Aust. J. Chem., 1984,37, 1895-902
`
`Micellar Catalysis of Organic Reactions. XIV*
`Hydrolysis of Some 1,4-Benzodiazepin-2-one
`Drugs in Acidic Solution
`
`Trevor J. ~ r o x t o n , * ' ~ Timothy a an* and Steven R. orriso on*
`A Department of Organic Chemistry, La Trobe University, Bundoora, Vic. 3083.
`TO whom correspondence should be addressed.
`
`Abstract
`Kinetic studies of the acidic hydrolysis of diazepam and nitrazepam were carried out in the presence
`of micelles of sodium dodecyl sulfate (sds). The hydrolysis of diazepam was shown to occur with
`biphasic kinetics. This is consistent with initial hydrolysis of the azomethine bond followed by very
`slow hydrolysis of the amide bond as found for hydrolysis in aqueous solution. Nitrazepam, however,
`was found to decompose with monophasic kinetics consistent with initial amide hydrolysis. Reactions
`involving the hydrolysis of the azomethine bond were shown to be independent of acid concentration
`and subject to inhibition by micelles of sds. Reactions involving amide hydrolysis were shown to
`be first order in acid concentration and subject to micellar catalysis. The mechanistic change for the
`hydrolysis of nitrazepam on transfer from water (initial azomethine cleavage) to micelles of sds
`(initial amide cleavage), was presumably the result of the inhibition of azomethine hydrolysis and
`the catalysis of amide hydrolysis by the micelles.
`
`Introduction
`The study of the effects of micelles on organic reactions is currently of great interest.
`In particular, there have been many reports of the catalysis of organic reactions.'
`Less widespread are reports of micelle-induced changes of mechanism.
`In some cases, different products have been reported for reactions carried out in
`the presence of micelles compared with reaction in water. This can occur when the
`product of the reaction in water is unstable in the presence of micelles. For example,
`the basic hydrolysis of alkyl and aryl N-(4-nitropheny1)carbamates in water at 25OC
`In the presence of
`yields N-(44trophenyl)carbamate ion as the stable product.
`micelles of cetyltrimethylammonium bromide (ctab), however, the product was
`4-nitroaniline. Decarboxylation of N-(4-nitropheny1)carbamate ion was shown to
`be strongly catalysed ( x 45) by micelles of ctab. Thus, a micelle-induced change of
`product, but not of mechanism, was ~ b s e r v e d . ~
`
`* Part XIII, Aust. J. Chem., 1984, 37, 977.
`Fendler, J. H., and Fendler, E. J., 'Catalysis in Micellar and Macromolecular Systems' (Academic
`Press: New York 1975).
`Broxton, T. J., Aust. J. Chem., 1984, 37, 47.
`
`Opiant Exhibit 2307
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00685, IPR2019-00688, IPR2019-00694
`Page 1
`
`

