Acyclic and Heterocyclic Azadiene Diels–Alder Reactions Promoted by Perfluoroalcohol Solvent Hydrogen Bonding: Comprehensive Examination of Scope

Abstract

Herein, the first use of perfluoroalcohol H-bonding in accelerating acyclic azadiene inverse electron demand cycloaddition reactions is described and its use in the promotion of heterocyclic azadiene cycloaddition reactions is generalized through examination of a complete range of azadienes. The scope of dienophiles was comprehensively explored, relative reactivity trends and solvent compatibilities were established with respect to both the dienophile as well as azadiene, H-bonding solvent effects that lead to rate enhancements, yield improvements, and impact regioselectivity and mode of cycloaddition are defined, new viable diene/dienophile reaction partners in the cycloaddition reactions are disclosed, and key comparison rate constants are reported. The perfluoroalcohol effectiveness at accelerating an inverse electron demand Diels–Alder cycloaddition is directly correlated with its H-bond potential (pK_a). Not only are the reactions of electron-rich dienophiles accelerated, but those of strained and even unactivated alkenes and alkynes are improved, including representative bioorthogonal click reactions.

Graphical Abstract

Introduction

The Diels–Alder cycloadditions of heterocyclic and acyclic azadienes are powerful reactions widely employed in organic synthesis, medicinal chemistry campaigns, and as tools in chemical biology.^1–3 By virtue of a nitrogen atom incorporation into the core diene systems, azadienes are intrinsically electron deficient and participate most effectively in inverse electron demand Diels–Alder reactions. In general, the greater the number of nitrogen atoms embedded in the diene core, the greater the cycloaddition reactivity (e.g., tetrazines > triazines > diazines > pyridines). In part, this may be attributed to their increasing electron-deficient character (lower LUMO) and, in the case of heterocyclic azadienes, their increasingly reduced aromaticity and smaller transition state distortion energies.⁴ Added electron-withdrawing substituents can further enhance this intrinsic reactivity as well as reinforce or alter their reaction regioselectivity. Many pairings of electron-rich or strained dienophiles with the electron-deficient azadienes participate in room temperature cycloaddition reactions, some even at extraordinarily dilute concentrations, and under conditions that are typically milder than traditional normal Diels–Alder reactions. However, unlike the normal Diels–Alder reaction and largely because of the low basicity of nitrogen atoms found in heterocyclic or acyclic azadienes, efforts to catalyze such reactions with Lewis acids have largely failed due to preferential coordination with and nonproductive consumption of the reactive electron-rich dienophiles. Although there are notable exceptions, including Wegner’s bidentate boron-based Lewis acid-catalyzed cycloaddition reactions of 1,2-diazines,^5–9 Rawal’s Ag-catalyzed reaction of phthalazines with siloxyalkynes,^10,11 and our recent ZnCl₂-promoted N1/N4 (vs C3/C6) [4+2] formal cycloaddition of selected 1,2,4,5-tetrazines,¹² each are limited in scope and the latter two examples feature altered mechanisms. Thus, no general approach for catalysis of the inverse electron demand Diels–Alder reactions of azadienes had emerged until we disclosed a potentially powerful solvent H-bonding acceleration of such reactions using perfluoroalcohols,^13,14 weakly acidic but non-nucleophilic protic solvents. However, beyond the initial studies with 1,2,3-triazines, the scope of its impact remains unknown, as was its potential utility with the less stable acyclic versus heterocyclic azadienes. As a result, we examined and herein generalize the solvent H-bonding catalysis of heterocyclic azadiene cycloaddition reactions with the complete range of azadienes and describe its first use in promoting acyclic azadiene cycloaddition reactions (Figure 1). In these studies, a full range of electron-rich dienophiles was explored, relative reactivity trends and solvent compatibilities were established with respect to both the dienophile as well as azadiene, H-bonding solvent effects leading to enhancements in the rate or impact on regioselectivity are defined, and key rate constants are quantitated. Substantial expansions in the scope of viable diene/dienophiles reaction partners in the cycloaddition reactions are disclosed. Even a selected instance of alteration of the mode of cycloaddition across the heterocyclic azadiene was observed.

Figure 1.

Perfluoroalcohol H-bonding promoted azadiene cycloaddition reactions.

Results and Discussion

N-Sulfonyl-1-aza-1,3-butadienes.

The participation of simple α,β-unsaturated imines in Diels–Alder reactions is rarely observed and suffers from low conversions, competitive 1,2-imine addition, imine/enamine tautomerization, or product enamine stability.^1,15 However, complementary introduction of N1 as well as C3 electron-withdrawing groups accentuate their electron-deficient character, promoting their participation in inverse electron demand Diels–Alder reactions, and the use of a bulky N1 substituent both decelerates competitive 1,2-addition and stabilizes the cycloaddition product enamine.¹⁶ N-Sulfonyl-1-azabutadienes have proven to be the synthetically most accessible, stable, and effective of the acyclic 1-azabutadienes examined^16,17 and an asymmetric variant of the reaction has been developed, enlisting a unique class of chiral enol ethers (>20:1 endo and >20:1 facial selectivity).¹⁸ Remarkable in the reactions is the often near exclusive endo diastereoselectivity (>20:1) derived from stabilizing secondary orbital interactions that is further enhanced by a transition state anomeric effect, involving the azadiene nitrogen lone pair and trans periplanar σ C–O bond of an enol ether dienophile in the cycloaddition boat transition state. However, the modest reactivity of the acyclic N-sulfonyl-1-azadienes that do not contain further activating substituents still leave much to be desired if they are to reach their full potential in synthesis beyond our own use, which has included total syntheses of streptonigrone, fredricamycin A, nothapodytine, (−)-mappicine, (+)-camptothecin, and (−)-piericidin A1.³

The potential solvent H-bonding promotion of the cycloaddition of acyclic N-sulfonyl-1-aza-butadienes was first surveyed with the reaction of 8a with ethyl vinyl ether (1) and the results are summarized in Figure 2. Of the solvents cable of H-bonding (entries 1–8), trifluoroethanol (TFE) (entry 2, 81%) proved to be the optimal solvent among those examined due to its moderate pK_a (12.4), striking a balance between activation the azadiene by H-bonding while avoiding protonation and consumption of the dienophile 1. Tetrafluoropropanol (TFP, entry 7, pK_a = 12.7) behaved similarly, whereas more acidic perfluoroalcohols proved less effective (entries 1 and 8). Water (pK_a = 14) as well as conventional protic solvents did not exhibit a significant enhancement in reactivity presumably due to weaker H-bonding with the azadiene (pK_a MeOH = 15.5, pK_a EtOH = 16.0) or due to protonation and non-productive consumption of the enol ether due to the solvent acidity (pK_a AcOH = 4.8). In this regard, an interesting feature emerged in the comparison of perfluoro t-BuOH (entry 8, pK_a = 5.4, 56%) with AcOH where the lower nucleophilicity and steric bulk of the former may help reduce the nonproductive enol ether consumption. Although non H-bonding solvents also support the reaction (entries 9–16), the cycloadditions proceed at a slower rate and provide a lower yield of 10 where, interestingly, the less polar solvents seem to be preferred. Previously reported acceleration of Diels–Alder reactions in aprotic fluorous solvents (n-C₆F₁₄ and FC-75 (perfluoro THF), entry 15 and 16) was not observed in the reaction of 8a, indicating that the perfluoroalcohol acceleration is not due to a fluorophobic effect.^19a Finally, conduct of the reaction in CH₂Cl₂ with TFE as an additive under the same standardized reaction conditions revealed that the yield of product 10 progressively increased with increasing amounts of TFE, where 7 equiv of added TFE essentially matched the conversions observed in TFE as solvent (Figure 2). We view these results as perhaps unanticipated, but welcomed, and indicative of a much broader role for such solvent H-bonding activation of even weakly Lewis basic substrates in an expanded range of reactions.

Figure 2.

Effect of solvent on the reaction of 8a with ethyl vinyl ether (1) under standardized reaction conditions (1.0 M 8a, 10 equiv 1, 25 °C, 12 h).

The H-bonding interaction with the diene was confirmed by ¹H NMR (Figure 3). Serial addition of 8a to a solution of TFE (1 equiv) in CDCl₃ led to a diagnostic and pronounced downfield chemical shift of the TFE alcohol proton (Δ0.53 ppm, 1.5 equiv 8a) and a smaller upfield shift in the TFE methylene proton (Δ0.03 ppm, 1.5 equiv 8a). By conducting this titration in a reverse manner with addition of TFE to 8a (1 equiv) in CDCl₃, an upfield shift in the azadiene C2 proton (Δ0.03 ppm, 1.5 equiv TFE) was observed indicative of a decrease in electron density on the N1 nitrogen. It is unclear whether this interaction with and activation of the N-sulfonyl-1-azadiene arises through direct H-bonding to N1 or indirectly through H-bonding to the sulfonyl group.

