`O>;,Wl1ll
`
`'3")l:t tll'1':l::1'3'!ll tll'\!l)ltl )l:;t\)
`, 111l/ ,1nw1 ~,:inin .w :,11:i ,,,-,,v
`
`, Nlil )>,->:i >:iso l11'17Y.l 111:'1;i 0),\/Y.l .,WN
`65202 :l'lN-'.m ,31 0)111 ,nN '111
`03-5660974 , '17i'!l ; 03-5670633 , '?\:J
`
`MERCK & CO., INC., U.S.A .
`1"1l/ ,'llWl 111:, nr.i,w 111, :,11::i. ,,,-,,v
`, Nl11 )'1->:J. >:ti,:, l1N:'1Y.l11 111:'1;> 0),\/Y.l 1WN
`62038 :i>jN '.m ,11490 :r.n ,124 !111>:il pN
`03-5272666, 'Oi'!l; 03-5271919, ;>\:J
`
`Dr. Leonard .J. Chy all '1l'"1 n11n
`
`I
`I
`. ' !
`
`I
`
`. '
`'
`
`: I
`
`'
`
`1)80044/5480/230154711
`
`Merck Exhibit 2192, Page 1
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`SECOND DECLARATION OF LEONARD J. CHYALL, PH.D.
`
`I, the undersigned, Dr. Leonard J. Chyall, U.S. Passpo1t No. 432624896, with a
`
`business address of Chyall Pharmaceutical Consulting LLC, 3000 Kent Avenue, Suite
`
`Dl-105, West Lafayette, Indiana 47906, USA, having been wamed that I must state the
`
`truth and that I shall be liable to the penalties prescribed by law should I fail to do so,
`
`hereby declare in writing as follows:
`
`1.
`
`I am the same Leonard J. Chyall who submitted a declaration dated
`
`August 3, 2010 (the "First Declaration"), in support of the position of Teva
`
`Pha1maceutical Industries Ltd. ("Teva") in the proceedings before the Honorable
`
`, I
`
`Registrar of Patents regarding Israel Patent Application No. 172563 ("the Patent
`
`Application"), filed by Merck & Co. Inc., U.S.A. ("the Applicant").
`
`2.
`
`This declaration was prepared in response to the declarations of Prof. Jerry
`
`L. Atwood, Dr. Robert M. Wenslow and Mr. Robert Di Vincenzo, submitted on behalf of
`
`Merck on June 1, 2011.
`
`3.
`
`The fact that I have not commented on any particular point in the
`
`Declarations of Prof. Atwood, Dr. Wenslow or Mr. Di Vincenzo should not be taken to
`
`imply that I accept or agree with that point. There is nothing in these declarations that
`
`causes me to change the views that I expressed in my First Declaration.
`
`I
`
`I.
`
`PROF. ATWOOD'S ASSERTION THAT ONE COULD PREPARE SALTS
`OTHER THAN SITAGLIPTIN DIHYDROGENPHOSPHATE IS NOT
`CREDIBLE
`
`4.
`
`Prof. Atwood performed experiments to demonstrate that salts of
`
`sitagliptin and phosphoric acid other than a sitagliptin dihydrogenphosphate salt (a l: 1
`
`Merck Exhibit 2192, Page 2
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`salt) 1 could be prepared. As explained below, it is my opinion that Prof. Atwood's
`
`experiments used poor experimental techniques and did not demonstrate the preparation
`
`of salts other than the reasonably expected sitagliptin dihydrogenphosphate salt. Prof.
`
`Atwood's poor experimental techniques defeat his argument that one can prepare salts of
`
`sitagliptin and phosphoric acid other than a 1: 1 salt. Indeed, in one instance Prof.
`
`Atwood's poor experimental techniques led him to conclude that he had prepared a
`
`bis(sitagliptin) phosphoric acid salt (a 2:1 salt), when X-Ray Powder Diffraction
`
`("XRPD") analysis of the reaction product proves that Prof. Atwood actually prepared
`
`the reasonably expected sitagliptin dihydrogenphosphate salt.
`
`' I
`
`/
`
`A.
`
`5.
`
`Prof. Atwood Used Poor Experimental Techniques
`
`All of Prof. Atwood's experimental protocols had significant flaws,
`
`including use of irregularly high reactant concentrations and failure to filter and wash
`
`reaction products. These significant protocol flaws render Prof. Atwood's experimental
`
`conclusions unreliable and without scientific credibility.
