`
`I, the undersigned, Dr. Leonard J. Chyall, U.S. Passport No. 432624896, with a business
`
`address of Chyall Pharmaceutical Consulting LLC, 3000 Kent Avenue, Suite D1-105, West
`
`Lafayette, Indiana 47906, USA, having been warned 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 Chyall Declaration”) and a declaration dated March 7, 2012 (the
`
`“Second Chyall Declaration”), in support of the position of Teva Pharmaceutical
`
`Industries Ltd. (“Teva”) in the proceedings before the Honorable Deputy Registrar
`
`of Patents regarding Israel Patent Application No. 172563, filed by Merck & Co.
`
`Inc., U.S.A ("Merck").
`
`2.
`
`This declaration was prepared in response to the Affidavit of Prof. Jerry L. Atwood
`
`submitted on behalf of Merck regarding an experiment that Prof. Atwood conducted
`
`in August 2012 (the “Second Atwood Affidavit”). I was advised by Teva's
`
`counsel that Merck does not rely on paragraphs 3, 6, 7, 8 and 9 of the Second
`
`Atwood Affidavit.
`
`3.
`
`The fact that I have not commented on any particular point in the Second Atwood
`
`Affidavit does not mean that I accept or agree with that point. There is nothing in
`
`the Second Atwood Affidavit that causes me to change the views that I expressed in
`
`the First and Second Chyall Declarations.
`
`1
`
`1 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`Prof. Atwood's New Experiment
`
`4.
`
`In the Second Atwood Affidavit, Prof. Atwood describes a new experiment, in
`
`which he claims to have "repeated the procedure for making the 2:1 phosphate salt
`
`in isopropanol and water", which was described in Paragraphs 31 and 68 of his First
`
`Affidavit, this time with a "filtration and washing" step ("the New Experiment").
`
`I shall refer below to the unwashed solids recovered by Prof. Atwood using the
`
`procedure described in Paragraphs 31 and 68 of his First Affidavit as "the Atwood
`
`Unwashed Solids," and to the solids recovered by Prof. Atwood after his "filtration
`
`and washing step" as "the Atwood Washed Solids". Based on the elemental
`
`analysis of the Atwood Washed Solids, Prof. Atwood claims that his New
`
`Experiment shows that "the same 2:1 phosphate salt was obtained in all procedures,
`
`with or without Dr. Chyall's suggested work up" (paragraph 5 of Prof. Atwood's
`
`Second Affidavit). Prof. Atwood described his alleged "2:1 salt" as having a
`
`characteristic X-Ray Powder Diffraction ("XRPD") pattern that can be used to
`
`identify that alleged salt.
`
`5.
`
`I note that Prof. Atwood's New Experiment attempts to rebut (but fails to do so, as
`
`explained below) only one of the criticisms that I raised in the Second Chyall
`
`Declaration with regard to Prof. Atwood's first set of experiments, i.e., Prof.
`
`Atwood's failure to filter and wash the solid reaction products that he recovered.
`
`The New Experiment does not attempt to address any of the other criticisms I raised
`
`in the Second Chyall Declaration with regard to Prof. Atwood's experiments, such
`
`as Prof. Atwood's use of irregularly high reactant concentrations and lack of
`
`analytical data capable of proving that Prof. Atwood recovered any phosphate salts
`
`of sitagliptin other than the expected dihydrogenphosphate salt ("DHP Salt"). In
`
`2
`
`2 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`fact, Prof. Atwood's New Experiment is also fraught with these very same flaws,
`
`which are not remedied by his washing step.
`
`6.
`
`In addition, Prof. Atwood's New Experiment attempts to rebut my criticism of his
`
`failure to filter and wash the reaction products with respect to only one of Prof.
`
`Atwood's experiments: his "procedure for making the 2:1 phosphate salt in
`
`isopropanol and water" (described in paragraphs 31 and 68 of his First Affidavit).