Figure 3.

Confirmation of H-bonding of TFE with 8a by ¹H NMR (CDCl₃).

The reaction profiles of the cycloaddition of 8a with 1 were established in TFE versus the aprotic solvent CH₂Cl₂ under otherwise identical conditions to quantitatively define the reaction rate acceleration. The two reaction profiles are depicted in Figure 4 and clearly show that reaction in TFE is substantially faster and more efficient than that in CH₂Cl₂. Calculated second-order rate constants (k = 1.1 × 10⁻² M⁻¹·min⁻¹ (TFE) and k = (9.9 ± 0.2) × 10⁻⁴ M⁻¹·min⁻¹ (CH₂Cl₂)) established an 11-fold increase in rate for the reaction conducted in TFE.

Figure 4.

Rate of the cycloaddition of 8a with 1 in TFE and CH₂Cl₂ (1.0 M 8a, 10 equiv 1, 25 °C).

The cycloaddition reaction of a series of N-sulfonyl-1-azadienes (8a–d), including the less reactive 4-substituted dienes 8b–d, with ethyl vinyl ether were examined in both TFE and CH₂Cl₂ under otherwise identical conditions. The reactions in TFE, which were not extensively optimized, were found to now proceed effectively even under mild thermal conditions (50–60 °C) but were largely ineffective in CH₂Cl₂ (Figure 5). Notably, many of these azadienes previously required the use of inconvenient high pressures to afford synthetically useful yields.¹⁷ Similarly, the reactions of 8a with less reactive dienophiles, including the previously unreactive styrene, were found to now be viable reaction pairs capable of cycloaddition under mild thermal reaction conditions. The latter dienophiles are also stable to hexafluoroisopropanol (HFIP), which now can serve as an even more powerful H-bond donor for promoting the cycloaddition reactions.^19b This representative range of examples illustrates that the solvent H-bonding of the perfluoroalcohols TFE and HFIP not only accelerates the rate of the reactions, but also substantially expands the scope of effective N-sulfonyl-1-azadienes as well as the range of viable dienophile partners for participation in such cycloaddition reactions.

Figure 5.

Expanded diene/dienophile scope of acyclic N-sulfonyl-1-azadienes cycloadditions.

Finally and without optimization, the reaction of 8a with a representative chiral enol ether displayed a slightly better conversion in TFE versus toluene and notably proceeded without significant erosion of the sensitive anomeric stereochemistry (eq 1).

(1)

1,2,4,5-Tetrazines.

The most widely used and most reactive class of heterocyclic azadienes is the 3,6-disubstituted 1,2,4,5-tetrazines. They have been widely used in organic synthesis,^1,2,20 natural product total synthesis,^3,20 medicinal chemistry,²¹ and chemical biology²⁰ for decades now. In fact, it was their discovery²² that led to the now classical definition of an inverse electron demand versus normal Diels–Alder reaction.²³ Today there are countless publications on their synthesis, cycloaddition reactions and reactivity patterns, and applications. Many of their reactions already proceed a room temperature or lower even under dilute reaction conditions required of bioconjugation studies for which they are presently the fastest and most efficient of all such bioorthogonal click reactions.^26–30 Electron-withdrawing and conjugating C3/C6 substituents accelerate their cycloaddition reactions and pair best with electron-rich or strained dienophiles. These were not examined in our survey. Rather, we examined representative less reactive 1,2,4,5-tetrazine/dienophile pairs that are less effective and may benefit from the H-bonding assistance, focusing primarily on unactivated dienophiles (simple alkenes and alkynes) or 1,2,4,5-tetrazines bearing modest electron-donating versus electron-withdrawing substituents (e.g., SMe). Given the low basicity of 1,2,4,5-tetrazines (pK_a < 0 for the parent 1,2,4,5-tetrazine), the key question was whether they could benefit from a solvent H-bonding effect.

We initiated the studies with the examination of the reaction of tetrazines with norbornadiene, a reaction that proceeds by initial Diels–Alder reaction of the alkene with the tetrazine, followed by sequential retro Diels–Alder loss of N₂ and then cyclopentadiene with 1,2-diazine generation. A kinetic study was conducted of the reaction between 3,6-diphenyl-1,2,4,5-tetrazine (9a, 5 mM) and norbornadiene (2) in CHCl₃ and HFIP and their reaction profiles are shown in Figure 6. The study clearly illustrates the modest reactivity of 9a with 2 in a conventional solvent, and the significantly enhanced reactivity observed in HFIP. Calculated second-order rate constants for the reactions are k = 1.1 × 10⁻² M⁻¹·s⁻¹ (HFIP) and k = 3.9 × 10⁻⁴ M⁻¹·s⁻¹ (CHCl₃), representing a large 28-fold acceleration of the reaction in HFIP.

Figure 6.

Reaction profile and kinetic parameters for the cycloaddition of 9a with 2 in HFIP and CHCl₃ (5 mM 9a, 10 equiv 2, 25 °C).

Given both the pronounced effect and the insensitivity of norbornadiene to acidic solvents, the reaction of 9a with 2 in a range of perfluoroalcohols was examined alongside the reaction in MeOH at room temperature (Figure 7). The reaction displayed a smooth trend where the rate of reaction increased as the acidity of the solvent increased. The reaction run in perfluoro t-BuOH (pK_a = 5.4, t_1/2 ca. 15 min) proved the fastest of those examined, exhibiting the greatest effect on the rate, and the commercially available solvent HFIP (pK_a = 9.3) also exhibited a large improvement (t_1/2 = 22 min). In comparison, the reaction in MeOH conducted under the same conditions provided little product after 30 min and was only 14% complete after 11 h. Both TFE (pK_a 12.4, not shown) and TFP (pK_a = 12.7) also proved effective, behaved similarly, and exhibited a t_1/2 of ca. 3 h under the same conditions.

Figure 7.

Solvent effect on the reaction of 9a with norbornadiene (2).

Summarized in Figure 8 are the results of a brief survey of the reactivity of three tetrazines, including not only 9a but also the more electron-rich 3,6-bis(thiomethyl)-1,2,4,5-tetrazine (9b) and more electron-deficient dimethyl 1,2,4,5-tetrazine dicarboxylate (9c), conducted with a set of dienophiles that typically display a modest reactivity. The previously unknown cycloadditions of tetrazine 9a and 9b with norbornadiene 2 provided much better yields of the desired pyridazines in HFIP compared with that in CHCl₃ under otherwise identical conditions. The cycloaddition of 9a with styrene (3a) was also faster in HFIP than CHCl₃ and was assessed following air oxidation of the dihydropyridazine product to give the corresponding pyridazine 20. With each the three tetrazines, the reaction with the unactivated dienophile phenylacetylene (5a) proceeded faster and better in HFIP than in CHCl₃ or toluene, with 9a and even 9c providing quantitative conversions. Most remarkable of these was the observation that the reaction of 9c with the electron-deficient dienophile 5a proceeded at room temperature (25 °C) and was complete in 35 min in HFIP (>99%) but provided only 13% of product when conducted in CHCl₃ for the same reaction time. When conducted at 50 °C, the reaction in HFIP was complete in less than 1 min and that in TFE required 8 min, whereas the reaction in CHCl₃ required 40 min. The otherwise less reactive candidate dienophiles 4e and indole (6) in their reactions with 9a also now proved much more efficient in HFIP. The cycloaddition between 9a and 4e highlights the use of 4e as a propiolate equivalent in Diels–Alder cycloadditions, benefitting from the use of HFIP solvent in which it proved stable. Finally, without examination in detail, the reaction of 9c with the bisindole substituted alkyne 5c, explored in initial efforts directed at the total synthesis of lycogalic acid and lycogarubin C,³¹ proved more effective in HFIP than toluene and did not require the sequential addition of tetrazine over a prolonged reaction time as previously reported. Finally, a larger scale reaction (1 mmol scale) of tetrazine 9b with 2 in HFIP was conducted under conditions identical to the smaller scale reaction in Figure 8 and afforded product 19 in 85% (eq 2, vs 88% on 0.05 mmol scale), which highlights the utility in preparative reactions.