`
`6.
`
`Irregularly High Reactant Concentrations: Prof. Atwood's
`
`experiments employed concentrations of sitagliptin and phosphoric acid that were so high
`
`that they caused his reaction mixtures to solidify. It is highly irregular and improper to
`
`conduct chemical reactions at concentrations that cause the entire reaction mixture to
`
`solidify. An organic chemistry reaction, such as the addition of an acid to an organic
`
`base, should take place under homogenous (i.e., uniform) conditions. Conducting the
`
`reaction with enough solvent to ensure adequate mixing throughout the experiment
`
`facilitates these conditions. One cannot ensure adequate mixing and homogeneous
`
`1
`
`A sitagliptin dihydrogenphosphate salt is a salt comprising one molecule of sitagliptin per one
`molecule of phosphoric acid.
`
`Page 2 of39
`
`Merck Exhibit 2192, Page 3
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`. \
`
`i
`
`\ !
`
`conditions if a reaction is conducted with too little solvent and the reaction mixture
`
`solidifies during the course of the reaction. The resulting inadequate mixing may lead to
`
`incomplete chemical reactions and may also lead to an increased formation of unwanted
`
`byproducts. For this reason, proper experimental technique would have entailed the use
`
`of enough solvent to suspend in solution any solids that formed from the chemical
`
`reaction. In my opinion, Prof. Atwood's use of too little solvent likely resulted in
`
`elevated levels of impurities, such as unreacted starting materials and/or reaction by(cid:173)
`
`products, in the solid products that he obtained.
`
`7.
`
`Failure To Filter And Wash: Filtration (with washing) is routinely used
`
`to separate reaction products from unreacted starting materials, impurities, and
`
`byproducts. Furthermore, such steps are essential if one seeks to use elemental analysis to
`
`detennine the identity of the product. In his experiments, Prof. Atwood failed to filter
`
`and/or wash the solid reaction products that he recovered to remove such unreacted
`
`starting materials, impurities and byproducts.
`
`8.
`
`In my opinion, even if Prof. Atwood had used the best possible protocols
`
`to conduct his experiments, the reaction products that he recovered would have required
`
`filtering and/or washing to remove unreacted starting materials, impurities and
`
`byproducts. Prof. Atwood, moreover, did not use the best possible protocols. As
`
`explained above, he used too little solvent, which likely resulted in the presence of
`
`elevated levels of unreacted starting materials, impurities and byproducts. The presence
`
`of such elevated levels of unreacted starting materials, impurities and byproducts
`
`increases the negative impact of Prof. Atwood's failure to filter and wash the reaction
`
`products that he recovered.
`
`Page 3 of39
`
`Merck Exhibit 2192, Page 4
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`9.
`
`Rather than use filtration ( and washing) to isolate and purify his solid
`
`reaction products, Prof. Atwood used vacuum drying to isolate solids from his reaction
`
`mixture. When vacuum drying is used to remove solvent from a reaction mixture, only
`
`the volatile materials are removed and the non-volatile materials are left behind in the
`
`product. The consequence of this drying protocol is that to the extent they are not
`
`volatile, unreacted starting materials, impurities and by-products will remain in the solid
`
`product after vacuum drying.
`
`10.
`
`Sitagliptin and phosphoric acid are not volatile. Accordingly, any
`
`unreacted sitagliptin or phosphoric acid in Prof. Atwood's experiments would have ended
`
`up in the solid product that Prof. Atwood recovered after vacuum drying.
`
`11.
`
`The absence of a filtration/washing step in Prof. Atwood's experiments
`
`virtually guaranteed that the solids that Prof. Atwood recovered at the end of his
`
`experiments contained the same ratio of sitagliptin to phosphoric acid as the starting
`
`materials, regardless of what type of salt (if any) formed. This is because Prof. Atwood's
`
`recovered solids would contain virtually all of the sitagliptin and phosphoric acid as
`
`either unreacted starting material or as a solid sitagliptin/phosphoric acid salt (i.e., the
`
`reaction product), regardless of the stoichiometry of that salt.
`
`B.
`
`Prof. Atwood's Identification Of The Products Of His Experiments
`Lacks Scientific Credibility
`
`12.
