
The Profarmaco Sodium Carbonate/Toluene Process
FF CP 42. Profarmaco conducts its process for man-
ufacturing bulk diltiazem in a [C] reactor vessel with a
volume of [C] cubic meters. Piselli, Tr. 1979-80.
FF CP 43. The reactor vessel contains a distillation
column, [C], and a variety of other equipment. The distil-
lation device which allows for azeotropically distilled
vapors to be cooled, condensed, and then either removed
from the system, or returned to the reactor vessel is ——
known as a Markusson trap. Piselli, Tr. 1978-1980; RX
3996.
FF CP 44. In the step immediately preceding
N-alkylation, Profarmaco carries out the [C]. Profarmaco
first charges [C]. Profarmaco then [C]. Following this
step, Profarmaco allows the contents of the reactor vessel
[C], which results in the formation of two phases: a lower
aqueous phase and an upper phase containing [C]. The
Profarmaco operator, following the separation, [C]. Fol-
lowing this procedure, the operator causes [C], thus
removing any last traces or droplets of water which may
have adhered to the sides of the reactor vessel. Any water
that is gathered as a resuit of [C] is then discharged by
the operator [C]. Piselli, Tr. 1979-1980; RX 3996(a).
App. 258
FF CP 45. The next step is the N-alkylation. That
process is carried out in the same reactor vessel. The
reactor vessel already contains DMC free base in a
toluene solution. To that solution Profarmaco adds
sodium carbonate which, by its specification, may not
contain more than [C]% water by weight. It also adds
TZP, which is prepared at Profarmaco, and which is
heated by Profarmaco to remove all water. Piselli, Tr.
1980-82, 1986; RX 3996(b).
FF CP 46. Once the TZP and sodium carbonate have
been added to the toluene solution containing DMC base,
the operator heats the reactor vessel as quickly as possi-
ble using the maximum amount of steam flowing through
the jackets surrounding the reactor vessel. At [C]°, the
operator reduces the steam flow so that the inside tem-
perature will reach about [C]°[C] without the reactor’s
contents overflowing. Through thermal inertia, the reac-
tion mixture increases in temperature to approximately
[C]° and the reaction mixture is then heated to [C]°. It
takes [C] for the reaction mixture to reach [C]°;
[C] for the reaction mixture to reach [C]°; and the
reaction mixture is then heated at a [C]° for [C] hours.
Piselli, Tr. 1982-83; RX 3996(b).
FF CP 47. At the [C]° range, the water/toluene solu-
tion begins to distill azeotropically. Profarmaco begins to
see carbon dioxide evolution at approximately [C]° and
begins to see water collecting in the Markusson trap also
at approximately [C]°. Because water is heavier than
toluene, the water collects in the Markusson trap while
the toluene returns to the reaction vessel. Piselli, Tr.
1983-84.
App. 259
FF CP 48. Profarmaco observed in the R&D labora-
tory a relationship between carbon dioxide evolution and
N-alkylation. Profarmaco has observed that the N-alkyla-
tion reaction takes place while carbon dioxide evolution
is occurring. Piselli, Tr. 1984-85.
FF CP 49. Profarmaco takes five separate steps to
prevent water from entering the reactor vessel and to
remove water created during N-alkylation step. RX 4024-
C; RX 3996(a); RX 3996(b); Piselli, Tr. 1985-87, 1977-81.
Specifically:
1. during the [C] step, which immediately pre-
cedes N-alkylation, the Profarmaco operator
[C] and to settle into two phases, the lower
of which is water. That aqueous phase is
then discharged [C]. RPX 4024-C; RX
3996(a); Piselli, Tr. 1980.
2. The operator then causes [C].” These last
traces of water are then discharged [C] to
make sure that there is no water in the reac-
tor. RPX 4024(c); RX 3996(a); Piselli, Tr. 1980.
3. During the N-alkylation step, Profarmaco
adds to the reactor vessel (which already
contains a toluene solution containing [C])
TZP which Profarmaco has dried by heating.
Piselli, Tr. 1981-82; RX 4024-C; RX 3996(a).
4. [C]. Piselli, Tr. 1981-82; RPX 4024-C; RPX
3996(a).
5. [C] Piselli, Tr. 1985-86, 1978-79; RX P4024-C.
App. 260
2. Differences Between the Profarmaco Pro-
cess and the ‘035 Process
FF CP 50. The differences between the Profarmaco
process currently employed to manufacture bulk
diltiazem in the N-alkylation step and the processes
claimed in the 035 patent (and in the Examples contained
in the patent) include the following:
a) Profarmaco uses sodium carbonate as a
base; the ‘035 patent specifies potassium
carbonate and potassium hydroxide;
b) Profarmaco uses toluene as a solvent; the
‘035 patent uses acetone and lower alkyl
acetates, or mixtures of those solvents and
water;
c) according to complainants’ theory of the
case, the ‘035 patent process operates in the
presence of water, and calls for the optional
addition of water; Profarmaco’s process
requires that water be removed constantly
throughout the N-alkylation reaction and no
water is specifically added.
d) Profarmaco’s process is conducted at a tem-
perature of approximately [C]°; the ‘035 pro-
cesses are conducted at a maximum of 77°.
CX 1 (“035 patent, col. 2, line 62).
e) in the Profarmaco process, Profarmaco
arrives at a solution in toluene of the inter-
mediate; Profarmaco is therefore ready to
conduct the subsequent acetylation reaction
in the same reactor vessel using the same
reactants. By contrast, in the ‘035 process
the intermediate is isolated.
See Piselli, Tr. 1986-87.
ee er
alia a
App. 261
a. Solvents
FF CP 51. In comparing the solvent system of
toluene to either acetone or ethyl acetate, “there are cer-
tainly differences that one can point to in the properties
of these solvents.” Gokel, Tr. 698.
FF CP 52. According to complainant’s expert Dr.
Gokel, “the key difference” of the Profarmaco process
from the 035 process is the use of toluene as the solvent.
Gokel, Tr. 764.
FF CP 53. The information reported in Dr. Gokel’s
report entitled “Fermion and Profarmaco Versions of Tan-
abe Diltiazem Synthesis” reflects what Dr. Gokel “con-
structed to aid [his] thinking at an early stage” in the
present litigation. In determining the equivalence
between the Fermion and Profarmaco processes with the
‘035 process, Dr. Gokel considered many parameters
relating to the solvents used for the N-alkylation reaction.
One of the parameters that Dr. Gokel considered and
thought might influence his opinion was solvent polarity
parameters, while another was a comparison of the water
miscibilities of the different solvents. Dr. Gokel also listed
and considered dipole moments and dielectric constants
for the different solvents. Dr. Gokel summarized all of the
different values in a solvent table on pages 13-15 of his
report. CX 606; Gokel, Tr. 810-15.
FF CP 54. When looking at the solvent tables
included in Dr. Gokel’s report eniitled “Fermion and
Profarmaco Versions of Tanabe Diltiazem Synthesis,” Dr.
Gokel compared the relevant values of one solvent with a
App. 262
second solvent, as determined by the same methodology,
same tester, same equipment, etc. for the different sol-
vents. CX 606; Gokel, Tr. 814-15.
Toluene is Substantially Different in Structure From the
‘035 Carbonyl Solvents
FF CP 55. The predominant structural feature of each
of the solvents claimed in the ‘035 patent is the presence
of a carbonyl group, which is shown enclosed by the
dotted lines in the following formulas:
[Diagram Omitted In Printing]
RRX 3983; Taber, Tr. 2058-59.
FF CP 56. Toluene is an aromatic hydrocarbon whose
structure is illustrated by the formula:
[Diagram Omitted In Printing]
RPX 3987.
FF CP 57. Toluene contains neither a carbonyl group
nor any structure analogous to a carbonyl. Toluene is not
a ketone (like acetone) or an ester (like an alkyl acetate).
Taber, Tr. 2145; Gokel, Tr. 702.
FF CP 58. The oxygen atoms in the carbonyls of the
solvents claimed in the ‘035 patent have two unbonded
pairs of electrons which can be donated to positively
charged species (“cations”), such as potassium ions (K+),
as depicted below:
[Diagram Omitted In Printing]
RX 3983; Taber, Tr. 2059.
App. 263
FF CP 59. The second oxygen atom in an alkyl acetate
also possesses two pairs of unbonded electrons that can
be donated. RX 3983; Taber, Tr. 2059.
The ‘035 Carbonyl Solvents Are Good Donor Solvents
Whereas Toluene Is Not
FF CP 60. A donor solvent is a solvent which can
donate electron density to stabilize an electron deficient
species such as a cation. A donor solvent provides stabi-
lization to an electron deficient species, such as a sodium
or potassium cation, which are both electron deficient.
Gokel, Tr. 702. 3
FF CP 61. The ’035 carbonyl solvents are good donor
solvents. RPX 3984; Taber, Tr. 2060; Taylor, Tr. 2606.
FF CP 62. Toluene is a very poor donor solvent. RPX
3984; Taber, Tr. 2060, 2145.
FF CP 63. A donor solvent is “a material that has a
polar functional group that can solvate a Lewis acid
entity.” A Lewis acid entity would include a potassium
ion. Gokel, Tr. 802.
FF CP 64. Acetone and lower alky] acetates are stron-
ger donors than toluene. Gokel. Tr. 806.
FF CP 65. Because ‘035 carbonyl solvents can donate
electrons, particularly when they contain water, they are
able to solvate (or solubilize) and thus stabilize cations of
inorganic bases, such as the potassium bases of the ‘035
patent. Taber, Tr. 2058, 2059.
App. 264
FF CP 66. Potassium bases are more soluble in car-
bonyl solvents than are sodium bases. RX 4038C; Taber,
Tr. 2147; Kende, Tr. 1455.
FF CP 67. Because of its poor donorability, toluene
cannot effectively solvate (or solubilize) and thus stabi-
lize cations of inorganic bases, such as potassium ions or
sodium ions. RX 4034; Taber, Tr. 2061-2065.
FF CP 68. Sodium carbonate is not soluble in toluene.
Taber, Tr. 2093.
The ‘035 Carbonyl Solvents Possess Medium Polarity
Whereas Toluene Is Nonpolar
FF CP 69. The ‘035 carbonyl solvents, methyl acetate,
acetone and ethyl acetate range in donor number from
16.4 to 17.1. RPX 3984; Taber, Tr. 2060.
FF CP 70. The donor number for toluene is 0.1. RPX
3984; Taber, Tr. 2060.
FF CP 71. The ‘035 carbonyl solvents are more than
160 times better donors than is toluene. RPX 3984
FF CP 72. The ‘035 carbonyl solvents, methyl acetate,
ethyl acetate and acetone, are of medium polarity, having
dipole moments ranging between 5.7 and 9.0 and
dielectric constants ranging from 6.0 to 20.56. RPX 3985
FF CP 73. Toluene is a non-polar solvent, having a
dipole moment of 1.0 and dielectric constant of 2.38. RPX
3985; Taber, Tr. 2145; Gokel, Tr. 810.
FF CP 74. Because water is soluble in the ‘035 car-
bony] solvents, it increases the dielectric constant of the
‘035 carbonyl solvents. Taber, Tr. 2067.
i ia 2
App. 265
FF CP 75. Ionic species are solvated and stabilized
better by polar solvents than by non-polar solvents.
Gokel, Tr. 810.
FF CP 76. Toluene, being a non-polar solvent, lacks
the ability to dissolve inorganic bases. Taber, Tr. 2064-65.
The ‘035 Solvents and Water Are Substantially Soluble
Within Each Other Whereas Toluene and Water Are Not
FF CP 77. The ‘035 carbonyl solvents are substantially
soluble in water, and water is substantially soluble in
those solvents. RPX 3986; Taber, Tr. 2068.
FF CP 78. Acetone is infinitely soluble in water, and
water is infinitely soluble in acetone. RPX 3986
FF CP 79. Methyl acetate is very soluble in water
(approximately 23%), and water is very soluble in methyl
acetate. RPX 3986
FF CP 80. Ethyl acetate is soluble in water at 2.94%,
and water is soluble in ethyl acetate at 8.08%. RPX 3986
FF CP 81. Toluene is soluble in water at 0.052% and
water is soluble in toluene at 0.033%. RPX 3986
FF CP 82. The solubility of water in the “wet toluene”
used in the Profarmaco process is .03%. Taber, Tr. 2068.
FF CP 83. It is “well-known” that water is immiscible
in toluene. Taylor, Tr. 2607.
FF CP 84. The least soluble of the ‘035 carbonyl
solvents (ethyl acetate) is more than 50 times more sol-
uble in water than toluene, and water is more than 200
times more soluble in ethyl acetate than in toluene. RPX
3986.
App. 266
Water Can Solvate Inorganic Bases
FF CP 85. The solubility of water in the ‘035 carbonyl
solvents, and vice versa, contributes to the ability of the
‘035 carbonyl solvents to solvate (or solubilize) the
potassium bases disclosed in the ‘035 patent. Taber, Tr.
2145-46.
FF CP 86. One of ordinary skill in the art in 1981
would have known that the solubility of potassium salts
in acetone would be enhanced by the addition of water.
Taylor. Tr. 2606.
FF CP 87. Water stabilizes and thus makes more
soluble negatively charged hydroxide and carbonate ions
(“anions”) in a reaction solution. RPX 3994; Taber, Tr.
2057-58.
The ‘035 Carbonyl Solvents Are Both Hydrophilic and
Lipophilic Whereas Toluene Is Only Lipophilic
FF CP 88. The ‘035 carbonyl solvents possess both
hydrophilic (water-loving) and lipophilic (oil-loving)
properties. RPX 3993; Taber, Tr. 2058; 2068-69.
FF CP 89. Because the ‘035 carbonyl solvents have
both hydrophilic and lipophilic properties, they are able
to bring together in solution TZP, the inorganic base, and
DMC (in the form of aziridinium). Taber, Tr. 2146.
FF CP 90. Because the ‘035 carbonyl solvents have
both hydrophilic and lipophilic properties, the TZP, inor-
ganic base, and DMC all dissolve in the organic-aqueous
phase surrounding the inorganic base particles. Taber, Tr.
2083, 2146.
App. 267
FF CP 91. Toluene has strongly lipophilic properties
with little or no hydrophilic properties. RX 3993; Taber,
Tr. 2069, 2077, 2146.
FF CP 92. Because Toluene has very little hydrophilic
properties, most water included in the Profarmaco pro-
cess is associated with the surface of the sodium carbo-
nate base particle. Taber, Tr. 2146.
b. Bases
Sodium and Potassium Bases Are Not Interchangeable
in the ‘035 Process, but Are Interchangeable in Profar-
maco’s Process
FF CP 93. Sodium carbonate is not equivalent to the
potassium bases claimed in claim 1 of the ‘035 patent.
Taber, Tr. 2150.
FF CP 94. The carbonyl containing solvents of the
‘035 patent are known to be able to solvate at least to
some degree potassium. This same phenomenon is not
known, however, with sodium, at least not to the same
degree. Thus, especially given the teachings of the ‘035
patent, a person skilled in the art at the time of the
alleged ‘035 invention would have believed that the spec-
ific carbonyl containing solvents of the ‘035 patent were
linked with and were capable of providing some sol-
ubility only to potassium hydroxide and potassium car-
bonate. Taylor, Tr. 2604-05.
FF CP 95. Potassium salts are generally more soluble
in solvating organic solvents than are sodium salts.
Gokel, Tr. 844; Taylor, Tr. 2604-05.
App. 268
FF CP 96. The sodium cation is more charge dense
than the potassium cation. Gokel, Tr. 844.
FF CP 97. Whereas sodium carbonate does not work
in the same way as the potassium bases in the ‘035 patent,
sodium carbonate would work in the same way as
potassium carbonate in the Profarmaco process. This is
because sodium and potassium bases are often inter-
changeable when dealing with reaction solutions contain-
ing pure water, as in the aqueous phase surrounding the
inorganic base particle in the Profarmaco process. Taber,
Tr. 2146-47.
FF CP 98. Potassium salts and sodium salts are inter-
changeable in aqueous systems. Taber, Tr. 2117-18.
c. Reaction Mechanisms
The Profarmaco and ‘035 Processes Operate with Sub-
stantially Different Reaction Mechanisms
FF CP 99. The Profarmaco process proceeds with a
different mechanism than does the ‘035 process. Taber, Tr.
2111-12.
FF CP 100. The ‘035 process is one in which the
inorganic base particles (potassium base particles) are
surrounded by a solvent-water mixture, wherein the con-
centration of water is greatest at the surface of the parti-
cle and decreases with distance from the particle. Some
water is dissolved in the bulk organic phase. Taber, Tr.
2069-71; Gokel, Tr. 705.
FF CP 101. Complainants’ expert, Dr. Gokel, has car-
ried out no experiments and is unaware of any experi-
ments carried out by others, comparing a surface solvent
App. 269
phase formed in the ‘035 process with a surface solvent
phase formed in any of the Respondents’ processes.
Gokel, Tr. 1053.
FF CP 102. Dr. Baldwin would expect to find more
dissolved base in the solvent system of the ‘035 process
than he would in the toluene phase of the Profarmaco
process. RX 3048C.
FF CP 103. The pH of the carbonate buffer contained
in the aqueous phase of the Profarmaco process is the
same, whether sodium carbonate or potassium carbonate
is used as the base. Because a dramatic difference is
obtained in the ‘035 process when using a sodium base
instead of a potassium base, this indicates that the ‘035
reaction system is a mixed solvent system, comprising
water, organic solvents and potassium base. This mixed
solvent system is further evidenced by the knowledge
that potassium bases are more soluble than sodium bases
in the ‘035 carbonyl solvents, due to the ability of the
carbonyl solvents to solvate potassium ions more effi-
ciently than sodium ions. Taber, Tr. 2147.
FF CP 104. The ratio of water to organic solvent in
the ‘035 process is a gradient or continuum extending
outward from the solid base particles of the ‘035 patent.
Taber, Tr. 2069-71. In the ‘035 system, there exists a
“phase boundary” between the ethyl acetate and water
phases “which is on the ethyl acetate side more like ethyl
acetate; on the water side, more like water. And in the
middle there is a progression from one to the other.
Gokel, Tr. 705.
FF CP 105. Complainants’ expert Dr. Gokel “would
certainly expect” that the difference between toluene and
App. 270
ethyl acetate would alter the phase boundary present in
the respective systems. Gokel, Tr. 707.
FF CP 106. The TZP in the ‘035 process is deproto-
nated by carbonate or hydroxide ions and the resulting
amide anion reacts with the aziridinium ion to yield the
alkylated product. Taber, Tr. 2074-75.
FF CP 107. No direct experimental evidence exists
that the claimed N-alkylation process of the ‘035 patent
using potassium carbonate as a base is hydroxide-medi-
ated. Gokel, Tr. 1028. If anything, there are indications
that it is not hydroxide-mediated. Taylor, Tr. 2674.