Figure 8.

Expanded diene/dienophile scope of the cycloadditions of 1,2,4,5-tetrazines.

(2)

The more electron-rich and unsymmetrical 1,2,4,5-tetrazine 9d also underwent an effective cycloaddition reaction with norbornadiene in HFIP under mild conditions (40 °C), whereas the reaction in 1,4-dioxane under otherwise identical conditions was not observed to an appreciable extent (Figure 9). Interestingly, the reaction of 9d with phenylacetylene 5a was not only much faster (62% vs < 10% yield at 100 °C, 5.5 h), but the reaction also displayed a slightly enhanced regioselectivity (28a:28b = 2.9:1) in HFIP compared with the reaction in 1,4-dioxane³² (28a:28b = 1.8:1). A good yield of 28a/28b could be obtained in 1,4-dioxane if the reaction time was extended to 70 h at 110 °C, providing a cycloaddition yield and regioselectivity identical to our prior report.³² Finally, the slow intramolecular cycloaddition of 9e was found to proceed both faster and in higher yield to provide 29 than the same reaction conducted in 1,4-dioxane. This latter reaction typically requires an activating amine protecting group for observation of cycloaddition under mild reaction conditions,^32–34 but now may be effectively conducted without the deliberate amine deactivation.

Figure 9.

Reactions of unactivated dienophiles with unsymmetrical and more electron-rich 1,2,4,5-tetrazines.

Because of the significant improvements in the reactions of the modestly strained norbornadiene with the 1,2,4,5-tetrazines, kinetic studies of the cycloaddition of 3-phenyl-1,2,4,5-tetrazine (9n) and cyclooctyne 5d were conducted to probe the potential HFIP acceleration of a representative example of a tetrazine ligation reaction. To date, the tetrazine ligation reactions are the most general of the bioorthogonal reactions because of their superb reactions rates and efficiencies at the needed low concentrations, and because of their effective use in not only cell free systems, but also in cell-based systems and in vivo.²⁸ Kinetic studies were conducted in THF, DMSO and HFIP (5 mM 9n, 25 °C), where the amount of cyclooctyne in HFIP was half that used with THF and DMSO (1.2 equiv vs 2.4 equiv). Even with this reduction in the amount of the cyclooctyne, the reaction profiles depicted in Figure 10 show the substantial increase in rate of reaction in HFIP. The calculated second-order rate constants of k = 29.2 M⁻¹·s⁻¹ (HFIP) versus k = 1.1 M⁻¹·s⁻¹ (THF) and k = 1.2 M⁻¹·s⁻¹ (DMSO) represent a 27-fold (vs THF) and 24-fold (vs DMSO) acceleration of the reaction in HFIP (Figure 10A). Cycloaddition with a stoichiometric amount of 5b at the relatively dilute concentration of 5 mM was also conducted and afforded 30 in essentially quantitative yield in <120 s (97%, Figure 10B). This substantial rate acceleration of the cyclooctyne cycloaddition in HFIP indicates that the tetrazine bioorthogonal reactions of a broader range of common cycloaddition partners, such as cyclopropenes, trans-cyclooctene and norbornenes, can be expected that will improve the efficiency of the reactions, facilitate relevant bioconjugation studies, and expand the scope of effective diene/dienophile reaction partners.

Figure 10.

A. Reaction profile and kinetic parameters for the cycloaddition of 9n with cyclooctyne 5b in HFIP, DMSO, and THF at 25 °C. B. Cycloaddition using a stoichiometric amount of 5b.

N1/N4 Cycloaddition of 1,2,4,5-tetrazines.

A unique impact of the solvent H-bonding activation of a 1,2,4,5-tetrazine was observed in the reaction of 3,6-bis(thiomethyl)-1,2,4,5-tetrazine (9b) with aryl-conjugated enamines, where both the rate of reaction was increased, and the mode of cycloaddition was altered from the conventional C3/C6 cycloaddition to N1/N4 cycloaddition. As a result, this reaction was investigated in detail and has been reported elsewhere (Figure 11).¹⁴ The study disclosed the key factors that result in this change in the reaction mode, and reported an extensive survey of conventional and perfluoroalcohol solvents, the reaction scope for 3,6-bis(thiomethyl)-1,2,4,5-tetrazine (9b), a definition of the enamine substate scope and simplification of its implementation through in situ enamine generation from aldehydes and ketones, and provided a survey of participating 1,2,4,5-tetrazines. Studies of the reaction mechanism were also reported and to date has only been observed in the reaction of aryl-conjugated enamines with the more electron-rich tetrazines. Whereas conventional protic solvents (e.g., MeOH) provide products of the normal C3/C6 cycloaddition, the solvent H-bonding of HFIP promoted the unprecedented N1/N4 cyclization, representing the first reported example of an altered cyclization mode (N1/N4 vs C3/C6) of a 1,2,4,5-tetrazine.

Figure 11.

N1/N4 Cycloaddition of 1,2,4,5-tetrazines.¹⁴

1,2,4-Triazines.

Like 1,2,4,5-tetrazines, 1,2,4-triazines participate effectively as electron-deficient heterocyclic azadienes in inverse electron demand Diels–Alder reactions with electron-rich dienophiles^35–38 and have been widely used in synthesis and advanced as candidate bioorthogonal click reactions by Prescher and others.³⁹ Typical dienophiles that react with 1,2,4-triazines (enol ethers, ketene acetals, enamines) are those that are also competitively protonated by HFIP and were not examined in detail. However, representative strained (e.g., norbornadiene) and unactivated (e.g., alkenes, alkynes) dienophiles that typically display modest to poor reactivity towards 1,2,4-triazines were found to exhibit substantially improved reaction rates and yields when conducted in HFIP (Figure 12). This represents a productive expansion of the viable cycloaddition reactions now available to the 1,2,4-triazines, especially those that might be intrinsically less reactive. For example, reactions with norbornadiene (31–33) which convert 1,2,4-triazines to the corresponding pyridines can now be conducted under relatively mild conditions or even room temperature while reactions in conventional solvents require high reaction temperatures (typically > 130 °C) for complete conversion. Interestingly, like the observations made with the unsymmetrical 1,2,4,5-tetrazine 9d, a switch or enhancement in the regioselectivity was also observed in HFIP. The modest regioselectivity displayed in the reaction of 9h with phenylacetylene (5a) in CHCl₃, which preferentially provides 34b (unambiguously determined by X-ray crystallography⁴⁰), is altered in HFIP to now provide a slight preference for 34a. More significantly, the reaction of 9f with alkene 3d that displays no regioselectivity, forming 35a and 35b (1:1) in toluene, now preferentially generates 35a (35a:35b = 4:1) along with the significantly improved overall yield.

Figure 12.

Expanded diene/dienophile scope of the cycloadditions of 1,2,4-triazines and impact on regioselectivity.

Given the electron-deficient nature of 1,2,4-triazine 9f and its lack of intrinsic basicity, it may appear to be incapable of H-bonding that leads to activation. In addition to an impact of H-bonding on the relative transition state energy, we suggest that this potentially may also occur by solvation of destabilizing lone pair/lone pair clashes in the ground state. Apparent H-bonding to N1 relieves the N1/N2 in plane lone pair destabilization as well as that of an in plane lone pair/lone pair clash of N1 with the C6 methyl ester carbonyl oxygen perhaps reinforcing its activating conjugation with the 1,2,4-triazine (Figure 13). Alternatively, or perhaps additionally, H-bonding to N4 relieves two in plane lone pair/lone pair clashes of N4 with the C3 and C5 methyl ester carbonyl oxygens perhaps reinforcing their conjugation with and activation of the 1,2,4-triazine. In addition to the productive impact either of these would have on the rate of reaction, the former would seem to favor formation of 34a and may account for the subtle impact on reaction regioselectivity as well. Perhaps analogous features are also responsible for the surprisingly effective reactions of dimethyl 1,2,4,5-tetrazine-3,6-dicarboxylate (9c) summarized in Figure 8.

Figure 13.

Potential H-bonding to 1,2,4-triazine 9h.

1,2,3-Triazines.