`
`Prof. Atwood used a number of analytical techniques to analyze the
`
`products of his experiments and to determine whether he prepared a salt other than
`
`sitagliptin dihydrogenphosphate. The techniques that Prof. Atwood used, and the results
`
`he obtained, prove in one instance that Prof. Atwood obtained a sitagliptin
`
`Page4 of39
`
`Merck Exhibit 2192, Page 5
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`i
`
`' !
`
`dihydrogenphosphate salt ( a 1: 1 salt), and in all other instances the analytical results are
`
`insufficient to identify the product of Prof. Atwood's experiments.
`
`1.
`
`Prof. Atwood's Experimental Procedures Render Unreliable
`Any Conclusions Drawn From His Elemental Analysis Data
`
`13.
`
`One of the techniques used by Prof. Atwood to identify the reaction
`
`products is elemental analysis. Elemental analysis measures the relative amounts of
`
`individual elements in a product, such as carbon, nitrogen, and hydrogen. Elemental
`
`analysis cannot, however, distinguish between a mixture of unreacted starting materials, a
`
`salt of the same materials, or a combination of starting materials and salts. The following
`
`different products, for example, would all result in the same elemental analysis data
`
`because they all contain the same elements in the same ratios:
`
`a)
`
`b)
`
`c)
`
`Sitagliptin dihydrogenphosphate salt (a 1: 1 salt) as a mixture with
`unreacted sitagliptin where the amount of unreacted sitagliptin is the same
`as the amount of sitagliptin in the 1: 1 salt;
`
`A simple 2: 1 mixture of sitagliptin and phosphoric acid where no salt was
`formed; and
`
`Bis(sitagliptin) phosphoric acid salt (a salt with a 2: 1 ratio of sitagliptin to
`phosphoric acid) if such a salt could be prepared.
`
`14.
`
`Elemental analysis aids in the structural elucidation of novel chemical
`
`species such as salts of organic compounds when proper protocols are used for the
`
`synthesis of these salts, such as filtering and/or washing a product to remove any
`
`unreacted starting materials and other impurities. Even then, other characterization
`
`techniques designed to probe the chemical structure of the material must be used to
`
`confirm that a salt was obtained, and to confirm the identify of that salt.
`
`15.
`
`By design, as explained above, the solid material obtained in Prof.
`
`Atwood's experiments contained sitagliptin and phosphoric acid in the same molar ratio
`
`Page 5 of39
`
`Merck Exhibit 2192, Page 6
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`as that of the sitagliptin and phosphoric acid starting materials, regardless of what type of
`
`salt fo1med. Prof. Atwood did not perform a routine filtration step in any of his
`
`experiments and, therefore, failed to remove non-volatile um-eacted starting materials,
`
`impurities, and byproducts.
`
`16.
`
`Prof. Atwood's failure to remove non-volatile um-eacted starting materials,
`
`impurities, and byproducts renders his elemental analysis data scientifically um-eliable for
`
`i
`.J
`
`purposes of proving that either a bis(sitagliptin) phosphoric acid salt or a sitagliptin
`
`bis(phosphoric acid) salt was obtained. Prof. Atwood's elemental analysis data merely
`
`confirms his chosen molar ratio of sitagliptin and phosphoric acid starting materials.
`
`When Prof. Atwood chose to perform an experiment with a 2:1 molar ratio of sitagliptin
`
`to phosphoric acid, elemental analysis of his product necessarily had a 2: I molar ratio of
`
`sitagliptin to phosphoric acid. Similarly, when Prof. Atwood chose to perform an
`
`experiment with a 1 :2 molar ration of sitagliptin to phosphoric acid, elemental analysis of
`
`his product necessarily had a 1 :2 molar ratio of sitagliptin to phosphoric acid. These
`
`results are not proof of the molar ratio of any obtained salt because elemental analysis
`
`cannot distinguish between salt products and mixtures of unreacted sitagliptin and/or
`
`phosphoric acid.
`
`17.
`
`Elemental analysis is a powerful tool when one uses appropriate
`
`experimental techniques to ensure the isolation of a purified product. Prof. Atwood
`
`failed to use appropriate experimental techniques to isolate his products, rendering his
`
`elemental analysis data unreliable for purposes of salt identification.
