Synthesis of 3,5-isoxazolidinediones and 1H-2,3-benzoxazine-1,4(3H)-diones from aliphatic oximes and dicarboxylic acid chlorides.

The synthesis of the title compounds was carried out by reacting dicarboxylic acid chlorides with oximes in the presence of excess triethylamine. Disubstituted malonyl chlorides gave 2-alkenyl-4,4-dialkyl-3,5-isoxazolidinediones (8a-f) and 2,2'-ethylidene-bis[4,4-dialkyl-3,5-isoxazolidinedione]s (9a-f). Compounds 9 were formed from 8 and its N-unsubstituted 3,5-isoxazolidinedione decomposition product. Phthaloyl chlorides reacted with acetone oxime to yield 3-(1-methylethenyl)-1H-2,3-benzoxazine-1,4(3H)-diones (16a-e). Products 16 spontaneously decomposed to give N-unsubstituted 1H-2,3-benzoxazine-1,4(3H)-diones (17a-e) at rates that were dependent on temperature and solvent. Kinetic studies showed that two of the compounds decomposed by zero-order kinetics under neutral conditions. Butanedioyl chloride did not produce a cyclic product.

Introduction

The 3,5-isoxazolidinediones (1) are an important pharmacologically active class of compounds. We have shown that they are potent hypolipidemic[1−5] and cytotoxic[6] agents, specific inhibitors of the type-2 isoform of IMPDH,[7] and inhibitors of aldose reductase.[8,9] They also act as antidiabetic agents.[10−15] Six-membered ring analogues of 1 include the 2H-1,2-oxazine-3,6-diones (2) and the 1H-2,3-benzoxazine-1,4(3H)-diones (3).[16−25] While there has been one report each on the fungicidal properties[23] and the β-Lactamase pro-inhibitory activity[26] of 3, the pharmacological properties of 2 and 3 have been largely unexplored. In contrast, the corresponding 2H-1,3-benzoxazine-2,4(3H)-diones (4) have a wide range of pharmacologic activities including antimycobacterial,[27−31] antituberculotic,[32−36] analgesic,[37,38] cardiovascular,[39] antimitotic,[40] antifungal,[41,42] cytotoxic,[43] and antifungal,[44] and the 2H-1,4-benzoxazine-2,3(4H)diones (5) are active as antimiotic,[40] antimicrobial,[45] antiallergic,[46,47] and renin inhibitory[48] agents. It appears likely then that 2 and 3 would also have useful pharmacological activities.

As part of our ongoing investigation into the pharmacological activities of 3,5-isoxazolidinediones and related ring systems, it was of interest to investigate the synthesis of N-alkenyl derivatives of 1, 2, and 3. We reported previously that addition of dimethyl (6a) and diethylmalonyl chloride (6b) to ether solutions of acetone oxime (7a) and triethylamine gave 2-(1-methylethenyl)-3,5-isoxazolidinediones (8a and 8b) and 2,2′-(1-methylethylidene)bis[4,4-dialkyl-3,5-isoxazolidinedione]s (9a and 9b). Also formed were small amounts of N-unsubstituted 3,5-isoxazolidinediones (10a and 10b) and O,O′-(2,2-dialkyl-1,3-dioxo-1,3-propanediyl)dioxime 2-propanones (11a and 11b) (Scheme 1).[49] Products 8a and 8b represent the only examples of 2-alkenyl-3,5-isoxazolidinediones reported to date. Mechanisms for the formation of products 8 and 9 were proposed. In simple terms products 8 can be considered to arise by the condensation of the oxime with 6 followed by base-promoted ring closure of the monooxime ester. Compounds 8b and 9b were subsequently shown to give moderate lowering of blood serum cholesterol and serum triglyceride levels in CF1 mice when administered for 16 days at a dosage of 20 mg/kg/day ip.[1,2]

Scheme 1

Reactions of Substituted Malonyl Chlorides with Oximes

Because only a limited number of derivatives of 8 and 9 were obtained, we explored in this study additional reactions between 6 and 7 in order to gain insight into the scope of the reaction. We also investigated the unreported reactions of phthaloyl and butanedioyl chlorides, respectively, with oximes to determine if they would react similarly to 6. Reported herein are the results of these investigations.

Results and Discussion

Reactions of malonyl chlorides with a variety of oximes were investigated. In our previous study we observed that the stability of compounds 8 was related to the steric size of the substituents at position-4 of the ring and followed the order Et > Me > H. Both 8a and 8b were stable to distillation under reduced pressure, but when stored at room temperature, 8b was indefinitely stable, whereas 8a slowly decomposed over several months. Compound 8 (R = H) was too unstable to be isolated. Higher yields of 8b and 9b were obtained when the reactions were carried out in the presence of a 50% excess of Et3N.[49] Accordingly, the reactions in the current study were generally carried out using this excess.

Dipropylmalonyl chloride (6c) was added to an ether solution of acetone oxime (7a) in the presence of excess Et3N at 0 °C, and the mixture was stirred for 20 h at room temperature to give 8c, 9c, and 10c, respectively (Table 1). When the reaction was carried out with inverse addition under the same conditions, i.e., by adding 7a to an ether solution of 6c and Et3N, the yields of 8c and 10c increased slightly, whereas that of 9c decreased. In a similar manner dibutylmalonyl chloride (6d) reacted with 7a to yield 8d, 9d, and 10d. The yield of 8d was negligible in the standard addition but increased substantially, along with a decreased yield of 9d and an increased yield of 10d, in the inverse addition. There was no evidence of bis-oxime ester formation (11c or 11d) in either of the two reactions. Compound 8c was stable to heating at 56 °C for 3.5 h. Both 8c and 8d were indefinitely stable at room temperature.