1,2,3-Triazines are the class of heterocyclic azadienes with which our discovery of solvent H-bonding catalysis was first made.¹³ They represent the least explored of the isomeric triazines due in part to initial challenges associated with their preparation combined with their poor cycloaddition scope in early studies.^13,41a We conducted the first systematic study of the impact of substituents on the reactivity of 1,2,3-triazines in inverse electron demand Diels–Alder reactions that led to the expansion of the scope to now include a wide range of electron-rich dienophiles, including ynamines, enamines, ketene acetals, enol ethers, and amidines.^41b In our earlier studies,¹³ the cycloaddition of the challenging conjugated enamine 4f, which is stable to both HFIP and TFE, was found to be promoted effectively in HFIP with a full range of 5-substituted 1,2,3-triazines, displaying a predicable reactivity trend (R = CO₂Me > Ph > H > SMe). Moreover, the less reactive 1,2,3-triazines were found to display the more pronounced improvements in yields (81−43%) and all conversions were found to be superior to those observed in CHCl₃ (58−0%) (Figure 14). Even 4,6-dicarboethoxy-1,2,3-triazine that bears two electron-withdrawing substituents bound at noncomplementary sites was found to benefit from TFE H-bonding catalysis with the simple non-conjugated pyrrolidine enamines derived from cyclopentanone (60% vs 31% in CHCl₃) and cyclohexanone (47% vs 12% in CHCl₃) under these same conditions (60 °C, 24 h), notably without impacting the intrinsic regioselectivity or C4/N1 mode of cycloaddition.¹³ Additional features disclosed in this earlier work include NMR confirmation of the azadiene H-bonding, a comprehensive solvent screen for the reaction with enamines, use of TFE versus HFIP for non-conjugated enamines, and stunning results for the key step in a concise total synthesis of methoxatin.¹³ Subsequent computational studies of Houk indicated that the enamine cycloaddition reactions span the range of concerted, asynchronous to stepwise addition-cyclization reactions, depending on the azadiene substitution, and that those computed with HFIP H-bond activation constitute stepwise addition-cyclization reactions.⁴² Herein, we provide complementary examples that cover 1,2,3-triazine/enamine pairs that previous studies did not address because of the then ongoing focus on specific applications that included total syntheses of pyrimidoblamic acid, P-3A⁴³ and dihydrolysergic acid.⁴⁴

Figure 14.

Selected HFIP-promoted cycloadditions of conjugated enamines with 1,2,3-triazines.

The expanded scope of the cycloaddition of 1,2,3-triazines with pyrrolidine enamines in HFIP (for stabilized enamines, namely 4d and 4e) or TFE (for non-conjugated enamines, namely 4a–4c) is presented in Figure 15. Reaction rates as well as yields are significantly improved when the reactions were conducted in either HFIP or TFE independent of 1,2,3-triazine substitution pattern. This is not only because the cycloaddition itself is accelerated, but also because the subsequent aromatization improves with perfluoroalcohol protonation of the cycloadduct pyrrolidine nitrogen, facilitating elimination. Pyrrolidine enamines derived from cyclic (4a and 4b), acyclic (4c), and conjugated (4d) ketones and aldehydes participated in more effective cycloadditions with a full range of substituted 1,2,3-triazines in HFIP/TFE than conventional solvents. Significantly, the typically unreactive vinylogous enamide 4e provided 41 and 42 in 23% and 68% yield upon reaction with 5-phenyl-1,2,3-triazine (9i) and methyl 1,2,3-triaizne-5-carboxylate (9j), respectively.

Figure 15.

Expanded scope of the cycloaddition reactions of 1,2,3-triazines and enamines.

Finally, phenylacetylene (5a) was found to react with methyl 1,2,3-triazine-5-carboxylate (9j) to provide pyridine 43 (35%) in an improved yield in TFE compared with CHCl₃ (11%, eq 3). To the best of our knowledge, this is the first reported example of cycloaddition of a 1,2,3-triazine with an unactivated alkyne. However, in general, the reaction of 1,2,3-triazines with unactivated alkynes (phenylacetylene), alkenes (e.g., styrene), and moderately strained alkenes (e.g., norbornadiene) remained largely ineffective even when conducted in HFIP or TFE.

(3)

1,3,5-Triazines.

1,3,5-Triazines participate in well-defined [4 + 2] cycloaddition reactions with electron-rich dienophiles, including ynamines, enamines, N,O-ketene acetals, and amidines.^35,45–49 As we previously reported, the conjugated pyrrolidine enamine 4f, which is stable in HFIP, underwent a productive reaction with 1,3,5-triazine, a reaction that was ineffective in the conventional solvent CHCl₃ (eq 4).¹³

(4)

Beyond the example in eq 3 and excluding the dienophiles reactive toward perfluoroalcohols that were not examined, little or no improvement was typically found in the cycloadditions of 1,3,5-triazines independent of dienophile category or 1,3,5-triazine substitution pattern. For example, unactivated alkenes (styrene), alkynes (phenylacetylene), and the moderately strained alkene norbornadiene failed to react with triethyl 1,3,5-triazine-2,4,6-tricarboxylate in both conventional solvents as well as perfluoroalcohols. However, it was discovered that HFIP can promote the direct cyclization of ketones or aldehydes with triethyl 1,3,5-triazine-2,4,6-tricarboxylate (9m) directly providing pyrimidines 45–48 in moderate yield (Figure 16). These were conducted without optimization efforts and are results that likely could be improved. It is possible that HFIP serves to not only activate the 1,3,5-triazine perhaps in a H-bonding solvation manner, but also acts as an acidic catalyst, promoting ketone/aldehyde enol formation for participation in the formal Diels–Alder reaction.⁵⁰

Figure 16.

Ketone/aldehyde enol cycloaddition with 1,3,5-triazine 9m.

Conclusions

A comprehensive examination of the scope of solvent H-bonding-promoted inverse electron demand Diels–Alder reactions was conducted with systematic evaluation of both acyclic and heterocyclic azadienes, resulting in an expanded diene/dienophile reaction scope, establishment of quantitative as well as qualitative azadiene rate and reactivity enhancements, and yield improvements with a broad array of electron-rich, strained, and even unactivated dienophiles. Predictable alterations or enhancements in the regioselectivity of the cycloaddition reactions are observed and even one instance of the change in the mode of cycloaddition (N1/N4 vs C3/C6 tetrazine reaction) was discovered. Most notably, we report that the observations extend to promoting the reactions of previously unreactive diene/dienophile pairs. Even the use of the perfluoroalcohol as an additive versus use as solvent is demonstrated. Disclosed is the activation of a full range of azadienes with appropriately chosen perfluoroalcohols that now include acyclic N-sulfonyl-1-aza-1,3-butadienes (e.g., 8a–8c) as well as the full suite of heterocyclic azadienes [1,2,4,5-tetrazines (e.g., 9a–9e), 1,2,4-triazines (e.g., 9f–9h), 1,2,3-triazines (e.g., 9i–9l), and 1,3,5-triazines (e.g., 9m)] in cycloaddition reactions with representative and often typically unreactive dienophiles. These now include not only enamines and enol ethers, but also norbornadiene, alkenes, alkynes, indole, and enols of ketones and aldehydes. Some notable dienophiles that were previously underexplored due to their low reactivity, such as N-vinylpyrrolidin-2-one (3c) and methyl 3-(pyrrolidin-1-yl)acrylate (4e), now participate in cycloadditions efficiently with the aid of perfluoroalcohol solvent H-bonding. The effectiveness of individual perfluoroalcohols to accelerate an inverse electron demand Diels–Alder cycloadditions is directly correlated with their H-bond potential (pK_a) but can be dependent on the stability of an electron-rich dienophile; conditions that can be tuned to avoid its nonproductive consumption (pK_a of perfluoroalcohol). In contrast, the reactions of a strained alkene/alkyne or unactivated dienophile are effectively accelerated, also directly correlate with the perfluoroalcohol H-bond potential, but do not display competitive nonproductive consumption by solvent protonation. Most notable of this latter feature is the enhanced reactivity displayed by the azadienes toward unactivated or strained alkenes/alkynes prominently featured in click chemistry and its bioorthogonal conjugation reactions (25–30 fold enhancement in rate). Finally, and notably unlike the activation of carbonyl containing dienophiles in normal Diels–Alder reactions,^51–53 the azadiene substrates detailed herein are not activated for reaction in water, common alcoholic solvents (e.g., MeOH), or by alcohol-bearing reagents or auxiliaries. The discovery, systematic evaluation, and generalization of the perfluoroalcohol H-bonding catalysis of heterocyclic azadiene cycloadditions now allows reactions to be predictably conducted under milder reaction conditions with a broader scope than previously realized, and substantial expansions in the scope of viable diene/dienophiles reaction partners in the cycloaddition reactions are disclosed.

EXPERIMENTAL SECTION

General methods.