`
`Prof. Atwood did not present any experiment to rebut any of my criticisms,
`
`including his failure to filter and wash, with regard to any of the other preparations
`
`of alleged non-DHP Salts described in the First Atwood Affidavit.
`
`7.
`
`There is nothing in the Second Atwood Affidavit that causes me to change the
`
`views that I expressed in the First and Second Chyall Declarations, i.e., that the only
`
`pharmaceutically suitable stable salt that will result from a reaction of sitagliptin
`
`free base and phosphoric acid is the DHP Salt, a salt containing a 1:1 ratio of
`
`sitagliptin to phosphoric acid.
`
`Prof. Atwood Used Poor Experimental Techniques
`
`8.
`
`Based on my review of Prof. Atwood’s Second Affidavit and laboratory notebook
`
`pages in which he describes the New Experiment, Prof. Atwood’s protocol included
`
`at least the following steps that may have resulted in inadequate washing of the
`
`recovered solids:
`
`A
`
`Prof. Atwood states that his reaction solution solidified. See Atwood Exhibit
`
`HH. In my experience, a reaction mixture that has solidified is more likely to
`
`contain entrapped impurities, such as unreacted starting materials, than a solid
`
`product that precipitated from the reaction solution.
`
`3
`
`3 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`B.
`
`Prof. Atwood states that he used a spatula to place his solidified material on a
`
`Büchner funnel fitted with an approximately 7 cm diameter piece of filter
`
`paper. See Atwood Exhibit HH. Prof. Atwood does not indicate whether he
`
`crushed or broke up his solidified reaction material before placing the
`
`solidified material on the Büchner funnel, a step that would have improved the
`
`effectiveness of Prof. Atwood’s washing.
`
`C.
`
`Prof. Atwood also does not indicate whether he evenly spread his recovered
`
`solids on the filter paper to minimize the possibility that washing solvent
`
`could pass through the filter with little or no contact with the solids to be
`
`washed. Any washing solvent that passed through the filter with little or no
`
`contact with the solids to be washed would not effectively wash the solids.
`
`D.
`
`Prof. Atwood used filter paper with an approximate diameter of 7 cm (see
`
`Atwood Exhibit HH), meaning that the Büchner funnel that Prof. Atwood used
`
`also likely had a diameter of approximately 7 cm. If Prof. Atwood recovered
`
`100% of his starting materials, he would only have recovered about 1.7 grams
`
`of solids. Depending on whether and/or how Prof. Atwood spread his
`
`recovered solids on the filter paper, the use of a filter with a 7 cm diameter
`
`with such a small amount of solids may leave parts of the filter bare, which
`
`would permit washing solvent to pass through the filter with little or no
`
`contact with the solids to be washed.
`
`E.
`
`Prof. Atwood states that prior to the actual washing, he drew air through his
`
`recovered solids and the filter paper for 5 minutes. See Atwood Exhibit HH.
`
`If no liquid was extracted from the solids by this procedure, then it likely did
`
`little more than dry the solids. Drying the solids, including any impurities,
`
`4
`
`4 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`byproducts and/or unreacted sitagliptin base dissolved in the reaction solvent
`
`and trapped inside the recovered solids, would make it more difficult to
`
`remove
`
`those now solidified
`
`impurities, byproducts and/or unreacted
`
`sitagliptin base through subsequent washing.
`
`F.
`
`Prof. Atwood states that he used 3 x 3 mL of isopropanol solvent to wash and
`
`filter the solids that he recovered. See Atwood Exhibit HH. This is a very
`
`small amount of solvent for washing Prof. Atwood’s recovered solids when
`
`using a Büchner funnel with a diameter of approximately 7 cm. Using too
`
`little solvent for washing would result in ineffective removal of impurities,
`
`byproducts and unreacted starting materials.
`
`9.