`

`T. J. Broxton, T. Ryan and S. R. Morrison
`
`We have previously demonstrated some examples where the fine details of the
`mechanism of basic hydrolysis of amides are affected by micelles of tab.^,^ For the
`hydrolysis of N-aryl-N-phenylbenzamides, a change of mechanism from rate determin-
`ing solvent-assisted C-N bond breaking in water to rate determining hydroxide ion
`attack in ctab was o b ~ e r v e d . ~ In the case of N,p-dimethyl-N-phenylbenzamide,
`basic hydrolysis in water occurred via a slow protonation of the nitrogen atom in
`the tetrahedral intermediate followed by fast C-N bond cleavage, whereas in ctab
`the rate-determining step was solvent-assisted C-N bond-breaking?
`We are now interested in the investigation of systems in which a more dramatic
`mechanistic change is possible. We looked for a system where two possible sites of
`attack were available. Such a system is the benzodiazepinone nucleus (I), where
`nucleophilic attack may occur at either the 2-position, leading to amide hydrolysis
`[(I) -+ (2)] or at the 5-position, leading to azomethine cleavage [(I) -+ (3)] (Scheme
`1).
`Previous studies, in water, of the acidic hydrolysis of benzodiazepines of thera-
`peutic interest have been reported. These include studies of the hydrolysis of 7-chloro-
`(diazepam) (la)'-'
`and
`l-methyl-5-phenyl-l,3-dihydro-2H-l,4-benzodiazepin-2-one
`7-nitro-5-phenyl-1,3-dihydro-2H-1,4-benzodiazepin-2-one
`(nitrazepam) (lb).839
`
`In each case, the hydrolytic mechanism involves cleavage initially of the 4,5 azo-
`methine bond and then of the 1,2 amide bond to give the appropriate aminobenzo-
`phenone (4) and glycine (Scheme 1). For other benzodiazepines, e.g., oxazepam5
`cleavage of the 1,2 amide bond has been reported to precede
`and chlordiazep~xide,~~
`
`Broxton, T. J., Fernando, D. R., and Rowe, J. E., J. Ovg. Chem., 1981, 46, 3522.
`Broxton, T. J., and Duddy, N. W., Aust. J. Chem., 1979,32, 1717.
`W a n , W. W., Yakatan, G. J., and Maness, D. D., J. Phavm. Sci., 1977, 66, 573.
`Nakano, M., Inotsume, N., Kohri, N., and Arita, T., Znt. J. Pharm., 1979, 3, 195.
`Mayer, W., Erbe, S., Wolf, G., and Voigt, R., Pharmazie, 1974, 29, 700.
`Han, W. W., Yakatan, G. J., and Maness, D. D., J. Pharm. Sci., 1977, 66, 795.
`Inotsume, N., and Nakano, M., J. Pharm. Sci., 1980, 69, 1331.
`lo Han, W. W., Yakatan, G. J., and Maness, D. D., J. Phavm. Sci., 1976, 65, 1198.
`
`Opiant Exhibit 2307
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00685, IPR2019-00688, IPR2019-00694
`Page 2
`
`

`

`Micellar Catalysis of Organic Reactions. XIV
`
`1897
`
`cleavage of the 4,5 azomethine bond. Thus, the mechanism of hydrolysis of benzo-
`diazepines is sensitive to substituents on the aromatic ring and on the amide nitrogen.
`Depending on the order of bond cleavage, intermediates (2) or (3) can be formed
`during the hydrolysis (Scheme 1). If the 4,5 azomethine bond is cleaved first, inter-
`mediate (3) is formed, while if the initial hydrolysis step occurs at the 1,2 amide bond,
`intermediate (2) is formed. In acidic solution, it has been proposed5,6 that intermediate
`(3) would accumulate in solution because the reversal of the first step would be
`inhibited since the nucleophilicity of the amino group would be reduced as a result
`of protonation. Thus, the reaction should display biphasic kinetics if initial azomethine
`hydrolysis occurs. Intermediate (2), however, is not expected to accumulate in acidic
`solution. This is consistent with many reports in the literature that lactam formation
`from amino acids like intermediate (2), is fa~ile.",'~ Thus, if initial amide hydrolysis
`occurs, the rapid equilibration should result in monophasic kinetics.
`The effects of micelles on the acidic hydrolysis of amide and azomethine bonds
`has been reported. Micelles of sodium dodecyl sulfate (sds) have been reported to
`inhibit both the acidic hydrolysis of amides13 and azomethine groups.'4 The magnitude
`of inhibition, however, varies and it was decided to investigate the effects of micelles
`of sds on the mechanism of hydrolysis of diazepam (la) and nitrazepam (lb). Since
`the site of initial hydrolysis of benzodiazepines is finely balanced and since amide
`and azomethine hydrolyses are subject to different degrees of inhibition by micelles
`of sds, we hoped to find examples where the mechanism of hydrolysis of benzodiaze-
`pines changed on transfer from water to the micelle.
`
`Results and Discussion
`(a) Diazepam
`Han5 has proposed that in aqueous solution the initial event is hydrolysis of the
`azomethine bond forming intermediate (3a) in equilibrium with diazepam (la).
`Subsequent hydrolysis of (3a) to (4a) occurs much more ~ l o w l y . ~ We measured the
`rate of loss of diazepam at 280 nm and determined the percentage of diazepam
`present at equilibrium (by measuring the absorbance at 280 nm initially, i.e. due to
`diazepam, and at equilibrium). In addition, we measured the absorbance at 280 nm
`for compound (5a) as a model for intermediate (3a). Compound (5a) and intermediate
`(3a) differ only in the replacement of the NH, group by H; it would thus seem reason-
`able that the ultraviolet spectra be similar. By this method, the percentage of diazepam
`at equilibrium was found to be 65%. This value was almost independent of acid
`
`(5b)
`
`NO,
`
`H
`
`l 1 Hendrickson, J. B., Cram, D. J., and Hammond, G. S., 'Organic Chemistry' 3rd Edn, p. 513
`(McGraw-Hill : New York 1970).
`lZ Sternbach, L. H., and Reeder, E., J. 01.g. Chem., 1961, 26, 4936.
`l 3 Linda, P., Stener, A,, Cipiciani, A., and Savelli, G., J. Chem. Soc., Perkin Trans. 2, 1983, 821.
`l4 Behme, M. T. A., and Cordes, E. H., J. Am. Chem. Soc., 1965,87,260.
`
`Opiant Exhibit 2307
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00685, IPR2019-00688, IPR2019-00694
`Page 3
`
`