All reagents and solvents were used as supplied without further purification unless otherwise noted. Anhydrous solvents were dried over 4 Å molecular sieves. Preparative thin-layer chromatography (PTLC) and column chromatography were conducted using Millipore SiO₂ 60 F254 PTLC (0.5 mm) and Zeochem ZEOprep 60 ECO SiO₂ (40−63 μm), respectively. Analytical TLC was conducted using Millipore SiO₂ 60 F254 TLC (0.250 mm) plates. ¹H and ¹³C NMR spectra were obtained on a Bruker AVANCE III HD 600 MHz spectrometer equipped with either a 5 mm QCI or a 5 mm CPDCH probe, a Bruker AVANCE III 500 MHz spectrometer equipped with a 5 mm BBFO probe, or a Bruker AVANCE III 400 MHz spectrometer equipped with a 5 mm BBFO probe. Chemical shifts are referenced to residual solvent (CDCl₃, δ 7.26 (¹H) and 77.0 (¹³C); CD₂Cl₂, δ 5.32 (¹H) and 54.2 (¹³C). IR spectra were obtained on a Thermo Nicolet 380 FT-IR with a SmartOrbit Diamond ATR accessory. Mass spectrometry analysis was performed by direct sample injection on an Agilent G1969A ESI-TOF mass spectrometer. The single-crystal X-ray diffraction study was carried out on a Bruker Kappa APEX-II CCD diffractometer equipped with Mo Kα radiation (λ = 0.71073 Å). Melting points (mp) are uncorrected. The ¹H NMR of compounds prepared herein but that represent previously characterized products prepared by related methods (same cycloaddition reactions in conventional solvents) are reported below whereas all new compounds are reported with complete characterization.

Cycloaddition of N-Sulfonyl-1-aza-1,3-butadienes.

A solution of N-sulfonyl-1-aza-1,3-butadienes^16,17 (0.1 mmol) in CH₂Cl₂ or HFIP (0.1 mL) was treated with ethyl vinyl ether 1, styrenes or 1-vinylpyrrolidin-2-one (1.0 mmol, 10 equiv). The reaction mixture was sealed in a 4 mL vial and stirred at the described conditions and concentrated under vacuum. For 10, 12, 13–16, the residue was purified by either column chromatography or preparative TLC (PTLC) to afford pure cycloaddition products. For 11, the yield was determined with quantitative ¹H NMR using mesitylene as the internal standard based on our previous reported ¹H NMR spectrum. Only yields with HFIP are listed below. ¹H NMRs of 10–13 exactly match those that we have previously reported.^16,17

2-Ethoxy-5-methyl-1-(phenylsulfonyl)-1,2,3,4-tetrahydropyridine (10).

22.8 mg, 81%. Purified by column chromatography (20% EtOAc–hexanes). White solid. ¹H NMR (600 MHz, CDCl₃) δ 7.83 − 7.74 (m, 2H), 7.59 − 7.54 (m, 1H), 7.51 (dq, J = 7.6, 5.3, 3.3 Hz, 2H), 6.29 (dt, J = 3.8, 1.9 Hz, 1H), 5.17 (s, 1H), 3.79 (tdd, J = 11.4, 8.4, 5.7 Hz, 1H), 3.68 − 3.55 (m, 1H), 2.12 (td, J = 16.9, 14.5, 6.0 Hz, 1H), 1.87 − 1.79 (m, 1H), 1.67 (s, 3H), 1.61 (dt, J = 17.8, 5.0 Hz, 1H), 1.18 (td, J = 7.1, 4.0 Hz, 3H), 0.93 (dddt, J = 13.8, 9.5, 6.7, 3.0 Hz, 1H).

2-Ethoxy-4-phenyl-1-(phenylsulfonyl)-1,2,3,4-tetrahydropyridine (12).

14.8 mg, 43%. Purified by column chromatography (20% EtOAc–hexanes). White solid. ¹H NMR (500 MHz, C₆D₆) δ 7.78 − 7.61 (m, 2H), 7.10 (d, J = 4.3 Hz, 3H), 7.06 − 6.88 (m, 5H), 6.87 (ddd, J = 8.4, 2.1, 1.4 Hz, 1H), 5.23 (td, J = 2.8, 1.4 Hz, 1H), 5.02 (ddd, J = 8.3, 4.7, 1.6 Hz, 1H), 3.56 (dq, J = 9.2, 7.0 Hz, 1H), 3.17 (dq, J = 9.2, 7.0 Hz, 1H), 2.86 (ddt, J = 8.1, 4.8, 1.5 Hz, 1H), 2.05 − 1.85 (m, 1H), 1.26 (ddd, J = 14.0, 8.2, 3.0 Hz, 1H), 0.74 (t, J = 7.0 Hz, 3H).

2-Ethoxy-4,6-diphenyl-1-(phenylsulfonyl)-1,2,3,4-tetrahydropyridine (13).

9.6 mg, 23%. Purified by PTLC (33% EtOAc–hexanes). White solid. ¹H NMR (500 MHz, CDCl₃) δ 7.80 − 7.73 (m, 2H), 7.67 − 7.61 (m, 1H), 7.56 −7.49 (m, 2H), 7.49 − 7.43 (m, 2H), 7.37 − 7.30 (m, 3H), 7.24 − 7.18 (m, 2H), 7.18 − 7.13 (m, 1H), 7.04 − 6.98 (m, 2H), 5.98 (d, J = 3.4 Hz, 1H), 5.62 (dd, J = 5.9, 4.1 Hz, 1H), 4.15 − 4.04 (m, 1H), 3.72 (dq, J = 9.5, 7.0 Hz, 1H), 2.54 (td, J = 7.5, 3.4 Hz, 1H), 2.20 (ddd, J = 13.6, 7.4, 5.9 Hz, 1H), 1.94 (ddd, J = 14.0, 7.6, 4.2 Hz, 1H), 1.26 (t, J = 7.1 Hz, 4H).

5-Methyl-2-phenyl-1-(phenylsulfonyl)-1,2,3,4-tetrahydropyridine (14).

14.1 mg, 45%. Purified by column chromatography (10% EtOAc–hexanes). White solid, mp 68–71 °C. ¹H NMR (600 MHz, CDCl₃) δ 7.83 − 7.74 (m, 2H), 7.60 − 7.53 (m, 1H), 7.48 (dd, J = 8.4, 7.2 Hz, 2H), 7.27 (dt, J = 6.9, 1.5 Hz, 2H), 7.25 − 7.19 (m, 3H), 6.70 (quint, J = 1.5 Hz, 1H), 5.17 (t, J = 3.7 Hz, 1H), 1.94 (ddt, J = 13.4, 5.2, 2.6 Hz, 1H), 1.68 (q, J = 1.1 Hz, 3H), 1.65 − 1.55 (m, 2H), 1.43 (dddd, J = 13.4, 11.5, 6.9, 4.6 Hz, 1H). ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 140.0, 139.2, 132.4, 128.9, 128.3, 126.92, 126.89, 125.8, 118.6, 118.0, 55.0, 26.0, 22.3, 20.9. IR (film) ν_max 1680, 1447, 1339, 1230, 1168, 1124, 1089, 1066, 1034, 1016, 997, 761, 725, 687, 667, 611 cm⁻¹. HRMS (ESI-TOF) m/z: [M + H⁺] calcd for C₁₈H₂₀NO₂S, 314.1215; found, 314.1222.

2-(4-Methoxyphenyl)-5-methyl-1-(phenylsulfonyl)-1,2,3,4-tetrahydropyridine (15).

23.7 mg, 69%. Purified by column chromatography (10% EtOAc–hexanes). White solid, mp 85–88 °C. ¹H NMR (500 MHz, CDCl₃) δ 7.80 − 7.69 (m, 2H), 7.60 − 7.50 (m, 1H), 7.51 − 7.40 (m, 2H), 7.17 − 7.02 (m, 2H), 6.86 − 6.73 (m, 2H), 6.65 (p, J = 1.4 Hz, 1H), 5.09 (t, J = 3.5 Hz, 1H), 3.77 (s, 3H), 1.93 − 1.83 (m, 1H), 1.65 (s, 3H), 1.62 − 1.56 (m, 2H), 1.43 − 1.32 (m, 1H). ¹³C{¹H} NMR (126 MHz, CDCl₃) δ 158.5, 139.2, 132.4, 132.2, 128.9, 127.0, 126.9, 118.5, 117.9, 113.7, 55.2, 54.5, 26.1, 22.3, 20.9. IR (film) ν_max 1511, 1446, 1343, 1248, 1161, 1092, 1074, 1034, 953, 836, 748, 719, 689 cm⁻¹. HRMS (ESI-TOF) m/z: [M + H⁺] calcd for C₁₉H₂₂NO₃S, 344.1320; found, 344.1329.

