A Merged Aldol Condensation, Alkene Isomerization, Cycloaddition/Cycloreversion Sequence Employing Oxazinone Intermediates for the Synthesis of Substituted Pyridines

Abstract

A domino reaction sequence has been evaluated that begins with union of novel dihydrooxazinone precursors with 2-alkynyl-substituted benzaldehyde components through aldol condensation. Ensuing operations, including alkene isomerization, Diels–Alder, and retrograde Diels–Alder with loss of CO₂ occurs in the same reaction vessel to provide polysubstituted tricyclic pyridine products.

The value of domino reactions, which consist of two or more chemical events that occur in a single flask, has been most clearly elucidated by Tietze,[2] and is also apparent to the practicing chemist: fewer time-consuming purification processes and mitigation of solvent waste is advantageous as compared to stepwise synthesis and is more aligned with the principles of the ideal synthesis.[3]

Previous efforts from our lab have employed a domino reaction process involving an aldol condensation of diketopiperazine precursors, alkene isomerization to an intermediate pyrazinone, and cycloaddition to construct the [2.2.2]diazabicyclic functionality (Scheme [1], eq. 1).[4] If the cycloaddition event intercepts an alkyne dienophile, the resulting diazabicycloalkene adduct (e.g., 2) can undergo further reaction: cycloreversion and extrusion of one bridging lactam function (as an isocyanate or cyanogen derivative) to afford 2-pyridone or pyridine structures.[5] We determined that intermediate [2.2.2]diazabicyclic structures resembling 2 could be reliably converted into the derived pyridone 4 through a three-step process that involved activation of one lactam bridge as the derived acetylated intermediate 3 prior to thermal extrusion (microwave, 300 W, 1 h, max. temp ca. 200 °C). While the three-step process for conversion of [2.2.2]diazabicycloalkene adducts such as 2 into pyridone products resembling 4 was general, we desired a more expedient overall sequence. We anticipated that an analogous domino reaction process initiated with a dihydrooxazinone precursor 5 might also undergo aldol condensation and isomerization and intercept oxazinone 7 (Scheme [1], eq. 2). Following cycloaddition of 7, the resulting intermediate cycloadduct 8, which bears a lactone-bridging function, would extrude CO₂ possibly at the temperatures necessary for the preceding cycloaddition and thereby deliver the tricyclic pyridine product 9. In this way, we anticipated that the cycloaddition and cycloreversion steps could be merged into a single operation. Although not widely explored, oxazinone cycloaddition/cycloreversion reactions are known from the work Hoornaert,[6] however, our desire to incorporate these pericyclic processes into more elaborate domino reaction pathways was not established. We reasoned that successful execution of the desired plan would advance new chemistry and provide complementary methods for the rapid assembly of polycyclic substituted pyridines, a widespread and privileged scaffold that is shared among many molecules that find application in medicine, agriculture, and material science.[7] Given the many valuable properties of pyridines, continued efforts directed toward the construction of pyridines are warranted and remain an active research area.[8] Metal-free domino reaction processes, such as the one described in this letter, offer the potential for direct synthesis of substituted pyridines with minimal environmental impact.[9]

In order to begin our study of the domino reaction sequence, we first needed to prepare the requisite dihydrooxazinone precursors (e.g., 5), which were apparently unknown. We determined that the glycine-glycolate derived dihydrooxazinone precursor 5a could be prepared in three steps starting from chloroacetic acid (Scheme [2]).[10] The lactim ether functionality in 5a was established by Staudinger reduction of the derived azide and cyclization of the intermediate aza-Wittig intermediate on the pendant methyl ester at 90 °C. The desired product 5a was separated from the stoichiometric phosphine oxide byproduct most conveniently by distillation using a Kugelrohr apparatus.

Although dihydrooxazinone 5a was somewhat prone to degradation (by putative polymerization to insoluble products with unresolved spectroscopic characteristics), material purified by Kugelrohr distillation was stable in the freezer on the duration of weeks. The C-5 methyl- and phenyl substituted dihydrooxazinones 5b and 5c were prepared using a modified sequence starting from lactate and mandelate precursors (Scheme [2]). The synthesis of C-5 phenyldihydrooxazinone 5c required more elevated temperatures (110 °C) and longer duration (24 h) in order to promote complete cyclization of the aza-Wittig intermediate on the methyl ester and deliver the lactim O-methyl ether functionality. The more robust stability of 5b and 5c was a favorable attribute that enabled purification on silica gel and improved the efficiency of the reaction sequence (31 and 48% yield over three steps).