`
`Page 6 of39
`
`Merck Exhibit 2192, Page 7
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`·,
`I
`
`2.
`
`Prof. Atwood's Reliance On X-Ray Powder Diffraction
`Analyses Do Not Prove That Prof. Atwood Obtained A
`Bis(Sitagliptin) Phosphoric Acid Salt Or A Sitagliptin
`Bis(Phosphoric Acid) Salt
`
`18.
`
`In addition to elemental analysis data, Prof. Atwood also offered XRPD
`
`data in an attempt to prove that he recovered salts other than sitagliptin
`
`dihydrogenphosphate. Prof. Atwood's XRPD results, however, do not prove that he
`
`synthesized any salt other than sitagliptin dihydrogenphosphate.
`
`19.
`
`Specific crystalline salts have unique XRPD patterns. In some
`
`circumstances, that characteristic of crystalline salts allows one to use the unique XRPD
`
`pattern of a salt to identify the salt. For example, if one prepared a crystalline salt while
`
`using proper experimental procedures, which would include purification and isolation by
`
`filtration, then one can deduce the identity of the obtained crystalline salt from the molar
`
`ratios of the reactants used. If an XRPD pattern is obtained for this crystalline salt then
`
`this pattern can be used as a reference pattern for the identification of this salt from other
`
`experiments.2 Therefore, a characteristic XRPD pattern can be used as a reference XR.PD
`
`pattern for pattern matching analysis to determine the chemical identity of an unknown
`
`salt. If the unknown salt prepared by combining sitagliptin and phosphoric acid has the
`
`characteristic XRPD pattern of a known crystalline form of a sitagliptin
`
`dihydrogenphosphate salt, then one may conclude that the unknown salt is actually
`
`sitagliptin dihydrogenphosphate salt.
`
`20.
`
`In my opinion, a person of ordinary skill in the art would not conclude that
`
`any of Prof. Atwood's XRPD patterns are characteristic ofXRPD patterns for a
`
`More accurately, a reference XRPD pattern for a crystalline material can be used for the
`identification of that particular crystalline form of the material.
`
`Page 7 of39
`
`Merck Exhibit 2192, Page 8
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`i . i
`
`i
`
`.!
`
`bis(sitagliptin) phosphoric acid salt or sitagliptin bis(phosphoric acid) salt. As far as I
`
`!mow, neither a bis(sitagliptin) phosphoric acid salt nor a sitagliptin bis(phosphoric acid)
`
`salt has ever been isolated and characterized, and I do not !mow of any published XRPD
`
`pattern for bis(sitagliptin) phosphoric acid salt or for sitagliptin bis(phosphoric acid) salt.
`
`Prof. Atwood's XRPD patterns therefore cannot be used to confirm whether the solids
`
`that he recovered in his experiments are actually a bis(sitagliptin) phosphoric acid salt or
`
`a sitagliptin bis(phosphoric acid) salt.
`
`21.
`
`XRPD analysis, by its nature, does not provide any information about the
`
`molar ratio of the components of the recovered crystalline solids. Accordingly, Prof.
`
`Atwood's XRPD analyses are insufficient to confirm that he prepared a bis(sitagliptin)
`
`phosphoric acid salt or a sitagliptin bis(phosphoric acid) salt as he contends. In contrast,
`
`as I explain in i!il 26-33, infra, Prof. Atwood's XRPD analysis for one of his experiments
`
`proves that he obtained a sitagliptin dihydrogenphosphate salt and not a bis(sitagliptin)
`
`phosphoric acid salt as Prof. Atwood erroneously concluded.
`
`3.
`
`Prof, Atwood's Thermogravimetric Analysis Data Does Not
`Prove That He Obtained A Bis(Sitagliptin) Phosphoric Acid Salt
`Or A Sitagliptin Bis(Phosphoric Acid) Salt
`
`22.
`
`Prof. Atwood's Thennogravimetric Analysis (TGA) data does not provide
`
`any information about the identity of the recovered salt. TGA is designed to measure
`
`changes in the weight of a sample as the sample is heated. Among other things, TGA
`
`measures whether a material loses weight due to loss of water or another volatile
`
`component upon heating. TGA cannot be used to determine the identity or structure of a
`
`previously unidentified salt or to confirm the structure or identity of a previously
`
`unidentified salt.
`
`Page 8 of39
`
`Merck Exhibit 2192, Page 9
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`4.