1-(5-Methyl-1-(phenylsulfonyl)-1,2,3,4-tetrahydropyridin-2-yl)pyrrolidin-2-one (16).

17.6 mg, 55%. Purified by column chromatography (33% EtOAc–hexanes). White solid, mp 90−93 °C. ¹H NMR (500 MHz, CDCl₃) δ 7.87 − 7.76 (m, 2H), 7.64 − 7.55 (m, 1H), 7.56 − 7.47 (m, 2H), 6.60 (s, 1H), 5.80 (t, J = 3.8 Hz, 1H), 3.37 (ddd, J = 9.6, 8.3, 6.5 Hz, 1H), 3.21 (ddd, J = 9.6, 8.3, 4.9 Hz, 1H), 2.37 − 2.24 (m, 2H), 2.07 − 1.95 (m, 1H), 1.95 − 1.76 (m, 4H), 1.70 (s, 3H), 1.46 − 1.32 (m, 1H). ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 174.8, 138.6, 133.0, 129.2, 127.0, 118.6, 116.2, 59.9, 45.3, 31.1, 26.4, 23.1, 20.8, 18.2. IR (film) ν_max 1672, 1446, 1419, 1349, 1285, 1161, 1094, 953, 756, 721, 690 cm⁻¹. HRMS (ESI-TOF) m/z: [M + H⁺] calcd for C₁₆H₂₁N₂O₃S, 321.1273; found, 321.1277.

Methyl (R)-2-(((R)-5-Methyl-1-(phenylsulfonyl)-1,2,3,4-tetrahydropyridin-2-yl)oxy)propanoate (17).

A solution of 8a¹⁶ (10.5 mg, 0.05 mmol) in TFE (0.1 mL) was treated with methyl (R)-2-(vinyloxy)propanoate (9.8 mg, 0.75 mmol, 1.5 equiv) under a N₂ atmosphere and the solution was stirred in a sealed vial at 60 °C (oil bath) for 68 h. The solution was concentrated under vacuum and the d.r. ratio was determined with quantitative ¹H NMR. The residue was purified by PTLC (33% EtOAc–hexanes) to afford 17 (overall yield 43%; major diastereomer: 5.6 mg, 33%) as a colorless oil. Major diastereomer: ¹H NMR (600 MHz, CDCl₃) δ 7.75 − 7.66 (m, 2H), 7.61 − 7.53 (m, 1H), 7.49 (td, J = 7.9, 7.5, 1.8 Hz, 2H), 6.29 (s, 1H), 5.30 (s, 1H), 4.61 (q, J = 7.0 Hz, 1H), 3.79 (s, 3H), 2.16 (ddd, J = 18.5, 13.1, 5.8 Hz, 1H), 1.98 − 1.90 (m, 1H), 1.67 (s, 3H), 1.60 (dd, J = 17.5, 6.2 Hz, 1H), 1.36 (d, J = 7.1 Hz, 3H), 0.82 (tdd, J = 13.4, 6.1, 2.7 Hz, 1H). ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 173.8, 138.7, 132.7, 129.1, 126.7, 120.2, 115.8, 79.5, 69.4, 51.9, 21.7, 20.9, 18.6. IR (film) ν_max 1746, 1446, 1350, 1207, 1162, 1137, 1110, 1070, 1023, 941, 837, 722, 689 cm⁻¹. HRMS (ESI-TOF) m/z: [M + H⁺] calcd for C₁₆H₂₂NO₅S, 339.1140; found, 339.1147. ${\alpha }_D^{20}$ − 88 ± 0.5 (c 0.34, CH₂Cl₂).

Cycloaddition of 1,2,4,5-Tetrazines.

3,6-Diphenylpyridazine (18).

A solution of 9a (11.7 mg, 0.05 mmol) in HFIP (0.5 mL) was treated with 2,5-norbornadiene (2, 9.2 mg, 10 μL, 0.1 mmol) and the reaction mixture was stirred at 25 °C for 3 h. The solution was concentrated under vacuum and the residue was purified by PTLC (33% EtOAc–hexanes) to afford 18 (8.5 mg, 73%) as a white solid. ¹H NMR (400 MHz, CDCl₃) δ 8.24 − 8.17 (m, 4H), 7.95 (s, 2H), 7.58 (dddt, J = 9.8, 8.7, 5.4, 1.3 Hz, 6H). The ¹H NMR exactly matches the previously reported data.⁵⁴

3,6-Bis(methylthio)pyridazine (19).

A solution of 9b³³ (8.7 mg, 0.05 mmol) in HFIP (0.5 mL) was treated with 2,5-norbornadiene (2, 18.4 mg, 21 μL, 0.2 mmol) and the reaction mixture was stirred at 60 °C (oil bath) for 17.5 h. The solution was concentrated under vacuum and the residue was purified by PTLC (33% EtOAc–hexanes) to afford 18 (7.6 mg, 88%) as a white solid. ¹H NMR (400 MHz, CDCl₃) δ 7.11 (s, 2H), 2.70 (s, 6H). The ¹H NMR exactly matches that which we previously reported.³³

Preparative reaction of 9b with 2.

A solution of 9b³³ (174 mg, 1.0 mmol) in HFIP (10 mL) was treated with 2 (369 mg, 0.4 mL, 4.0 mmol) and the reaction mixture was warmed to 60 °C (oil bath) and stirred for 17.5 h. The reaction mixture was concentrated under vacuum and the residue was purified by column chromatography (20% EtOAc–hexanes) to afford 19 (146 mg, 85%) as a white solid identical to the material above.

3,4,6-Triphenylpyridazine (20).

Using 3a as the dienophile. A solution of 9a (24.3 mg, 0.1 mmol) in HFIP (0.1 mL) was treated with styrene (3a, 31.2 mg, 35 μL, 0.3 mmol) and the reaction mixture was stirred at 60 °C (oil bath) for 12.5 h. The solution was concentrated under vacuum, redissolved in EtOAc (ca. 0.3 mL) and exposed to air for 48 h until dihydropyridazine was fully converted to pyridazine (judged by TLC and LCMS). The solution was then concentrated under vacuum and the residue was purified by PTLC (33% EtOAc–hexanes) to afford 20 (17.6 mg, 57%) as a white solid. ¹H NMR (400 MHz, CDCl₃) δ 8.25 − 8.15 (m, 2H), 7.86 (s, 1H), 7.61 −7.46 (m, 5H), 7.41 − 7.26 (m, 8H). The ¹H NMR exactly matches that of material previously reported.⁵⁵ Using 5a as the dienophile. A solution of 9a (24.3 mg, 0.1 mmol) in HFIP (0.1 mL) was treated with phenylacetylene (5a, 30.6 mg, 33 μL, 0.3 mmol) and the reaction mixture was stirred in a sealed vial at 80 °C (oil bath) for 13 h. The solution was concentrated under vacuum and the residue was purified by column chromatography (20% EtOAc–hexanes) to afford 20 (30.5 mg, 99%) as a white solid.

Methyl 3,6-Diphenylpyridazine-4-carboxylate (21).

A solution of 9a (11.7 mg, 0.05 mmol) in HFIP (0.1 mL) was treated with methyl 3-(pyrrolidin-1-yl)acrylate (4e, 15.5 mg, 0.1 mmol) and the reaction mixture was stirred in a sealed vial at 100 °C (oil bath) for 4.5 h. The solution was concentrated under vacuum and purified by PTLC (33% EtOAc–hexanes) to afford 21 (10.8 mg, 74%) as a white solid. ¹H NMR (400 MHz, CDCl₃) δ 8.21 − 8.16 (m, 2H), 8.14 (s, 1H), 7.73 − 7.69 (m, 2H), 7.57 − 7.54 (m, 2H), 7.52 − 7.50 (m, 3H), 3.80 (s, 3H). ¹³C NMR (151 MHz, CDCl₃) δ 167.1, 157.9, 156.9, 136.4, 135.2, 130.5, 130.0, 129.6, 129.2, 128.8, 128.4, 127.1, 123.3, 53.0. IR (film) ν_max 1688, 1611, 1458, 1416, 1359, 1307, 1246, 1187, 1150, 1039, 786, 616 cm⁻¹. HRMS (ESI-TOF) m/z: [M + H⁺] calcd for C₁₈H₁₅N₂O₂, 291.1134; found, 291.1142.