Table 1

Reactions of Dialkylmalonyl Chlorides (6) with Oximes (7)

		percent yielda
acid chloride	oxime	8	9	10	11	13	14
6a	7a	8a (29b,c,d)	9a (25b,c,d)	10a (2b,c)	11a (trb,c)
6b	7a	8b (23b)	9b (51b)	10b (trb)	11b (trb)
6c	7a	8c (22)	9c (28)	10c (5)	11c –
		8c (28e)	9c (18e)	10c (11e)
6d	7a	8d (tr)	9d (40)	10d (12)	11d –
		8d (22e)	9d (30e)	10d (17e)
6b	E-7b	8e (45)	9e (16)	10b –	11e –
6c	E-7b	8f (18)	9f −	10c –	E,E′-11f (40)
6b	12a					13a (62f)	14a –f
						13a (2g)	14a (35g)
6b	12b					13b –	14b (16)

Except where noted, the reactions were carried out by adding 6 to an ether solution of 7 and a 50% excess of Et3N at 0 °C and then stirring at rt for 20 h.

Reference (42).

A stoichiometric quantity of Et3N was used.

Yield revised from reference (42).

7a was added to an ether solution of 6 and Et3N.

Initial yield by HPLC.

Yield by HPLC after 72 h.

The reaction of 6b with (E)-acetophenone oxime (7b) produced both 8e and 9e but no 10b or 11e. Compound 8e was stable at room temperature but thermally unstable to vacuum distillation, giving a complex mixture that showed no vinyl hydrogen signals in its 1H NMR spectrum. The yields of 8e and 9e varied with the amount of base used. When the reaction was carried out with a stoichiometric quantity of Et3N, the yields of 8e and 9e were 36% and <1%, respectively. However, using a 50% excess of Et3N caused the yield of 8e to increase slightly (39%) and the yield of 9e to increase to 10%, similar to that observed for 8b and 9b. Compound 9e showed separate 1H NMR signals for the isoxazolidinyl ethyl groups due to its asymmetry. Its mass spectrum was typical of compounds 9 in general in that it did not show a molecular ion peak but gave instead a peak corresponding to cleavage of one of the isoxazolidine rings.[50] The corresponding reaction of 6c with E-7b produced a little 8f but no 9f or 10c. Instead E,E′-11f was obtained as the major product. Apparently the presence of the larger propyl groups retards ring closure of the initially formed monooxime ester to 8f and promotes diesterification of 6c. The stereochemistry of 11f has been assigned as E,E′ because it was formed from E-7b and its 13C NMR spectrum shows the oximyl methyl groups at δ 14.4. The corresponding methyl groups of the Z,Z′-isomer or one of the methyl groups of the E,Z-isomer are expected to appear near δ 26.[51]

The reaction of 6b with cyclic oximes was studied. Cyclohexanone oxime (12a) reacted with 6b to yield initially 13a (62%) and no 14a as shown by 1H NMR and HPLC analysis of the isolated product mixture. Compound 13a was thermally unstable at room temperature and gradually decomposed over 72 h to give partial conversion to 14a (35%) and several unidentified decomposition products. Only 2% of 13a remained. It is probable that 10b was an intermediate in the conversion.[49] Intermediate 10b was likely formed in the same manner as compounds 17 (see discussion below). The overall process would involve partial decomposition of 13a to 10b followed by addition of the NH bond of 10b across the alkenyl double bond of 13a. The exact mechanism for the addition step is not known.

In a similar manner cyclopentanone oxime (12b) reacted with 6b to produce 14b, but 13b was not obtained. Based on the behavior observed in the corresponding reaction of 12a, it is likely that 13b was formed but was not stable enough for isolation or detection.

Phthaloyl chloride (15a) reacted with 7a in the presence of triethylamine to yield 3-(1-methylethenyl)-1H-2,3-benzoxazine-1,4(3H)-dione (16a) and the N-unsubstituted 1H-2,3-benzoxazine-1,4(3H)-dione (17a)[52,53] (Scheme 2). There was no evidence for the presence of the 2,2-bis(2,3-benzoxazine-1,4-dione)propane 18 or bis-acetoneoxime phthalate 19 (Table 2). An attempt to prepare 16a by reacting phthalic anhydride with 7a gave only unreacted starting materials. The reaction between 4-methylphthaloyl chloride (15b) and 7a gave a mixture containing the 6-methyl- and 7-methyl-3-alkenyl-1H-2,3-benzoxazine-1,4-diones (16b and 16c). No 17b or 17c was isolated. The presence of the two isomers was demonstrated in the 1H NMR spectrum, which showed four vinyl hydrogen signals and two methyl singlets (58:42) at δ 2.58 and δ 2.56, respectively. Attempts to separate the two isomers by column chromatography or analytical HPLC were unsuccessful. In a similar manner 4-methoxyphthaloyl chloride (15c) reacted with 7a to produce a nearly equimolar mixture containing the 6-methoxy- and 7-methoxy-3-alkenyl-1H-2,3-benzoxazine-1.4-diones (16d and 16e). Two methoxy singlets (52:48) appeared in the 1H NMR spectrum at δ 3.89 and 3.87, respectively. 3-Nitrophthaloyl chloride (15d) reacted with 7a to give a complex product mixture. The presence of 16f and 16g was not detected in the 1H NMR spectrum of the mixture, which showed no vinyl hydrogens, and there was no evidence for the formation of 17f and 17g. It is possible that the presence of the strong electron-withdrawing nitro group caused the initially formed N-alkenyl products 16f and 16g to be too unstable for isolation.

Scheme 2

Reaction of Phthaloyl Chlorides with 7a

Table 2

Reaction of Phthaloyl Chlorides (15) with 7a

		percent yielda
acid chloride	oxime	16	17
15a	7a	16a (74)	17a (11)
15b	7a	16b, 16c (76)	17b, 17c (0)
15c	7a	16d, 16e (30)	17d, 17e (<1)
15d	7a	16f, 16g −	17f, 17g –

The reactions were carried out by adding 15 to an ether solution of 7a and a 50% excess of Et3N at 0 °C and then stirring at rt for 24 h.