FF CP 108. The particular base-solvent combinations
of the ‘035 patent result in the reactants coming together
in solution and thus allow the reaction to proceed at
relatively low temperatures with good yields. Kende, Tr.
1194.
FF CP 109. Complainants’ expert, Dr. Gokel, was
unsure whether the actual alkylating agent in either the
Profarmaco or the ‘035 processes is the aziridinium ion.
Specifically, the only thing Dr. Gokel knows is that in
both reactions some aziridinium ion is formed, but he
does not know whether the aziridinium ion is the actual
alkylating agent or not. Although he “think(s], it is rea-
sonable that it could be, . . . [he] can’t rule out the other
possibility.” That the alkylation of TZP occurs predom-
inately through the aziridinium ion would be a “guess”
to Dr. Gokel. Gokel, Tr. 849-51.
App. 271
FF CP 110. Dr. Gokel also agreed that the aziridinium
ion would be likely involved in the ’257 process, in addi-
tion to its likely involvement in both the ‘035 and Profar-
maco processes. Gokel, Tr. 850-51.
FF CP 111. Dr. Baldwin’s labeling experiments in
JEB1-JEB4 do not prove that the aziridinium ion is the
alkylating species. RX 3963; RX 4038C.
FF CP 112. Profarmaco’s expert testified that in the
Profarmaco reaction, the aziridinium ion acts as a phase
transfer agent between the thin water layer surrounding
the inorganic base particle and the bulk toluene phase.
Taber, Tr. 2081, 2109-12; RPX 4000, 4001.
Substantially Lower Amounts of Water Are Present in
the Profarmaco Process
FF CP 113. In the Profarmaco process, the amounts of
water present are much smaller than the amounts of
water present in the ‘035 process. Specifically, in Example
3 of the 035 patent, the amount of water associated with
potassium carbonate is about 1.1 moles of water per mole
of potassium carbonate. RX 1688 (Liotta Dep. Tr. 977-78).
In contrast, in the Profarmaco process wherein the water
concentration of the water-extracted toluene is [C]%, the
molar ratio of water to sodium carbonate is about [C]% or
[C] times less than in the ’035 process. RX 1688.
FF CP 114. The small amount of dissolved water in
the toluene of the Profarmaco process exists in aggregates
of molecules. Taber, Tr. 2077; RPX 4000.
App. 272
FF CP 115. The minuscule amounts of water in the
toluene associate with the surface of the sodium carbo-
nate base in the Profarmaco process, whereas in the pro-
cess of the ‘035 patent significant amounts of water are
dissolved in the carbonyl solvent and the water partici-
pates in the solvation and dissolution of the inorganic
potassium base. RPX 4000; Taber, Tr. 2069-71, 2076-78.
FF CP 116. No mention is made of azeotropic removal
of water in any of the examples of the ‘035 patent.
Instead, the ‘035 patent teaches in the examples that the
reaction is carried out under reflux conditions, meaning
that the vapors of solvent released from the reaction
mixture during boiling are condensed to a liquid in a
reflux condenser and returned to the reaction vessel.
Gokel, Tr. 904-06.
The Profarmaco Process Is More Like the ’257 Process
Than the ‘035 Process
FF CP 117. The Profarmaco process is much more like
the ‘257 process than the ‘035 process because where
reversible deprotonation of the TZP starting material
occurs in the ‘035 process, the TZP starting material in the
Profarmaco process is directly alkylated after deprotona-
tion occurs. Thus, no equilibrium (or reversibility) exists
in the Profarmaco process between the TZP starting mate-
rial and its anion. Taber, Tr. 2112-14; RPX 4001.
More Energy Is Required to Carry Out Profarmaco Pro-
cess
FF CP 118. A higher amount of energy is required for
the reaction occurring in the Profarmaco process than for
App. 273
the reaction occurring in the process of the 035 patent, as
reflected by the higher reaction temperatures required for
obtaining good yields in the Profarmaco process. Taber,
Tr. 2109.
FF CP 119. The Profarmaco process is carried out at a
temperature of [C] °[C] versus 77°[C] or less in the ‘035
process. CX 1; Piselli, Tr. 1987.
Evolution of Carbon Dioxide Occurs in the Profarmaco
Process, But Not in the ‘035 Process
FF CP 120. Because of the higher reaction tempera-
tures necessary to carry out the Profarmaco process, car-
bon dioxide is evolved during the Profarmaco process,
thereby also producing sodium hydroxide. Taber, Tr.
2088.
FF CP 121. Nothing in the ‘035 patent indicates that
carbon dioxide is evolved during the ‘035 process. Taber,
Tr. 2088; Gokel, Tr. 1043-1044.
FF CP 122. The fact that the evolution of carbon
dioxide in the Profarmaco process coincides with the
production of the product indicates that the hydroxide
ion formation, which occurs simultaneously with carbon
dioxide evolution, is important to the Profarmaco pro-
cess. Taber, Tr. 2089.
Different pH Levels Exist in Profarmaco and ’035 Pro-
cesses
FF CP 123. While the pH in the ‘035 process drops
from an initial pH of 11.5 to 8.5, the pH in the Profarmaco
App. 274
process is maintained at a minimum level of 11.5. Taber,
Tr. 2088-89.
FF CP 124. One pH interval level represents a differ-
ence in hydroxide ion concentration of a factor of 10.
Thus, there is 1000 times more hydroxide ion present in
the Profarmaco process than in the ‘035 process. Taber, Tr.
2089.
FF CP 125. The Profarmaco process operates at a
much higher pH level than the process of the ‘035 patent,
due to the higher hydroxide ion concentration and signif-
icantly lower amount of water present in the Profarmaco
process. Taber, Tr. 2089-90.
d. Experimental Evidence Demonstrates
That the Profarmaco Process Works in A
Substantially Different Way Than the ‘035
Process
FF CP 126. At least four different sets of experiments
have been made of record in the present investigation
demonstrating that when the solvent toluene is substi-
tuted for the ‘035 carbonyl] solvents in the ‘035 process the
reaction proceeds very differently: (1) the experiments
underlying the EPO Comparative Test Report submitted
by Tanabe during the prosecution of the European appli-
cation corresponding to the ‘035 patent; (2) experiments
conducted by Tanabe scientists in the early 1980's; (3)
experiments conducted by complainants’ expert Dr. Bald-
win; and (4) experiments conducted by Profarmaco. RX
3929, 3494, 3963, 3936; Taber, Tr. 2120-23, 2131-2145.
App. 275
FF CP 127. At least four different sets of experimental
data demonstrate that sodium carbonate reacts quite dif-
ferently than potassium carbonate in the ‘035 process: (1)
the EPO Comparative Test Report submitted by Tanabe
during the prosecution of the European application corre-
sponding to the ‘035 patent; (2) research reports by the
‘035 inventors; (3) experimental tests by complainants’
expert Dr. Baldwin; and (4) experiments conducted by
Profarmaco. RX 3929, 3361, 3963, 3936; Taber, Tr. 2119-23,
2131-2145.
Comparative Test Report
FF CP 128. In Experiment No. 1 of Table 1, a yield of
86.2% was reported when using potassium hydroxide and
acetone at a reaction temperature of 50°[C] for a reaction
period of 7 hours. RX 3929C. The reaction temperature of
50-60°[C] was within the range described in the ‘035
patent specification. RX 3048C.
FF CP 129. In Experiment No. 2 of Table 1, a yield of
94.5% was reported when using potassium carbonate and
acetone under reflux conditions for a reaction period of 9
hours. RX 3929C.
FF CP 130. In experiment No. 3 of Table 1, a yield of
90.2% was reported when using potassium carbonate and
ethyl acetate under reflux conditions for a reaction period
of 23 hours. RX 3929C.
FF CP 131. In Experiment No. 4 of Table 1, a yield of
90.7% was reported when using potassium carbonate and
acetone/water under reflux conditions for a reaction
period of 3 hours. RX 3929C.
App. 276
FF CP 132. In Experiment No. 5 of Table 1, a yield of
92.7% was reported when using potassium carbonate and
ethyl acetate/ water under reflux conditions for a reaction
period of 6 hours. RX 3929C.
FF CP 133. In Experiment No. 6 of Table 1, a yield of
87.3% was obtained when using potassium carbonate and
methyl acetate/water under reflux conditions for a reac-
tion period of 30 hours. RX 3929C.
FF CP 134. The base and solvent systems, reactions
temperatures, reaction times and yields obtained in
Experiment Nos. 1-6 in Table 1 of the EPO Comparative
Test Report correspond with the bases and solvents, reac-
tion conditions and yields obtained in Examples 1, 4, 5, 2,
3 and 7, respectively, of the ‘035 patent. CX 1; RX 3929-C.
FF CP 135. In Experiment No. 8 of Table 1, the yield
of product was reported as “no reaction” when potassium
hydroxide and toluene were used at a reaction tempera-
ture of 50-60°[C] for a reaction period of 7 hours. RX
3929C. The reaction temperature of 50-60°[C] was within
the range described in the ‘035 patent. RX 3048C.
FF CP 136. In Experiment No. 11 of Table 1, a yield of
10% was reported when using sodium hydroxide and
acetone under reflux conditions for a reaction period of
15 to 20 hours. RX 3929C.
FF CP 137. In Experiment No. 12 of Table 1, the yield
of product was reported as “no reaction” when sodium
carbonate and acetone were used under reflux conditions
for a reaction period of 15 to 20 hours. RX 3929C.
FF CP 138. Where “no reaction” is reported in Experi-
ments 8 and 12 of the Comparative Test Report, Tanabe
paw eee ee
App. 277
did not necessarily mean zero yield but rather meant a
poor yield. Taber, Tr. 2281.
FF CP 139. Complainants’ expert, Dr. Gokel, pro-
vided as an explanation for why a 86.2% yield was
obtained in Experiment No. 1 (KOH / Acetone) of Table 1,
whereas only a 10% yield was obtained in Experiment
No. 11 (NaOH/Acetone) of Table 1, was that “[i]t can be
explained by the fact this is sodium hydroxide and that’s
potassium hydroxide. . . . ” Another possible explanation
for why different yields were obtained in Experiments
Nos. 1 and 11 is that in a theoretical sense there exists a
difference in the solubilities of potassium hydroxide and
sodium hydroxide, which thus could account for the dif-
ferences in the reaction yields obtained. Dr. Gokel recog-
nized this to be a “well-known phenomenon.” Gokel, Tr.
845-47.
Experiments Performed by Tanabe Scientists
Toluene Did Not Work in ‘035 Process
FF CP 140. In a Tanabe technology department
report, dated October 1981 (approximately two months
prior to the December 1981 date of the Japanese priority
patent application upon which the ‘035 patent is based),
Tanabe scientists reported that the N-alkylation of TZP at
a reaction temperature of [C] °[C] using [C] as the base
and [C] as the solvent did not work. Tanabe repeated the
reaction several times, varying the reaction temperature
and amount of water added, but were unable to obtain an
appreciable product. Taber, Tr. 2136-37; Gokel. Tr. 779-84-
RX 3494.
App. 278
FF CP 141. Dr. Gaino, one of the co-inventors named
on the face of the ‘035 patent, reported in his notebook
that N-alkylating TZP using [C] as the base and [C] as the
solvent failed to work. Taber, Tr. 2138-39; RX 3368C.
FF CP 142. The Tanabe research reports reflect the
Tanabe scientists’ finding that toluene is not a useful
solvent for the ‘035 process. Taber, Tr. 2139.
Sodium Carbonate Did Not Work in the ‘035 Process
FF CP 143. Tanabe performed experiments about
three months prior to the December 1981 date of the
Japanese priority patent application upon which the ‘035
patent is based wherein [C] and [C] were substituted for
[C] as the base for N-alkylating TZP in [C] . When using
either [C] or [C] , tanabe scientists were unable to make
the N-alkylation reaction work. The experiment using
sodium carbonate and acetone corresponded with Experi-
ment No. 12 in Table 1 of the European Comparative Test
Report using sodium carbonate and acetone wherein “no
reaction” is reported. Thus, Experiment 12 may be based
on this test. RX 3362C, RX 3361C; Taber, Tr. 2123, 2131-35.
FF CP 144. Tanabe Research Reports reflect the find-
ing of Tanabe scientists that sodium carbonate is not
useful as a base in the ‘035 process. Taber, Tr. 2139.
Experiments Performed by Complainants’ Expert, Dr.
Baldwin, During the Investigation
FF CP 145. Experiments carried out by Complainant's
expert Dr. Baldwin demonstrate that toluene does not
Se TE ci aR ee e
App. 279
work as a solvent in the ‘035 process. Taber, Tr. 2143-44;
RX 3963, RPX 3991C.
FF CP 146. Experiments carried out by Complainant's
expert Dr. Baldwin demonstrate that sodium carbonate
does not work efficiently as a base in the ‘035 process, as
that process is taught in the patent's examples. Taber, Tr.
2143-2144; RX 3963, RPX 3991C.
FF CP 147. In Experiment JEB15, which was designed
to simulate a process of the ‘035 patent in which toluene
was interchanged for a solvent of the ‘035 patent, a yield
of 24% was obtained when N-alkylating TZP with
potassium hydroxide as the base and toluene as the sol-
vent at a reaction temperature of 111°[C]. RX 3963; RX
3048C.
FF CP 148. In Experiment JEB16, which was designed
to simulate a process of the ‘035 patent in which toluene
was interchanged for a solvent of the ‘035 patent, a yield
of 56% was obtained when N-alkylating TZP with
potassium carbonate as the base and toluene as the sol-
vent at a reaction temperature of 111°[C]. RX 3963; RX
3048C.
FF CP 149. In Experiment JEB17, the yield of product
obtained dropped from 90.7% to 35% when sodium car-
bonate was substituted for potassium carbonate under
some reaction conditions of Example 2 of the ‘035 patent.
RX 3963; RPX 3991C.
FF CP 150. In Experiment JEB18, the yield of product
obtained dropped from 92.7% to 65% when sodium car-
bonate was substituted for potassium carbonate under
App. 280
some reaction conditions of Example 3 of the ‘035 patent.
RX 3963; RPX 3991C.
FF CP 151. In Experiment JEB19, the yield of product
obtained dropped from 90.7% to 10% when sodium car-
bonate was substituted for potassium carbonate under
some reaction conditions of Example 2 of the ‘035 patent.
RX 3963; RPX 3991C.
FF CP 152. In Experiment JEB20, a yield of 97% was
obtained whe. sodium carbonate was substituted for
potassium carbonate under some reaction conditions of
Example 3 of the ‘035 patent, but only after heating the
reaction mixture at reflux temperature for 23 hours
(almost 4 times the reaction time in Example 3). RX 3963;
RPX 3991C.
FF CP 153. Experiment No. 12 in Table 1 of the
Comparative Test Report, in which “no reaction” was
reported for an N-alkylation reaction using sodium car-
bonate and acetone, is consistent with the 10% yield that
the complainants’ expert Dr. Baldwin obtained in JEB 19,
wherein sodium carbonate and acetone also were used.
Taber, Tr. 2281.
FF CP 154. Experiment No. 8 in Table 1 of the Com-
parative Test Report, in which “no reaction” was reported
for an N-alkylation reaction using potassium hydroxide
and toluene, is consistent with the 24% yield that the
complainants’ expert Dr. Baldwin obtained in JEB 15,
wherein potassium hydroxide and toluene also were
used. Taber, Tr. 2281.
App. 281
FF CP 155. Dr. Baldwin’s experiments also demon-
strated the importance of water removal during the Pro-
farmaco process. RPX 3992; RX 3963.
FF CP 156. In Experiment JEB2 (which sought to
mimic the Profarmaco process), Dr. Baldwin’s assistants
failed to follow his instruction that steps be taken to
remove water during the reaction. RX 4038C; Gokel, Tr.
866-67.
FF CP 157. Without taking steps to remove water in
JEB2, a yield of only 32% was obtained. RX 3963.
FF CP 158. When Dr. Baldwin repeated Experiments
JEB2 with azeotropic water removal, the yield increased
from 32% to 98%. RX 3963.
FF CP 159. Complainants’ expert, Dr. Gokel, agreed
that “an effort was made to remove water” during Dr.
Baldwin’s repeat of JEB2, including transferring the reac-
tion mixture to a clean Wheaton vile after the neutraliza-
tion step, as well as using a heat gun to heat the
distillation head to ensure that any water adhering to its
walls was driven over into the condenser. In addition, a
clean condenser was attached to the reaction system prior
to completing the reaction. The effort made to remove
water in the repeat of Experiment JEB2 was consistent
with Profarmaco’s effort to remove water during its pro-
cess by azeotropic distillation. Gokel, Tr. 867-69.
FF CP 160. No attempt also was made to remove
water azeotropically during Experiments JEB 15 and JEB
16. Gokel, Tr. 801; RX 3048C.
FF CP 161. Complainants’ expert, Dr. Gokel, had no
ideal whether Dr. Baldwin’s Experiment Nos. JEB1-JEB20
App. 282
had been optimized; in other words, they may have been
or they may not have been. Gokel, Tr. 616-617.
FF CP 162. If Dr. Baldwin’s experiments were not
already optimized, they could have been optimized if
complainants’ counsel chose to have it done. Gokel, Tr.
1113-15.
Experiments Performed by Profarmaco During Investi-
gation
FF CP 163. During the course of this investigation,
Dr. Piselli of Profarmaco conducted certain experiments.
Those experiments are summarized at page 703 of RX
3936. Trial Tr. at 1987-88.
FF CP 164. In Experiments 1-3, Dr. Piselli repeated
Example 3 of the ‘035 process. These experiments were
run in triplicate. Piselli, Tr. 1988; RX 3936.
FF CP 165. The yields obtained by Dr. Piselli in these
three repetitions of Example 3 of the ‘035 patent were
virtually identical to the yield indicated in the ‘035 patent
itself. Similarly, the product produced, based on TLC
analysis and melting point range, appears to be identical
to that indicated in Example 3 of the ‘035 patent. Piselli,
Tr. 1988; RX 3936; CX 1.
FF CP 166. In Experiments 4-6, Dr. Piselli used Exam-
ple 3 of the ‘035 patent as a starting point for three
experiments, in which he substituted sodium carbonate
for potassium carbonate. Dr. Piselli ascertained that after
a period of time that was slightly longer than that speci-
fied in Example 3 of the ‘035 patent, each of the three
App. 283
experiments provided a low yield. Piselli, Tr. 1989-1990;
RX 3936.
FF CP 167. In Experiment 4, Dr. Piselli therefore
extended the reaction time to 15 hours, and in Experi-
ment 6 extended the reaction time to 30 hours and
changed a number of other factors. In each instance, the
yield remained low. Piselli, Tr. 1989; RX 3936.
FF CP 168. Also, in Experiments 5 and 6, the purity of
the product obtained was poor, as characterized by a
“NEG” indication in the “Purity by TLC” column. RX
3936 at 703.
FF CP 169. Similarly, the melting points of the prod-
uct obtained in Experiments 5 and 6 were significantly
lower than the melting range for the product obtained by
a simple replication of the Example 3 of the ‘035 patent.