`

`T. J. Broxton, T. Ryan and S. R. Morrison
`
`concentration within the range 0.013-0.5 M HC1 and is within 10% of the value
`obtained by extrapolation of Nakano's results6 to 68°C. Nakano determined the
`percentage of diazepam at equilibrium by the chemical separation of the equilibrium
`mixture. This result gave us some confidence in the U.V. method and in our assumption
`about the similarity of the U.V. spectra of (5a) and (3a).
`In an equilibrium such as (la)
`(3a) the rate constant obtained by monitoring
`(la) loss is in fact the sum offorward and reverse reactions (k, +k-,).I5 By using
`the value for the measured percentage of diazepam at equilibrium k,, the rate of the
`forward reaction (la) -+ (3a) was calculated from the observed rate. These results
`are in Table 1.
`
`Table 1. Observed rate constants for the first phase of the acidic hydrolysis of
`diazepam (la) at 68.5"C
`Loss of diazepam was followed at 280 nm. A very slow subsequent reaction
`was observed but it was too slow to obtain reliable rate constants. Rate
`constants for (la) -+ (3a) calculated as described in text
`
`[ H a
`(M)
`
`0 A
`
`104kl/s-1 for [sds]/m~ of
`100B
`2OOB
`
`40
`
`300B
`
`A In absence of sds 65 % diazepam at equilibrium.
`In presence of sds 40 % diazepam at equilibrium.
`
`The rate of cleavage of the azomethine bond is independent of acid concentration
`within the range 0.013-0.26 M HCI. Tight isosbestic points were obtained at 224,
`256 and 270 nm, indicating the absence of any accumulation of an intermediate
`during the establishment of the equilibrium between (la) and (3a).
`The percentage of diazepam present in equilibrium was similarly calculated in the
`presence of sds. By using these results, rate constants k for the forward reaction
`(la) -+ (3a) in the presence of sds were calculated from the observed rates. These
`results (Table 1) show that the reaction is inhibited by micelles of sds. The rate-sds
`profile showed a plateau between 200 and 400 mM sds, which indicates complete
`solubilization of diazepam in sds micelles at 200 mM sds. Tight isosbestic points were
`obtained at 225, 252 and 272 nm as for reaction in water. Furthermore, the rate of
`conversion (la) -, (3a) in the presence of 200 mM sds was independent of acid
`concentration within the range 0.013-0.26 M HCI.
`In both aqueous solution and in the presence of sds, it was found that when the
`pH of the equilibrium mixture (la) + (3a) was increased to pH 7, diazepam (la)
`was reformed in quantitative yield.
`This suggests that hydrolysis of diazepam follows the (la) e (3a) -, (4a) pathway,
`both in aqueous solution and in the presence of sds. Thus, initial attack occurs at
`C 5, leading to azomethine cleavage in both cases.
`The position of the equilibrium between diazepam and intermediate (3a) was
`displaced slightly in favour of the intermediate in the presence of 200 mM sds relative
`
`l5 Pannetier, G., and Souchay, P., 'Chemical Kinetics' p. 166 (Elsevier: Amsterdam 1967).
`
`Opiant Exhibit 2307
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00685, IPR2019-00688, IPR2019-00694
`Page 4
`
`