Compound 16a was found to be thermally unstable. A sample of pure 16a gradually and completely decomposed in a sealed glass vial at room temperature over 3 months to yield 17a. Therefore, the decomposition was spontaneous at room temperature. Similarly, a mixture of 6-methyl- and 7-methyl-1H-2,3-benzoxazine-1,4-dione 17b and 17c was obtained by allowing the mixture of 16b and 16c to partially decompose at room temperature over 2 weeks (33% conversion by HPLC). Two methyl singlets (52:48) appeared in the 1H NMR spectrum at δ 2.60 and 2.58, respectively. Compound 16a decomposed to give 17a (98%) and N-hydroxyphthalimide (20) (2%) when it was heated under nitrogen at 130 °C for 1 h. Continued heating for an additional 2 h at 130 °C yielded 20 (99%) and 17a (1%) (Scheme 3). Attempts to purify 16a by recrystallization from hot cyclohexane led to partial conversion to 17a, and this continued each time the recrystallization was repeated. The formation of 17 by the thermal decomposition of 16 represents an improved and efficient method for its formation compared to the previously reported synthetic route.[52,53] It has an advantage in that the loss of the N-alkenyl substituent of 16 occurs under neutral and relatively mild thermolytic conditions, so the use of acid, base, or high temperature is avoided. Thus acetone oxime appears to be a good reagent for converting phthaloyl chlorides to 17. It should be noted that compounds 17 cannot be prepared directly from 15 and hydroxylamine; N-hydroxyphthalimide is formed instead.[52−54]

Scheme 3

Thermal Decomposition of 16

Compound 16a was found to be stable in absolute EtOH/CH2Cl2 (80:20) at room temperature for 30 h. A solution (4.9 × 10–2 M) of 16a in MeCN/water (75:25) at room temperature showed virtually no decomposition after 1 h and slight decomposition to 17a (8%) after 10 h as shown by HPLC analysis. The decompositions were much more rapid when the solution was heated at reflux, giving 17a (93%) and 20 (7%) after 1 h and 17a (7%) and 20 (89%) after 10 h. The decomposition behavior observed in the refluxing solvent system was very similar to that observed from heating 16a at 130 °C under N2. This suggests that the decompositions occurred by the same mechanism under the two different conditions.

The relative stabilities of 16a, 16b,c, and 16d,e were compared in refluxing MeCN/water solutions under neutral conditions [MeCN/water (70:30)], acidic conditions [MeCN/water/chloroacetic acid (68:29:3)], and basic conditions [MeCN/water/pyridine (67:29:4)]. The reactions were monitored by HPLC, and data were taken at 15 min intervals. Kinetic plots of concn vs time, ln concn vs time, and 1/concn vs time were made to determine the reaction orders. The kinetic plots showed that under the neutral conditions 16a and 16b,c decomposed by zero-order kinetics, whereas 16d,e decomposed by first-order kinetics. All of the compounds decomposed by first-order kinetics under the acidic and basic conditions.[55] Table 3 shows the decomposition half-lives and the square of the linear correlation coefficients (r2) for the three sets of compounds under the different decomposition conditions. All of the r2 values are high, indicating a high probability for a linear correlation. Alkaline and acidic conditions enhanced the rate of decomposition, with basic conditions having a more pronounced effect. Overall the three compounds were most stable under neutral conditions and least stable under alkaline conditions. Under neutral conditions the order of stability of the three compounds appears to parallel their expected hydrophilicities. Factors such as polarity, hydrogen bonding, hydrophilicity, acid–base interactions, and other factors might make contributions to the observed relative rates. We will not speculate on the relative importance of these factors among the three compounds. However, it is important to emphasize that compounds 16 have N-alkenyl substituents on a resonance-stabilized amide type of nitrogen, so it is not surprising that these compounds did not show behavior consistent with those of enamines. Enamines would be expected to be more stable under alkaline and neutral conditions and less stable under acidic conditions.

Table 3

Thermal Decomposition of 3-(1-Methylethenyl)-1H-2,3-benzoxazine-1,4(3H)-diones (16) in Refluxing MeCN/Water Solutionsa

		t1/2 (min) (r2)
compound	concn (M × 103)a	neutral conditionsb	acidic conditionsc	basic conditionsd
16a	4.9	135e (0.993)	38f (0.978)	9f (0.999)
16b,c	4.6	480e (0.984)	50f (0.982)	10f (0.996)
16d,e	4.3	57f (0.998)	40f (0.998)	35f (0.991)

The solutions were prepared by dissolving 0.020 g of each compound in 20 mL of the appropriate solvent system.

MeCN/water (70:30).

MeCN/water/chloroacetic acid (68:29:3).

MeCN/water/pyridine (67:29:4).

Zero-order reaction.

First-order reaction.

Reactions exhibiting zero-order kinetics are relatively uncommon. This order is most often observed in photochemical reactions, heterogeneous reactions, and enzyme-catalyzed reactions.[56] Heterogeneous reactions involving a solid catalyst and enzyme-catalyzed reactions are found to be first order with respect to the catalyst or enzyme and zero order with respect to the substrate. This is because the rate of these reactions is determined by the concentration of the catalyst or enzyme, provided that the substrate is in sufficient excess. The zero-order rates exhibited by 16a and 16b,c are significantly slower than those of the first-order reactions. These compounds are less reactive than 16d,e under neutral conditions, implying different mechanisms. A plausible explanation for the zero-order reactions is that they occurred by catalysis on active sites on the surface of their glass reaction containers, and the number of these active sites was relatively small compared to the concentration of the compounds. In contrast, the first-order reactions are spontaneous thermally induced reactions. They are enhanced under acid or base catalysis or in the case of 16d,e by the electron-releasing methoxy group.