Piselli, Tr. 1988-1990; RX 3936 at 703.
FF CP 170. With respect to Experiments 4-6, Dr. Pis-
elli testified that the yields reflected in those experiments
“is not a process. It’s something that should be aban-
doned.” Piselli, Tr. 1990.
FF CP 171. In Experiments 7 and 8, Dr. Piselli repli-
cated Example 4 of the ‘035 patent. Piselli, Tr. 1990; RX
3936.
FF CP 172. The yields and quality obtained in Experi-
ments 7 and 8 compare favorably with the yields and
quality reflected in Example 4 of the ‘035 patent. Piselli,
Tr. 1990-1991; RX 3996 at 703.
FF CP 173. In Experiments 10 and 11, Dr. Piselli
replicated Example 4 of the ‘035 patent, except that he
App. 284
substituted sodium carbonate for potassium carbonate.
Piselli, Tr. 1991; RX 3936.
FF CP 174. Experiments 10 and 11 produced a low
yield and a poor quality product, as reflected by the
“neg” comment in the purity by TLC column. These
results were not improved by continuing the reaction for
18 hours. Piselli, Tr. 1991; RX 3936 at 703.
FF CP 175. Experiments 4-6 and 10-11 demonstrate
that sodium carbonate is not a useful base in the ‘035
process. Taber, Tr. 2142.
FF CP 176. In Experiment 15, Dr. Piselli replicated
Example 3 of the ‘035 patent except that he substituted
toluene as the solvent in the place of ethyl acetate. Dr.
Piselli used the reflux temperature of ethyl acetate, as
used in the patent example. The yield was extremely low
and the product was not pure. Piselli, Tr. 1991-1992, 2031;
RX 3936 at 703.
FF CP 177. In Experiment 16, Dr. Piselli replicated
Example 3 of the ‘035 patent except that he substituted
toluene for acetone. There was no yield in this reaction at
all. Dr. Piselli used the boiling temperature of acetone.
Piselli, Tr. 1992, 2031; RX 3936 at 703.
Optimization Defined
FF CP 178. Optimization is not the same thing as
experimentation. A process is optimized only after it has
been found to be a “consistent” process, specifically a
process from which “a well-defined product with a well-
defined yield” is obtained. It is only at this point in time
6 OE ES OPS Oa
App. 285
that “technological optimization” (i.e., trying “small vari-
ations in the operating parameters”) of a process is con-
ducted. For example, if an experiment using the same
reactants and reaction conditioner was repeated ten
times, any variation in yield would be minimal. Piselli, Tr.
2017-18, 2021.
FF CP 197. Dr. Piselli testified that one could not
“optimize” the Profarmaco experiments: “Optimization is
one thing, the process is another. You optimize a process
that has a certain consistency and is valid, but if the
process doesn’t exist, you don’t optimize it.” Piselli, Tr.
1990.
e. The Profarmaco Process Also Performs A
Substantially Different Function and
Achieves A Substantially Different
Result Than the ’035 Process
FF CP 180. Because claim 1 of the ‘035 patent does
not include a recovery step, the product of claim 1 is
N-alkylated TZP in a reaction mixture containing water, a
polar, water-miscible carbonyl solvent and various salts.
CX 1; Taber, Tr. 2183.
FF CP 181. The Profarmaco process produces a solu-
tion of N-alkylated TZP in toluene taber, TR. 2183-2184.
FFCP 182. The examples of the ‘035 patent teach that the
N-alkylated TZP obtained as the product of claim 1 of the
‘035 patent must be isolated, purified, and transferred to
another reaction vessel before the manufacture of
diltiazem can proceed. CX 1; Piselli, Tr. 1987.
App. 286
FF CP 183. The product of Profarmaco’s sodium car-
bonate/toluene N-alkylation process provides commer-
cial advantages, for example, convenience, unobtainable
using the product of the N-alkylation process claimed in
the claim 1 of the ‘035 patent. Taber, Tr. 2183-2185.
FF CP 184. Because toluene is immiscible with water,
the solution of N-alkylatedTZP in toluene produced from
the N-alkylation step of the Profarmaco process can be
directly washed with water to remove byproducts and
unreacted DMC, leaving behind a solution in which one
can directly carry out the subsequent acetylation reaction.
Taber, Tr. 2185.
FF CP 185. The function of the process of claim 1 of
the ‘035 patent is to produce an organic reaction mixture
containing N-alkylated TZP in a carbonyl solvent-water
mixture. This product will contain water, dissolved base
and salts along with alkylated TZP. To utilize the sol-
ubilized N-alkylated TZP, the reaction mixture must be (i)
extracted, (ii) washed, (iii) filtered, (iv) concentrated, (v)
redissolved, and (vi) transferred to another reactor prior
to the subsequently applied steps, including, inter alia,
the acetylation and salt-forming steps. RX 3348; Taber, Tr.
2183; Piselli, Tr. 1987.
B. The Abic Process Does Not Infringe Claim 1 of
the ‘035 Patent
1. The Abic Process Is Not Equivalent to Any
Process Disclosed Or Claimed By The ‘035
Patent
FF CA 1. The Abic process does not employ either of
the two bases or either of the two organic solvents or any
App. 287
of the five specific base-solvent combinations identified
in the ‘035 patent. See RX 1194; RX 1195.
FF CA 2. Abic’s commercial process uses barium
hydroxide octahydrate as a base, a biphasic solvent sys-
tem of methylene chloride and water, and tri-
ethylbenzylammonium chloride (TEBA) and is similar to
the process described in example 4 of its United States
Patent No. 4,466,995 (“the ‘995 patent”). RX 1701C.
FF CA 3. All the processes used by Abic to manufac-
ture diltiazem hydrochloride were disclosed in the Abic
DMF. [C]. RX 1701.
The Bases Are Not Equivalent
FF CA 4. Abic’s base is barium hydroxide octahy-
drate, and not potassium hydroxide or potassium carbo-
nate. RX 1194; Taylor Tr. 2638; RX 1701C.
FF CA 5. The ‘035 patent does not disclose or suggest
barium hydroxide to a person of ordinary skill. Taylor Tr.
2609.
FF CA 6. Barium is an alkaline earth metal, and as
such forms divalent cations. Taylor Tr. 2609.
FF CA 7. Among other things, barium hydroxide is
less soluble than potassium hydroxide in carbonyl sol-
vents such as acetone and lower alkyl acetates. Taylor Tr.
2610.
FF CA 8. If one of ordinary skill in the art were
investigating the interchangeability of other bases with
the potassium bases of the ‘035 patent, one would likely
first try sodium hydroxide (NaOH) because sodium
App. 288
hydroxide is more common and substantially less expen-
sive than potassium hydroxide. Taylor Tr. 2626.
FF CA 9. Tanabe tried and abandoned sodium
hydroxide in combination with DMSO. RX 1589C.
FF CA 10. As is shown by the Comparative Test
Report, Tanabe did try the base/solvent combination of
sodium hydroxide and acetone, and concluded that it was
not part of the method of the invention of the ‘035 patent.
Tr. 2636; RPX 1146; RX 1096; RX 1344; RX 3225.
FF CA 11. Barium hydroxide would be expected to be
less effective than sodium hydroxide in the ‘035 process
because barium is even less soluble than sodium in the
carbonyl solvents acetone or ethyl acetate of the ‘035
patent. Taylor Tr. 2627.
FF CA 12. Consequently, if sodium hydroxide were
found to be not as good as potassium hydroxide, one of
ordinary skill in the art would not be led to try barium
hydroxide, since barium hydroxide would be expected to
be even worse in the ‘035 process, which discloses solva-
tion of the solid base in a carbonyl solvent. Taylor Tr.
2627.
FF CA 13. Accordingly, one of ordinary skill in the
art, knowing that even sodium hydroxide was not inter-
changeable with potassium hydroxide would not have
expected that barium hydroxide would be interchange-
able with potassium hydroxide or potassium carbonate.
Taylor Tr. 2609-10.
App. 289
The Organic Solvents Are Not Equivalent
FF CA 14. The organic solvent in Abic’s process is
methylene chloride. RX 1195; Taylor Tr. 2641; RX 1701C.
FF CA 15. The ‘035 patent does not teach the use of
methylene chloride as an oroganic solvent to be used in
the N-alkylation of the ‘035 process. See Taylor Tr. 2612.
FF CA 16. The ‘035 patent disclosed chloroform, a
chlorinated hydrocarbon like methylene chloride, but did
not disclose its use or the use of any other chlorinated
hydrocarbon solvent in its N-alkylation process. CX 1;
Gokel Tr. 769-70.
FF CA 17. Acetone is a ketone. Taylor Tr. 2621.
FF CA 18. One of ordinary skill in the art looking to
investigate the scope of potentially interchangeable sol-
vents to replace acetone in the ‘035 process would have
looked for solvents which shared the important structural
and functional characteristics of the carbonyl solvents of
the ‘035 patent. i.e, would have looked at oxygen-con-
taining, cation-solvating, water-miscible solvents. Taylor
Tr. 2624.
FF CA 19. Some common solvents which one might
have investigated include methyl ethyl ketone, dioxane,
methanol, and DMSO. Taylor Tr. 2621-24; RPX 1155.
FF CA 20. Methylene chloride would not be one of
the solvents one would first try since it does not solvate
cations well, has no oxygen atoms to act as donors, and is
nearly totally immiscible with water. Taylor Tr. 2624.
FF CA 21. If solvents such as dioxane, methanol, and
methyl ethyl ketone were not as effective as acetone in a
App. 290
reaction, one would not be led to try methylene chloride,
since that would be going in the “wrong direction,” to
even more inferior water-immiscible solvents. Taylor Tr.
2624-25.
Tanabe Tried to Extend the Scope of It’s Invention in
1981 and Failed
FF CA 22. Tanabe tried sodium carbonate instead of
potassium carbonate, but it did not work as well as the
‘035 bases. RX 3368; Taber Tr. 2134-36.
FF CA 23. Tanabe tried structurally and functionally
similar solvents, such as dioxane, methyl ethyl ketone,
methanol, and DEMSO, and they didn’t work as well as
the ‘035 solvents. RX 3361, Gokel Tr. 928-31; RX 2046;
Gokel Tr. 676-79; RX 3368; Taber Tr. 2138; RX 1272C;
Taylor Tr. 2629-31.
FF CA 24. In fact, Tanabe tested and abandoned the
base-solvent combination of potassium carbonate-methyl
ethyl ketone (the combination currently employed by
fermiomion) because Tanabe could not get that combina-
tion to work. Taylor Tr. 2629; Gokel Tr. 676-79;
RPX-10161a; and RX-2046.
FF CA 25. Tanabe did not try barium hydroxide as a
base, methylene chloride as a solvent, or the combination
of barium hydroxide and methylene chloride. See Taylor
Tr. 2633; RX 1703.
App. 291
The Liquid-Solid Process of the ‘035 Patent and the
Abic Liquid-Liquid Process Are Not Equivalent
FF CA 26. Each of the organic solvent-water mixtures
of the 035 patent forms a single liquid phase, i.e., it is a
solution of water in the acetone or lower alkyl acetate
organic solvent. Taylor 1804, 2666-68.
FF CA 27. Abic’s solvent system is a biphasic solvent
system — it consists of two distinct liquid phases — meth-
ylene chloride and water. RX 1195; Liotta Tr. 1804-05;
Taylor Tr. 2664, 2666-67.
FF CA 28. There is no solid phase present in the Abic
process. Schwartz Tr. 2520-21.
FF CA 29. Liquid-liquid phase transfer catalyst pro-
cesses such as that used by Abic and solid-liquid pro-
cesses such as those of the’035 patent are not chemically
equivalent. Taylor Tr. 2667; RX-3969 at 108.
FF CA 30. Dr. Liotta wrote in his book that, even
when a solid-liquid process and a liquid-liquid process
are both phase transfer catalyzed, they are by no means
equivalent. RX 3969 at 108.
Abic’s Process Is Phase Transfer Catalyzed by TEBA
FF CA 31. Dr. Charles Liotta, complainants’ expert
with respect to infringement of the ‘035 patent by the
Abic process, submitted a declaration in this investiga-
tion in which he stated that:
Convincing evidence has not been presented by
Abic to date that (TEBA) is operating as a phase
App. 292
transfer catalyst in their N-alkylation process
step.
RX 1333C, { 6d.
FF CA 32. At his deposition, Dr. Liotta explained how
such evidence could be generated:
Q If Ihave two experiments, one where I have
a phase transfer catalyst present and one
where it is exactly the same but I leave out
the phase transfer catalyst, and the rate in
the one with the phase transfer catalyst is
faster than the one without it, does that
indicate that a phase transfer catalysis is
taking place?
A_ If you have repeated the experiments so you
have reproductibility and stirring speeds are
the same in both and everything was the
same, the indication is that you have evi-
dence for the operation of phase transfer
catalysis.
Liotta Tr. 1832-33.
FF CA 33. Abic carried out a series of such experi-
ments, as described by Dr. Liotta, which are identified as
experiments 1, 10, and 14-20 on RPX 105la. Taylor Tr.
2642-45.
FF CA 34. In experiments 1, 14, 17, 18, the N-alkyla-
tion was performed in the presence of TEBA catalyst;
experiments 10, 13, 15, 16, 19 and 20 did not use the
TEBA phase transfer catalyst. RPX 1051a.
FF CA 35. The TZP and DMC were allowed to react
in the methylene chloride-water biphasic solvent system
for 1.5, 3, 6, or 12 hours. RPX 1051la.
Te See ee es
Se I ae Oe Ee eee Ret Pep
2
a eee
App. 293
FF CA 36. Experiments 14 and 17, 15 and 16 are
repeats of each other, and the similar results indicate that
the results were reproducible. Taylor Tr. 2645-47.
FF CA 37. The stirring speeds were carefully con-
trolled, and were essentially the same in all experiments.
Schwartz Tr. 2524, 2532.
FF CA 38. The Abic experiments met the criteria set
forth by Dr. Liotta. Taylor Tr. 2650.
FF CA 39. The results were plotted on a graph which
shows that the phase-transfer catalyzed reaction was at
least twice as fast as the uncatalyzed reaction. Taylor Tr.
2647-49; RPX 1058.
The Abic Process Functions in a Different Way Than the
‘035 Process
FF CA 40. In the Abic liquid-liquid biphasic solvent
process, the TZP and DMC are in the organic phase (the
methylene chloride layer), while the barium hydroxide
remains dissolved in the aqueous phase. TEBA, the phase
transfer catalyst, is soluble in both phases. The
aziridinium ion which, as a cation, is insoluble in meth-
ylene chloride, remains in the aqueous layer. Taylor, Tr.
2638; RPX 1156.
FF CA 41. The hydroxide ion in the water layer
cannot efficiently deprotonate the TZP in the methylene
chloride layer, because they are in separate layers,
organic and aqueous. The TEBA phase transfer catalyst
carries the hydroxide ion as TEBA hydroxide into the
methylene chloride layer, where it deprotonates the TZP.
App. 294
The TZP anion can then react with DMC to N-alkylate the
TZP N-aryl amide. Taylor Tr. 2638; RPX 1099.
FF CA 42. The methylene chloride phase has an addi-
tional function in the Abic process over and above the
function of acetone in the ‘035 process. Taylor Tr. 2641.
FF CA 43. The water-immiscible solvent methylene
chloride keeps the TZP and DMC separate from dissolved
aqueous barium hydroxide. Taylor Tr. 2641
FF CA 44. Kugita I discloses that TZP can hydrolyze
in the presence of aqueous sodium hydroxide, especially
at higher temperatures. RX 3806; Kende Tr. 1452-54.
FF CA 45. Abic’s experience with biphasic toluena-
aqueous hydroxide processes is in accord; at higher tem-
peratures, hydrolysis of TZP was seen. RX 1007C; Haber
Tr. 2417-20.
FF CA 46. In the Abic process, the methylene chlo-
ride, operating at 40° C, protects the TZP and DMC from
hydrolysis by aqueous barium hydroxide. Taylor Tr. 2641.
FF CA 47. Because there is no aqueous hydroxide in
the ’035 solid-liquid potassium hydroxide/acetone pro-
cess, hydrolysis is not a problem in the ‘035 process.
Taylor Tr. 2666; Taber Tr. 2173.
Abic’s Process Is Not Equivalent to the Potassium
Hydroxide-Acetone Process of the ‘035 Patent
FF CA 48. Abic’s expert compared the potassium
hydroxide (KOH) /acetone system of the ‘035 patent and
the barium hydroxide/methylene chloride system of
Abic’s process. Taylor Tr. 2664-65; RPX 1157.
EE EE eee
App. 295
FF CA 49. The bases and solvents are different. Id.
FF CA 50. There is no water in the ‘035 potassium
hydroxide process; thee is a separate aqueous layer in the
Abic barium hydroxide process. Id.
FF CA 51. There is one solvent phase in the ‘035
process; there are two solvent phases in the Abic process.
Id.
FF CA 52. The alkylating agent in the ‘035 process is
aziridinium ion, and is located in the acetone single sol-
vent phase in the ‘035 process. CPX 14; RPX 1157; Taylor
Tr. 2664.
FF CA 53. In Abic’s process, the active alkylating
agent, DMC itself, is located in the methylene chloride
phase. Taylor Tr. 2665.
FF CA 54. In the ‘035 process, potassium hydroxide is
able to deprotonate TZP because it is somewhat soluble
in the acetone solvent. In the Abic process, barium
hydroxide, which is not soluble in methylene chloride, is
able to deprotonate TZP via phase transfer catalysis. Tay-
lor Tr. 2656, 2668.
FF CA 55. The potassium hydroxide in the ‘035 sys-
tem and barium hydroxide in Abic’s system do not func-
tion the same way to deprotonate TZP. Taylor Ir. 2665.
FF CA 56. The organic solvents in the ‘035 and the
Abic processes have different functions: the function of
acetone in the ‘035 process is to solubilize the KOH base
in the organic phase; the solvents in Abic’s process have
the important function of separating the reagents. Taylor
Tr. 2665-66.
App. 296
FF CA 57. The ‘035 process is a solid-liquid system
while the Abic process is a liquid-liquid system. Taylor Tr.
2666.
FF CA 58. The Abic process uses the phase transfer
catalyst TEBA, but there is no phase transfer catalyst in
the ‘035 process. Taylor Tr. 2665.
FF CA 59. Solid-liquid phase-transfer-catalyzed sys-
tems and liquid-liquid phase transfer catalyzed systems
are not regarded as equivalent in the art. Taylor Tr.
2666-67; RX 3969.
FF CA 60. Abic’s liquid-liquid phase-transfer-cata-
lyzed system is even less equivalent to a simple solid-
liquid system, i.e., a solid-liquid system without a phase
transfer catalyst, such as that of the ‘035 process. Taylor
Tr. 2667; RX 3969.