`
`Prof. Atwood's Reliance On Differential Scanning Calorimetry
`Analyses Does Not Prove That He Obtained A Bis(Sitagliptin)
`Phosphoric Acid Salt Or A Sitagliptin Bis(Phosphoric Acid) Salt
`
`23.
`
`Prof. Atwood's Differential Scanning Calorimetry(DSC) data does not
`
`provide any information about the identity of the products he recovered. DSC is a
`
`the1moanalytical technique that measures the transfer ofthennal energy (i.e. heat) either
`
`into or from a sample as it is heated. DSC may be used to distinguish different crystalline
`
`forms due to differences in melting points. DSC cannot, however, be used to determine
`
`the stmcture of a previously unidentified salt or to confirm the stmcture or identity of a
`
`I
`
`.1
`
`previously unidentified salt.
`
`5.
`
`Prof. Atwood Did Not Use Analytical Tests That Are
`Diagnostic Of Chemical Structure
`
`24.
`
`Prof. Atwood claims that he prepared a bis(sitagliptin) phosphoric acid
`
`salt and a sitagliptin bis(phosphoric acid) salt, both of which have not been previously
`
`described in the published literature. Well-respected organic chemistry publications
`
`require specific types of analytical data before they will publish an article claiming the
`
`preparation of a new organic compound. For example, The Journal Of Organic
`
`Chemistry guidelines provide that "[f]or all new compounds, evidence adequate to
`
`establish both identity and degree of purity (homogeneity) must be provided" (original
`
`emphasis).3 Those guidelines further specify that the "[e]vidence for documenting the
`
`identity of new compounds should include both proton and carbon NMR [nuclear
`
`'
`
`The Journal Of Organic Chemistry, Guidelines for Authors, Section 2.2.2. (Updated June 20 I!),
`included herewith as Exhibit S.
`
`Page 9 of39
`
`Merck Exhibit 2192, Page 10
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`magnetic resonance] data and either MS accurate mass (HRMS) or elemental analysis
`
`data" ( emphasis in original).4
`
`25.
`
`Despite claiming to have prepared new chemical compounds, i.e. a
`
`bis(sitagliptin) phosphoric acid salt and a sitagliptin bis(phosphoric acid) salt, Prof.
`
`Atwood did not offer any proton or carbon NMR data to document or confirm the
`
`identity of those compounds. Such data would have been particularly informative here
`
`because a person having ordinary skill in the art would have expected only sitagliptin
`
`dihydrogenphosphate to form as a stable salt of sitagliptin and phosphoric acid.
`
`Countering that expectation would have required a more vigorous analysis of the product.
`
`Furthermore, the products of Prof. Atwood's experiments are· highly suspect because he
`
`used poor experimental techniques, as discussed herein. For example, Prof. Atwood did
`
`not perform filtration to clean his products and separate them from unreacted sitagliptin
`
`and phosphoric acid and from reaction byproducts. This is another reason why NMR
`
`data would have been very useful in determining the nature of the products that Prof.
`
`Atwood obtained.
`
`C.
`
`There Is No Credible Evidence That Prof. Atwood Prepared A
`Bis(Sitagliptin) Phosphoric Acid Salt
`
`1.
`
`Prof. Atwood's XRPD Data Proves That The Alleged
`Bis(Sitagliptin) Phosphoric Acid Salt That He Prepared Using
`Methanol As The Reaction Solvent Is Actually A Sitagliptin
`Dihydrogenphosphate Salt
`
`26.
`
`Prof. Atwood incorrectly asserts that he prepared a bis(sitagliptin)
`
`phosphoric acid salt (a 2:l salt)5 using methanol as the reaction solvent. Atwood
`
`4
`
`Id.
`A bis( sitagliptin) phosphoric acid salt, if it could be prepared, would be a salt comprising two
`molecules of sitagllptin per molecule of phosphoric acid.
`
`Page 10 of39
`
`Merck Exhibit 2192, Page 11
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`Declaration, ,I4] 34-37. As explained below, Prof. Atwood's data demonstrates that he
`
`actually prepared a sitagliptin dihydrogenphosphate salt (a 1:1 salt), not a bis(sitagliptin)
`
`phosphoric acid salt (a 2:1 salt).
`
`27.