It is likely that either propyne or allene was evolved during the decompositions. Attempts to form a mercuric derivative of the evolved gas by the procedure of Johnson and McEwen gave a solid whose melting point indicated that dipropynyl mercury might be present.[57] When 0.3 mmol of 16b,c was allowed to stand at 70 °C for 1 h in a sealed IR gas cell, no IR discernible absorptions were observed. However, when the gas cell was kept at 70 °C for 24 h, 13% decomposition occurred, and a FTIR spectrum identical to that of pure acetone vapor was obtained. There was no evidence for the presence of either propyne or allene. In a separate experiment 16a was allowed to decompose in a sealed flask that was first purged with nitrogen at room temperature for 1 h and then evacuated to 55 Torr. The flask was submerged in an oil bath maintained at 125 °C while 16a partially decomposed over 30 min. After cooling, a volume of the headspace gas was removed and analyzed by GC–MS. In addition to atmospheric gases only acetone was observed to be present.

The decomposition of 16a can be depicted as an intramolecular β-elimination of either propyne from 21A or allene from 21B (Scheme 4). In terms of thermodynamic stability propyne is expected to preferentially form because its experimental heat of formation is more favorable by 5.5–5.6 kJ/mol than that of allene.[58,59] Although difficulties have been experienced in obtaining accurate C–H bond dissociation energies of simple alkanes and alkenes, the C–H bond dissociation energy of ethylene is reported to be 91–116 kJ/mol greater than that of the allylic C–H bond of propene at 298 K.[60,61] No experimental values for bond dissociation energies for the vinyl C–H bonds of propene have been published, but ab initio STO-3G calculations have suggested that the bond dissociation energies of the vinylic CH2 bonds of propene are within a few percent of the ethylene value.[62−64] These bond dissociation energy values argue for the preferred formation of allene via 21B provided that the decomposition is governed by kinetic control. It is well established that propyne and allene exist as an equilibrium mixture. At 298 K the equilibrium constant (Keq) for the isomerization of allene to propyne is 24.74.[63] This corresponds to a gas phase mixture containing 96.1% propyne and 3.9% allene. In the decompositions carried out in the refluxing solvent systems and at 130 °C equilibrium was probably established fairly rapidly, and if allene was initially or partially formed, it likely isomerized quickly to propyne. It is possible that the acetone was formed by hydration of propyne (or allene) with water that was adsorbed on the surface of the glass. Glasses are known to contain dissolved hydroxyl species and water.[65−68] Propyne and other alkynes are known to react with hydroxyl radical at 253–343 K by an addition mechanism.[69,70] Alkenes can also be hydrated in the gas phase.[71−73] Although it is not known with certainty how the acetone was formed, its presence is consistent with the loss of a three-carbon atom fragment from 16a. In the catalyzed decomposition reactions carried out in MeCN/water, the presence of pyridine would serve to enhance the rate of β-elimination from 21A or 21B by offering a competing pathway for its occurrence. The presence of chloroacetic acid would allow for the formation of 16A by β-elimination via 22.

Scheme 4

Decomposition Modes of 3-(1-Methylethenyl)-1H-2,3-benzoxazine- 1,4(3H)-diones

The reaction between 7a and butanedioyl chloride (23) in the presence of triethylamine was investigated to determine if the 3-alkenyl-2,3-oxazine-1,4-dione (25) would be produced. Neither the 3-alkenyl-2,3-oxazine-1,4-dione (25) nor the N-unsubstituted bis-2,3-2H-1,2-oxazine-3,6-dione (26) was formed (Scheme 5). The bis-dimethylketoximyl butanedioate (24) was formed instead. Compound 23 apparently reacted in its preferred anti conformation. The initially formed monoester did not undergo cyclization but underwent diesterification to yield 24. Diacid chlorides having carbon chains longer than four carbons would also be expected to give diesterified products.

Scheme 5

Reaction of Butanedioyl Chloride and 7a

Scheme 6 suggests the reaction pathways that are followed in the reactions that we have investigated involving acetone oxime with diacid chlorides. Initial reaction between 27 and 7a gives the monoxime ester 28, which can either lose methyl hydrogen to triethylamine and undergo ring closure to 30 or react with a second molecule of 7a to form 29. When X corresponds to one disubstituted sp3 hybridized carbon atom or ortho-disubstituted aromatic carbons, 30 is produced. In the former instance the reactions are sensitive to the size of the substituents both at position-2 of the malonyl chlorides and on the oxime. When X corresponds to two (and presumably more than two) sp3 hybridized carbons, dioxime esters 29 are formed. Compounds 30 are thermally unstable and spontaneously decompose at varying rates with loss of propyne and possibly allene to yield 31. Once formed compounds 31 can add to 30 to produce 32. This can occur under the reaction conditions or after 30 has been isolated as shown by the conversion of 13a to 14a. Compounds 32 are stable. The production of 32 does not occur in the reactions involving phthaloyl chlorides.

Scheme 6

Reaction Pathways Followed in the Reaction of Diacid Chlorides with Oximes

Experimental Section

General Procedure for the Reaction of Dialkylmalonyl Chlorides (6) with Oximes (7)

To a solution of 25 mmol of oxime (7 or 12) and 75 mmol of Et3N in 70 mL of anhydrous Et2O at 0 °C was added dropwise over 15 min 25 mmol of dialkylmalonyl chloride (6). The reaction mixture was allowed to warm to rt and then stirred under N2 for 20 h. The white precipitate was filtered, and the filtrate was washed with HCl (10%, 3 × 25 mL) and then dried (MgSO4). The Et2O solution was dried (MgSO4) and evaporated under reduced pressure. The residue was purified by column chromatography over silica gel (230–400 mesh) using hexane/EtOAc (97:3).