With three dihydrooxazinone precursors (5a–c) in hand, we selected the most simple 2-alkynyl benzaldehyde derivative 10 with which to initiate our studies of the domino reaction sequence (Scheme [3]). When 10 and 5a were heated in toluene at 110 °C with DBU (1.5 equiv) we observed a small amount the desired tricyclic 2-methoxypyrdine product 13a (ca. 10–20%). Slow addition of the aldehyde via syringe pump improved the yield of product 13a to 77%.[11] No intermediate products resulting from aldol addition, condensation, isomerization, or cycloaddition were evident, and the unpurified reaction mixture appeared to contain only 13a and unreacted aldehyde component (in some cases). It was remarkable to us that DBU is sufficiently basic to promote enolization and aldol condensation of dihydrooxazinone substrates, which contrasts with our previous work using diketopiperazine starting materials (e.g., 1), which required a stronger base (NaOMe or LiHMDS) to promote enolization and aldol addition (see Scheme [1], eq. 1).[3] [4]

Two additional benzaldehyde derivatives 11 and 12 (extending either n-butyl or phenyl residues from the alkyne terminus) were prepared in order to explore this domino sequence in more detail. Using the unsubstituted dihydrooxazinone precursor 5a with either 11 or 12, the resulting 2-methoxypyridine products 13b and 13c were obtained, however, the isolated yield for these products was low (10% and 13%). Attempts to improve the reaction yield efficiency by varying temperature, stoichiometry, or base were met without success.

Reactions performed with the methyl-substituted dihydrooxazinone precursor 5b gave more consistent results with each of the three benzaldehyde reaction components (10–12), and the desired products 14a, 14b, and 14c were obtained in 58, 54, and 61% yield. Use of 5-phenyl dihydrooxazinone 5c was also successful in the reaction sequence with the alkynyl benzaldehyde derivatives 10–12 and afforded the respective tricyclic products 15a, 15b, and 15c in 32%, 22%, and 50% yield. Overall, reactions with the 5-alkyl dihydrooxazinones 5b or 5c did not benefit from slow addition of aldehyde; as such these reactions did not require the syringe pump and were thus more convenient to execute. In all reactions performed, the unpurified reaction products were uncomplicated and contained desired products and unreacted starting materials and only trace impurities. Accordingly, we attribute the modest yields in several cases to poor mass recovery, a feature consistent with the observed propensity of the dihydrooxazinone precursors to degrade and polymerize.

The reaction sequence described in this letter demonstrates the proof of principle of a domino reaction sequence that features several base-promoted and pericyclic bond-forming and bond-cleaving processes. The sequence is initiated by condensation of novel dihydrooxazinone starting materials with aromatic aldehyde precursors. The ensuing alkene isomerization to the intermediate oxazinone precedes merged cycloaddition and cycloreversion sequence (evolving CO₂) to give 2-methoxypyridine products. The overall multicomponent domino reaction process is promoted with mild organic base (DBU) and occurs at conveniently accessible temperatures (110 °C).

No conflict of interest has been declared by the author(s).

Acknowledgment

The authors acknowledge support from the National Institutes of Health (R15 GM107702 to J.R.S.).