FF CA 61. Phase transfer of the hydroxide ion could
be effected by heating Abic’s system. Unlike catalysis,
which selectively increases the rate of hydroxide ion
transfer, heating indiscriminately speeds up everything
that is going on, thus increasing the potential for side
reactions. Taylor Tr. 2668-69; RX 3969.
FF CA 62. Examples of side reactions which could
occur in the Abic system as a result of heating to speed
up the reaction are: dimer formation, hydrolysis of DMC,
and hydrolysis of TZP. Taylor Tr. 2669-70; RPX 1051a.
FF CA 63. In Abic’s process, dimer formation is
inhibited by use of a phase transfer catalyst. Taylor Tr.
2669-70; RPX-1051a.
App. 297
FF CA 64. Use of a phase transfer catalyst in Abic’s
system also enables a reduction in the volume of solvent
used. Taylor Tr. 2670-71; RX-1024.
FF CA 65. The elements of Abic’s barium hydroxide
process, i.e., the base, the two solvents, the alkylating
agent, and the phase transfer catalyst, are not the same as
the elements of the potassium hydroxide/acetone base/
solvent combination of the ‘035 patent, and the two pro-
cesses do not function in the same way and are not
equivalent. See Taylor Tr. 2672.
FF CA 66. Even if the alkylating agents were the
same, and phase transfer catalysis were not taking place,
the two processes would still not be equivalent. See Tay-
lor Tr. 2672.
FF CA 67. That the ‘035 process starts with the same
substrate, employe the same alkylating agent, and obtains
the same product has no bearing on the equivalence of
the base/solvent combinations to the claimed process,
because the starting material, alkylating agent and the
product are the prior art, not the elements of the claimed
improvement. See Taylor Tr. 2672-73.
The Abic Process System Is Not Equivalent to the
Potassium Carbonate/Acetone or Potassium Carbonate/
Ethyl Acetate Processes of the ‘035 Patent
FF CA 68. The Abic process is even less similar to the
potassium carbonate/acetone or potassium carbonate /
ethyl acetate processes of the ‘035 patent, because the
Abic base, barium hydroxide (Ba(OH),), is not a carbo-
nate base, but a hydroxide base. RPX 1157; Taylor Tr.
2673.
App. 298
FF CA 69. The Abic process is not equivalent to the
potassium carbonate/acetone or potassium carbonate/
lower alkyl acetate processes of the ‘035 patent because
those processes cannot be made to work as they are
disclosed by the ‘035 patent. See Taylor Tr. 2675.
FF CA 70. Attempts by Abic to replicate Examples 4
and 5 of the ‘035 patent (powdery potassium carbo-
nate=acetone and powdery potassium carbonate-ethyl
acetate, with no added water) failed. Taylor Tr. 2675;
RPX-1051A.
FF CA 71. Abic found that the potassium carbonate/
acetone and potassium carbonate/ethyl acetae base/sol-
vent combination is do not work without added water.
Taylor Tr. 2676, 2680-81; RX 1272.
FF CA 72. The Tanabe laboratory notebooks pro-
duced by Tanabe and in evidence showed that Tanabe
knew from experiments conducted in 1981, before it filed
its patent application on December 7, 1991, that the
potassium carbonate/acetone and potassium carbonate/
ethyl acetate combinations did not work consistently
without added water. Taylor Tr. 2677; Taber Tr. 2099, 2148.
FF CA 73. Abic’s process, which consistently works,
is not equivalent to the ‘035 systems of potassium carbo-
nate/acetone and potassium carbonate/ethyl acetate,
which do not work. Taylor Tr. 2678; RX 1272C.
FF CA 74. If, contrary to the evidence, it were
assumed that the ‘035 processes of potassium carbonate/
acetone and potassium carbonate/lower alkyl aceate did
work, the Abic process would not be equivalent to either
process. See Taylor Tr. 2673, 2680.
App. 299
FF CA 75. Barium hydroxide and potassium carbo-
nate bases are different, and are non-equivalent ways of
deprotonating TZP, because the deprotonating species is
hydroxide (OH’) in the one case, and carbonate ion (CO,>)
in the other. Taylor Tr. 2673-74.
FF CA 76. To make the potassium carbonate and
barium hydroxide bases seem more similar, complainants
have postulated that the potassium carbonate processes
are “hydroxide mediated”; that is, that potassium carbo-
nate forms some hydroxide ion, which then deprotonates
TZP. CPX 14; Gokel Tr. 709.
FF CA 77. There is no evidence that the potassium
carbonate processes of the ‘035 patent are hydroxide
mediated. Taylor Tr. 2674.
FF CA 78. Dr. Kende testified that he didn’t know
whether the processes were hydroxide mediated or not,
and didn’t even know how one would carry out an exper-
iment to determine this. Kende Tr. 3415-16.
FF CA 79. Abic’s process is not equivalent to the
potassium carbonate/acetone combination; and for the
additional reasons that these ‘035 processes employ a
different base, and are not hydroxide mediated. See Tay-
lor Tr. 2679.
Abic’s Process Is Not Equivalent to the Potassium Car-
bonate/Acetone-Water and Potassium Carbonate/Lower
Acetate-Water System of the ‘035 Patent
FE CA 80. Potassium carbonate processes with added
water are similar to the potassium carbonate/acetone of
potassium carbonate/ethyl acetate processes of the ‘035
App. 300
patent, except that with the small added amount of water
they produce satisfactory results. Taylor Tr. 2681-82; RPX
1051A.
FF CA 81. Although there is a small amount of water
in the potassium carbonate/acetone-water and potassium
carbonate/ethyl acetate-water processes, they are still
solid-liquid processes, because the water (a) reacts with
potassium carbonate to form solid potassium car-
bonatesesquihydrate, or (b) dissolves in the acetone or
lower alkyl acetate solvent. Taylor Tr. 2682-3; 2687-98.
FF CA 82. For the same reasons that the Abic process
is not equivalent to the KOH/ACETONE processes, or
the potassium carbonate/acetone and potassium carbo-
nate-lower alkyl acetate processes, it is not equivalent to
the potassium carbonate-acetone-water or potassium car-
bonate-lower alkyl acetate-water processes. Taylor
2681-82.
FF CA 83. Abic repeated Examples 3 and 7 of the ‘035
patent, with yields similar to those reported by the ‘035
patent, and observed only one solid phase and one liquid
phase. Schwartz Tr. 2513-15; RX 1175; Taylor Tr. 2692-82;
RPX1051a.
FF CA 84. Abic also performed some experiments
which showed that the ‘035 solid-liquid processes, even
with added water, and the Abic liquid-liquid processes
were fundamentally different. RPX 1051a; Taylor Tr.
2682-86.
FF CA 85. Abic performed an experiment repeating
Example 3 of the ‘035 patent (potassium carbonate/ethyl
App. 301
acetate-water) but in which methylene chloride was sub-
stituted for ethyl acetate. After 6 hours (as specified in
the ‘035 example), there was 21.5 percent of unreacted
TZP and 77 percent of product. Taylor Tr. 2684;
RPX-1051a.
FF CA 86. The yield of ’035 Example 3 is reported to
be 92.7%. RX 1194.
FF CA 87. The yield of the process is important in
determining the equivalence of commercial processes,
which require high yields. Gokel Tr. 1116-17.
FF CA 88. Abix performed as identical experiment
with the sale difference that a small amount of TEBA
phase transfer catalyst was added. After 6 hours, there
remained 77 percent of product, but only 2.7 percent of
unreacted starting material and a large amount of the
dimer side product. Taylor Tr. 2684; RPX1051a.
FF CA 89. Abix also repeated Example 4 of Abic’s
‘995 patent, with and without the TEBA phase transfer
catalyst. In Abic’s liquid-liquid biphasic solvent process,
the results were exactly the reverse, a dimer was formed
in the absence of phase transfer catalyst, but no dimer
was formed in the presence of the phase transfer catalyst.
Taylor Tr. 2685-86; RPX 1051a.
FF CA 90. It is not surprising that the two systems
behave in opposite ways, since they are completely differ-
ent systems. Taylor Tr. 2686.
FF CA 91. Complainants assert that the small amount
of water present in these base/solvent combinations of
the ‘035 patent is present as a “surface solvent phase”
which complainants attempt to liken to the aqueous
App. 302
phase of Abic’s biphasic solvent process. See Taylor Tr.
2698-99.
FF CA 92. Complainants introduced no evidence of
such a surface solvent phase under the base-solvent con-
ditions of the ‘035 patent. See Liotta Tr. 1762-64, and
1779-81.
FF CA 93. The “surface solvent phase” is postulated
by complainants to be a very thin layer, approximately
100 angstroms (A) thick, of indeterminate composition
associated with the solid. Gokel Tr. 707, 1053.
FF CA 94. It would take 50,000 of the postulated 100
A surface solvent phases laid one on top of the other to
make up the thickness of a pencil line. The surface sol-
vent phase would be invisible to the naked eye. Wrighton
Tr. 1610.
FF CA 95. Even if there were a surface solvent phase,
the two processes would still be non-equivalent, for the
reasons set forth above. See Taylor Tr. 2698-99.
FF CA 96. Professor Wrighton of M.I.T., an expert on
surface chemistry, testified that the term “surface solvent
phase” was not customarily used in surface chemistry,
and had no recognized meaning. Wrighton Tr. 1608.
FF CA 97. Before this investigation, Dr. Liotta had
never used the term “surface solvent phase” in any pub-
lication. Liotta Tr. 1749.
FF CA 98. Before this investigation, Dr. Liotta had
never called a surface solvent phase a “biphasic solvent
system.” It appears that he used that terminology as and
analogy to Abic’s biphasic solvent system. Liotta Tr.
1705-06.
— U!U!Cd@Y
App. 303
FF CA 99. The postulated surface solvent phase film,
of indeterminant composition, and only 1/50,000th of the
thickness of a pencil line, is not the chemical equivalent
of the aqueous phase of the Abic methylene chloride-
water biphasic solvent system. Taylor Tr. 2700.
FF CA 100. Dr. Liotta performed experiments
attempting to prove the presence of a surface solvent
phase when water was added to a mixture of potassium
carbonate and ethyl acetate. Liotta Tr. 1812-13; CX 636.
FF CA 101. Dr. Liotta’s experiments established that
the potassium carbonate and water did not form a “sur-
face solvent phase”, but formed the well known solid
compound potassium carbonate sesquihydrate
(K,CO,1.5H,O). Taylor Tr. 2687-2697.
FF CA 102. Potassium carbonate sesquihydrate is not
a surface solvent phase. Gokel Tr. 1052; Wrighton Tr. 1622;
Taylo: Tr. 2688.
FF CA 103. Abic repeated Dr. Liotta’s experiments,
and tested the products by differential scanning calorime-
try (DSC). RX 1702.
FF CA 104. Dr. Wrighton testified that the DSC
results established that when potassium carbonate was
treated with water in ethyl acetate as Dr. Liotta had done,
the products were potassium carbonate sesquihydrate, or
potassium carbonate sesquihydrate with some residual
potassium carbonate. Wrighton Tr. 1602-22.
FF CA 105. Dr. Ronald Jenkins of the International
Center for Diffraction Data testified as to analyses he had
performed on the same material that Dr. Wrighton testi-
fied about. Dr. Jenkins concluded that the products were
App. 304
potassium carbonate sesquihydrate, or potassium carbo-
nate sesquihydrate, with some residual potassium carbo-
nate. Jenkins Tr. 2945-2953.
2. Abic Independently Developed Its Own
Process - It Did Not Copy the Process of the
‘035 Patent
Abic’s Initial Experiments
FF CA 106. In 1982 Abic decided to market a calcium
channel blocker, and chose diltiazem as its goal. Haber Tr.
1401.
FF CA 107. Although thee was no patent on diltiazem
in Isreal, Abic sought a license from Tanabe for sale to
other countries. Haber Tr. 2402.
FF CA 108. Tanabe refused to license Abic, or to
supply raw material. Haber Tr. 2402-2403.
FF CA 109. As a consequence, Abic began research
and development of the overall process for synthesizing
diltiazem hydrochloride in December 1982 or very early
in 1983. Habert Tr. 2402.
FF CA 110. By early 1983, Abic knew of the ‘257
patent and its foreign counterparts, but not of the ‘035
patent or any foreign counterpart of it. Haber Tr. 2403-04.
FF CA 111. There are about seven or eight steps in
Abic’s procedure to manufacture diltiazem hydro-
chloride, and the N-alkylation is the fifth or sixth step.
Haber Tr. 2404.
FF CA 112. Therefore, it wasn’t until May of 1983
that Abic had the starting material in hand to enable it to
begin working on the N-alkylation step. Haber Tr. 2408.
App. 305
FF CA 113. Abic’s work on the synthesis of
diltiazem hydrochloride, particularly the N-alkylation
step, is reflected in Abic’s internal periodic reports for
1982 and 1983. Haber Tr. 2409-10; RX 1007C; RX 1008C;
RX 1009C; RX 1010C; RX 1013C; RX 1015C.
History of Abic Work on N-Alkylation
Date Event
May 1983 First N-alkylation NaH/DMSO
August 1983 First phase transfer catalyst work
KOH + Methylene Chloride +
Water + TBABr
Fix on methylene chloride as sol-
vent
E Fix on TEBA as PTC
3 Try NaOH as base
September - NovemberNaOH + Methylene
1983 Chloride + Water + TEBA
December 1983 Try other than alkali metal
bases
Mg(OH),
Ca(OH),
NH,OH
CaCO,
MEt,
Knowledge of EPO ‘035 coun-
terpart
FORA BA inal Eatbad tng
February 1984 Test Ba(OH),-H,O
a RPX 1144
eee wewet se ge
BE
App. 306
FF CA 114. Based on the literature, Abic believed
that alkylating the nitrogen on the seven-membered ring
would be a straightforward procedure. Haber Tr. 2406-07.
FF CA 115. Because Abic needed diltiazem precur-
sor for study of the acetylation step, and for further
pharmaceutical testing, Abic began alkylating TZP under
the conditions already reported in the ‘257 patent and the
Kugita publications, sodium hydride and DMSO. Haber
Tr. 2411.
The Early Development of Abic’s Phase-Transfer Cata-
lyzed, Methylene-Chloride-Water Processes
FF CA 116. Abic quickly moved away from the
sodium hydride-DMSO base-solvent combination by
replacing DMSO with DMF. Haber Tr. 2411-12.
FF CA 117. Abic then began to look for alkylation
processes which did not employ sodium hydride. Haber
Tr. 2412.
FF CA 118. It was known at that time, the summer
of 1983, that one could alkylate carbon atoms (C-alkyla-
tion) using either harsh conditions or the milder condi-
tions of phase transfer catalysis, in the presence of water,
and it was felt that those milder phase transfer catalysis
methods could be adapted to the N-alkylation of TZP.
Haber Tr. 2412.
FF CA 119. Abic believed that alkylating under mil-
der conditions would minimize the possibility of side
reactions. Haber Tr. 2412-13.
FF CA 120. Although there is a large amount of
water present in classical phase transfer conditions, Abic
App. 307
was not concerned with potential hydrolysis of the TZP.
Haber Tr. 2413-14.
FF CA 121. Similarly, Abic was not concerned with
the potential retro-Michael reactions under phase transfer
catalyzed conditions because of the particular structure of
the TZP molecule. Haber Tr. 2414-15.
FF CA 122. Because Tanabe had not observed
O-alkylation at the 3-hydroxyl group of TZP under the
harsher conditions of the ‘257 patent, Abic was not con-
cerned that such O-alkylation was likely to take place
under the milder phase-transfer conditions. Haber Tr.
2415.
FF CA 123. Abic was not concerned about the
potential for alkylation at the carbonyl oxygen because in
the presence of a base, alkylation occurs almost exclu-
sively at the nitrogen. Haber Tr. 2416.
FF CA 124. Abic was not concerned that DMC
would be unstable under Abic’s phase transfer conditions
because, as with all alkylating agents, conditions can be
‘modified to minimize instability. Haber Tr. 2416-17.
FF CA 125. In August, 1983, Abic tried two phase
transfer catalyzed processes, one using potassium
hydroxide-methylene chloride-water and the other using
potassium hydroxide-toluene-water, both with TEBA bro-
mide as the phase transfer catalyst. Haber Tr. 2413.
FF CA 126. In August, 1983 Abic was not aware of
the ‘035 patent or any of its foreign counterparts. Haber
Tr. 2413.
FF CA 127. The phase transfer catalyzed reaction in
toluene did not work at low temperatures, and at high
App. 308
temperatures there was some hydrolysis of the lactam.
Haber Tr. 2418-19; RX 1007C.
FF CA 128. However, the phase transfer catalyzed
reaction in methylene chloride worked well at low tem-
perature, so hydrolysis which could occur at high tem-
perature was not a problem. Haber Tr. 2418-19; RX 1007C.
FF CA 129. Although Abic tried a number of other
solvents, none were as good as methylene chloride, so
from August, 1983, Abic concentrated on developing the
methylene chloride-water phase transfer catalyzed pro-
cess. Haber Tr. 2422-23.
FF CA 130. Abic also experimented with several
phase transfer catalysts, but rapidly settled on TEBA
because it gave the best results. Haber Tr. 2423-24.
FF CA 131. The product of the alkylation using
potassium hydroxide as the base was somewhat impure.
Haber Tr. 2423-25.
FF CA 132. In August 1983, therefore, Abic tried
sodium hydroxide as the base, since it was the most
similar base to potassium hydroxide. RX 1008; RPX 1144;
Haber Tr. 2424-25.
FF CA 133. Sodium hydroxide gave a purer prod-
uct, but it still contained about 10% of the unidentified
impurity. Haber Tr. 2425.
FF CA 134. Abic continued using sodium hydroxide
as a base for two to three months to make precursor for
use in studying the subsequent acetylation, hydro-
chlorination and purification processes. Haber Tr.
2425-26.
App. 309
Abic’s Development of Its Phase-Transfer-Catalyzed
Barium Hydroxide-Methylene Chloride-Water Process
FF CA 135. Eventually Abic discovered that the
impurity obtained with potassium hydroxide and sodium
hydroxide was “dimer”, which was formed by an alkyla-
tion reaction between the solvent methylene chloride and
two molecules of TZP. Haber Tr. 2426.
FF CA 136. Abic also was aware of British Patent
No. 1,236,467, a counterpart of the ‘257 patent, as well as
other counterparts, claiming alkylation processes employ-
ing alkali metal salts. RX 1010C; RX 1700; Haber Tr. 2432.
FF CA 137. In an effort to avoid the formation of
dimer, and in an attempt to develop a process using bases
other than alkali metal bases which would not infringe
the ‘257 foreign counterpart process patents, Abic in
December, 1983 began experimenting with ammonium
hydroxide, magnesium hydroxide and calcium hydrox-
ide. Haber Tr. 2429-31, 2434.
FF CA 138. Abic also apparently became aware in
December of 1983, for the first time, of the European
patent application that was the counterpart of the ’035
patent. Haber Tr. 2429-30.