4,4-Dipropyl-2-(1-methylethenyl)-3,5-isoxazolidinedione (8c)

Reaction of 27 mmol of diacid chloride gave a colorless oil: yield 1.35 g, 22%, IR (neat) 1821, 1726, 1641 cm–1; 1H NMR (300 MHz; CDC13) δ 5.02 (s, 1H), 4.60 (s, 1H), 2.21 (s, 3H), 1.82 (m, 4H), 1.31 (m, 4H) and 0.92 (t, 6H); 13C NMR (75 MHz; CDCI3) δ 13.8, 18.0, 19.6, 37.9, 53.1, 100.3, 136.4, 167.7, 172.2; HRMS m/z [M + H]+ calcd for C12H20NO3 226.1444, found 226.1446; HRMS m/z [M + Na]+ calcd for C12H19NO3Na 248.126, found 248.126. Reaction of 25 mmol of diacid chloride using inverse addition gave 1.56 g, 28% of 8c.

2,2′-(1-Methylethylidene)bis[4,4-dipropyl-3,5-isoxazolidinedione] (9c)

Reaction of 27 mmol of diacid chloride using inverse addition gave a colorless viscous liquid which solidified on standing: yield 1.54 g, 28%; mp 74–76 °C (LB pet ether); IR (neat) 1817 (s), 1738 cm–1 (s); 1H NMR (300 MHz; CDC13) δ 2.09 (s, 6H), 1.74 (m, 8H), 1.31 (m, 8H), and 0.92 (t, 12H); 13C NMR (75 MHz; CDCI3) δ 13.9, 17.9, 25.7, 37.7, 53.2, 78.8, 170.5, 171.9; HRMS m/z [M + K]+ calcd for C21H34N2O6K 449.2055, found 449.2037. Anal. Calcd for C21H34N2O6: C, 61.44; H, 8.35; N, 6.82. Found: C, 61.37; H, 8.24; N, 6.71. Reaction of 25 mmol of diacid chloride using inverse addition gave 0.90 g, 18% of 9c.

4,4-Dipropyl-3,5-isoxazolidinedione (10c)

Reaction of 27 mmol of diacid chloride gave a colorless viscous liquid: yield 0.25 g, 5%; IR (neat) 1817, 1738 cm–1; 1H NMR (300 MHz; CDC13) δ 8.54 (br s, 1H), 1.80 (m, 4H), 1.31 (m, 4H), and 0.92 (t, 6H); 13C NMR (75 MHz; CDCI3) δ 13.8, 17.9, 37.4, 51.6, 174.6, 175.7; HRMS m/z [M + Na]+ calcd for C9H15NO3Na 208.0950, found 208.0948. Reaction of 25 mmol of diacid chloride using inverse addition gave 0.50 g, 11% of 10c.

4,4-Dibutyl-2-(1-methylethenyl)-3,5-isoxazolidinedione (8d)

Reaction of 20 mmol of diacid chloride using inverse addition gave a colorless viscous liquid: yield 1.12 g, 22%; IR (neat) 1820, 1728, 1642 cm–1; 1H NMR (300 MHz; CDC13) δ 5.04 (s, 1H), 4.60 (s, 1H), 2.22 (s, 3H), 1.83 (m, 4H), 1.28 (m, 8H) and 0.90 (t, 6H); 13C NMR (75 MHz; CDCI3) δ 13.7, 19.6, 22.5, 26.6, 35.7, 53.0, 100.3, 136.5, 167.8, 172.2; HRMS m/z [M + H]+ calcd for C14H24NO3 254.1751, found 254.1755; HRMS m/z [M + Na]+ calcd for C14H23NO3Na 276.156, found 276.158.

2,2′-(1-Methylethylidene)bis[4,4-dibutyl-3,5-isoxazolidinedione] (9d)

Reaction of 18 mmol of diacid chloride gave a colorless viscous liquid that solidified on standing: yield 1.70 g, 40%; mp 67–68 °C (LB pet ether); IR (neat) 1822, 1735 cm–1 (s); 1H NMR (300 MHz; CDC13) δ 2.10 (s, 6H), 1.79 m, 8H), 1.30 (m, 16H), and 0.89 (t, 12H); 13C NMR (75 MHz; CDCI3) δ 13.6, 22.6, 25.7, 26.5, 35.4, 53.0, 78.8, 170.5, 171.8; HRMS m/z [M + H]+ calcd for C25H43N2O6 467.3116, found 467.3105; HRMS m/z [M – C11H18NO3]+ calcd for C14H24NO3 254.1757, found 254.1766; HRMS m/z [M + K]+ calcd for C25H42N2O6K 505.2674, found 505.2676. Reaction of 20 mmol of diacid chloride using inverse addition gave 1.39 g, 30% of 9d.

4,4-Dibutyl-3,5-isoxazolidinedione (10d)

Reaction of 18 mmol of diacid chloride gave a colorless viscous liquid: yield 0.46 g, 12%; 230–400 mesh silica gel (hexane/EtOAc (80:20)); IR (neat) 1817, 1737 cm–1; 1H NMR (300 MHz; CDC13) δ 10.25 (br s, 1H), 1.84 m, 4H), 1.30 (m, 8H), and 0.89 (t, 6H); 13C NMR (75 MHz; CDCI3) δ 13.6, 22.5, 26.6, 35.1, 51.5, 174.3, 176.0; HRMS m/z [M + Na]+ calcd for C11H19NO3Na 236.1263, found 236.1248; HRMS m/z [M + K]+ calcd for C11H19NO3K 252.1003, found 252.0985. Reaction of 20 mmol of diacid chloride using inverse addition gave 0.73 g, 17% of 10d.

4,4-Diethyl-2-(1-phenylethenyl)-3,5-isoxazolidinedione (8e)

Reaction of 20 mmol of diacid chloride gave a colorless viscous liquid: yield 2.32 g, 45%; 60–200 mesh silica gel (CHCl3/hexane (98:2)) or 50 cm ODS-2 HPLC column (MeCN/water (60:40)); IR(neat) 1816(s), and 1721(s) cm–1 (s); 1H NMR (60 MHz; CDC13) δ 7.30 (s overlapping a small m, 5H), 5.47 (s, 1H), 5.34 (s, 1H), 1.83 (q, 4H), and 0.91 (t, 6H); MS m/z (rel intensity) 259 (44), 215 (48), 199 (85), 108 (89) and 106 (100); HRMS m/z [M]+ calcd for C15H17NO3 259.1208, found 259.1209. Anal. Calcd for C15H17NO3: C, 69.48; H, 6.61; N, 5.40. Found: C, 69.64; H, 6.91; N, 5.04.