`
`In this experiment, Prof. Atwood started with a 2:1 ratio ofsitagliptin to
`
`phosphoric acid. Prof. Atwood used unusually high concentrations of the reactants that
`
`caused the reaction mixture to solidify. As explained above, this causes inadequate
`
`mixing, which may lead to incomplete chemical reactions and possibly to an increase in
`
`the formation of byproducts. Moreover, Prof. Atwood did not filter or wash the solid
`
`product from this reaction to remove any non-volatile unreacted starting materials,
`
`impurities, and byproducts.
`
`28.
`
`Prof. Atwood obtained elemental analysis data for the product of his
`
`reaction and that data provided the expected result of a 2: 1 ratio of sitagliptin to
`
`phosphoric acid in the recovered product. 6 That elemental analysis data is of little
`
`probative value because without a filtration or washing step to remove unreacted starting
`
`materials, impurities, and byproducts, one would have expected the recovered solids to
`
`have an elemental analysis consistent with the molar ratio of sitagliptin and phosphoric
`
`acid starting materials, which was a 2: 1 molar ratio.
`
`29.
`
`Furthennore, as discussed above, Prof. Atwood's DSC and TGA data
`
`cannot be used to identify the type of salt that he recovered in his methanol experiment.
`
`30.
`
`Significantly, in this particular case, Prof. Atwood's XRPD data proved to
`
`be very informative because it matches the XRPD pattern of a previously identified salt
`
`•
`
`Prof. Atwood also stated that the alleged 2: 1 salt product is a hydrate. However, Prof. Atwood's
`analysis is inconclusive as he did not perform a chemical identity test on the product that is specific
`for water.
`
`Page 11 of39
`
`Merck Exhibit 2192, Page 12
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`of sitagliptin and phosphoric acid. Specifically, Prof. Atwood's XRPD pattern for the
`
`solids recovered in his methanol experiment matches the XRPD pattern of a previously
`
`characterized sitagliptin dihydrogenphosphate salt ( a I: 1 salt) disclosed in US
`
`2010/0041885 ("the '885 publication")/ Example 5 and Figure 5a. See '885 publication,
`
`'il'il 20, 39, 127-129, and Fig. 5a. The match between Prof. Atwood's XRPD pattern and
`
`the XRPD pattern in the '885 publication proves that Prof. Atwood recovered a sitagliptin
`
`dihydrogenphosphate salt ( a 1: I salt) in his methanol experiment, not the bis( sitagliptin)
`
`I
`
`I
`'
`
`phosphoric acid salt ( a 2: l salt) which he contends that he recovered. Unlike Prof.
`
`Atwood, the applicants of the '885 publication used routine and proper experimental
`
`procedures in conducting the experiment given in Example 5 of this publication, such that
`
`the solids could be isolated from the solution using vacuum filtration. Based on the
`
`stoichiometry of the reaction mixture and the techniques used to isolate the solids, it is
`
`my opinion that the applicants of the '885 publication prepared sitagliptin
`
`dihydrogenphosphate salt (a 1:1 salt) in this Example.
`
`31.
`
`Furthermore, Prof. Atwood's XRPD data is nearly identical to the XRPD
`
`pattern of sitagliptin dihydrogenphosphate salt disclosed in Figure 16 of WO
`
`2005/020920 ("the WO '920 pub!ication").8 See the WO '920 publication, pp. 3, 16, 25,
`
`and Fig. 16. The WO '920 publication describes the XRPD pattern provided in Figure 16
`
`as that of an ethanol solvate (i.e., ethanolate) of sitagliptin dihydrogenphosphate. See id.
`
`The WO '920 publication characterizes the ethanol solvate of sitagliptin
`
`dihydrogenphosphate using XRPD, DSC, TGA, and solid-state carbon-13 and fluorine-19
`
`7 A copy of the '885 publication was provided as Exhibit F to my First Declaration.
`
`a A copy of the WO '920 publication was provided as Exhibit G to my First Declaration.