`
`Prof. Atwood’s laboratory notebook, Exhibit HH of his Second Affidavit, by itself,
`
`does not provide enough detail to determine whether one or more of the above steps
`
`rendered Prof. Atwood’s washing steps inadequate. Therefore, unlike my criticism of
`
`Prof. Atwood's previous experiments – which did not include any filtration and
`
`washing and therefore did not require that I conduct experiments to conclude that
`
`Prof. Atwood’s assertions regarding the solids that he recovered were unreliable and
`
`without scientific merit – I could only prove the misleading nature of Prof. Atwood's
`
`New Experiment, which included a "filtration and washing" step, by conducting
`
`experiments.
`
`My Experiments Prove That Prof. Atwood's "Filtration And Washing" Was
`
`Ineffective
`
`10.
`
`I received from Teva a sample container labeled lot no. D6655070112, which I
`
`understand to contain Sitagliptin Free Base. The sample was assigned LIMS No.
`
`308390, and I characterized the material using XPRD (see Exhibit A). The XRPD
`
`5
`
`5 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`pattern obtained for the material confirmed that the material was crystalline sitagliptin
`
`base as disclosed in PCT Publication No. WO 2009/070314 A2.
`
`11.
`
`I first replicated as closely as possible the New Experiment described in the Second
`
`Atwood Affidavit. I conducted this replication to ensure that it was possible, based
`
`on the procedure described in Prof. Atwood's laboratory notebook, Exhibit HH, to
`
`obtain crystalline solids with the characteristic XRPD pattern of Prof. Atwood's
`
`alleged "2:1 salt".
`
`12. However, my solution did not solidify overnight like Prof. Atwood's solution. In
`
`order to precipitate the reaction product, I cooled the reaction mixture using an ice
`
`bath with stirring. I filtered and washed my recovered solids using the same
`
`"filtration and washing" protocol that Prof. Atwood used to filter and wash his
`
`recovered solids. I analyzed the recovered solids by XRPD. The recovered solids
`
`had the same characteristic XRPD pattern as that of Prof. Atwood's solids
`
`(Exhibit B), which demonstrates that my use of an ice bath to precipitate solids did
`
`not affect the final product and was not a material deviation from Prof. Atwood’s
`
`procedure. I refer to the solids that I recovered from this experiment as "the
`
`Replicated Atwood Wash Solids". A detailed description of how I obtained the
`
`Replicated Atwood Wash Solids is set forth in my laboratory notebooks, (Exhibit C).
`
`13.
`
`I next conducted an experiment to see the effect of progressively more thorough
`
`washings than employed by Prof. Atwood’s "filtration and washing" protocol. To do
`
`so, I again replicated Prof. Atwood’s New Experiment, except at double the scale, so
`
`as to obtain a sufficient amount of material. This time the solution solidified
`
`overnight, like Prof. Atwood's solution. I analyzed the recovered solids by XRPD.
`
`The solids recovered on the Büchner funnel after washing once with isopropanol had
`
`6
`
`6 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`the same characteristic XRPD pattern (Exhibit D) as that obtained for Prof. Atwood's
`
`solids. My washing procedure, which also included additional washing steps, differed
`
`from that used by Prof. Atwood. The differences between Prof. Atwood's “washing
`
`and filtering” procedure and my washing and filtering procedure are explained below:
`
`a) Filter Paper: I used 45 mm diameter filter paper, which has approximately
`
`59% less surface area than the 70 mm (7 cm) diameter filter paper used by
`
`Prof. Atwood. I used a smaller diameter filter and filter paper to give my solids
`
`more intimate contact with the washing solvent than there would have been if I
`
`had followed Prof. Atwood’s washing protocol.
`
`b)
`
`Scale: I started my experiment at double the scale of Prof. Atwood’s New
`
`Experiment. Consequently, I started my experiment with approximately twice
`
`the solids recovered and filtered by Prof. Atwood. By using twice the solids in
`
`my first series of washings, a reduced filter surface area, and as explained
`
`below, larger volumes of wash solvent, I ensured that my solids had more
`
`intimate contact with the washing solvent than there would have been if I had
`
`followed Prof. Atwood’s washing protocol.