2,2′-(1-Phenylethylidene)bis[4,4-diethyl-3,5-Isoxazolidinedione] (9e)

Reaction of 20 mmol of diacid chloride gave white crystals that precipitated from the crude product mixture: yield 0.66 g, 16%; mp 143–145 °C (LB pet ether); IR (Nujol) 1760(s) and 1672(s) cm–1; 1H NMR (60 MHz, CDC13) δ 7.43 (m, 5H), 3.75 (s, 3H), 2.03 (q, 4H), 1.63 (q, 4H), 1.06 (t, 6H), and 0.47 (t, 6H); MS m/z (rel intensity) 260 (3) (M – C7H17NO3), 247 (45), 246 (100), 219 (52), and 105 (99); HPLC purity (254 nm, Partisil 10, hexane/EtOAc (80:20): 99.6%.; HRMS m/z [M – C7H17NO3]+ calcd for C15H18NO3 260.1286, found 260.1286.

4,4-Dipropyl-2-(1-phenylethenyl)-3,5-isoxazolidinedione (8f)

Reaction of 22 mmol of diacid chloride gave a colorless viscous liquid: yield 1.14 g, 18%; IR (neat) 1817, 1737 cm–1 (s); 1H NMR (300 MHz; CDC13) δ 7.39 (s overlapping a small m, 5H), 5.56 (s, 1H), 5.44 (s, 1H), 1.87 (m, 4H), 1.40 (m, 4H), and 0.96 (t, 6H); 13C NMR (75 MHz; CDCI3) δ 13.9, 18.3, 38.1, 53.0, 109.1, 126.7, 128.5, 129.4, 133.1, 139.7, 169.0, 172.5; HRMS m/z [M + H]+ calcd for C17H21NO3H 288.1601, found 288.1617; HRMS m/z [M + Na]+ calcd for C17H21NO3Na 310.1420, found 310.1498.

O,O′-(2,2-Dipropyl-1,3-dioxo-1,3-propanediyl)-E,E′-dioxime 1-Phenylethanone (11f)

Reaction of 22 mmol of diacid chloride gave a white solid: yield 1.87 g, 40%; mp 104–105 °C (absolute EtOH); IR (Nujol) 1778, 1755 cm–1 (s); 1H NMR (300 MHz; CDC13) δ 7.75 (m, 4H), 7.43 (m, 6H), 2.34 (s, 6H), 2.09 (m, 4H), 1.40 (m, 4H), and 0.99 (t, 6H); 13C NMR (75 MHz; CDCI3) δ 14.4, 14.5, 17.5, 35.1, 57.2, 127.1, 128.6, 130.8, 134.5, 163.8, 168.5. Anal. Calcd for C25H30N2O4: C, 71.07; H, 7.14; N, 6.63. Found: C, 70.98; H, 7.14; N, 6.51.

2-(1-Cyclohexenyl)-4,4-diethyl-3,5-isoxazolidinedione (13a)

Reaction of 20 mmol of diacid chloride gave 3.4 g of an unstable brown oil; yield 62%, initial yield gradually decomposed to 2% after 72 h by HPLC; ODS-2 using H2O/MeCN (40:60) at 254 nm; IR(neat) 1817 (s), 1728 (s) cm–1; 1H NMR (60 MHz, CDC13) δ 5.96 (m, br, 1H)), 2.6 (m, 4H), 2.1–1.4 (m, 8H), and 0.90 (t, 6H).

1,l-Bis(2-(4,4-Diethyl-3,5-isoxazolidinedione)cyclohexane (14a)

Reaction of 20 mmol of diacid chloride gave a white crystalline solid that precipitated over 72 h from the product mixture: yield 1.38 g, 35%; mp 102–103 °C (HB pet ether); IR (Nujol) 1810(s) and 1715 cm –1 (s); 1H NMR (300 MHz; CDC13) δ 2.60 (br t, 4H), 1.86 (m, 8H), 1.69 (m, 4H), 1.57 (m, 2H), and 0.95 (t, 12H); 13C NMR (75 MHz; CDCl3) δ 9.1, 22.3, 24.4, 28.7, 33.6, 54.5, 82.5, 70.2, 171.9; MS m/z (rel intensity) 238 (66%) (M – C7H10NO3), 193 (10%), 164 (11%), 98 (26%), 97 (57%), 95 (100%). HRMS m/z [M – C7H10NO3]+ calcd for C13H20NO3 238.1443, found 238.1444. Anal. Calcd for C20H30N2O6: C, 60.89; H, 7.67. Found: C, 60.91; H, 7.62.

1,l-Bis(2-(4,4-Diethylisoxazolidine-3,5-dione)cyclopentane (14b)

Reaction of 20 mmol of diacid chloride gave a white crystalline solid precipitated over 24 h from the crude product mixture: yield 0.62 g, 16%; mp 100–101 °C (LB pet ether); IR (Nujol) 1810 (s) and 1715 cm –1 (s); 1H NMR (300 MHz; CDC13) δ 2.63 (m, 4H), 2.2–1.5 (m, 12H), and 0.95 (t, 12H). Anal. Calcd for C19H28N2O6: C, 59.98; H, 7.42; N, 7.37. Found: C, 59.75; H, 7.35; N, 7.29.

General Procedure for the Preparation of the 3-(1-Methylethenyl)-1H-2,3-benzoxazine-1,4(3H)-diones (16)

In a three-neck flask fitted with an addition funnel and a condenser was prepared a solution containing acetone oxime (7a) (20 mmol) and Et3N (60 mmol) in 70 mL of anhydrous Et2O. The flask was swept with N2, and the solution was chilled in an ice–water bath to 0 °C with stirring. To the cold solution was added dropwise over a 15 min period 20 mmol of the phthaloyl chloride (15) in 50 mL of anhydrous Et2O. The solution was stirred at rt under N2 for 24 h. The white precipitate was filtered, and the filtrate was washed with HCl (10%, 3 × 50 mL) and then dried (MgSO4). The filtrate was concentrated to give crude 16. Purification by column chromatography over silica gel gave pure 16.