`
`Page 12 of39
`
`Merck Exhibit 2192, Page 13
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`NMR spectroscopy. See id., Figures 16-20, pp. 3, 16-18. The WO '920 publication lists
`
`Robert M. Wenslow as the first named inventor, the same Dr. Robert M. Wenslow who
`
`submitted a declaration on Merck's behalf in this opposition. It is my expert opinion that
`
`the slight differences between Prof. Atwood's XRPD and the XRPD pattern of the WO
`
`'920 publication are attributable to differences in the solvate fonus. The sitagliptin
`
`dihydrogenphosphate form obtained by Prof. Atwood is most likely a methanol solvate
`
`(i.e., methanolate) or perhaps a hydrate, while the WO '920 publication provided the
`
`XRPD pattern for an ethanolate. The nearly identical match between Prof. Atwood's
`
`XRPD pattern and the XRPD pattern in the WO '920 publication further proves that Prof.
`
`Atwood recovered a sitagliptin dihydrogenphosphate salt ( a 1: 1 salt) in his methanol
`
`experiment, not a bis(sitag!iptin) phosphoric acid salt (a 2:1 salt).
`
`32..
`
`Based on Prof. Atwood's XRPD data, the product of the experiment
`
`i
`!
`
`disclosed in '1]'1] 34-37 of the Atwood Declaration is a mixture of crystalline sitagliptin
`
`dihydrogenphosphate salt (a 1:1 salt) and unreacted sitagliptin, and there is no credible
`
`evidence that any bis(sitagliptin)phosphoric acid salt was prepared. To the contrary, the
`
`XPRD data indicates that Prof. Atwood synthesized a sitagliptin dihydrogenphosphate
`
`salt.
`
`33.
`
`Prof. Atwood's attempt to prepare a bis(sitagliptin) phosphoric acid salt in
`
`a methanol solvent, and the data that he generated from that experiment, demonstrate that
`
`his synthesis procedures can result in the recovery of solids that have his desired ratio of
`
`sitagliptin to phosphoric acid as determined by elemental analysis, even though the
`
`recovered solids contain only a sitagliptin dihydrogenphosphate salt (a 1:1 salt) as
`
`confirmed by XRPD.
`
`Page 13 of39
`
`Merck Exhibit 2192, Page 14
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`2.
`
`There Is No Credible Evidence That Prof. Atwood Prepared A
`Bis(Sitagliptin) Phosphoric Acid Salt Using A Mixture Of
`Isopropanol And Water As The Reaction Solvent
`Prof. Atwood's declaration (ilil 31-33) describes another experiment that
`
`34.
`
`allegedly resulted in a bis(sitagliptin)phosphoric acid salt (a 2: 1 salt). As explained
`
`below, Prof. Atwood's experimental procedure was flawed and his analytical data does
`
`not support his conclusions.
`
`35.
`
`Prof. Atwood asserts that he prepared a bis(sitagliptin) phosphoric acid
`
`salt (a 2: 1 salt) using a mixture of isopropanol and water as the reaction solvent. Atwood
`
`Declaration, ilil 31-33. Prof. Atwood's evidence/data is incomplete, does not prove that
`
`the product he obtaine<,l is a bis(sitagliptin}phosphoric acid salt, and is insufficient to
`
`conclude what kind of salt was prepared. In my opinion, Prof. Atwood more likely
`
`prepared a sitagliptin dihydrogenphosphate salt ( a 1: 1 salt) in combination with unreacted
`
`sitagliptin. I base that opinion on my education, extensive experience in salt screening
`
`experimentation, and on the following: (1) the theoretical predictions outlined in the
`
`Expert Opinion of Prof. Abu Serajuddin; (2) the results of my salt screening experiments;
`
`and (3) the shortcomings in Prof. Atwood's experimental protocol.
`
`36.
`
`In this experiment, Prof. Atwood mixed the desired ratios of sitagliptin
`
`and phosphoric acid (here a 2: 1 ratio) in an isopropanol/water solvent. Prof. Atwood
`
`employed unusually high reactant concentrations that led to the solution solidifying under
`
`the reaction conditions, which, as discussed above, likely resulted in an increased level of
`
`impurities in the solid product. Furthermore, Prof. Atwood did not filter or wash the
`
`solid product from this reaction to remove any impurities or unreacted starting materials.
`
`Prof. Atwood removed the solvent (but not the non-volatile starting materials, impurities
`
`Page 14 of39
`
`Merck Exhibit 2192, Page 15
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`or other components) by vacuum drying the reaction product. As explained above,
`
`conducting the experiment in this manner necessarily resulted in recovering solids having
`
`the same ratio of sitagliptin molecules to phosphoric acid molecules as that of the starting
`
`materials. See 'il'il 7-11, supra.