FF CA 139. Abic was not concerned with potential
infringement in Europe, because the European applica-
tion was restricted to potassium bases, and Abic at the
time was using sodium hydroxide. Haber Tr. 2435.
FF CA 140. Ammonium hydroxide and magnesium
hydroxide bases did not work in the Abic phase transfer
catalyzed methylene chloride-water process. Haber Tr.
2431.
App. 310
FF CA 141. However, calcium hydroxide in the
phase transfer catalyzed methylene chloride-water pro-
cess gave high purity product without the formation of
the unwanted dimer. Haber Tr. 2431-32.
FF CA 142. Neither calcium hydroxide nor calcium
carbonate in acetone resulted in N-alkylation. RX 1010C;
Haber Tr. 2430-31.
FF CA 143. In February 1984, Abic tested barium
hydroxide with its methylene chloride-water-TEBA sys-
tem and found that barium hydroxide gave good yields,
practically no formation of dimer, and fewer side reac-
tions with DMC. Haber Tr. 2436.
FF CA 144. Abic repeated tests of other bases in the
methylene chloride-water system with a phase transfer
catalyst, and confirmed that potassium hydroxide and
potassium carbonate yielded large amounts (20%-30%) of
the dimer under those conditions. Haber Tr. 2436-37; RX
1015C.
FF CA 145. For comparative purposes, Abic also
tried some of the base-solvent combinations of the EPO
81234 application, but Abic did not pursue those combi-
nations because Abic has a policy of not infringing valid
patents. Haber Tr. 2438-39; RX 1013C.
FF CA 146. Abic tried numerous bases in the meth-
ylene chloride-water-TEBA system, including, in chrono-
logical order, potassium hydroxide, sodium hydroxide,
potassium carbonate, sodium bicarbonate, calcium carbo-
nate, magnesium hydroxide, ammonium hydroxide, tri-
ethylamine, and alumina, but in every case Abic obtained
either low yields of N-alkylated product or high yields of
App. 311
the unwanted dimer formation. Haber Tr. 2439-40;
RPX-1145.
FF CA 147. [C] Haber Tr. 2440, 2443-43; RX 1145, RX
1024C.
FF CA 148. [C] Haber Tr. 2443-44; RX 1024C.
FF CA 149. [C] Haber Tr. 2445-46; RX 1024C.
FF CA 150. [C] Haber Tr. 2447-48; RX 1024C.
FF CA 151. [C] Haber Tr. 2449; RX 1024C.
FF CA 152. Abic’s process is not a copy of the
Tanabe process. Haber Tr. 2452.
FF CA 153. Abic’s effort to develop its own process
was wholly independent, and was not stimulated by
knowledge of the ‘035 patent or of any counterpart.
Haber Tr. 2452.
FF CA 154. Abic obtained patents on its process in
the United States of America, Israel, Japan, Canada and
Europe. Haber Tr. 2450.
FF CA 155. In the United States, Abic’s patent appli-
cation was examined by Examiner Bond, who cited the
‘257 patent and the ‘035 patent as prior art, and con-
cluded that Abic’s process was patentable over those
references. Haber Tr. 2450; RX 1195 [CX 632].
FF CA 156. Professor Taylor is an organic chemist
specializing in organic synthesis, synthetic methodology,
and heterocyclic chemistry with emphasis on the devel-
opment of new synthetic methods in heterocyclic and
medicinal chemistry. Taylor Tr. 2586-87.
App. 312
FF CA 157. Professor Taylor has been a professor of
organic chemistry at Princeton for 41 years. He is cur-
rently the A. Barton Hepburn Professor of Organic Chem-
istry at Princeton. Taylor Tr. 2586.
FF CA 158. Professor Taylor has published over 400
articles, written three books in the field of heterocyclic
chemistry, and is editor of a 60-volume series called
Chemistry of Heterocyclic Compounds that is generally rec-
ognized as the reference series in heterocyclic com-
pounds. He has obtained about 40 patents, and is a past
Chairman of the Organic Division of the American Chem-
ical Society. Taylor Tr. 2587-89.
FF CA 159. Professor Taylor is a consultant to chem-
ical companies in the area of process development. Taylor
Tr. 2592.
FF CA 160. Professor Taylor was awarded the
American Chemical Society Award for Creative Work in
Synthetic Organic Chemistry, the Gowland Hopkins
Medal, the Fifth International Award in Heterocyclic
Chemistry, and recently the C.Cope Scholar Award for his
work in heterocyclic chemistry. Taylor Tr. 2592.
FF CA 161. Professor Taylor was accepted as an
expert in the field of organic chemistry, including the
subfields of organic chemistry and heterocyclic chemistry,
including the seven-membered ring of heterocyclic amide
compounds. Taylor Tr. 2593-94.
FF CA 162. Professor Mark S. Wrighton is an expert in
surface chemistry and in the interpretation of differential
scanning calorimetry curves. Wrighton Tr. 1607; RX 1679.
i
é
0]
f+
ie
App. 313
FF CA 163. Professor Wrighton is Provost and Pro-
fessor of Chemistry at the Massachusetts Institute of Tech-
nology. He has been on the MIT faculty since 1972, and has
been a full professor since 1977. Wrighton Tr. 1605.
FF CA 164. Professor Wrighton’s general area of
research involves surface chemistry, photochemistry and
electric chemistry. Virtually all of his current research
concerns the property of surfaces. Wrighton Tr. 1605-06.
FF CA 165. Professor Wrighton has published more
than 400 scientific papers. Wrighton Tr. 1605.
FF CA 166. Professor Wrighton has published
papers which contain the results of differential scanning
calorimetry analyses. Wrighton Tr. 1606-07.
FF CA 167. Professor Wrighton was accepted as an
expert in surface chemistry and in the interpretation of
DSC curves. Wrighton Tr. 1607.
3. The Experts
Professor Liotta
FF CA 168. Dr. Liotta submitted a declaration in this
investigation, which admittedly had some inaccuracies.
The declaration also confused the claims of the patent
with the examples cited in the specification. Liotta Tr.
1710-17. In an effort to show that the yield of Abic’s
process is not affected by the use of TEBA, the phase
transfer catalyst, he reported incorrect yields (id. at
1713-15). Although Dr. Liotta declared that he had
reviewed the prior art relating to the claims, he had only
examined the prior art of phase transfer catalysis. Id. at
Tr. 1712-13.
App. 314
FF CA 169. Dr. Liotta referred to his postulated
“surface solvent phase” as a “biphasic solvent system”
only for purposes of this litigation; he had never referred
to it before in those terms, and did it only to draw a
comparison between Abic’s solvent system and examples
in the ‘035 patent. Liotta Tr. 1705-06.
FF CA 170. Dr. Liotta first concluded that the KOH-
DMSO process was not equivalent to the ‘035 processes;
then concluded that the KOH-DMSO process was equiva-
lent; and then changed his mind again and said that they
were not equivalent. Liotta Tr. 1708-09.
FF CA 171. Dr. Liotta obtained a patent on his the-
ory of the “omega phase,” a “surface solvent phase”
which contained a phase transfer catalyst. Id. at 1750. In
his patent he distinguished such omega phase systems as
patentably distinct from conventional phase transfer cata-
lyst processes. Id. at 1750-52. He admitted at his deposi-
tion that the Abic process is en example of such a
conventional phase transfer catalyst process. Id. at 1753.
But he also alleged that his theory of the omega phase
was wrong, and that his patent was wrong. Id. at 1755.
FF CA 172. Dr. Liotta could not bring himself to
admit that TEBA is a phase transfer catalyst, and kept
calling it a “quat” and a “surfactant”. Not only did Dr.
Taylor testify that it is a phase transfer catalyst but Dr.
Baldwin, an expert for complainant, referred to it as a
phase transfer catalyst, as did Dr. Kende. CX 635; Liotta
Tr. 1703-04; Kende Tr. 1340.
FF CA 173. Eventually Dr. Liotta admitted that
TEBA is one of the most commonly used phase transfer
catalysts (Liotta Tr. 1702), because he had written that in
App. 315
his book. RX 3969; Taylor Tr. 2654; RX 3969; Liotta Tr.
1676, 1701; Taylor Tr. 2653.
FF CA 174. Dr. Liotta was forced to admit that there
is no experimental data supporting his surfactant theory
(Liotta Tr. 1676). He further admitted that he had not
done any experiments of his own to show that TEBA is
not acting as a phase transfer catalyst in the Abic system,
and refused to accept Abic’s data which shows increased
yield when TEBA is used. Liotta Tr. 1835-37.
FF CA 175. For the above reasons, the Administra-
tive Law Judge declines to accept the opinions of Dr.
Liotta in support of complainants’ positions.
Professor Atwood
FF CA 176. Dr. Atwood is primarily a single-crystal
x-ray crystallographer. Atwood Tr. 3311.
FF CA 177. None of Dr. Atwood’s publications is
about x-ray powder diffraction. Atwood Tr. 3315, 3260.
FF CA 178. Unlike Dr. Jenkins, he hasn’t written
text-books about x-ray powder diffraction. Atwood Tr.
3260.
FF CA 179. Unlike Dr. Jenkins, Dr. Atwood does not
belong to any organization devoted to x-ray powder dif-
fraction. Atwood Tr. 3311.
FF CA 180. Unlike Dr. Jenkins, Dr. Atwood does not
have any patents on x-ray powder diffraction equipment.
Atwood Tr. 3317-18, 3260.
App. 316
FF CA 181. Unlike Dr. Jenkins, Dr. Atwood does not
design his own x-ray powder diffraction equipment.
Atwood Tr. 3322-3323.
FF CA 182. Because the materials which are of inter-
est in this investigation are powders, x-ray powder dif-
fraction is the more pertinent expertise. Atwood Tr.
3313-14.
Dr. Jenkins’ Qualifications
FF CA 183. Dr. Jenkins is currently employed as the
principle scientist at the International Center for X-Ray
Product Direction Data, a not-for-profit data base organi-
zation that archives and supplies about 65,000 powdered
refraction patterns. Jenkins Tr. 2936-2937.
FF CA 184. Dr. Jenkins has been involved in X-ray
powder refraction for about 35 years. Jenkins Tr. 2937.
FF CA 185. Dr. Jenkins worked for ESSO Research,
Limited for 10 years, the last years spent in X-ray powder
refraction analysis. He then worked at North American
Phillips, where he was involved with the development of
instrumentation and software for X-ray powder refrac-
tion. He also was involved in the application of the
technique and teaching of the method of powder refrac-
tion. Jenkins Tr. 2937-2938.
FF CA 186. Dr. Jenkins obtained 6 patents for devel-
opmental work at North American Phillips, of which
three were on equipment for X-ray powder refraction.
Jenkins Tr. 2938
FF CA 187. Dr. Jenkins has published about 80
papers on X-ray powder refraction, and written three
as a. x Ss
App. 317
books targeted at an audience of material scientists (geol-
ogists, mineralogists, chemists). The first book was writ-
ten in 1972, and sold about 8,000 copies. The second was
an audio course done in connection with the American
Chemical Society. The third is coming out in March.
Jenkins Tr. 2939
FF CA 188. Dr. Jenkins was accepted as an expert in
X-ray powder refraction. Jenkins Tr. 2940.
4. TEBA/Phase Transfer Catalysis
FF CA 189. Dr. Liotta does not have any experimen-
tal evidence to support his opinion that TEBA in Abic’s
process acts as a surfactant. Liotta Tr. 1676.
FF CA 190. There is no documentary evidence of
record in this investigation that TEBA acts as a surfactant
or is regarded as a surfactant by people of ordinary skill
in the art. Liotta Tr. 1701.
FF CA 191. TEBA is one of the most commonly used
phase transfer catalysts in chemistry. Liotta Tr. 1701.
FF CA 192. Dr. Baldwin regards TEBA in Abic’s
system as a phase transfer catalyst. Liotta Tr. 1704-03.
FF CA 193. When TEBA is used in Abic’s system, no
dimer is formed, whereas when TEBA is not used, dimer
is formed. Liotta Tr. 1801-03.
FF CA 194. In example 3 of the ‘035 patent, if meth-
ylene chloride is substituted for ethyl acetate, and no
TEBA is present, no dimer forms, whereas, if TEBA is
present, dimer forms. Liotta Tr. 1803-04.
App. 318
FF CA 195. Solid-liquid phase transfer catalysis is
different than liquid-liquid phase transfer catalysis. Lio-
tta Tr. 1838.
C. The Fermion Process
FF CF 1. In February 1983, Dr. Lindholm (a Fermion
development manager) assigned the project of develop-
ing a process for manufacturing diltiazem to Mr. Hyténen
(who was then a product development chemist). Hyt6nen
Tr. 2992-2993; Lindho!m Tr. 3067.
FF CF 2. Dr. Lindholm testified at the hearing in
part as an expert witness. Dr. Lindholm was accepted as
an expert in chemical process design. Lindholm Tr.
3068-3069.
FF CF 3. Fermion’s process development effort
involved more than the single N-alkylation step. The
Fermion diltiazem synthesis procedure today includes
nine processing steps. The development effort com-
menced with the early steps in synthesis. Hyt6nen Tr.
2994.
FF CF 4. In early September 1983, Mr. Hytdnen
began working on the N-alkylation step of the diltiazem
process. RPX 2022C, RX 2114C and RX 2115C, identified
at Hytdnen Tr. 2994-2995.
FF CF 5. The first N-alkylation experiment con-
ducted by Mr. Hytdnen was using ‘257 conditions so that
he could obtain some N-alkylated TZP and learn how the
‘257 process worked. Hyténen Tr. 2994-2995.
cal FE MEL SEE BAR MR SAT ST ARTE.
App. 319
FF CF 6. Mr. Hyténen conducted many tests with
various base/solvent combinations for about one year. In
_ September and October 1984, Fermion experimented with
2-butanone or methy ethyl ketone (MEK), [C] in Septem-
ber and October 1984. Hyt6nen Tr. 2995-3004; RX 2114C
and RX 2115C.
FF CF 7. Mr. Hytdnen read the ‘035 patent to
exclude the use of MEK as a solvent. He testified as
follows:
Q When you reviewed the ‘035 patent, was
there anything which suggested the use of
MEK?
A No, there was not.
Why not?
—O
A They had only mentioned acetone, and they
had — not even ketone mentioned. On the
other hand, they had determined that the
lower alkaline acetate, in the same way they
could have determined - described also the
lower ketones, if they had known that the
reaction can be done with other ketones
than acetone.
—O
Is MEK a homolog of acetone?
A Yes.
Q Do I understand that despite the facts that
MEK is a homolog of acetone, the ‘035 pat-
ent still do not suggest the use of MEK to
you?
* * *
THE WITNESS: It did not suggest any other
ketone than acetone.
App. 320
BY MR. KELLEY:
Q Why not?
* * *
THE WITNESS: I understand that Tanabe’s
researchers thought that other ketones cannot be
used in this reaction.
Hytonen Tr. 3014-3015.
FF CF 8. Butanone is a homolog of acetone, 1.e.,
butanone has an additional methylene group.’ It is not a
lower alkyl acetate. Because of the additional methylene
group, butanone has different properties from acetone. At
the least, it can be said that butanone is more solvable in
hydrocarbons than acetone; it is less solvable in water
than acetone, it has a higher boiling point than acetone; it
is less polar than acetone. Gokel Tr. 635, 637.
FF CF 9. Normally, if one of ordinary skill in the art
wanted to seek how far one could extend the N-alkyla-
tion of ‘035 patent, one would try another ketone besides
acetone, possibly 2-butanone (or MEK). By the same
token, one of ordinary skill in the art familiar with a
range of ketones would read the ‘035 patent, notice the
specificity and exclusivity of the claim to the use of
acetone, and conclude that other ketones were not
1 A homolog(ue) is defined as:
Member of a series of compounds whose structure
differs regularly by some radical, e.g., =CH2, from
that of its adjacent neighbors in the series.
R. Grant & C. Grant, eds. Grants & Hackh’s Chemical Dictionary
287 (1987).
:
P
; os
4
=.
sl
ql
)
App. 321
included because they did not work. Taylor Tr. 2620-2621,
2784-85.
FF CF 10. While at university, Mr. Hytdnen tried
alternatives to acetone in an alkylation with potassium
carbonate. One of the substitutions he tried was MEK for
the acetone. He later had better success with another
chemical. Although he tried these substitutions, Mr.
Hytonen believes that many researchers (including his
professor) think that certain alkylation reactions are spec-
ific to acetone. Hytdnen Tr. 3015-3016, 3045-3046.
FF CF 11. Mr. Hytonen had experience with MEK at
Fermion before starting his diltiazem development work.
Hyt6nen Tr. 3016.
FF CF 12. Fermion conducts pilot tests on a scale
that is smaller than industrial scale yet larger than labora-
tory scale, using instruments similar to those used for
commercial manufacturing. Hytdnen Tr. 3005.
FF CF 13. In October 1984, Fermion conducted a
pilot plant test on the N-alkylation process using
potassium carbonate/MEK as the base/solvent combina-
tion. This pilot plant test was a failure. RPX 2022C;
Hyténen Tr. 3004-3005.
FF CF 14. After the October pilot plant failure with
MEK, Fermion conducted additional experiments with a
mixture of [C] as the solvent and [C] as the base; [C] and
either [C], with and without [C]; [C] and either [C]; [C]
with either [C]; [C] with [C]; [C] with [C]; and [C] with
[C]. RX 2114C and RX 2115C (experiments B2449, B2454,
B3367(a) and (b), B3371(a) and (b), B3372(a) and (b),
B1048, B1049 and B3370).
App. 322
FF CF 15. In 1985 Fermion conducted additional
testing with the combination of [C] as the base/solvent
combination for the N-alkylation step. RX 2114C and RX
2115C (experiments B1054, B3275(a) and (b), B1058(a) and
(b), B3376(a), (b), (c) and (d), B1061, B1062(a), (b) and (c),
B1064 and B1065(a), (b), (c) and (d)).
FF CF 16. Mr. Hyténen conducted further experi-
ments in 1985 with the MEK and potassium carbonate
process to determine why it had failed in the pilot plant.
Hyténen Tr. 3007.
FF CF 17. Fermion also developed an alternative
process using potassium carbonate/ethyl acetate, and
produced diltiazem for commercial use. Hyténen Tr.
3005-3007.
FF CF 18. Fermion found the potassium carbonate/
ethyl acetate combination to be unreliable on a commer-
cial scale because potassium carbonate had to be added
twice and one had to follow the reaction to completion,
and to add DMC-HC1 in different amounts based upon an
analysis of the reaction. These procedures required the
presence of a skilled chemist. Hyténen Tr. 3006-3007,
3047-3048.
FF CF 19. The operators of Fermion’s process are not
chemists, or laboratory technicians. They are individuals
without formal chemical education. The skill of the oper-
ators is taken into account when developing a commercial
process. Lindholm Tr. 3089-3090.