• References and Notes

• 1 Present address: J. B. Williamson, Department of Chemistry, The University of North Carolina, Chapel Hill, NC 27599, USA.
• 2a Tietze LF, Brasche G, Gericke KM. Domino Reactions in Organic Synthesis . Wiley-VCH; Weinheim: 2006
  Search in Google Scholar
• 2b Tietze LF, Modi A. Med. Res. Rev. 2000; 20: 304-304
  CrossrefPubMedSearch in Google Scholar
• 2c Tietze LF. Chem. Rev. 1996; 96: 115-115
  CrossrefPubMedSearch in Google Scholar
• 3 Gaich T, Baran PS. J. Org. Chem. 2010; 75: 4657-4657
  CrossrefPubMedSearch in Google Scholar
• 4a Margrey KA, Chinn AJ, Laws SW, Pike RD, Scheerer JR. Org. Lett. 2012; 43: 2458-2458
  Search in Google Scholar
• 4b Laws SW, Scheerer JR. J. Org. Chem. 2013; 78: 2422-2422
  CrossrefPubMedSearch in Google Scholar
• 5a Margrey KM, Hazzard AD, Scheerer JR. Org. Lett. 2014; 16: 904-904
  CrossrefPubMedSearch in Google Scholar
• 5b Leibowitz MK, Winter ES, Scheerer JR. Tetrahedron Lett. 2015; 56: 6069-6069
  CrossrefPubMedSearch in Google Scholar
• 6a Rogiers J, Wu XJ, Toppet S, Compernolle F, Hoornaert GJ. Tetrahedron 2001; 57: 8971-8971
  CrossrefSearch in Google Scholar
• 6b De Borggraeve W, Rombouts F, Van der Eycken E, Hoornaert GJ. Synlett 2000; 713-713
• 6c Wu XJ, Dubois K, Rogiers J, Toppet S, Compernolle F, Hoornaert GJ. Tetrahedron 2000; 56: 3043-3043
  CrossrefSearch in Google Scholar
• 6d Van der Eycken E, Deroover G, Toppet SM, Hoornaert GJ. Tetrahedron Lett. 1999; 40: 9147-9147
  CrossrefSearch in Google Scholar
• 6e Medaer BP, Hoornaert GJ. Tetrahedron 1999; 55: 3987-3987
  CrossrefSearch in Google Scholar
• 6f Medaer BP, Vanaken KJ, Hoornaert GJ. Tetrahedron 1996; 52: 8813-8813
  CrossrefSearch in Google Scholar
• 6g Medaer B, Vanaken K, Hoornaert G. Tetrahedron Lett. 1994; 35: 9767-9767
  CrossrefSearch in Google Scholar
• 6h Vanaken KJ, Lux GM, Deroover GG, Meerpoel L, Hoornaert GJ. Tetrahedron 1994; 50: 5211-5211
  CrossrefSearch in Google Scholar
• 6i Fannes C, Meerpoel L, Toppet S, Hoornaert G. Synthesis 1992; 705-705
  Thieme Connect
• 6j Fannes CC, Hoornaert GJ. Tetrahedron Lett. 1992; 33: 2049-2049
  CrossrefSearch in Google Scholar
• 6k Tutonda MG, Vandenberghe SM, Vanaken KJ, Hoornaert GJ. J. Org. Chem. 1992; 57: 2935-2935
  CrossrefSearch in Google Scholar
• 6l Vanaken KJ, Meerpoel L, Hoornaert GJ. Tetrahedron Lett. 1992; 33: 2713-2713
  CrossrefSearch in Google Scholar
• 6m Meerpoel L, Hoornaert G. Tetrahedron Lett. 1989; 30: 3183-3183
  CrossrefSearch in Google Scholar
• 7a Roughley SD, Jordan AM. J. Med. Chem. 2011; 54: 3451-3451
  CrossrefPubMedSearch in Google Scholar
• 7b Michael JP. Nat. Prod. Rep. 2005; 22: 627-627
  CrossrefPubMedSearch in Google Scholar
• 7c Henry GD. Tetrahedron 2004; 60: 6043-6043
  CrossrefSearch in Google Scholar
• 7d Joule JA, Mills K. Heterocyclic Chemistry . Blackwell Publishing; Oxford: 2000. 4th ed. 63
  Search in Google Scholar

Selected reviews:
• 8a Varela JA, Saá C. Chem. Rev. 2003; 103: 3787-3787
  CrossrefPubMedSearch in Google Scholar
• 8b Gulevich AV, Dudnik AS, Chernyak N, Gevorgyan V. Chem. Rev. 2013; 113: 3084-3084
  CrossrefPubMedSearch in Google Scholar
• 8c Hill MD. Chem. Eur. J. 2010; 16: 12052-12052
  CrossrefPubMedSearch in Google Scholar
• 8d Bull JA, Mousseau JJ, Pelletier G, Charette AB. Chem. Rev. 2012; 112: 2642-2642
  CrossrefPubMedSearch in Google Scholar
• 8e Nakao Y. Synthesis 2011; 3209-3209
  Thieme Connect
• 9a Allais C, Grassot JM, Rodriguez J, Constantieux T. Chem. Rev. 2014; 114: 10829-10829
  CrossrefPubMedSearch in Google Scholar
• 9b Kral K, Hapke M. Angew. Chem. Int. Ed. 2011; 50: 2434-2434
  CrossrefPubMedSearch in Google Scholar
• 10 See Supporting Information for procedures that accompany Scheme 2 and the spectra and corresponding characterization data for all new compounds (including 5a–c, 13b,c, 14a–c, 15a–c).
• 11 Representative Procedure for the Domino Reaction Leading to Tricyclic Pyridine Product 13a Dihydrooxazinone 5a (50 mg, 0.38 mmol) was dissolved in toluene (3.0 mL, 0.12 M) and DBU (85 μL, 0.57 mmol, 1.5 equiv) was added. The reaction vessel was heated in an oil bath to a gentle reflux (110 °C) and 2-alkynyl benzaldehyde 10 (74 mg, 1.5 equiv) in toluene (1.0 mL) was introduced slowly to the reaction over 2 h (using a syringe pump). After stirring for 18 h at 110 °C, the reaction was cooled to r.t., transferred to a separatory funnel, and partitioned between sat. aq NH₄Cl (10 mL) and EtOAc (10 mL). The organic layer was removed, and the aqueous portion was extracted with EtOAc (3 × 10 mL). The combined organic layers were washed with brine (10 mL), dried (Na₂SO₄), filtered through Celite, and concentrated in vacuo. The resulting residue (82 mg) was purified by flash column chromatography on silica gel (gradient elution: 20% → 80% of CHCl₃ in hexane) to afford compound 13a (57 mg, 77% yield) as a light yellow oil; R_f  = 0.20 (50% CHCl₃–Hex). IR (film): 1586, 1463, 1307, 1029, 772 cm^–1. ¹H NMR (400 MHz, CDCl₃): δ = 7.88 (d, J = 8.6 Hz, 1 H), 7.62 (d, J = 7.8 Hz, 1 H), 7.52 (d, J = 7.4 Hz, 1 H), 7.34 (t, J = 7.4 Hz, 1 H), 7.26 (t, J = 7.4 Hz, 1 H), 4.00 (s, 3 H), 3.87 (s, 2 H). ¹³C NMR (100 MHz, CDCl₃): δ = 163.9, 162.5, 140.4, 139.8, 130.0, 128.4, 126.9, 126.0, 125.0, 119.2, 108.7, 53.7, 38.6. HRMS: m/z calcd for C₁₃H₉NONa [M + Na]⁺: 198.0913; found: 198.0913.