`
`37.
`
`Prof. Atwood concludes that he obtained a bis(sitagliptin) phosphoric acid
`
`salt based on the results of elemental analysis, X-ray powder diffraction (XRPD),
`
`thermogravimetric analysis (TGA), and differential scanning calorimetry (DSC). Prof.
`
`Atwood's analytical data using those techniques does not prove that he actually obtained
`
`i
`
`a bis(sitagliptin) phosphoric acid salt.
`
`38.
`
`Elemental Analysis: As discussed above, the elemental analysis data
`
`obtained by Prof Atwood merely confirms the molar ratio of sitag!iptin and phosphoric
`
`acid present in the starting materials. Prof. Atwood's poor experimental technique
`
`renders that data useless for salt identification.
`
`39.
`
`XRPD: In my opinion, a person of ordinary skill in the art would not
`
`conclude that Prof. Atwood's XRPD pattern is characteristic of an XRPD pattern for a
`
`bis(sitagliptin) phosphoric acid salt. Prof. Atwood's XRPD data does not provide any
`
`information about the molar ratio of the components of the crystalline solids that he
`
`recovered. I have not found any published XRPD pattern for a sitagliptin / phosphoric
`
`acid salt that matches the XRPD pattern of the solids that Prof. Atwood recovered his
`
`isopropanol/water experiment. I also do not know of any published XRPD pattern for
`
`bis(sitagliptin) phosphoric acid salt. As far as I know, a bis(sitagliptin) phosphoric acid
`
`salt has never been isolated and characterized. Prof. Atwood's XRPD pattern therefore
`
`cannot be used to confom whether the solid that he recovered in his isopropanol/water
`
`Page 15 of39
`
`Merck Exhibit 2192, Page 16
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`' !
`'
`
`i
`
`experiment is actually a bis(sitagliptin) phosphoric acid salt. As explained above, in my
`
`opinion it is more likely that Prof. Atwood recovered a mixture of crystalline I: I salt and
`
`unreacted sitagliptin than that he recovered a bis(sitagliptin) phosphoric acid salt (a 2: 1
`
`salt).
`
`40.
`
`TGA/DSC: As discussed above, Prof. Atwood's TGA, and DSC data do
`
`not, and cannot, prove that he synthesized a bis(sitagliptin) phosphoric acid salt. .
`
`41.
`
`For all of the foregoing reasons, the data presented by Prof. Atwood does
`
`not prove that he synthesized a bis(sitagliptin) phosphoric acid salt in his
`
`isopropanol/water experiment. The tme identity of the product of Prof. Atwood's
`
`isopropanol/water experiment remains unknown. In my opinion, the product is likely a
`
`mixture of a crystalline 1: 1 salt and unreacted sitagliptin.
`
`D.
`
`42.
`
`Prof. Atwood's Alleged Preparation Of A Sitagliptin Ammonia
`Phosphoric Acid Salt Is Not Relevant
`
`Prof. Atwood's example of the alleged preparation a of sitagliptin
`
`ammonia phosphoric acid has nothing to do with whether sitagliptin and phosphoric acid
`
`can or do form stable pharmaceutically suitable salts at molar ratios other than 1: 1.
`
`43.
`
`Prof. Atwood asserts that preparation ofa sitagliptin ammonia phosphoric
`
`acid salt demonstrates that "the second proton of phosphoric acid can be donated to
`
`sitagliptin." Atwood Declaration at 'I! 38. Even if phosphoric acid donates two protons in
`
`sitagliptin ammonia phosphoric acid salt, that does not prove that two molecules of
`
`sitagliptin can remove two protons from a single molecule of phosphoric acid. A
`
`strongly ionic ammonium ion is likely to have an influence on the interaction between
`
`sitagliptin and phosphoric acid. The interaction between one molecule of ammonium,
`
`one molecule of sitagliptin, and one molecule of phosphoric acid would not be expected
`
`Page 16 of39
`
`Merck Exhibit 2192, Page 17
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`
`
`to be the same as an interaction between two molecules of sitagliptin and one molecule of
`
`phosphoric acid.
`
`44.
`
`Prof. Atwood's experiment with ammonium dihydrogenphosphate also
`
`cannot be used to support his position that he ob