`

`Micellar Catalysis of Organic Reactions. XIV
`
`to the situation in water. Since the rate of conversion (la) -+ (3a) is reduced in the
`presence of sds, an increase in the proportion of (3a) at equilibrium must be the result
`of a greater reduction in the rate of conversion (3a) -+ (la). This may be the result of
`different conformations of (3a) in aqueous solution and in the negatively charged
`micelle. However, recyclization of (3a) requires the amino group to be unprotonated.
`In acidic solution, (3a) is in equilibrium with the corresponding protonated form.
`This protonated form would be favoured in micelles of sds because of the electrostatic
`stabilization provided by the micelle. Thus, the recyclization of (3a) would be inhibited
`relative to the situation in water.
`
`(b) Nitrazepam
`Hans proposed that in aqueous solution initial hydrolysis occurs at the azomethine
`linkage. We have determined the percentage of nitrazepam present at equilibrium
`as a function of acid concentration. This was done as for diazepam, compound
`(5b) being used as a model for intermediate (3b). We found that above 0.1 M HCI,
`the equilibrium was displaced completely towards intermediates (3b), while at lower
`acid concentrations, significant amounts of nitrazepam existed in equilibrium with
`(3b). These range from 13 % nitrazepam at 0.05 M HCl, through 25 % at 0.025 M HC1
`to 38 % at 0.012 M HCl.
`The variation of the percentage of nitrazepam at equilibrium as a function of acid
`concentration probably reflects the lower basicity of nitrazepam, which contains a
`nitro group, compared with diazepam which has a chloro substituent. Furthermore,
`less reactant is present at equilibrium for nitrazepam than for diazepam. This can
`be explained by the fact that the positive charge in protonated nitrazepam is conjugated
`with the nitro group whereas in intermediate (3b) the posltive charge is not conjugated
`with the nitro group. Thus, the destabilizing effect of the strong electron-withdrawing
`nitro group is more severe for nitrazepam than for intermediate (3b). For diazepam,
`the chloro group is electron-releasing by resonance and hence stabilizes the protonated
`starting material more than intermediate (3a) for which the positive charge and the
`chloro group are not conjugated.
`
`Table 2. Observed first-order rate constants for the acidic hydrolysis of nitrazepam (lb) at 68.5"
`The loss of nitrazepam was followed at 280 nm. Rate constants for (1) -+ (3) as described in
`text, percentage of nitrazepam at equilibrium being used
`
`[HC1IA
`(MI
`
`0
`
`OB
`
`20
`
`-
`
`104kl/s-I for [sds]/m~ of
`40
`100
`
`-
`
`ZOOC
`
`300
`
`400
`
`9.60 (9.56)
`6.36
`6.95
`0.527 (0)
`6.02 (6.26)
`6.44
`7.08
`0.264 (0)
`4.40 (4.43)
`6.61
`7.72
`0.132 (0)
`3.05 (3.33)
`6.0
`7.70
`0.066 (13)
`2.20 (2.02)
`5.30
`6.90
`0.026 (25)
`3 . 9
`5.5
`0.013 (38)
`1.70
`2.32
`2.99
`3.51
`A Percentage of nitrazepam at equilibrium in aqueous acidic solution in parenthesis.
`Constant ionic strength, 0.527 M (KCI).
`Values in parenthesis followed production of benzophenone (4b) at 364 nm.
`
`1.7
`
`1 . 8
`
`By using these results, the rate of conversion (lb) -+ (3b) (Table 2), was calculated
`from the observed rate of establishment of equilibrium. The rate of hydrolysis of
`
`Opiant Exhibit 2307
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00685, IPR2019-00688, IPR2019-00694
`Page 5
`
`