• References and Notes

• 1 Present address: J. B. Williamson, Department of Chemistry, The University of North Carolina, Chapel Hill, NC 27599, USA.
• 2a Tietze LF, Brasche G, Gericke KM. Domino Reactions in Organic Synthesis . Wiley-VCH; Weinheim: 2006
  Search in Google Scholar
• 2b Tietze LF, Modi A. Med. Res. Rev. 2000; 20: 304-304
  CrossrefPubMedSearch in Google Scholar
• 2c Tietze LF. Chem. Rev. 1996; 96: 115-115
  CrossrefPubMedSearch in Google Scholar
• 3 Gaich T, Baran PS. J. Org. Chem. 2010; 75: 4657-4657
  CrossrefPubMedSearch in Google Scholar
• 4a Margrey KA, Chinn AJ, Laws SW, Pike RD, Scheerer JR. Org. Lett. 2012; 43: 2458-2458
  Search in Google Scholar
• 4b Laws SW, Scheerer JR. J. Org. Chem. 2013; 78: 2422-2422
  CrossrefPubMedSearch in Google Scholar
• 5a Margrey KM, Hazzard AD, Scheerer JR. Org. Lett. 2014; 16: 904-904
  CrossrefPubMedSearch in Google Scholar
• 5b Leibowitz MK, Winter ES, Scheerer JR. Tetrahedron Lett. 2015; 56: 6069-6069
  CrossrefPubMedSearch in Google Scholar
• 6a Rogiers J, Wu XJ, Toppet S, Compernolle F, Hoornaert GJ. Tetrahedron 2001; 57: 8971-8971
  CrossrefSearch in Google Scholar
• 6b De Borggraeve W, Rombouts F, Van der Eycken E, Hoornaert GJ. Synlett 2000; 713-713
• 6c Wu XJ, Dubois K, Rogiers J, Toppet S, Compernolle F, Hoornaert GJ. Tetrahedron 2000; 56: 3043-3043
  CrossrefSearch in Google Scholar
• 6d Van der Eycken E, Deroover G, Toppet SM, Hoornaert GJ. Tetrahedron Lett. 1999; 40: 9147-9147
  CrossrefSearch in Google Scholar
• 6e Medaer BP, Hoornaert GJ. Tetrahedron 1999; 55: 3987-3987
  CrossrefSearch in Google Scholar
• 6f Medaer BP, Vanaken KJ, Hoornaert GJ. Tetrahedron 1996; 52: 8813-8813
  CrossrefSearch in Google Scholar
• 6g Medaer B, Vanaken K, Hoornaert G. Tetrahedron Lett. 1994; 35: 9767-9767
  CrossrefSearch in Google Scholar
• 6h Vanaken KJ, Lux GM, Deroover GG, Meerpoel L, Hoornaert GJ. Tetrahedron 1994; 50: 5211-5211
  CrossrefSearch in Google Scholar
• 6i Fannes C, Meerpoel L, Toppet S, Hoornaert G. Synthesis 1992; 705-705
  Thieme Connect
• 6j Fannes CC, Hoornaert GJ. Tetrahedron Lett. 1992; 33: 2049-2049
  CrossrefSearch in Google Scholar
• 6k Tutonda MG, Vandenberghe SM, Vanaken KJ, Hoornaert GJ. J. Org. Chem. 1992; 57: 2935-2935
  CrossrefSearch in Google Scholar
• 6l Vanaken KJ, Meerpoel L, Hoornaert GJ. Tetrahedron Lett. 1992; 33: 2713-2713
  CrossrefSearch in Google Scholar
• 6m Meerpoel L, Hoornaert G. Tetrahedron Lett. 1989; 30: 3183-3183
  CrossrefSearch in Google Scholar
• 7a Roughley SD, Jordan AM. J. Med. Chem. 2011; 54: 3451-3451
  CrossrefPubMedSearch in Google Scholar
• 7b Michael JP. Nat. Prod. Rep. 2005; 22: 627-627
  CrossrefPubMedSearch in Google Scholar
• 7c Henry GD. Tetrahedron 2004; 60: 6043-6043
  CrossrefSearch in Google Scholar
• 7d Joule JA, Mills K. Heterocyclic Chemistry . Blackwell Publishing; Oxford: 2000. 4th ed. 63
  Search in Google Scholar