FF CF 20. The Finnish counterpart of the ‘035 patent
was not in effect during the time period that Fermion
UNS ETT ttt
App. 323
used the potassium carbonate/ethyl acetate combination.
Hytonen Tr. 3007, 3047-48.
FF CF 21. In late 1985, Mr. Hytodnen thought he had
solved the problem with the MEK and potassium carbo-
nate system, and conducted a second pilot plant test.
Hyténen Tr. 3007.
FF CF 22. The second MEK and potassium carbo-
nate pilot plant test was also a failure. Hytonen Tr. 3007.
FF CF 23. Fermion conducted further experiments
with the MEK and potassium carbonate process, and
conducted a third pilot plant test in January 1986 which
was a success. Hytdnen Tr. 3008.
FF CF 24. In the accused Fermion process, the
N-alkylation step uses TZP, DMC-HCI, K,CO,, butanone
(or “MEK”) and water. Gokel Tr. 624; Cx 192¢.
FF CF 25. Fermion learned that the amount of [C]
Present in the MEK and potassium carbonate process was
critical. Mr. Hytodnen testified to this point as follows:
4 ‘ f a li ad on ees Te yes os " 2
7 - - . sa Sa . i patna 4 nai ane pitas ~ ae oF ‘
le Si a a as is i rg -: ae
; Q Did you determine what the cause of the
problem was with your MEK process?
) ; A Yes, we found out.
What did you determine caused the failure
on the first pilot run?
A In the first pilot experiment, we had too
much [C] in the reaction mixture.
App. 324
Q Did you determine what the problem was in
the second experiment?
A_ Yes. It had too little [C].
Hytoénen Tr. 3008.
FF CF 26. The ‘035 patent contains no teachings that
the amount of [C] in the process is critical. For example,
complainants’ expert testified as follows:
Q Nowhere in the ‘035 patent is there any
explicit teaching that [C] is critical to the
success of the reaction, is there?
A I think that’s a fair statement.
Kende Tr. 1428.
FF CF 27. The ‘035 patent provided no guidance to
Fermion and Mr. Hyténen in solving the problems
encountered with the MEK and potassium carbonate pro-
cess. In this regard, Mr. Hyt6nen testified as follows:
Q You testified a few minutes ago that the ‘035
patent did not provide any help in develop-
ing the present Fermion process. Do you
recall that?
A That is true.
Q Why didn’t it provide you with any help?
A For instance, there was no mention about —
they did not mention the critical nature of
[C]. On the contrary, they had examples in
which there were reactions that had no [C]
in them.
> oi c See ee ee ee ee ee BRE Siates 6 -e re tee ses Re hte Sy eee wiras wigs
a Sas P a Aad ata ie. seks Sebamed ete Ret te
hey cig A hoy *s ¥.
App. 325
Q Did the ’035 patent provide you with any
help in determining the cause of the failures
in the first two pilot plant runs with MEK?
A No, it did not.
Hytdnen Tr. 3013-3014.
FF CF 28. Fermion learned that the MEK and
potassium carbonate process did not work with either too
much or too little added [C]. Hytdnen Tr. 3008.
FF CF 29. After Fermion’s success with the MEK and
potassium carbonate N-alkylation process in the pilot
plant, Mr. Hytonen began to experiment with making the
Process less sensitive to the amount of [C] present, and
therefore “more reliable.” Hyténen Tr. 3009-3010.
FF CF 30. Mr. Hyt6nen discovered that by reducing
the ratio of [C], it was possible to reduce the sensitivity of
the MEK and potassium carbonate to the amount of [C]
present. Hytdnen Tr. 3010.
FF CF 31. The present Fermion process has [C] added
to the MEK. Hytédnen Tr. 3008-3009, Lindholm Tr. 3090.
FF CF 32. Fermion discovered that its present process
is extremely reliable, always proceeding to completion,
i.e., all the TZP is consumed. Hyténen Tr. 3011; Lindholm
Tr. 3075.
FF CF 33. The ’035 patent does not provide any
teaching that the [C] [C] will have any effect on the
reaction. Mr. Hyténen testified as follows:
Q During the time that you were improving
the MEK process, did the ’035 patent pro-
vide any help?
App. 326
A No, it did not.
Q Does the ‘035 patent teach a chemist what
the effect of varying the relative amount of
[C] will have on the process?
JUDGE HARRIS:You better ask that again.
BY MR. KELLY:
Q Let me break it into pieces and try to ask
that again. Is there any description or teach-
ing in the ‘035 patent that you will get a
more reliable process by reducing the [C]
[C]?
A No.
Tr. 3011-3012.
FF CF 34. The ‘035 patent provided no guidance to
Fermion in reducing the ratio of iC) . Tr. 3011.
FF CF 35. Fermion’s experts, Dr. Lindholm and Pro-
fessor Magnus, testified that Fermion’s process works in
a different way than the process of the ‘035 patent claims.
Lindholm Tr. 3075, Magnus Tr. 3178.
FE CF 36. Fermion conducted over [C] N-alkylation
experiments between October 1983 and early 1986 in
developing its process. Lindholm Tr. 3101; RX 2114C.
FF CF 37. In 1981, Tanabe attempted to use MEK as
the solvent in the N-alkylation of TZP. RX 2047C, RX
2046C, RX 2094C and RX 3494C.
FF CF 38. In 1981, Tanabe tested and rejected MEK as
the solvent for the N-alkylation of TZP. RX 2047C, RX
2046C, RX 2094C, RX 3494C.
App. 327
FF CF 39. In 1981, Tanabe’s experiments with MEK
either resulted in no product or impure product. RX
2047C, RX 2046C, RX 2094C and RX 3494C.
FF CF 40. Fermion introduced evidence that a process
which produced impure product would be considered a
failure. See Hyténen Tr. 2999.
FF CF 41. The adverse consequence of too little or too
much [C] in the ‘035 patent’s potassium carbonate /ethyl
acetate process was experienced by Tanabe in its commer-
cial process, and is demonstrated by Tanabe’s later 1991
testing. RX 2237C; Nakao Tr. 384, 420.
FF CF 42. Complainants’ expert, Professor Gokel,
testified that the way in which the process of the ‘035
patent claims worked was through a so-called “surface
solvent phase” or a boundary phase. Gokel Tr. 941-942,
982.
FF CF 43. Complainants offered no evidence that a
surface solvent phase was present in the Fermion process.
The only tests offered by complainants were not
N-alkylations and lacked both the TZP and DMC.HC1.
CX 636, CX 680C and CX 681C.
FF CF 44. Dr. Lindholm testified as follows:
Q Have you reviewed any experiments which
were conducted by Professor Gokel and Lio-
tta in connection with this litigation?
A Yes, I have.
Q In Your opinion are those experiments rele-
vant in any way to the matters involved in
this litigation?
App. 328
>
They are not in my opinion.
Why not?
@)
A_ In these experiments there were no starting
material, no reagent that’s the TZP or DMC.
So in my opinion these experiments does not
show anything about what happens when
you do the real reaction.
Q Do you recall the testimony by Dr. Gokel
that in his experiments he saw or observed a
clump of potassium carbonate forming in
the experiments with MEK?
>
Yes, I remember.
Q Based upon your experience in the Fermion
process, does the potassium carbonate form
a hard clump upon the addition of water?
A_ No, it does not. We have not seen the thing
occur either in laboratory or on the indus-
trial scale.
Dr. Lindholm Tr. 3095-3096.
FF CF 45. In all of its experiments and commercial
experience, Fermion has never observed the results
obtained by complainants in complainants’ tests in which
neither TZP nor DMC.HC1 were present. Lindholm Tr.
3095-3096.
FF CF 46. Fermion’s expert, Dr. Lindholm, testified
that complainants’ testing provides no basis for conclud-
ing that the Fermion process works in the same way as
the process of the ‘035 patent. Lindholm Tr. 3095-3096.
FF CF 47. Fermion duplicated examples found in the
‘035 patent, and compared the results obtained with the
is erent
App. 329
‘035 patent solvents to those obtained when MEK was
used as the solvent. CX 394
FF CF 48. Dr. Lindholm, one of Fermion’s experts,
concluded that based on these experiments, MEK was not
equivalent to the ’035 patent claim solvents in the
N-alkylation of TZP. Dr. Lindholm testified as follows:
Q What is it about Cx 394 that further sup-
ports your opinion?
A This CX 394 is a test series we run in the
laboratory using the ‘035 patent examples,
just as they are described in the patent. And
for comparison, these examples, we substi-
tuted the solvent mentioned in the examples
with MEK. And if we look at page 3, we see
here a comparison of patent example 2 using
either acetone or MEK.
From these experiments, we see that these
two reactions run rather Similarly. The
butanone is a little bit faster than acetone.
If we then go two Pages forward and look at
patent Example 4, this is now a dry system.
We can see here that the acetone process
works. After nine hours, there is really
nearly no starting material left in the reac-
tion.
But on the other hand, we can see that using
MEK, we have Virtually no reaction.
Taking these two examples together, as we
should do, if these two solvents were equiv-
alent, we would get the same results. We
would even have a slightly faster reaction in
MEK, as the boiling poin, is higher.
But as we do not have that, this means that
these solvents can’t be equivalent. If we turn
one page back, we have patent Example 3.
App. 330
This is with ethyl acetate. From the numbers
here, we can see that the reaction grows
nicely in ethyl acetate.
But on the other hand, we can see that using
MEK, the reaction proceeds, but to a very
small extent. After six hours reaction time,
we have still roughly about 60 percent of the
starting material left.
This indicates that or at least shows that
MEK and ethyl acetate are not equivalent.
Q When you said the reaction went rather
nicely in ethyl acetate, what did you mean?
A I mean that the consumption of starting
material was nearly complete.
Lindholm Tr. 3077-3078.
FF CF 49. Dr. Lindholm testified, that with the excep-
tion of ‘035 patent example 2, in all cases the substitution
of MEK for the solvent of the patent example provided
substantially different results. Lindholm Tr. 3077-3078.
FF CF 50. In patent example 3, the use of ground
potassium carbonate in combination with ethyl acetate
resulted in an N-alkylation reaction which proceeded to
completion as described in the patent, i.e., within six
hours. In contrast, with MEK and ground potassium car-
bonate, after six hours roughly 60% of the starting TZP
remained. Lindholm Tr. 3077-3078, Hyténen Tr. 3138-3139.
FF CF 51. In patent example 4, Fermion was able to
duplicate the results reported for acetone and powdery
potassium carbonate, essentially complete TZP conver-
sion within nine hours. However, when MEK was substi-
tuted for the acetone, the reaction did not proceed at all,
App. 331
essentially all TZP remained unreacted. Hyténen Tr.
3077-3078.
FF CF 52. Patent example 5 did not Proceed as writ-
ten. CX 394.
N-alkylation which Proceeded to completion in two
hours, while with methyl acetate the reactior took thirty
hours. CX 394.
FF CF 54. Fermion chose to use patent example 3 as
exemplary of the ‘035 patent claim process because this
was the sole ‘035 patent example repeated by complain-
ants’ expert, Professor Baldwin. Lindholm Tr 3086.
FF CF 55. Subsequently, Fermion learned that patent
example [C] closely resembles [C] process. Lindholm Tr.
3087.
FF CF 56. Complainants had testing conducted by
Professors Baldwin, Gokel and Liotta. Cx 635, CX 636 Cx
680C and CX 681C.
FF CF 57 Professor Baldwin testified that he could
not have predicted the results he obtained with MEK,
2-butanone:
Q Is there any way before this experiment you
could have predicted this result?
A What result?
App. 332
Q That increasing [C] [with 2-butanone] would
increase yield.
A In the particular butanone system, no, I
couldn’t have predicted that.
RX 2238C.
FF CF 58. The only tests offered by complainants
concerning the so-called “surface solvent phase” were not
N-alkylations and lacked both the TZP and DMC.HC1.
CX 636, CX 680C and CX 681C; Lindholm Tr. 3095-3096.
FF CF 59. Professor Baldwin’s tests demonstrate that
the mere substitution of MEK for the ‘035 patent’s sol-
vents did not result in an N-alkylation process which
achieved yields and/or productivity comparable to that
achieved by the Fermion commercial process. RPX 2004C-
A. Lindholm Tr. 3092-3093.
IV. INVALIDITY - CLAIM 1 OF THE ‘035 PATENT IS
INVALID UNDER 35 U.S.C. § 103
A. Background
FF D 1. During the United States prosecution of the
‘035 patent, Tanabe submitted to the PTO a Statement of
Art in which Tanabe described its alleged invention as
involving only certain specified base/solvent combina-
tions, specifically stating that the “Applicants’ invention”
was N-alkylation:
“[In] the presence of potassium hydroxide in
acetone or potassium carbonate in acetone,
SY MEE MEE pect SAG OPA As SOLE
OEE eI eT I A A Re a Te, ok Re Sey A
EE ee en a an a ee
App. 333
lower alkyl acetate, water-acetone, or water-
lower alkyl acetate.”
CX 2, paper 4 at 2-3.
FF D 2. In the Statement of Art Tanabe distinguished
its alleged invention over the prior art by arguing that the
invention produced yields of no less than 87% as opposed
to 65-70% obtained with prior art processes.2 XC 2, paper
4 at 2-3.
FF D 3. On December 21, 1981, Tanabe filed an appli-
cation in Europe corresponding to the ‘035 patent appli-
cation. RX 3325. On June 1, 1984, Tanabe’s EPO
application was rejected over the ’967 and ‘257 patents.
RX 3325.
FF D 4. The ‘967 patent taught the use of “alkali
metal hydroxides” in a variety of solvents including
toluene to attach DMC to a benzothiazepinone differing
from TZP by only a single R group at the three Position of
the seven-membered ring. Potassium hydroxide and
sodium hydroxide are examples of alkali metal hydrox-
ides. RX 1103; Gokel, Tr. 870-82, 877-78, 881; Taber, Tr.
2159-60. In response to that rejection, on October 1, 1984,
Tanabe submitted a twelve page document urging that
the application be approved over the ‘257 and ’967 pat-
ents. Included in this response were five pages of attor-
ney argument and what Tanabe referred to as a
“Comparative Test Report” which, Tanabe claimed, dem-
onstrated the patentability of its claimed process over
* Tanabe did not disclose Krapcho’s U.S. Patent 3,075,967
(the “ ‘967 patent”) to the U.S. Patent and Trademark Office
(“PTO”) during the original prosecution of the ‘035 patent. CX 2.
App. 334
prior art disclosures that used, inter alia, sodium carbo-
nate and toluene. RX 3929C. Specifically, Tanabe repre-
sented that the Comparative Test Report
“[S]hows that the specific combination of bases and
solvents, i.e., potassium hydroxide in acetone or
potassium carbonate a solvent selected from
acetone, lower alkyl acetate, a mixture of ace-
tone and water and a mixture of lower alkyl
acetate and water, leads to surprising results of
the method according to the invention.”
RX 3929C at 705 (emphasis added).
FF D 5. Tanabe’s submission analyzed the experimen-
tal data reflected in the Comparative Test Report and
explained as follows to the EPO:
For example, when the condensation reaction
was carried out by the use of sodium hydroxide
or sodium carbonate as the base, the yield of the
product was less than 10% even if acetone was
used as the solvent. Moreover, when sodium
amide or sodium hydride was used as the base,
the yield of the product was not more than 12.4%
in the case where toluene or dioxane was used as
the solvent. Furthermore, even if potassium
hydroxide or potassium carbonate was used as
the base, the yield of the product was less than
30% in the case where dioxane, toluene or meth-
anol was used. (RX 3929C at 706.)
Tanabe further explained that:
[judging from the facts (i) that the refer-
ences cited by the Examining Division teach
neither the use of potassium carbonate as
the base nor the use of the specific base-
solvent combinations to be employed in the
method of the present invention; (ii) that,
App. 335
when the condensation reaction was carried
out by the use of sodium hydroxide or
sodium carbonate as the base, the yield of
the product was less than 10%; and (iii) that,
even if potassium hydroxide or potassium
carbonate was used as the base, the yield of
the product was less than 30% in the case
where dioxane, toluene or methanol was
used, it is obvious that the above-mentioned
advantageous features of the present inven-
tion have never been taught nor suggested
in the references cited by the Examining
Division. Thus, the replacement of sodium
hydride and dimethylsulfoxide by the speci-
fic base-solvent combinations of the present
invention is unobvious over the cited refer-
ences. Due to the superior and surprising
results obtained by the method of the invention,
the present application possesses level of
inventiveness necessary for its patentability.
RX 3929C (emphasis added).
FF D 6. Tanabe made identical arguments to
numerous other patent offices, including those of Israel
and Finland. 3338C; 3233C.
FF D 7. Reexamination of the ‘035 patent was
requested by Respondent Abic, and supported by a decla-
ration of Dr. Taylor. RX 1085 (See RX 2193).
FF D 8. The examiner initially agreed with Dr. Taylor,
and rejected all the claims of the ’035 patent. RX 1603 (RX
2193, RX 2204).
The bases for the rejection included:
(a) The ’257 patent showed a conventional
method of N-alkylation;
App. 336
(b) Pachter and subsequent references showed
the “widely used” N-alkylation of N-aryl
amides with the same bases (potassium hydrox-
ide, potassium carbonate) and the same solvent
(acetone) as the ‘035 patent;
(c) British ‘119 and Nagarajan showed dimet-
hylaminoethylation (i.e., reaction with DMC) of
dibenzoxazepinones;
(d) Johnstone “further illustrat[ed] the value of
the Pachter et al. technique.”
RX 1603.
FF D 9. In response, Tanabe submitted declarations
by Drs. Baldwin and Kende which argued that:
(a) The Pachter technique was not widely
known;
(b) Pachter did not render the ‘035 patent
obvious because it (i) disclosed only amides
which were not cyclic and (ii) did not disclose
DMC as an alkylating agent;
(c) British ‘119 and Nagarajan were limited to
“nitro-substituted” amides, and thus not rele-
vant to N-alkylation of TZP which had no nitro
substituent;
(d) _Johnstone was not pertinent;
(e) A number of potential side reactions,
including retro-Michael reaction, ring-cleavage,
O-alkylation, and carbonyl O-alkylation might
occur, and might prevent high yields of the ‘035
patent.
RX 1658, RX 1329.
SAW 7
App. 337
FF D 10. During the reexamination of the ‘035 patent,
the examiner was presented with about 172 prior art
references. RX 1603.
FF D 11. The examiner accepted some of Tanabe’s
arguments, and in deciding to issue a reexamination cer-
tificate, held that the Prior art then of record did not
establish the obviousness of the ‘035 patent. RX 1653; RX
1654; Taylor Tr. 2714-15, 2924-26; RPX 1149; RPX 1151.
FF D 12. The N-alkylation process described in the
‘035 patent is typical of the types of Projects that process
development chemists would have undertaken in 1981.
Pachter Tr. 1495-96,
FF D 13. It is generally accepted that when potassium
hydroxide is used to deprotonate an amide, including
TZP, water is produced as a side product. Gokel Tr.
103940.