`
`c) Volume of Wash Solvent:
`
`Prof. Atwood used 3 mL portions of
`
`isopropanol to wash his reaction products that were prepared on a 1.5 g scale. I
`
`used at least triple the relative amount of solvent for my washings. For example
`
`when I ran the reaction on a 3 g scale, I used 18 mL portions of isopropanol for
`
`the washing step.
`
`d) Number Of Washes: Prof. Atwood washed his recovered solids three times. I
`
`conducted three series of three washes each, as follows:
`
`7
`
`7 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`a.
`
`First Series of Washes - I washed my recovered solids three times
`
`during the first series of washes. I refer to the solids remaining after the
`
`third wash as “the 3X Washed Solids.”
`
`b.
`
`Second Series of Washes - I took approximately two-thirds of the 3X
`
`Washed Solids and washed them an additional three times in a second
`
`series of washes. I refer to the solids remaining after the second series of
`
`three washings as “the 6X Washed Solids.”
`
`c.
`
`Third Series of Washes - I took approximately one-half of the 6X
`
`Washed Solids and washed them an additional three times in a third
`
`series of washes. I refer to the solids remaining after the third series of
`
`three washings as “the 9X Washed Solids.”
`
`e) Washing Solvent Volume:
`
`a.
`
`The 3X Washed Solids: I conducted my first series of three washes
`
`using 18 mL of isopropanol per wash. The reason for this is as follows.
`
`Prof. Atwood washed his solids three times with 3 mL of isopropanol.
`
`Since I had doubled the scale of Prof. Atwood's New Experiment, I
`
`would have needed to use 6 mL of isopropanol per wash to achieve the
`
`same wash solvent to solids ratio as Prof. Atwood. However, to ensure
`
`that my washing was more thorough than Prof. Atwood’s washing, I
`
`used 18 mL of isopropanol per wash. This afforded a wash solvent to
`
`solids ratio of about three times the ratio that Prof. Atwood used in his
`
`New Experiment.
`
`8
`
`8 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`b.
`
`The 6X Washed Solids: After removing approximately 1/3 of the 3X
`
`washed solids from the Büchner funnel (to be used for subsequent
`
`analytical characterization) I washed the remaining 2/3 of the material in
`
`a second series of three washings with 12 mL of isopropanol in each
`
`washing. This afforded, relative to the weight of solids, a wash solvent
`
`to solids ratio of about three times the ratio that Prof. Atwood used in his
`
`New Experiment. Accordingly, the solids that I recovered after the
`
`second series of three washings (i.e., the 6X Washed Solids) were
`
`washed a total of six times, using a total of about six times more washing
`
`solvent (relative to the weight of solids) than Prof. Atwood used to wash
`
`the solids that he recovered.
`
`c.
`
`The 9X Washed Solids: After removing approximately 1/2 of the 6X
`
`washed solids from the Büchner funnel (to be used for subsequent
`
`analytical characterization) I washed the remaining 1/2 of the 6X
`
`Washed Solids in a third series of three washings with 12 mL of
`
`isopropanol in each washing. This afforded, relative to the weight of
`
`solids, a wash solvent to solids ratio of about six times the ratio that
`
`Prof. Atwood used in his New Experiment. Accordingly, the solids that
`
`I recovered after the third series of three washings (i.e., the 9X Washed
`
`Solids) were washed a total of nine times, using a total of about 12 times
`
`more washing solvent (relative
`
`to
`
`the weight of solids)
`
`than
`
`Prof. Atwood used to wash the solids that he recovered.
`
`14.
`
`I analyzed the 3X Washed Solids, the 6X Washed Solids and the 9X Washed Solids
`
`by XRPD. The XRPD patterns for the 3X and 6X Washed Solids (Exhibits E and
`
`9
`
`9 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`F, respectively) were the same as the XRPD pattern for Prof. Atwood's alleged
`
`"2:1 salt". The XRPD pattern of the 9X Washed Solids did not match the XRPD
`
`pattern of Prof. Atwood's alleged "2:1 salt" (Exhibit G). This indicates that during the
`
`third series of washings, the crystal structure that had previously existed, i.e., the
`
`crystal structure of Prof. Atwood's alleged "2:1 salt", was destroyed. This
`
`demonstrates that Prof. Atwood's alleged "2:1 salt" is not stable.