Selected reviews:
• 8a Varela JA, Saá C. Chem. Rev. 2003; 103: 3787-3787
  CrossrefPubMedSearch in Google Scholar
• 8b Gulevich AV, Dudnik AS, Chernyak N, Gevorgyan V. Chem. Rev. 2013; 113: 3084-3084
  CrossrefPubMedSearch in Google Scholar
• 8c Hill MD. Chem. Eur. J. 2010; 16: 12052-12052
  CrossrefPubMedSearch in Google Scholar
• 8d Bull JA, Mousseau JJ, Pelletier G, Charette AB. Chem. Rev. 2012; 112: 2642-2642
  CrossrefPubMedSearch in Google Scholar
• 8e Nakao Y. Synthesis 2011; 3209-3209
  Thieme Connect
• 9a Allais C, Grassot JM, Rodriguez J, Constantieux T. Chem. Rev. 2014; 114: 10829-10829
  CrossrefPubMedSearch in Google Scholar
• 9b Kral K, Hapke M. Angew. Chem. Int. Ed. 2011; 50: 2434-2434
  CrossrefPubMedSearch in Google Scholar
• 10 See Supporting Information for procedures that accompany Scheme 2 and the spectra and corresponding characterization data for all new compounds (including 5a–c, 13b,c, 14a–c, 15a–c).
• 11 Representative Procedure for the Domino Reaction Leading to Tricyclic Pyridine Product 13a Dihydrooxazinone 5a (50 mg, 0.38 mmol) was dissolved in toluene (3.0 mL, 0.12 M) and DBU (85 μL, 0.57 mmol, 1.5 equiv) was added. The reaction vessel was heated in an oil bath to a gentle reflux (110 °C) and 2-alkynyl benzaldehyde 10 (74 mg, 1.5 equiv) in toluene (1.0 mL) was introduced slowly to the reaction over 2 h (using a syringe pump). After stirring for 18 h at 110 °C, the reaction was cooled to r.t., transferred to a separatory funnel, and partitioned between sat. aq NH₄Cl (10 mL) and EtOAc (10 mL). The organic layer was removed, and the aqueous portion was extracted with EtOAc (3 × 10 mL). The combined organic layers were washed with brine (10 mL), dried (Na₂SO₄), filtered through Celite, and concentrated in vacuo. The resulting residue (82 mg) was purified by flash column chromatography on silica gel (gradient elution: 20% → 80% of CHCl₃ in hexane) to afford compound 13a (57 mg, 77% yield) as a light yellow oil; R_f  = 0.20 (50% CHCl₃–Hex). IR (film): 1586, 1463, 1307, 1029, 772 cm^–1. ¹H NMR (400 MHz, CDCl₃): δ = 7.88 (d, J = 8.6 Hz, 1 H), 7.62 (d, J = 7.8 Hz, 1 H), 7.52 (d, J = 7.4 Hz, 1 H), 7.34 (t, J = 7.4 Hz, 1 H), 7.26 (t, J = 7.4 Hz, 1 H), 4.00 (s, 3 H), 3.87 (s, 2 H). ¹³C NMR (100 MHz, CDCl₃): δ = 163.9, 162.5, 140.4, 139.8, 130.0, 128.4, 126.9, 126.0, 125.0, 119.2, 108.7, 53.7, 38.6. HRMS: m/z calcd for C₁₃H₉NONa [M + Na]⁺: 198.0913; found: 198.0913.

• Supplementary Material