FF D 14. The patent examiner was not told that the
prior art disclosed hydrous systems for the N-alkylation
of benzothiazepinones. Further, the patent examiner said
nothing during the interview which Dr. Baldwin attended
to indicate he was aware of the Prior art teaching the
N-alkylation of benzothiazephine type compounds in
hydrous conditions. RX 4038C; Kende Tr. 1431-32.
FF D 15. In his declaration submitted to the examiner
during the Reexamination (RX 3132), Dr. Kende did not
advise the examiner that in 1981, Yamawaki disclosed
hydrous systems involving the use of potassium carbo-
nate, sodium hydroxide and potassium hydroxide as
bases which are excellent alternatives to the conventional
bases used in N-alkylation methods. Kende Tr. 1355-57.
App. 338
FF D 16. In granting the request for reexamination,
the examiner stated in part as follows:
The arguments concerning the possibility of side
reactions by Taylor, Baldwin in [sic] Kende are
not seen as having great weight in this particu-
lar case one way or another. Nor are the argu-
ments concerning the use of DMC.HC1.
RX 1653.
FF D 17. The chemistry of organic compounds
revolved around the chemistry of functional groups.
Kende Tr. 526; Taylor Tr. 2780-81.
FF D 18. Functional groups are more important than
ring structure in determining chemical reactivity. Taylor
Tr. 2780.
FF D 19. The functional group known as an “N-aryl
amide,” which is part of the structure of TZP, has the
following general structure:
[Diagram Omitted In Printing]
Kende Tr. 1457-58; RPX 4023.
FF D 20. One can choose reaction conditions by
focusing on the functional group on which one wishes to
carry out the chemical transformation. The ring frame-
work to which the functional group is attached plays a
minor role, if any, in the functional group chemistry. In
organic chemistry, synthetic reactions depend upon and
are focused on the properties of the functional group.
Taylor Tr. 2780.
FF D 21. The inventors of the ‘035 patent knew prior
to 1981 that the N-alkylation reaction worked and that
App. 339
commercially feasible methods existed. The only question
faced by the inventors was whether cheaper, easier to
handle bases or solvents could be used. RX 3739-C; RX
3737-C.
FF D 22. The process claimed in the ‘035 patent
involves the conversion of an N-aryl amide, i.e., TZP or
acetyl-TZP, to an N-alkylated amide. RPX 4023; RPX 4015;
Kende Tr. 1457-1458; Pachter Tr. 1497.
FF D 23. In attempting to improve on the ’257 patent,
a person of ordinary skill in the art would have first
looked for art related to benzothiazepines. Then, one
would have looked for references to work in related
Systems, which in this case in N-aryl amides, because that
is the reactive part of the TZP molecule for alkylation.
The person of ordinary skill would have found references
such as Pachter et al., “Methylation of Some Amides in
Acetone,” 74 J. Am. Chem. Soc. 1321-22 (1952) (“Pachter
reference”); Worley et al., “2-Dialkylphosphony]- and
2-Alkylidene-3, 4-dihydro-3-0xo0-2H-1,4-ben-
zothiazepines,” 40 J. Org. Chem. 1731-34 (1975)
(“Worley”); Clark et al., “Synthesis and Analgesic Activ-
ity of 1,3-Dihydro-3-(substituted Phenyl) imidazo[4,5-
b]pyridin-2-ones and 3-(substituted pheny])-1,2,3-triaz-
olo[4,5-b] pyridines,” 21 J. Med Chem. 965-78 (1978)
(“Clark”); Nagarajan et al., “Condensed Hetrerotricycles:
Amino & Aminoalkyridibenz [b,f] [1,4] oxazepin-II(IOH)-
ones,” 12 Indian J. Chem. 236-46 (1974) (“Nagarajan”);
and Latif and Sattar, “A Note on the Alkylation of
Amides,” 32 J. Indian Chem 489-90 (1955) (“Latif”). Pach-
ter Tr. 1496-97.
App. 340
FF D 24. In connection with its efforts to synthesize
the diltiazem molecule, Abic assembled as many refer-
ences on benzothiazepines as it could. Significantly, Abic
found the field of benzothiazepines relevant, and did not
limit its research only to the N-alkylation of TZP. Abic
also conducted a chemical structure search. A chemical
structure search is a search based on a chemical nucleus,
regardless of what structure is attached to it. For this
search, Abic selected a six-member ring with a seven-
member ring attached to it, the seven-member ring con-
taining sulphur and nitrogen. This search would have
inchided the Krapcho patents, but not the Pachter or
Worley references because they do not have the ring
structures associated with benzothiazepines. Haber Tr.
2453-59.
FF D 25. It is not clear from the record whether Abic
personnel already had knowledge of the Pachter refer-
ence before the Abic literature search was conducted. No
one at Abic, especially Dr. Haber, is the hypothetical
person of ordinary skill in the art who is presumed to
have knowledge of all prior art. See Haber Tr. 2453-59.
B. The Prior Art
1. The Kugita ‘257 Patent
FF D 26. Complainants’ expert, Dr. Kende, distin-
guished the ‘035 patent over the ‘257 patent by stating:
It teaches the use of milder bases than the ‘257
in that sense certainly it’s mild conditions. Mild
ie ACL TF NORE a hc Fe) OE Soy
oe Eg ey ee ae oe ie & j oe
App. 341
in this sense means a base which is not so strong
that it will be irreversibly deprotonated.
Kende Tr. 1156.
FF D 27. During the Reexamination of the ‘035 patent,
the Examiner issued an Office Action wherein he stated
that “Kugita [’257] show the conventional process of pro-
duction of benzothiazepinones such as diltiazem by
alkylation... . ” Rx 1603 ({4); Taylor Tr. 2706.
FF D 28. In determining to accept Abic’s petition for
reexamination the examiner concluded that:
It would be obvious for one of ordinary skill in
the art to use the Pachter et al technique in the
Kugita et al. [ '257] process. Since the desirability
of Pachters’ technique has been long estab-
lished, it would be obvious to use it in a Process
such as that of Kugita et al. One would be
cess would be excellent.
RX 1603.
FF D 29. The ’257 patent teaches the N-alkylation of
the identical substrate of the ‘035 patent, TZP, using the
alkyl halide DMC.HC1 to yield the identical alkylated
Product. RX 3652: Taber Tr. 2166; RX 4038C.
FF D 30. The ’257 patent disclosed a process for the
N-alkylation of the identical substrate as the ‘035 patent
using as a base an alkali metal, alkali metal hydride, or
alkali metal amide, and as a suitable solvent, for example,
dioxane, toluene, xylene, or DMSO. RX 3652.
App. 342
FF D 31. The N-alkylation reaction disclosed in the
‘257 patent probably proceeds through the aziridinium.
Taber Tr. 2166; Gokel Tr. 850; RX 4038-C.
FF D 32. The ‘257 patent teaches that ben-
zothiazepinones, which are N-aryl amides, could be alky-
lated under rigorous conditions. Pachter Tr. 1499.
FF D 33. In looking to improve upon the ‘257 process,
a process development chemist would rapidly realize that
the reactive portion of the TZP molecule is what is known
as an “N-aryl amide”. RPX 4023; Pachter Tr. 1496.
FF D 34. The ‘257 patent (the only patent cited in the
‘035 patent) teaches that benzothiazepinones, which are
N-aryl amides, can be alkylated under rather rigorous
conditions (e.g., NaH /DMSO) using somewhat dangerous
bases that can result in explosions, solvents that are
inconvenient, and which result in low yields. Pachter Tr.
1499; RX 1229, 1460.
FF D 35. Given only the ‘257 patent and the Pachter
reference, a person would have had an “excellent” chance
(90 percent) of success, i.e., producing some yield even
with the possibility of side reactions. Pachter Tr. 1504,
1508-09, 1511-12; Taylor Tr. 2703.
2. The Pachter Reference
FF D 36. The Pachter reference was published in 1952
as a result of work done by Dr. Pachter towards his Ph.D.
thesis under the tutelage of Dr. Kloetzel. Pachter Tr.
1499-51; RX 3770; RX 3769.
oe
"ae ee fins “ta rages: Ap x diets hd i steno ik: ety PAS tiie ia yey
praia eh
PERE TE COT, Cae eee
ae Wretegd aa Oe
aed ile
App. 343
FF D 37. The N-aryl amide Structure, which is a par’
of TZP, is also a Part of each of the substrates alkylated by
Dr. Pachter in 1952. Kende Tr. 1458; RPX 3770B; RPX 4023.
FF D 38. RPX 4023 depicts an N-aryl amide. Pachter
Tr. 1497,
FF D 39. The Gabriel synthesis is the alkylation of an
activated amide. It is not an N-aryl amide. Every first
year organic chemistry student gets to learn what the
Gabriel synthesis is. In Studying the literature Pachter
found that sometimes the Gabriel syntheses is carried out
in acetone. RX 3769. The Pachter reference applied the
known Gabriel synthesis conditions to N-aryl amides. It
worked and he concluded that it appears to have general
application to N-aryl amides. Pachter Tr. 1529-30, 1601-02.
FF D 40. Pachter’s Process used the same base-so]-
vent combination as the ’035 patent, namely, potassium
hydroxide and acetone, in hydrous reaction conditions.
Pachter Tr. 1502-03; Kende Tr. 1277 and 1286-1287.
FF D 41. The Pachter System is a hydrous system.
Kende Tr. 1461.
FF D 42. Pachter disclosed the following N-alkylation
of an N-aryl amide using KOH and acetone:
[Diagram Omitted In Printing]
RX above represents an alkyl halide. RPX 3770B; Pachter
Tr. 1504-05; Kende Tr. 1456-57.
FF D 43. Pachter disclosed successful N-alkylation
using methyl iodide as the alkyl halide, in which R was
methyl (~-CH,) or phenyl [Diagram Omitted In Printing].
RPX 3770B; Pachter Tr. 1504-05.
ca Se
App. 344
FF D 44. Pachter’s KOH/acetone reaction conditions
are hydrous reaction conditions. Kende Tr. 1286-87.
FF D 45. It was important to have as much of the base
in solution as possible in order for the reaction to take
place rapidly. Therefore, Dr. Pachter chose potassium
hydroxide as the base because potassium bases are more
soluble in acetone than sodium bases. Pachter Tr. 1527-28.
FF D 46. Dr. Pachter chose acetone as the solvent
since he knew he could get his compounds into the
acetone solution quite readily and because it would pro-
vide a good medium for the reaction. Pachter Tr. 1528.
FF D 47. The Pachter reference discloses the
N-alkylation of several N-aryl amides using the base/
solvent combination KOH/acetone. Each of the N-alky-
lated amides has an aryl, or benzene ring, and a carbonyl.
The benzene ring and the carbonyl flank the nitrogen,
which is to be alkylated. Pachter Tr. 1504-05; CX 638A; RX
3770; RPX 3770B.
FF D 48. Pachter investigated the N-alkylation of
N-aryl amides over a range of conditions, including those
in which the amide was activated toward alkylation,
deactivated, and neither activated nor deactivated, thus
demonstrating the general applicability of his reaction
procedure. Pachter 1504-07; Kende Tr. 1458-59; RX 3770;
RPX 3770B.
FF D 49. Using the same KOH/acetone base-solvent
combination claimed in the ‘035 patent, Pachter discov-
ered and disclosed that alkylation can be accomplished
“conveniently and in good yield.” RX 3770.
ee
App. 345
FF D 50. In the five specific examples described by
Pachter, the yields of N-alkylated amides were from 81%
to 90%. RPX 3770B.
FF D 51. Although certain of the N-aryl amides alky-
lated by Dr. Pachter had potential alternative reaction
Sites, they did not interfere with the desired reaction of
the amide. Pachter Tr. 1508, 1532-33; RPX 3770B.
FF D 52. Some of the compounds alkylated by Pach-
ter were complex. Pachter Tr. 1568.
FF D 53. The (N-methylbenzamido) diphenylamine
compounds that Pachter alkylated had two Possible sites
for alkylation, the amide nitrogen and the amine nitro-
gen. Alkylation of the amine did not interfere with
: alkylation of the amide. Pachter Tr. 1508.
FF D 54. Dr. Pachter decided to use potassium
hydroxide as the base because he knew it was more
soluble in organic solvents like acetone or ethyl acetate
than the sodium base. Pachter Tr. 1527.
FF D 55. Dr. Pachter used potassium hydroxide and
acetone based upon the teachings of the prior art that
alkylation of an amide in what is known as the Gabriel]
synthesis succeeded with potassium hydroxide, but failed
under ‘257 conditions. RX 3769 at 46-47; Pachter Tr.
1528-30.
FF D 56. In 1952, when Dr. Pachter applied the condi-
tions of the Gabriel synthesis to his own N-alkylation
reactions, he was a little less than someone skilled in the
art because he had not yet received his Ph.D. degree.
Taylor Tr. 2798.
App. 346
FF D 57. The person of ordinary skill in the art
would do exactly what Dr. Pachter did in 1952 — if you
want to carry out a reaction on a substrate, you look at
what’s been done that’s analogous and see if it can be
applied to the system. The closer the analogy, the closer
the example from the literature, perhaps the greater con-
fidence one has. But there is a standard way of doing
organic chemistry and this is the way people skilled in
the art do it. Taylor Tr. 2798-99.
FF D 58. In 1952, Dr. Pachter concluded that the
alkylation procedure with potassium hydroxide and ace-
tone seems to have “general application.” Pachter Tr.
1530.
FF D 59. Following the publication of the Pachter
reference, the Pachter base-solvent combination of KOH/
acetone for the N-alkylation of aryl amides became well-
known and well-recognized by those of ordinary skill in
the art as a generally applicable procedure for the
N-alkylation of aryl amides. Taylor Tr. 2737, 2740.
FF D 60. In his declaration submitted to the Patent
office during the Reexamination, Dr. Baldwin suggested
that a paper by Yamawaki suggested that the Pachter
method is not general. RX 1658; Taylor Tr. 2740-2741.
FF D 61. Yamawaki’s experiments were not limited
to N-aryl amides. Taylor Tr. 2740-2741.
FF D 62. Respondents presented over a dozen refer-
ences in this investigation which describe Pachter-type
N-alkylation of N-aryl amides. Taylor Tr. 2744-2745.
FF D 63. In explaining why the N-alkylation reac-
tion occurs at one nitrogen rather than another in one of
App. 347
the amides discussed in his reference, Dr. Pachter
explained that under neutral conditions both nitrogens
are extremely weak bases. However, under basic condi-
FF D 64. Every attempt known to Dr. Pachter to
N-alkylate an N-aryl amide using Pachter conditions has
succeeded. Dr. Pachter knows of about 100 such N-alkyla-
tions. Pachter Tr. 1565, 1567.
FF D 65. The Prior art showed that Pachter condi-
tions worked for the N-alkylation of all N-aryl amides
and some others. Taylor Tr. 2702.
FF D 66. In 1981, no reference was known of in
which Pachter’s conditions did not work for the alkyla-
tion of an N-aryl amide. Today, no reference is known of
in which the use of Pachter’s conditions not to work for
the alkylation of an N-aryl amide. See Pachter Tr. 1565;
Kende Tr. 1286.
FF D 67. Complainants’ expert agreed with the
remark made by the examiner that by 1981, the “desir-
ability” of the Pachter KOH/acetone technique has long
been established. Complainants’ expert did not, of
Course, take the position that the “applicability” of the
technique had been established. Kende Tr. 1280.
FF D 68. In Pachter’s Process, water is formed in the
reprotonation step. Kende Tr. 1287.
FF D 69. The alkylation of an amide under Pachter
conditions produces water as a side product. Gokel Tr.
1039-40; RPX 1096.
App. 348
FF D 70. The Pachter reference teaches one of ordi-
nary skill in the art that one can alkylate an amide under
hydrous conditions. Kende Tr. 1286-87, 1461; CX 638A.
FF D 71. During the reexamination of the ‘035 pat-
ent, the examiner stated that “Pachter et al. show the
widely used alkylation of aryl amides.” RX 1603 (PTO
Office Action); Taylor Tr. 2706-07.
FF D 72. Several references describe Pachter condi-
tions in general terms, e.g., Worley (RPX 1094), Johnstone
(RX 1137), Latif and Sattar (RX 1605), Clark (RPX 1093).
Kende Tr. 1290-99, 1306-08, 1380; Pachter Tr. 1509.
FF D 73. The process taught in the Pachter reference
was an improvement over earlier processes because it
achieved the N-alkylation reaction by switching the
known bases and solvents (later disclosed in the ‘257
patent) to potassium hydroxide/acetone (those later
described in the ‘035 patent). In his paper, Dr. Pachter
showed that in relatively short reaction times under very
convenient conditions, one could rapidly and in good
yield produce the necessary compound. Indeed, Dr. Pach-
ter’s paper teaches that some compounds are inactive
under ’257 conditions, but easily alkylated under Pachter
conditions. Kende Tr. 1284-85, 1460-61; Pachter Tr.
1502-04, 1524; CX 1, 638A; RX 3770; RPX 3770B.
FF D 74. The Pachter reference disclosed that the
usual method for alkylating N-aryl amides until his pub-
lication included the use of dangerous metals, metallic
sodium, or sodium hydride in inert solvents (i.e., ‘257
conditions). Pachter Tr. 1501-02.
App. 349
FFD 75. Pachter, in 1952, had taught that the substi-
tution of KOH/acetone for the base-solvent combinations
used in the ’257 patent would avoid the dangers and
inconveniences of such bases and solvents and could
actually increase yields. RX 3770; Pachter Tr. 1501-03.
FFD 76. The specification of the ‘035 patent is simi-
lar to the first few Paragraphs of the Pachter reference,
€.g., both describe Previous methods as inconvenient,
dangerous, and resulting in low yields. Indeed, Dr. Pach-
ter initially thought the ‘035 patent drafters “copied Ppara-
graph 1” of his Paper. Pachter Tr. 1503-04; Cx 1; RX 3770.
FF D 77. Pachter recognized the problem which
according to complainants, the 035 patent is said to have
solved. Pachter disclosed in his 1952 article that, “[t]he
usual method for the alkylation of amides, involving
metallic sodium and an inert solvent is at best a rather
inconvenient and somewhat dangerous Procedure.” Pach-
ter then suggested replacing the sodium, i.e., a ‘257 base,
with the KOH/acetone system, the same substitution pro-
posed by the ’035 patent. RX 3770; CX 1; RX 3652; Pachter
Tr. 1503-04; Kende Tr. 1284-85.
FF D 78. Using the same KOH / acetone base-solvent
combination claimed in the ‘035 patent, Pachter disclosed
that the N-alkylation can be “accomplished conveniently
and in good yield” in a relatively short Period of time. RX
3770.
FF D 79. Given the Pachter reference, all the prior
art that discusses Pachter as a general procedure, and
ignoring the Possibility of side reactions, complainants’
€xpert admitted that it would have been obvious that
TZP can be alkylated with methyl iodide under Pachter
ee eae ee
App. 350
conditions to give at least a yield of 10% of desired
product. Kende Tr. 1318-19.