3,6-Bis(methylthio)-4-phenylpyridazine (22).

A solution of 9b³³ (17.4 mg, 0.1 mmol) in HFIP (0.1 mL) was treated with phenylacetylene (5a, 30.6 mg, 33 μL, 0.3 mmol) and the reaction mixture was stirred in a sealed vial at 100 °C (oil bath) for 60 h. The solution was concentrated under vacuum and the residue was purified by column chromatography (20% EtOAc–hexanes) to afford 22 (16.4 mg, 66%) as a white solid. ¹H NMR (400 MHz, CDCl₃) δ 7.45 (dtdd, J = 7.9, 5.4, 2.7, 1.6 Hz, 5H), 7.03 (s, 1H), 2.73 (s, 3H), 2.64 (s, 3H). The ¹H NMR exactly matches that of material we previously reported.³³

Dimethyl 4-Phenylpyridazine-3,6-dicarboxylate (23).

A solution of 9c⁵⁶ (19.8 mg, 0.1 mmol) in HFIP (0.1 mL) was treated with phenylacetylene (5a, 30.6 mg, 33 μL, 0.3 mmol) and the reaction mixture was stirred in a sealed vial at 25 °C for 35 min. Color disappearance can be observed in 5 min. The solution was concentrated under vacuum and the residue was purified by column chromatography (40% EtOAc–hexanes) to afford 23 (27.2 mg, 100%) as a white solid. ¹H NMR (400 MHz, CDCl₃) δ 8.29 (s, 1H), 7.53 (dd, J = 5.1, 1.9 Hz, 3H), 7.49 − 7.43 (m, 2H), 4.13 (s, 3H), 3.89 (s, 3H). The ¹H NMR exactly matches that of material we previously reported.⁵⁶

Trimethyl Pyridazine-3,4,6-tricarboxylate (24).

A solution of 9c⁵⁶ (9.9 mg, 0.1 mmol) in HFIP (0.5 mL) was treated with methyl propiolate (5b, 8.4 mg, 9 μL, 0.1 mmol) and the reaction mixture was stirred in a sealed vial at 80 °C (oil bath) for 40 h. The solution was concentrated under vacuum and the residue was purified by PTLC (50% EtOAc–hexanes) to afford 24 (6.5 mg, 51%) as a pale-yellow solid. ¹H NMR (400 MHz, CDCl₃) δ 8.58 (s, 1H), 4.13 (s, 3H), 4.09 (s, 3H), 4.02 (s, 3H). ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 164.5, 163.3, 163.2, 153.2, 152.6, 129.0, 127.4, 53.84, 53.83, 53.77. IR (film) ν_max 1736, 1440, 1254, 1195, 1136, 1101, 986, 767 cm⁻¹. HRMS (ESI-TOF) m/z: [M + H⁺] calcd for C₁₀H₁₁N₂O₆, 255.0617; found, 255.0627.

Dimethyl 4,5-Bis(1-(methoxycarbonyl)-1H-indol-3-yl)pyridazine-3,6-dicarboxylate (25).

A solution of 5c³¹ (14.9 mg, 0.04 mmol) in HFIP (0.15 mL) was treated with 9c⁵⁶ (15.6 mg, 0.08 mmol) and the reaction mixture was stirred in a sealed vial at 80 °C (oil bath) for 40 h. The solution was concentrated under vacuum and the residue was purified by PTLC (33% EtOAc–hexanes) to afford 25 (7.8 mg, 36%) as a white solid. ¹H NMR (400 MHz, CDCl₃) δ 8.06 (d, J = 8.3 Hz, 2H), 7.53 (d, J = 7.3 Hz, 1H), 7.41 (s, 1H), 7.29 (d, J = 2.5 Hz, 1H), 7.19 (dd, J = 7.3, 4.3 Hz, 2H), 7.02 (q, J = 8.1, 6.4 Hz, 2H), 3.99 (s, 3H), 3.94 (s, 3H), 3.73 (s, 6H). The ¹H NMR exactly matches that of material we previously reported.³¹

1,4-Diphenyl-5H-pyridazino[4,5-b]indole (26).

A solution of 9a (11.7 mg, 0.05 mmol) in HFIP (0.1 mL) was treated with indole (6, 11.7 mg, 0.2 mmol) and the reaction mixture was stirred at 100 °C (oil bath) for 20 h. The solution was concentrated under vacuum, redissolved in EtOAc (ca. 0.3 mL) and exposed to air for at least 48 h until the dihydropyridazine was fully oxidized (judged by LCMS). The solution was then concentrated under vacuum and the residue was purified by PTLC (50% EtOAc–hexanes) to afford 26 (8.8 mg, 55%) as a white solid. Mp 145−148 °C.¹H NMR (400 MHz, CDCl₃) δ 8.09 − 8.03 (m, 2H), 8.01 − 7.95 (m, 2H), 7.90 (dt, J = 8.3, 1.0 Hz, 1H), 7.66 − 7.56 (m, 7H), 7.56 − 7.49 (m, 1H), 7.26 − 7.21 (m, 1H). ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 146.4, 135.5, 134.7, 129.6, 129.5, 129.4, 129.3, 128.9, 128.6, 128.4, 123.4, 121.6, 120.3, 118.1, 112.1. IR (film) ν_max 2923, 2853, 1494, 1453, 1406, 1383, 1324, 1222, 1134, 1112, 747, 698 cm⁻¹. HRMS (ESI-TOF) m/z: [M + H⁺] calcd for C₂₂H₁₆N₃, 322.1344; found, 322.1352.

6-(Methylthio)pyridazin-3-amine (27).

A solution of 9d³² (9.3 mg, 0.05 mmol) in HFIP (0.5 mL) was treated with 2,5-norbornadiene (2, 18.4 mg, 21 μL, 0.2 mmol) and the reaction mixture was stirred at 40 °C (oil bath) for 25 h. The solution was concentrated under vacuum and the residue was purified by PTLC (33% EtOAc–hexanes) to afford 27 (3.5 mg, 52%) as a white solid. ¹H NMR (600 MHz, CDCl₃) δ 7.11 (d, J = 9.2 Hz, 1H), 6.64 (d, J = 9.1 Hz, 1H), 4.61 (br s, 2H), 2.64 (s, 3H). The ¹H NMR exactly matches that of material previously reported.⁵⁷

3-(Methylthio)-6,7-dihydro-5H-pyrrolo[2,3-c]pyridazine (29).

A solution of 9e³⁴ (9.8 mg, 0.05 mmol) in HFIP (0.2 mL) was sealed in a 4 mL vial, warmed to 100 °C (oil bath) and stirred for 2 h. The solution was concentrated under vacuum and the residue was purified by PTLC (40% EtOAc–hexanes) to afford 29 (6.9 mg, 82%) as a white solid. ¹H NMR (600 MHz, CDCl₃) δ 6.92 (t, J = 1.7 Hz, 1H), 3.69 (t, J = 8.0 Hz, 2H), 3.08 (ddd, J = 8.4, 7.7, 1.7 Hz, 2H), 2.60 (s, 3H). The ¹H NMR was identical to that of material we previously reported.³⁴

Dimethyl 6-Phenylpyridine-2,5-dicarboxylate (31).

A solution of 9f⁵⁸ (13.7 mg, 0.05 mmol) in HFIP (0.5 mL) was treated with 2,5-norbornadiene (2, 18.4 mg, 21 μL, 0.2 mmol) and the reaction mixture was stirred at 25 °C for 4 h. The solution was concentrated under vacuum and the residue was purified by PTLC (33% EtOAc–hexanes) to afford 31 (8.6 mg, 63%) as a white solid. ¹H NMR (500 MHz, CDCl₃) δ 8.20 (d, J = 8.0 Hz, 1H), 8.13 (d, J = 8.0 Hz, 1H), 7.62 − 7.52 (m, 2H), 7.49 − 7.38 (m, 3H), 4.01 (s, 3H), 3.71 (s, 3H). The ¹H NMR exactly matches that of material previously reported.⁵⁸

2-Phenylpyridine (32).