`

`T. J. Broxton, T. Ryan and S. R. Morrison
`
`nitrazepam (lb) -+ (3b) was constant within experimental error at the higher acid
`concentrations, but decreased at lower acid concentrations. This also reflects the
`greater difficulty of protonation of nitrazepam compared with diazepam. Similar
`behaviour was obtained at constant ionic strength (0.527 M with KCI).
`The rate constants obtained for the second phase of reaction (3b) + (4b) showed
`that this reaction was first order in acid concentration. At 0.527 M HCI, k, was
`0.2 x
`s-'. At these acid con-
`s-I, while at 0.264 M HC1, k, was 0.12 x
`centrations, the equilibrium of step 1 was displaced completely to favour intermediate
`(3b). At 0.264 M HCl, the second phase of hydrolysis was 59 times slower than the
`first phase.
`In the presence of 200 mM sds, we found that the rate of hydrolysis increased almost
`sixfold as the acid concentration was increased from 0.013 to 0.53 M HCI.
`At 0.013 M HCI, the rate-sds profile showed that the reaction was inhibited by
`sds and that nitrazepam was completely solubilized at 200 mM sds. Thus, at 0.013 M
`HCI, it appears that the major reaction pathway involves azomethine cleavage, as
`for reaction in water.
`
`Table 3. Observed first-order rate constants for the
`acidic hydrolysis of 2-acetylamino-54trobenzophenone
`at 68.5"C
`The rate of production of 2-amino-5-nitrobenzo-
`phenone was followed at 364 nm
`
`WClI
`(M)
`
`-
`
`1O4k1/s-I for [sds]/m~ of
`100
`200
`
`0
`
`0.527
`0.264
`0.013
`A Reaction was too slow to measure accurately.
`
`1.85
`0.82
`A
`
`11.5
`8.7
`0.99
`
`10.5
`7.0
`0.67
`
`However, on the basis of our results for diazepam, the observed increase in rate
`of reaction in 200 mM sds, as the acid concentration was increased, is not consistent
`with azomethine hydrolysis. Furthermore, at 0.527 M HCI, the rate of hydrolysis
`was faster in the presence of 200 mM sds than in water. Thus, it appears possible
`that, at 0.527 M HCI, the predominant hydrolysis pathway involves initial amide
`cleavage [(lb) -+ (2b) + (4b)l. To test this mechanistic hypothesis, we used 2-acetyl-
`amino-5-nitrobenzophenone (5b) as a model for amide hydrolysis studies on a
`nitrazepam-like compound. Rate constants for the acidic hydrolysis of this compound
`are in Table 3. It can be seen that the rate of amide hydrolysis was first order in
`acid concentration and, furthermore, the hydrolysis was catalysed by micelles of
`sds. Maximum catalysis was obtained at 100 mM sds. At 0.26 M HCI, the hydrolysis
`was more than 10 times faster in sds than in water.
`The possibility of initial amide hydrolysis of nitrazepam at high acid concentrations
`in the presence of sds was further indicated by the observation of monophasic kinetics
`for the reaction. As discussed previously, initial amide hydrolysis resulting in
`formation of intermediate (2b), should display monophasic kinetics, since the inter-
`mediate should not accumulate in acidic solution. The observation that the rate of
`production of the substituted benzophenone (4b) at 364 nm was identical to the rate
`
`Opiant Exhibit 2307
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00685, IPR2019-00688, IPR2019-00694
`Page 6
`
`