3-(1-Methylethenyl)-1H-2,3-benzoxazine-1,4(3H)-dione (16a)

Reaction of 20 mmol of diacid chloride gave a white crystalline solid: yield 3.02 g, 74%; 60–200 mesh silica gel (CHCl3); mp 102–103 °C; IR (Nujol) 1748, 1667 cm–1; 1H NMR (300 MHz; CDCI3) δ 7.8–8.4 (m, 4H), 5.40 (s, 1 H), 5.18 (s, 1 H), 2.2 (s, 3H); HRMS m/z [M]+ calcd for C11H9NO3 203.0582, found 203.0585. Anal. Calcd for C11H9NO3: C, 65.02; H, 4.46; N, 6.89. Found: C, 64.80; H, 4.62; N, 6.82. Compound 16a gradually decomposed to 17a over several months when stored in a capped vial at room temperature.

1H-2,3-Benzoxazine-1,4-dione (17a)

Recrystallization of crude 11a (see above) from cyclohexane and then CHCl3 gave 17a as a white crystalline solid: yield 0.37 g, 11%; mp 231–232 °C (lit. mp 226 °C[52]); IR (Nujol) 1760(s), 1685(m) cm–1; 1H NMR (300 MHz; CDCI3) δ 7.95 (m); HRMS m/z [M]+ calcd for C8H5NO3 163.0269, found 163.0266. Anal. Calcd for C8H5NO3: C, 58.89; H, 3.09; N, 8.59. Found: C, 58.89; H, 3.19; N, 8.45.

6- and 7-Methyl-3-(1-methylethenyl)-1H-2,3-benzoxazine-1,4(3H)-dione (16b) and (16c)

Reaction of 20 mmol of diacid chloride gave a white crystalline solid: yield 3.78 g, 76%; 230–400 mesh silica gel (EtOAc/LB petroleum ether (5:95)); mp 70–110 °C; IR (Nujol) 1747, 1664 cm–1; 1H NMR (300 MHz; CDCI3) δ 7.64–8.21 (m, 6H), 5.49 (s, 1 H), 5.48 (s, 1 H), 5.24 (s, 1 H), 5.23 (s, 1 H), 2.58 (s, 3H), 2.56 (s, 3H), 2.21 (s, 6H); 13C NMR (75 MHz; CDCI3) δ 19.8, 19.9, 22.1, 22.2, 112.4, 112.9, 121.1, 122.0, 128.2, 128.8, 129.6, 130.0, 130.2, 130.3, 135.0, 135.3, 140.0, 140.2, 147.5, 147.8, 156.2, 156.6, 159.9, 160.4; MS m/z (rel intensity) 217 (0.1), 173 (33), 55 (100). Anal. Calcd for C12H11NO3: C, 66.35; H, 5.11; N, 6.45. Found: C, 66.13; H, 5.23; N, 6.33. The 1H NMR peak integration ratio of the two methyl singlets at δ 2.58 and 2.56, respectively, was 58:42.

6- and 7-Methyl-1H-2,3-benzoxazine-1,4(3H)-dione (17b) and (17c)

A mixture (52:48) of 16b and 16c weighing 1.91 g (8.8 mmol) was exposed to the atmosphere at rt for 2 weeks. Analysis of the solid by HPLC showed 33% decomposition of the alkene mixture, The solid was stirred in 20 mL of cold CHCl3 for 15 min and filtered to yield a mixture of 17b and 17c as a white solid: yield 0.48 g, 31%; mp 188–194 °C dec (CHCl3); IR (Nujol) 1759 (s), 1659 (s) cm–1; 1H NMR (300 MHz; acetone-d6) δ 7.78–8.08 (m, 6H), 2.60 (d, 3H) and 2.58 (s, 3H); 13C NMR (75 MHz; DMSO-d6) δ 17.1, 17.4, 116.7, 119.1, 120.3, 121.4, 121.5, 122.8, 124.3, 124.6, 130.9, 132.5, 140.9, 142.9, 154.5, 154.6, 157.5, and 157.8. The peak integration ratio of the two 1H NMR methyl singlets at δ 2.60 and 2.58, respectively, was 52:48. Anal. Calcd for C9H7NO3: C, 61.01; H, 3.99; N, 7.91. Found: C, 60.74; H, 3.73; N, 7.68.

6- and 7-Methoxy-3-(1-methylethenyl)-1H-2,3-benzoxazine-1,4(3H)-dione (16d) and (16e)

Reaction of 11 mmol of diacid chloride gave a white crystalline solid: yield 0.77 g, 30%; 100–200 mesh silica gel (CH2Cl2); mp 93–105 °C; IR (Nujol) 1747, 1656 cm–1; 1H NMR (300 MHz; CDCI3) δ 7.20–8.13 (m, 6H), 5.39 (s, 1 H), 5.37 (s, 1 H), 5.13 (s, 1 H), 5.11 (s, 1 H), 3.89 (s, 3H), 3.87 (s, 3H), 2.12 (s, 6H); 13C NMR (75 MHz; CDCl3) δ 18.8, 18.9, 55.1, 55.2, 109.1, 110.2, 111.1, 111.4, 111.5, 114.9, 120.9, 122.0, 122.6, 124.3, 129.1, 130.7, 131.5, 138.8, 154.7, 154.9, 158.5, 158.9, 162.9, 164.6; MS m/z (rel intensity) 233 (0.3), 189 (39), 55 (100). The 1H NMR peak integration ratio of the two methoxy singlets at δ 3.89 and 3.87, respectively, was 52:48. Anal. Calcd for C12H11NO4: C, 61.60; H, 4.76; N, 6.01. Found: C, 61.56; H, 4.63; N, 5.64.