`
`15. A detailed description of my experiments and analyses are provided in my laboratory
`
`notebooks, Exhibit C.
`
`16. The fact that the XRPD patterns of the Replicated Atwood Wash Solids, the 3X
`
`Washed Solids and the 6X Washed Solids were the same as Prof. Atwood's XRPD
`
`patterns of his unwashed and washed solids, indicates that they all included the same
`
`crystalline product.
`
`17. While an XRPD pattern provides information about the crystalline portion of a solid,
`
`it does not provide information about the non-crystalline portion of a solid in the case
`
`where mixtures are present. The non-crystalline portions of a solid, such as
`
`amorphous impurities, byproducts and/or unreacted sitagliptin base in the samples at
`
`issue in this proceeding, would show up if at all in an XRPD pattern of substantially
`
`crystalline material only as background noise. Therefore, in order for the crystalline
`
`product to be a "2:1 salt" irrespective of filtration and washing, as alleged by Prof.
`
`Atwood, it must maintain both the same characteristic XRPD pattern and the same
`
`relative amounts of the elements after different washing protocols. If, however, the
`
`crystalline product maintains the same characteristic XRPD pattern after different
`
`washing protocols, but the recovered solids have different ratios of the elements, then
`
`this evidences that different washing protocols have different capacities to remove
`
`10
`
`10 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`amorphous impurities from the crystalline product. It also demonstrates that the
`
`crystalline product cannot be a "2:1 salt" and must have included amorphous
`
`impurities, such as unreacted sitagliptin base, prior to the effective washing.
`
`18. The ratio of nitrogen to phosphorus in the different samples is particularly informative
`
`as nitrogen is only present in the "sitagliptin" component of the solids while
`
`phosphorus is only present in the "phosphate" component of the solids. As I
`
`demonstrate below, both elemental analysis and solution phase NMR spectroscopy
`
`show that my more thoroughly washed samples contain relatively less nitrogen with
`
`respect to the phosphorus content (i.e., a lower nitrogen to phosphorus molar ratio)
`
`than "the Atwood Unwashed Solids" and "the Atwood Washed Solids". This
`
`demonstrates that Prof. Atwood's "filtration and washing" step was ineffective, and,
`
`accordingly, that the crystalline product cannot be a "2:1 salt".
`
`My Analytical Data Proves That Prof. Atwood's Alleged "Filtration And Washing
`
`Step" Was Ineffective
`
`19.
`
`I used common analytical techniques to determine the ratio of elements in the
`
`Replicated Atwood Wash Product, the 3X Washed Solids and the 6X Washed Solids.
`
`I employed elemental analysis to examine the Replicated Atwood Wash Solids, the
`
`3X Washed Solids and the 6X Washed Solids. The ratio of nitrogen to phosphorus in
`
`sitagliptin dihydrogenphosphate (N/P ratio) is 2.261. The ratio of nitrogen to
`
`phosphorus for a theoretical bis(sitagliptin) phosphate structure is 4.522. For samples
`
`that contain between one and two molecules of sitagliptin per molecule of phosphoric
`
`acid, the N/P ratio will fall between these two values. The experimentally determined
`
`N/P ratio is linearly correlated with the sitagliptin to phosphoric acid ratio in the
`
`sample such that division of the experimental N/P ratio by 2.261 will provide the
`
`11
`
`11 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`number of molecules of sitagliptin for each molecule of phosphoric acid (See Exhibit
`
`H).
`
`20.