`

`Micellar Catalysis of Organic Reactions. XIV
`
`of loss of nitrazepam at 280 nm is also consistent with this explanation. The produc-
`tion of benzophenone (4b) at 364 nm was first order throughout with no trace of an
`induction period.
`An alternative explanation for the above result is reaction through fast initial
`azomethine cleavage and subsequent slow conversion of intermediate (3b) into
`product (4b). To test this possibility, we first allowed nitrazepam to hydrolyse to
`intermediate (3b) in aqueous acid. Then sufficient sds and acid were added to mimic
`the normal reaction conditions, and the conversion of intermediate (3b) into product
`(4b) was followed. It was found to be much slower than the reaction of nitrazepam
`in 0.26 M HC1/200 mM sds. Thus, the predominant pathway for the hydrolysis of
`nitrazepam at high acid concentrations in the presence of sds is initial amide hydrolysis
`
`The micellar catalysis expected for amide cleavage of nitrazepam is masked
`because the rate constant observed in aqueous solution is for the faster azomethine
`hydrolysis. The reason for the mechanistic changeover on transfer from water to
`sds, is the concurrent inhibition of azomethine cleavage and catalysis of amide cleavage.
`
`Experimental
`Materials
`Diazepam and nitrazepam were provided by Roche Products Pty Ltd.
`2-Acetylamino-5-nitrobenzophenone, m.p. 155-157" (lit.16 157-15g0), was prepared from 2-amino-
`5-nitrobenzophenone (Aldrich) by acetylation (Ac20/HOAc). Sodium dodecyl sulfate (BDH
`biochemicals) was purified by the method of Duynstee and Grunwald.17
`
`Kinetic Studies
`Rate measurements were carried out at 68.5", in a cuvette kept at constant temperature in the
`cell compartment of a Varian 635 ultraviolet-visible spectrophotometer. Stock solutions of the
`substrates (0.01 M in methanol), sodium dodecyl sulfate (0.4 M in water) and hydrochloric acid
`(1 M in water) were prepared in purified solvents.
`The required volumes of sds and hydrochloric acid were pipetted into a 50-ml volumetric flask
`and were diluted to the mark. This solution was then placed into the cuvette which was stoppered
`and placed in the jacketed cell holder of the spectrophotometer. The reaction mixture was allowed
`at least 30 min to attain temperature. The temperature within the cell was measured with a Jenco
`700C Thermistor thermometer. The reaction was initiated by the addition of 5 fi1 of substrate solution
`directly into the cuvette. The solution was shaken and absorbance-time measurements were recorded
`at the appropriate wavelengths for 10 half-lives on a National VP 651 1A X-T recorder. Rate constants
`were calculated from a graphical treatment [log(A, -A,)
`against time for reactions where substrate
`loss was being followed and log(A, - A,) against time where product formation was being followed)].
`For slow reactions, points were collected for the first two half-lives and the infinity value was
`calculated by using a computer program designed to give the best straight-line fit to the data. Good
`agreement was obtained between rate constants obtained by the two methods.
`Repetitive scans to demonstrate the existence of isosbestic points were obtained on a Hewlett-
`Packard 7041A X-Y recorder.
`The percentage of diazepam or nitrazepam present at equilibrium was determined by measurement
`of the absorbance of the model compound, the reactant (A,) and the equilibrium mixture (A,) at
`280 nm. The percentage reactant R(%) at equilibrium was calculated from equation ( I )
`
`l6 Mudry, C. A., and Frasca, A. R., Tetrahedron, 1973, 29, 603.
`l 7 Duynstee, E. F. J., and Grunwald, E., J. Am. Chem. Soc., 1959, 81, 4540.
`
`Opiant Exhibit 2307
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00685, IPR2019-00688, IPR2019-00694
`Page 7
`
`

`

`T. J. Broxtod, T. Ryan and S. R. Morrison
`
`The rate k of the forward reaction (1) -+ (3) was obtained from the observed rate of establishment
`of equilibrium by means of equation (2), in which I ( % ) is the percentage of the intermediate
`
`Acknowledgment
`We would like to thank Roche Products Pty. Ltd. (Sydney, Australia) for providing
`samples of diazepam and nitrazepam.
`
`Manuscript received 8 March 1984
`
`Opiant Exhibit 2307
`Nalox-1 Pharmaceuticals, LLC v. Opiant Pharmaceuticals, Inc.
`IPR2019-00685, IPR2019-00688, IPR2019-00694
`Page 8
`
`