Reaction of 1,4-Butanedioyl Chloride (23) with Acetone Oxime (7a)

A solution of 1.83 g (25 mmol of acetone oxime (7a) in 50 mL of anhydrous Et2O was cooled to 0 °C with stirring. To the solution was added 5.1 g (50 mmol) of Et3N in 40 mL of dry ether. A solution of 3.9 g (25 mmol) of 1,4-butanedioyl chloride (23) was added dropwise over 15 min with stirring. The mixture was stirred at rt for 24 h. The mixture was filtered, dried (MgSO4), and concentrated under reduced pressure to give a tan solid. Chromatography over silica gel (60–200 mesh) with EtOAc (100%) yielded O,O′-(1,4-dioxo-1,4-butanediyl)dioxime 2-propanone (24) as a pale yellow oil that solidified on standing: yield 1.05 g (37%); mp 60–63 °C (CCl4/HB petroleum ether (2:1)); IR (Nujol) 1765 (s) cm–1 (s); 1H NMR (60 MHz, CDC13) δ 2.8 0 (s, 4H), 2.08 (s, 6H), 2.03 (s, 6H); HRMS m/z [M – C3H6NO]+ calcd for C7H10NO3 156.0060, found 156.0060. Anal. Calcd for C10H16N2O4: C, 52.62; H, 7.07; N, 12.28. Found: C, 52.37; H, 7.35; N, 12.05.

Stability of 3-(1-Methylethenyl)-1H-2,3-benzoxazine-1,4(3H)dione (16a)

Thermal Stability

A 0.50 g portion of 15a was heated under N2 for 1 h at 130 °C to yield a mixture of 17a (98%) and N-hydroxyphthalimide (20) (2%): mp 231–232 °C. Heating 15a for 3 h at 130 °C yielded a mixture of 20 (99%) and 17a (1%): mp 233–234 °C (dec). (lit. mp 232–133 °C[74])

In Absolute EtOH for 30 h

A solution of 0.20 g of 16a in 15 mL of absolute EtOH/CH2Cl2 (80:20) was allowed to stand at rt for 30 h. The solution was concentrated to give unchanged 16a.

In MeCN/Water (75:25) at rt

A 4.9 × 10–2 M solution containing 0.50 g of 16a in 50 mL of MeCN/water (75:25) was stirred at rt for 1 h. The MeCN was removed under reduced pressure, and the aqueous solution was extracted with CHCl3 (2 × 50 mL). The CHCl3 washings were dried (MgSO4) and concentrated to give unchanged 16a. A separate reaction was carried out for 10 h under the same conditions to give a white fluffy solid: mp 101–105 °C. Analysis by HPLC showed 8% decomposition.

In Refluxing MeCN/Water (75:25)

A 4.9 × 10–2 M solution containing 0.50 g of 16a in 50 mL of MeCN/water (75:25) was heated at reflux for 1 h with stirring. Upon cooling to rt the MeCN was removed under reduced pressure, and the remaining aqueous solution was extracted with CHCl3 (2 × 25 mL). The CHCl3 washings were dried (MgSO4) and concentrated to give a mixture containing 17a (93%) and 20 (7%) as shown by HPLC analysis. A separate reaction was carried out under the same conditions for 10 h to give a mixture containing 17a (7%) and 20 (89%).

Stability of 3-(1-Methylethenyl)-1H-2,3-benzoxazine-1,4(3H)-diones (16a) in Refluxing MeCN/Water (70:30). General Procedure for the Determination of the Relative Stabilities of 16a and mixtures of 16b and 16c and 16d and 16e

A solution of 0.020 g (0.10 mmol) of 3-(1-methylethenyl)-2,3-benzoxazine-l,4-dione (16a) in 20 mL of MeCN/water (70:30) was heated under reflux for 90 m. A small aliquot was removed every 15 min and analyzed by reverse phase HPLC using a solvent system of MeCN/water (70:30), a flow rate of 2.0 mL/min, and a detector wavelength of 280 nm. The peak areas were corrected for the weight response factors of the components. In a similar manner the decomposition of 16a in refluxing MeCN/water/chloroacetic acid (68:29:3) and in refluxing MeCN/water/pyridine (67:29:4) were carried out. Also studied in each of the three refluxing solvent systems were a mixture of 16b and 16c (58:42) and a mixture of 16d and 16e (52:48).

Analysis of the Gas Evolved during Thermal Decomposition of 3-(1-MethyletHenyl)-1H-2,3-benzoxazine-l,4(3H)-dione (16a)

Preparation of Mercuric Derivative

A procedure similar to that of Johnson and McEwen was used.[57] An ethanolic solution of the gas evolved in the thermal decomposition of 16a at 130 °C was treated with aqueous alkaline mercuric iodide with stirring for 10 m. The precipitate was removed by suction filtration to yield a yellow-green powder: mp 190–210 °C partial melting. Johnson and McEwen reported that dipropynyl mercury melted at 203–204 °C.[57]

FTIR Analysis of the Gas Evolved during Thermal Decomposition

The gas evolved during the thermal decomposition of 16a at 130 °C was collected in an IR gas cell and sealed: FTIR 2932 (w), 1737 (m), 1441(w), and 1365 (w) cm–1. The FTIR spectrum was identical with that obtained with vaporized acetone.

GC–MS Headspace Analysis of the Gas Evolved during Thermal Decomposition

A 250 mL three-neck round-bottom flask was charged with 0.30 g (1.5 mmol) of 16a. The flask was purged with N2 for 1 h. The pressure was reduced to 55 Torr, and the sealed flask was submerged in an oil bath heated to 125 °C. Gas was evolved over 30 min. The flask was allowed to cool, and a sample of the headspace gas was analyzed by GC–MS. The gas was found to contain 5% acetone plus atmospheric components.