`
`I also used solution phase carbon-13 Nuclear Magnetic Resonance (13C NMR)
`
`spectroscopy to examine the Replicated Atwood Wash Solids and the 6X Washed
`
`Solids. Certain 13C NMR resonances of sitagliptin are sensitive to the protonation
`
`state of sitagliptin. For NMR solutions of mixtures of sitagliptin and phosphoric acid,
`
`the 13C NMR resonances represent an averaged value for protonated and unprotonated
`
`sitagliptin due to rapid exchange of the available protons between sitagliptin
`
`molecules. The ratio of sitagliptin to phosphoric acid in a sample may be determined
`
`by comparing the 13C NMR spectrum for the sample to NMR spectra for reference
`
`standards that contain known ratios of sitagliptin to phosphoric acid. By using this
`
`methodology I independently confirmed the ratio of sitagliptin to phosphoric acid in
`
`my Replicated Atwood Wash Solids and 6X wash solids. Details of how the
`
`13C NMR spectra were obtained and the resulting data are provided in Exhibit I.
`
`21. The molar ratios of sitagliptin to phosphoric acid that I determined based on 13C NMR
`
`spectra and elemental analysis data that I obtained, along with the elemental analysis
`
`data that Prof. Atwood obtained, are summarized in the table below. The elemental
`
`analysis results are in agreement with the 13C NMR results within experimental error.
`
`The data evidences that successively more thorough washing of the recovered solids
`
`progressively removed more unreacted starting materials,
`
`impurities, and/or
`
`byproducts and that the crystalline product cannot be a 2:1 salt.
`
`
`
`
`
`
`
`12
`
`12 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`(cid:6)(cid:7)(cid:15)(cid:18)(cid:14)(cid:10)(cid:1)
`(cid:24)(cid:6)(cid:17)(cid:22)(cid:19)(cid:8)(cid:10)(cid:25)(cid:1)
`
`(cid:1)
`(cid:2)(cid:27)(cid:29)(cid:23)(cid:23)(cid:15)(cid:1)(cid:9)(cid:22)(cid:29)(cid:12)(cid:26)(cid:18)(cid:16)(cid:15)(cid:1)(cid:8)(cid:23)(cid:20)(cid:19)(cid:15)(cid:26)(cid:1)
`(cid:34)(cid:4)(cid:19)(cid:25)(cid:26)(cid:27)(cid:1)(cid:2)(cid:27)(cid:29)(cid:23)(cid:23)(cid:15)(cid:1)(cid:2)(cid:17)(cid:17)(cid:19)(cid:15)(cid:12)(cid:28)(cid:19)(cid:27)(cid:35)(cid:1)
`(cid:2)(cid:27)(cid:29)(cid:23)(cid:23)(cid:15)(cid:1)(cid:10)(cid:12)(cid:26)(cid:18)(cid:16)(cid:15)(cid:1)(cid:8)(cid:23)(cid:20)(cid:19)(cid:15)(cid:26)(cid:1)
`(cid:34)(cid:8)(cid:16)(cid:14)(cid:23)(cid:22)(cid:15)(cid:1)(cid:2)(cid:27)(cid:29)(cid:23)(cid:23)(cid:15)(cid:1)(cid:2)(cid:17)(cid:17)(cid:19)(cid:15)(cid:12)(cid:28)(cid:19)(cid:27)(cid:35)(cid:1)
`(cid:7)(cid:16)(cid:24)(cid:20)(cid:19)(cid:14)(cid:12)(cid:27)(cid:16)(cid:15)(cid:1)(cid:2)(cid:27)(cid:29)(cid:23)(cid:23)(cid:15)(cid:1)(cid:10)(cid:12)(cid:26)(cid:18)(cid:1)(cid:8)(cid:23)(cid:20)(cid:19)(cid:15)(cid:26)(cid:1)