3. Worley
FF D 80. Worley describes the successful N-alkyla-
tion of a N-aryl amide lactam under Pachter conditions
using alkylating agents methyl iodide and ethyl bromo
acetate, reporting a 73% yield. Kende Tr. 1290-92; Pachter
Tr. 1510-11; RPX 1094.
FF D 81. Lactams, including TZP, are cyclic amides.
Kende Tr. 1290.
FF D 82. Worley taught that Pachter conditions can
be applied to a lactam (a cyclic amide) as well as to
Pachter’s cyclic amides. Kende Tr. 1291; Pachter Tr.
1510-11.
FF D 83. The compound alkylated by Worley had a
sulfur atom which, like the sulfur atom of TZP, can trans-
mit its effects through the aromatic ring down to the
nitrogen. If the sulfur atom of TZP were to affect the
N-alkylation reaction of Pachter, such a deleterious effect
would have been seen in Worley. Worley obtained a good
yield when using Pachter conditions. Pachter Tr. 1511.
FF D 84. Worley provided assurance that the sulfur
atom in the TZP ring would not inhibit the N-alkylation
reaction. Taylor Tr. 2703.
FF D 85. The amide group in Worley has approxi-
mately the same acidity as the amide group in TZP.
Kende Tr. 1304.
App. 351
FF D 86. Worley is a six-membered ring. In terms of
ease of alkylation, a distinction between the six-mem-
bered ring of Worley and the seven-membered TZP ring
is not necessary. On size alone, complainants’ expert
would not draw any distinction between six mer.ber
rings and seven member rings. Kende Tr. 1310.
FF D 87. In 1975, Worley used KOH/ acetone in the
following N-alkylation reaction, Stating that the pro-
cedure used was the “general procedure of Pachter and
Kloetzel for the alkylation of [N-aryl] amides with
potassium hydroxide in acetone” (the N-aryl amide struc-
ture shown in bold type):
[Diagram Omitted In Printing]
RX 3824; RPXx 3824A; Kende Tr. 1291-93; Pachter Tr.
1509-10.
FF D 88. The Worley compound is a very good
model for TZP. Both compounds are N-aryl amides; both
compounds have heterocyclic ring systems; both com-
pounds are aromatic and both compounds have sulfur in
the same position. Taylor Tr. 2738-2739,
FF D 89. Worley teaches using the “general pro-
cedure of Pachter.” Pachter Tr. 1509-10; RX 3824 at 1733.
FF D 90. Worley does not report any reaction (or
side reaction) of the sulphur atom. Kende Tr. 1303; RPX
1094,
FFD 91. Dr. Taylor believes that Worley, which uses
a substrate having a 6-membered heterocyclic ring, is
closer prior art to the ’035 patent than Nagarajan, which
uses a substrate with a 7-membered heterocyclic ring
(like TZP) but is an Oxazepinone. Taylor Tr. 2708-10.
App. 352
FF D 92. Given the ‘257 patent, the Pachter refer-
ence, as well as Worley, it would have been even more
obvious that one could alkylate TZP under the general
Pachter conditions - chances of success would have
increased to 95% since the Worley compound is more
similar to TZP in that it is a lactam and it also contains a
sulfur atom. Pachter Tr. 1511-12; Taber Tr. 2181-82; Taylor
Tr. 2703.
4. Johnstone
FF D 93. In 1969, a technical article, Johnstone et al.,
“A Rapid Method of N-alkylation of Amides,” 16 J.
Chem. Soc. 2223-24 (1969) (“Johnstone”), reported the use
of Pachter conditions to alkylate a substrate that is not an
N-aryl amide, calling Pachter “a singular example of easy
alkylation-of an amide. .. . ” RX 3848; Kende Tr. 1294.
FF D 94. In Johnstone, a base/solvent combination
of potassium hydroxide/acetone worked, whereas
sodium carbonate/acetone did not work. Kende Tr.
1296-97; RX 1137.
FF D 95. The Johnstone use of the Pachter reference
and the use of Pachter conditions is one further indica-
tion that people working on amides looked to reactions
performed on other amides, even if they involved very
different substrates. The Johnstone authors managed very
successfully to use what they termed an “easy alkyla-
tion.” Taylor Tr. 2710-11.
FF D 96. Johnstone shows an appreciation of the
potential generality of the Pachter technique, and that the
S's
App. 353
Pachter technique was in fact used as a general technique.
Taylor Tr. 2711,
5. Clark
FFD 97. In 1978, Clark et al. reported the following
N-alkylation reaction, where the N-aryl amide “was alky-
lated with alkyl halide and refluxing acetone solution in
the presence of Powdered potassium hydroxide accord-
ing to the method of Pachter and Kloetzel”:
[Diagram Omitted In Printing]
RX 3841 (the N-aryl amide Structure is in bold type);
Kende Tr. 1306-07.
FF D 98. Clark discloses 40 examples of hydrous
reactions on 40 compounds using Pachter conditions (a
tone), and reports Satisfactory yields. Kende Tr. 1306-08;
Taylor Tr. 2772-73; RPX 1093.
FF D 99. The N-alkylation reaction reported in
Clark is hydrous. Kende Tr. 1306.
FF D 100. Clark reports the use of Pachter condj-
tions with dialkylaminoethyl chloride: a dia-
{kylaminoethylating agent which, like DMC, reacts
through the aziridinium ion. Kende Tr. 1308; Taylor Tr.
2772-73.
App. 354
6. Latif
FF D101. Latif refers to Pachter as a general process
for the alkylation of amides that is applicable for almost
all types of alkyl halides. RX 1605.
FF D 102. Latif observed some limitations for use of
the Pachter technique, but not with respect to any N-aryl
amide. RX 1605; Pachter 1573-74.
7. Nagarajan
FF D 103. In the May 31, 1994 Notice of Intent to
Issue Reexamination Certificate, the examiner stated:
Perhaps the most pertinent references are the
British Patent and Nagarajan et al. Both of these
references show the aminoalkylation of lactams
which bears some structural relationship to
those of Kugita et al. using a process similar to
Pachter et al. However, all of the compounds
which are amino alkylated contain an activating
nitro group when the Pachter-type process is
employed. Nagarajan et al. shows that where no
activating nitro group is present that the more
harsh methods, similar to those of Kugita et al.,
must be employed. This indicates that where,
activating nitro group is not present, the Pachter
et al. technique is not operable. This teaches
away from the process of Gaino et al. the patent
being reexamined here.
RX 1654.
FF D 104. Nagarajan used ‘257 conditions on the
unsubstituted, i.e., no nitro-substitution, compounds
EOS’~S'S ‘-
App. 355
because these were the conditions everyone was using.
Pachter Tr. 1526.
Tr. 1361. Dr. Kende submitted a declaration during the
reexamination wherein he Suggested that the procedure
FF D 106. Under ’257 conditions, a nitro group is a
deactivating group, Pachter Tr. 1588.
FF D 107. To state that the presence of the nitro
8Toup could make alkylation easier or harder would be
speculation. Kende Tr. 1365.
FF D 109. Nagarajan started with ‘257 conditions
and then switched to Pachter conditions. Pachter Tr.
1587-88. Nagarajan went to Pachter conditions because
alkylation did not process under ’257 conditions. Pachter
Tr. 1542,
FFD 110. Dr. Kende, in his declaration submitted to
the examiner during reexamination, on Page 17 refer-
enced Nagarajan’s discussion of ting cleavage. The ring
App. 356
cleavage discussed in Nagarajan would not occur with
TZP. RX 3132; Kende Tr. 1367-68.
FF D111. Dr. Baldwin, in his declaration submitted
to the examiner during the reexamination, stated that
“the Nagarajan reference suggested that ring cleavage
was a distinct possibility under Pachter base/solvent con-
ditions of seven membered oxazepines. (See, Experimental,
page 245(d)). This too would have taught away from the
process of the ‘035 patent.” RX 1658. Dr. Kende stated
that he did not see the connection between Nagarajan and
the ‘035 substrate. RX 1658; Kende Tr. 1368.
FF D 112. There is no connection between ring
opening reported in Nagarajan and alleged ring opening
in TZP. Kende Tr. 1375; Taylor Tr. 2720.
FF D113. As expressly stated in Nagarajan, the ring
cleavage in Nagarajan depends upon the presence of the
nitro group. There is no nitro group present in TZP.
Taylor Tr. 2712, 2720-22.
FF D 114. In his declaration submitted to the PTO,
Dr. Baldwin identifies ring cleavage as a consequence of
the nitro group. RX 1658; Taylor Tr. 2719. Dr. Baldwin’s
declaration was misleading regarding Nagarajan, the
teaching of ring cleavage, and the possibility that it
would suggest ring cleavage in the ‘035 case where there
is no nitro group. Taylor Tr. 2723.
FF D 115. Nagarajan did not report ring cleavage
with potassium carbonate and acetone. Kende Tr. 1369.
FF D 116. Nagarajan reported ring cleavage with
’257 conditions and no ring cleavage under Pachter con-
ditions. Taylor Tr. 2725-27, 2729-31; Kende Tr. 1369. If
App. 357
there is no nitro group, as there is none with TZP, there is
no problem of ring cleavage, and also, under Pachter
conditions there is no problem of ring cleavage. Kende Tr.
1369-70; Taylor Tr. 2720-22.
FF D 117. Contrary to the examiner’s Opinion, the
Nagarajan paper does not teach that a hydrous system for
N-alkylation will not work with compounds that are
unsubstituted with the nitro substituent. Pachter Tr. 1523.
FF D 118. Page 245, Procedure D of Nagarajan
shows that Nagarajan, like Fones, attempted to alkylate
with DMC using ’257-like conditions, heating for a long
time (4 hours), resulting in 9% yield, 35% Starting mate-
rial and the remainder as decomposed material. RX 3820;
RPX 3820A; Pachter Tr. 1524-25.
FF D 119. Nagarajan (RX 3820) methylated (N-alky-
lated) with methyl iodide. After Succeeding with methyl
iodide, Nagarajan used DMC. Pachter Tr. 1534.
8. Burton
FFD 120. The N-alkylation of aryl amides using the
base-solvent combination K,CO,-acetone was taught as
early as 1968 by Burton et al., “Halogeno-o-phe-
nylenediamines and Derived Heterocycles Part I. Reduc-
tive Fission of Benzotriazoles to O-Phenylenediamines,”
10 J. Chem. Soc., 1268-73 (1968) (“Burton”). Burton dis-
closed the following N-alkylation reaction using K,CO,-
acetone:
[Diagram Omitted In Printing]
RX 3794; Taylor Tr. 2273-74.
App. 358
9. The Branca ’522 Patent
FF D121. Based upon Dr. Baldwin’s declaration, the
examiner stated that the Pachier reference would have
provided little if any guidance, regarding the use of DMC
or its hydrochloride salt in the N-alkylations of the ‘035
patent using a Pachter-type base/solvent combination.
The examiner stated that the equivalence of methyl
iodide and DMC has not been demonstrated using the
conditions of the ‘035, but only under the harsher condi-
tions employed in the ‘257 patent. Taylor Tr. 2764-65.
FF D 122. In connection with the Order Granting
Request for Reexamination, dated May 2, 1994, the exam-
iner stated in part as follows:
The equivalence of methyl iodide and DMC has
not been demonstrated using the conditions of
‘035 but only where more strongly forcing con-
ditions are employed in ‘257.
RX 1653.
FF D 123. Dr. Taylor’s opinion is that prior art not
of record before the PTO shows that using the conditions
of the ‘035 patent, and for the purposes of alkylating TZP,
methyl iodide and DMC are equivalent. Taylor Tr.
2764-65.
FF D 124. U.S. Letters Patent 4,377,522, issued to
Quirico Branca in 1983, is prior art based on its filing date
before the Japanese counterpart to the ‘035 patent. The
examiner did not have the ‘522 patent during the reex-
amination of the ‘035 patent. RX 1657; Kende Tr. 1312.
App. 359
FFD 125. The ’522 patent discloses the alkylation of
a seven member ring N-aryl amide using potassium car-
bonate/acetone and the alkylating agents DMC, DEC, or
methyl iodide. Taylor Tr. 2757-59; Kende Tr. 1311-13; RX
1657; RPX 1091.
FF D 126. Branca ’522 discloses the alkylation of a
seven member ring lactam using DMC with a weak inor-
ganic base, such as alkali metal carbonate (¢.g., potassium
FF D 127. Branca is a N-aryl amide seven-mem-
bered ring Structure, a benzodiazepine. It is a seven-
membered ring benzodiazepinone where the N-ary]
amide linkage is the same as it is in TZP. Taylor Tr.
2755-56.
FF D 128. RPX 1027, 1019, 1091, and 1092 are all
examples of benzodiazepines. Gokel Tr. 1012-14.
FF D 129. Benzodiazepines are related to ben-
zothiazepines in that they have a six member ring fused
to a seven member ring and they have the amide, how-
ever, they lack the sulfur. Gokel Tr. 1014.
FF D 130. The Branca ‘522 patent discloses an
alkylation reaction of a seven member ring N-aryl amide
using DEC, DMC or methyl iodide with potassium carbo-
nate and acetone. Kende Tr. 1311-1316; Taylor Tr. 2757-58;
RPX 1095.
FFD 131. The Branca patent provides an example of
the kind of art the examiner said was not before him,
Showing the equivalence of methyl iodide and DMC.
Branca provides an example of a substrate similar to TZP
App. 360
that is alkylated under ‘035 conditions with a dia-
Ikylaminoethyl halide and methyl] iodide. Taylor Tr. 2759.
FF D 132. Prior to the alleged invention of the ‘035
patent, Burton (RX 3794), Fischli (RX 2130), Branca (RX
1657), Bebenburg (RX 1655), Nagarajan (RX 3820) and
Nadzan (RX 3834) disclosed the use of the ‘035 base-
solvent combinations to alkylate N-aryl amides. Taylor Tr.
2773-74.
10. The Bebenburg ‘887 Patent
FF D 133. U.S. Letters Patent 3,910,887, which
issued to Walter von Bebenburg in 1975, was not of
record during the reexamination on the ‘035 patent. It
discloses a seven member N-aryl amide ring alkylation
using DMC as the alkylating agent, and potassium carbo-
nate/acetone as the base/solvent combination. Kende Tr.
1313-14; Taylor Tr. 2757-58; RX 1655; RPX 1092.
FF D 134. The ’887 patent suggests that one can
N-alkylate a seven-membered ring using methyl iodide
or DMC and a base/solvent of potassium carbonate/
acetone. Kende Tr. 1313-14; Taylor Tr. 2761-63; RX 1655;
RPX 1092.
11. The ‘338 Schenker Patent
FF D 135. U.S. Letter Patent 3,644,338, which issued
in 1972 to Karl Schenker, is not of record in the reexam.
See Taylor Tr. 2747. Schenker discloses the alkylation of a
compound which, although not an aryl amide, is an
amide with a seven-membered ring. The reaction uses
App. 361
DMC as one of the Possible alkylating agents, potassium
carbonate as a Possible base, and acetone as a possible
solvent. Taylor Tr. 2747-49, 2751, 2754; RX 1656; RPX 1095.
FF D 137. Dr. Pachter testified that benzamides
were substrates with which his conditions did not always
work. Taylor Tr. 2747-48.
The Schenker patent teaches that:
the kind of solvents that are capable of solvating
Potassium, such as a polar solvent. Taylor Tr. 2749.
FF D 139. Schenker also teaches that one should
avoid strong bases (such as those found in the ‘257 pat-
ent) because their use results in low yields. Taylor Tr.
2749-50; Kende Tr. 1315-16; RX 1656.
App. 362
FF D 140. Example 2 of Schenker discloses finely
grounded potassium carbonate in acetone, with
DMC-HCI. Taylor Tr. 2751-52; RX 1656.
FF D 141. Example 5 of Schenker discloses finely
ground potassium carbonate in acetone, with DMC-HCI.
Taylor Tr. 2752-53; RX 1656.
FF D142. Abic’s expert is of the opinion that Schen-
ker is closer art to the ‘035 patent than the Nagarajan
article. Schenker discloses the same base/solvent combi-
nation and alkylating agent disclosed in the ‘035 patent,
whereas Nagarajan used sodium hydroxide (a different
base) and acetone in a homogenous solution. Taylor Tr.
2754-55.
FF D143. Abic’s expert is of the opinion that Schen-
ker is closer to the ‘035 patent than any reference of
record. Taylor Tr. 2754-55.
12. The Krapco [sic] ‘006, ‘889, ‘967 and ‘902
Patents
FF D 144. The ’006 and ’889 patents, which contain
an identical Example 1B, teach the N-alkylation of ben-
zothiazepinones using sodium hydroxide and toluene in
a system that generates water. RX 3669; RX 3673; Taber Tr.
2160; Gokel Tr. 886-93; RX 4038C; Kende Tr. 1430-31.
FF D 145. The reaction described in Example 1B of
the ‘006 and ’889 patent is illustrated as follows:
[Diagram Omitted In Printing]
RPX 3673A; Gokel Tr. 887.
ee ee ee a ey Se,
App. 363
FFD 146. An organic chemist of Ordinary skill back
in 1980-81 would have recognized that the N-alkylation
taken to dry the toluene, glassware and other equipment
of the reaction system. RX 4038C.
at the two position of TZp Taber Tr. 2161; Gokel Tr. 888.
FFD 148. Dr. Baldwin believes that if TZp was used
as the starting material in the N-alkylation Process of
Example 1B of the ‘006 and ’889 patents, that process
would be equivalent to the Process of the ‘035 patent. RX
4038C.
FF D 149. Dr. Baldwin doesn’t “know one way or
the other” whether a chemist with the ’889 patent in front
of him would have tried the reaction in Example 1B with
TZP. Dr. Baldwin agreed that “[iJt is a Possibility” that the
chemist would have tried the reaction with TZP, having
seen that the Starting substrate in Example 1B was “anal-
ogous” to TZP. RX 4038C.
FF D 150. The following question and answer
Occurred at Dr. Baldwin’s October 6, 1994 deposition:
Q So if you had a series of reactions in which
various analogous substrates were treated
with the same reaction conditions, same
base, same solvent, same alkylating agent,
App. 364
same temperature, same stirring conditions,
then you would be able to hazard a guess as
to what another one of the set of analogous
structures would do under the same condi-
tions?
A I think you might be in a position to make
some guesses, yes.
RX 4038C
FF D 151. The reaction disclosed in Example 1B of
the ‘889 and ’006 patents would involve the aziridinium
ion as the alkylating agent. Gokel Tr. 895; RX 4038C;
Taylor Tr. 2775.
FF D 152. The reaction system in Example 1B of the
‘006 and 889 patents is a “reversible” reaction system. RX
4038C.
FF D 153. The alkylating agent used in Example 1B
of the ‘889 and ‘006 patents, dimethyla
This text is long and has been trimmed here. Open the source document for the complete record.
This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.