`(cid:34)(cid:27)(cid:18)(cid:19)(cid:26)(cid:1)(cid:25)(cid:16)(cid:24)(cid:23)(cid:25)(cid:27)(cid:35)(cid:1)
`(cid:39)(cid:11)(cid:1)(cid:10)(cid:12)(cid:26)(cid:18)(cid:16)(cid:15)(cid:1)(cid:8)(cid:23)(cid:20)(cid:19)(cid:15)(cid:26)(cid:1)
`(cid:34)(cid:27)(cid:18)(cid:19)(cid:26)(cid:1)(cid:25)(cid:16)(cid:24)(cid:23)(cid:25)(cid:27)(cid:35)(cid:1)
`(cid:42)(cid:11)(cid:1)(cid:10)(cid:12)(cid:26)(cid:18)(cid:16)(cid:15)(cid:1)(cid:8)(cid:23)(cid:20)(cid:19)(cid:15)(cid:26)(cid:1)
`(cid:34)(cid:27)(cid:18)(cid:19)(cid:26)(cid:1)(cid:25)(cid:16)(cid:24)(cid:23)(cid:25)(cid:27)(cid:35)(cid:1)
`
`(cid:6)(cid:13)(cid:21)(cid:7)(cid:11)(cid:14)(cid:13)(cid:18)(cid:21)(cid:13)(cid:16)(cid:23)(cid:4)(cid:12)(cid:17)(cid:20)(cid:18)(cid:12)(cid:17)(cid:19)(cid:13)(cid:8)(cid:1)(cid:2)(cid:8)(cid:13)(cid:9)(cid:1)(cid:3)(cid:17)(cid:14)(cid:7)(cid:19)(cid:1)(cid:5)(cid:7)(cid:21)(cid:13)(cid:17)(cid:1)
`
`(cid:1)(cid:15)(cid:16)(cid:27)(cid:16)(cid:25)(cid:21)(cid:19)(cid:22)(cid:16)(cid:15)(cid:1)(cid:13)(cid:30)(cid:1)(cid:16)(cid:20)(cid:16)(cid:21)(cid:16)(cid:22)(cid:27)(cid:12)(cid:20)(cid:1)
`(cid:12)(cid:22)(cid:12)(cid:20)(cid:30)(cid:26)(cid:19)(cid:26)(cid:1)
`(cid:38)(cid:32)(cid:36)(cid:45)(cid:31)(cid:37)(cid:1)
`(cid:37)(cid:32)(cid:44)(cid:39)(cid:31)(cid:37)
`(cid:37)(cid:32)(cid:44)(cid:37)(cid:31)(cid:37)
`(cid:37)(cid:32)(cid:42)(cid:40)(cid:31)(cid:37)
`(cid:37)(cid:32)(cid:41)(cid:45)(cid:31)(cid:37)
`
`(cid:15)(cid:16)(cid:27)(cid:16)(cid:25)(cid:21)(cid:19)(cid:22)(cid:16)(cid:15)(cid:1)(cid:13)(cid:30)(cid:1)(cid:37)(cid:39)(cid:3)(cid:1)(cid:6)(cid:5)(cid:7)(cid:1)
`(cid:26)(cid:24)(cid:16)(cid:14)(cid:27)(cid:25)(cid:23)(cid:26)(cid:14)(cid:23)(cid:24)(cid:30)(cid:1)
`(cid:33)
`(cid:33)
`(cid:37)(cid:32)(cid:43)(cid:42)(cid:31)(cid:37)(cid:1)(cid:27)(cid:23)(cid:1)(cid:37)(cid:32)(cid:43)(cid:45)(cid:31)(cid:37)(cid:1)
`(cid:33)
`(cid:37)(cid:32)(cid:41)(cid:36)(cid:31)(cid:37)(cid:1)(cid:27)(cid:23)(cid:1)(cid:37)(cid:32)(cid:41)(cid:41)(cid:31)(cid:37)
`
`22. The data summarized above clearly demonstrates that as the recovered solids are
`
`subjected to progressively more thorough washing, the molar ratio of sitagliptin to
`
`phosphoric acid drops from about 1.8:1 (in the Replicated Atwood Wash Product) to
`
`less than 1.6:1 (in the 6X washed Solids) without any change in the structure of the
`
`crystalline component of the solids. This proves that Prof. Atwood's "filtration and
`
`washing" step was ineffective, and that the more thorough washings that I performed
`
`removed a significant amount of unreacted starting materials, impurities and/or
`
`byproducts that were still present in the product after Prof. Atwood's "filtration and
`
`washing" step. Accordingly, Prof. Atwood's New Experiment is misleading, and the
`
`conclusions he draws therefrom are incorrect.
`
`23. Had Prof. Atwood more thoroughly washed his solids in his New Experiment, he
`
`would have found that his crystalline product has a ratio of less than 1.6 moles of
`
`sitagliptin per 1 mole of phosphoric acid. The progressive washing experiments that I
`
`conducted provided solids that approach a ratio of 1.5 moles of sitagliptin per 1 mole
`
`of phosphoric acid. Such a product cannot be a non-DHP salt. This is because
`
`phosphoric acid cannot transfer one half of a proton to a sitagliptin molecule to make
`
`a 1.5:1 sitagliptin to phosphoric acid salt.
`
`13
`
`13 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`24. The most likely explanation for my results is that the crystalline product that I
`
`recovered is a co-crystal containing a ratio of two molecules of sitagliptin DHP Salt to
`
`one molecule of unreacted sitagliptin free base.
`
`25.
`
`I conducted additional analyses to examine the products of my experiments, such as
`
`indexation, DSC, TGA, 1H NMR, solid-state 15N and 13C NMR, and Karl Fischer
`
`analysis. The data obtained in these additional analyses (see Exhibits C and J) are
`
`supportive of the conclusions presented herein. However, as the analyses described
`
`above clearly rebut Prof. Atwood's New Experiment and the conclusions he draws
`
`therefrom, I did not find it necessary to rely on these additional analyses.
`
`26. Attached as Exhibit K is a Project Sample Report identifying all of the analyses that I
`
`conducted in relation to this declaration. The Project Sample Report identifies the
`
`analyses on which I rely as well as analyses on which I do not rely.
`
`Conclusion
`
`27. Upon review of the Second Atwood Affidavit, it is apparent that several aspects of
`
`Prof. Atwood’s "filtration and washing" step may have rendered it inadequate for
`
`removing impurities, byproducts and/or unreacted starting materials in solids that he
`
`recovered from his New Experiment.
`
`28. The experiments that I conducted, which are described above, prove that Prof.
`
`Atwood's "filtration and washing step" was indeed ineffective in removing impurities,
`
`byproducts and/or unreacted starting materials in the solids that he recovered from his
`
`New Experiment. Accordingly, Prof. Atwood's conclusion that his New Experiment
`
`shows that "the same 2:1 phosphate salt was obtained in all procedures, with or
`
`without Dr. Chyall's suggested work up" (paragraph 5), is incorrect.
`
`14
`
`14 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`
`
`29. Had Prof. Atwood more thoroughly washed his solids in his New Experiment, he
`
`would have found that the Atwood Washed Solids still contained a significant amount
`
`of unreacted starting materials, impurities and/0r byproducts, and that his crystalline
`
`product cannot be a "2:1 salt” of sitagliptin and phosphoric acid, or any non-DHP salt.
`
`30.
`
`I therefore remain fully convinced that the only pharmaceutically suitable stable salt
`
`that will result from a reaction of sitagliptin free base and phosphoric acid is the DHP
`
`Salt, a salt containing a 1:1 ratio of sitagliptin to phosphoric acid.
`
`Date: February 19, 20l 3
`
`President, Chyall Pharmaceutical Consulting LLC
`
`15
`
`15 0f 15
`
`|PR2020-01045, Teva Ex. 1022
`|PR2020-01060, DRL Ex. 1022
`
`|PR2020-01072, Sun Ex. 1022
`
`15 of 15
`
`IPR2020-01045, Teva Ex. 1022
`IPR2020-01060, DRL Ex. 1022
`IPR2020-01072, Sun Ex. 1022
`
`