A solution of 9g (3.1 mg, 0.02 mmol) in HFIP (0.2 mL) was treated with 2,5-norbornadiene (2, 3.7 mg, 4 μL, 0.04 mmol) and the reaction mixture was sealed in a 4 mL vial and stirred at 90 °C (oil bath) for 5 h. The solution was concentrated under vacuum and the residue was purified by PTLC (25% EtOAc–hexanes) to afford 32 (1.6 mg, 51%) as a colorless oil. ¹H NMR (400 MHz, CDCl₃) δ 8.75 − 8.65 (m, 1H), 8.05 − 7.93 (m, 2H), 7.81 − 7.69 (m, 2H), 7.54 − 7.45 (m, 2H), 7.44 − 7.39 (m, 1H), 7.25 − 7.19 (m, 1H). The ¹H NMR exactly matches that of material previously reported.⁵⁹

Trimethyl Pyridine-2,3,6-tricarboxylate (33).

A solution of 9h⁶⁰ (11.8 mg, 0.04 mmol) in HFIP (0.4 mL) was treated with 2,5-norbornadiene (2, 7.4 mg, 8 μL, 0.08 mmol) and the reaction mixture was stirred at 25 °C for 1 h. The solution was concentrated under vacuum and the residue was purified by PTLC (33% EtOAc–hexanes) to afford 33 (7.6 mg, 64%) as a colorless oil. ¹H NMR (500 MHz, CDCl₃) δ 8.34 (d, J = 8.1 Hz, 1H), 8.24 (d, J = 8.1 Hz, 1H), 4.48 (dq, J = 12.3, 7.2 Hz, 4H), 4.41 (q, J = 7.2 Hz, 2H), 1.46 − 1.37 (m, 9H). The ¹H NMR exactly matches that of material we previously reported.⁶¹

Triethyl 5-Phenylpyridine-2,3,6-tricarboxylate (34a) and Triethyl 4-Phenylpyridine-2,3,6-tricarboxylate (35b).

A solution of 9h (19.2 mg, 0.065 mmol) in HFIP (0.5 mL) was treated with phenylacetylene (5a, 13.3 mg, 14 μL, 0.13 mmol) and the reaction mixture was sealed in a 4 mL vial and stirred at 100 °C (oil bath) for 48 h. The solution was concentrated under vacuum and the residue was purified by PTLC (33% EtOAc–hexanes) to afford 34a (4.6 mg, 19%) as a white solid and 34b (4.1 mg, 17%) as a white solid. The regioselectivity was determined with a single crystal X-ray structure determination of 34b. For 34a: mp 64–67 °C. ¹H NMR (500 MHz, CDCl₃) δ 8.25 (s, 1H), 7.45 (dd, J = 5.0, 1.9 Hz, 3H), 7.41 − 7.34 (m, 2H), 4.48 (q, J = 7.2 Hz, 2H), 4.41 (q, J = 7.1 Hz, 2H), 4.19 (q, J = 7.1 Hz, 2H), 1.42 (t, J = 7.2 Hz, 3H), 1.39 (t, J = 7.2 Hz, 3H), 1.06 (t, J = 7.2 Hz, 3H). ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 166.5, 164.4, 164.0, 150.5, 148.3, 147.1, 136.4, 132.8, 129.4, 128.7, 128.2, 127.9, 62.7, 62.6, 62.1, 14.2, 14.1, 13.6. IR (film) ν_max 1722, 1377, 1345, 1282, 1246, 1213, 1142, 1111, 1068, 1019, 767, 701 cm⁻¹. HRMS (ESI-TOF) m/z: [M + H⁺] calcd for C₂₀H₂₂NO₆, 372.1447; found, 372.1450. For 34b: mp 47–50 °C. ¹H NMR (500 MHz, CDCl₃) δ 8.26 (s, 1H), 7.46 (dt, J = 4.0, 2.9 Hz, 3H), 7.44 − 7.38 (m, 2H), 4.49 (dq, J = 16.2, 7.1 Hz, 4H), 4.19 (q, J = 7.1 Hz, 2H), 1.45 (dt, J = 8.3, 7.2 Hz, 6H), 1.07 (t, J = 7.1 Hz, 3H). ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 165.7, 165.6, 164.5, 151.2, 149.4, 139.7, 137.7, 136.4, 128.8, 128.7, 128.2, 127.1, 62.44, 62.37, 62.1, 14.03, 14.01, 13.6. IR (film) ν_max 1728, 1322, 1259, 1218, 1150, 1125, 1073, 1017, 858, 756, 701 cm⁻¹. HRMS (ESI-TOF) m/z: [M + H⁺] calcd for C₂₀H₂₂NO₆, 372.1447; found, 372.1452.

Dimethyl 3-Methyl-4,6-diphenylpyridine-2,5-dicarboxylate (35a) and Dimethyl 4-Methyl-3,6-diphenylpyridine-2,5-dicarboxylate (35b).

Cycloaddition. A solution of 9f⁵⁸ (9.4 mg, 0.034 mmol) in HFIP (0.35 mL) was treated with (E)-prop-1-en-1-ylbenzene (3c, 20.1 mg, 22 μL, 0.17 mmol) and the reaction mixture was sealed in a 8 mL vial and stirred at 100 °C (oil bath) for 67 h. The solution was concentrated under vacuum and the residue was purified by PTLC (33% EtOAc–hexanes) to afford an inseparable mixture of 35a and 35b (3.5 mg, 28% in total, 35a : 35b = 1:1 determined by ¹H NMR integration) as a white solid and dimethyl 3-methyl-4,6-diphenyl-1,4-dihydropyridine-2,5dicarboxylate 35a’ (5.2 mg, 43%) as a colorless oil. Characteristic peaks for 35b: ¹H NMR (400 MHz, CDCl₃) δ 3.69 (s, 3H), 3.62 (s, 3H), 2.18 (s, 3H). Oxidation of 35a’ to 35a. A premixed suspension of activated MnO₂ (28.7 mg, 0.33 mmol) in CH₂Cl₂ (0.15 mL) was treated with a solution of 35a’ (5.2 mg, 0.014 mmol) in CH₂Cl₂ (50 μL). The reaction mixture was stirred at 25 °C for 24 h, filtered, and concentrated to afford 35a (3.7 mg ,73%) as a white solid. Mp 71–74 °C.¹H NMR (400 MHz, CDCl₃) δ 7.67 (dd, J = 7.6, 2.0 Hz, 2H), 7.51 − 7.40 (m, 6H), 7.25 (dd, J = 7.8, 1.7 Hz, 2H), 4.02 (s, 3H), 3.36 (s, 3H), 2.29 (s, 3H). ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 168.1, 167.1, 153.1, 150.1, 149.8, 138.7, 136.2, 130.8, 129.8, 129.0, 128.5, 128.4, 128.3, 52.8, 52.1, 16.2. IR (film) ν_max 1732, 1439, 1408, 1242, 1209, 1134, 1083, 1017, 795, 700 cm⁻¹. HRMS (ESI-TOF) m/z: [M + H⁺] calcd for C₂₂H₂₀NO₄, 362.1392; found, 362.1386.

3-Phenyl-6,7-dihydro-5H-cyclopenta[b]pyridine (36).

A solution of 9i^41b (7.9 mg, 0.05 mmol) in HFIP (0.5 mL) was treated with 1-(cyclopent-1-en-1-yl)pyrrolidine (4a, 20.6 mg, 22 μL, 0.15 mmol) and the reaction mixture was stirred at 25 °C for 20 h. The solution was concentrated under vacuum and the residue was purified by PTLC (25% EtOAc–hexanes) to afford 36 (6.5 mg, 67%) as a white solid. ¹H NMR (400 MHz, CDCl₃) δ 8.56 (d, J = 2.1 Hz, 1H), 7.69 (d, J = 1.6 Hz, 1H), 7.59 − 7.51 (m, 2H), 7.50 − 7.43 (m, 2H), 7.42 − 7.34 (m, 1H), 3.07 (t, J = 7.7 Hz, 2H), 3.00 (t, J = 7.5 Hz, 2H), 2.19 (p, J = 7.6 Hz, 2H). The ¹H NMR exactly matches that of material we previously reported.^41b

3-Phenyl-5,6,7,8-tetrahydroquinoline (37).

A solution of 9i^41b (7.9 mg, 0.05 mmol) in HFIP (0.5 mL) was treated with 1-(cyclohex-1-en-1-yl)pyrrolidine (4a, 20.6 mg, 22 μL, 0.15 mmol) and the reaction mixture was stirred at 25 °C for 17 h. The solution was concentrated under vacuum and the residue was purified by PTLC (20% EtOAc–hexanes) to afford 37 (8.5 mg, 81%) as a pale-yellow oil. ¹H NMR (400 MHz, CDCl₃) δ 8.59 (s, 1H), 7.56 (dd, J = 6.3, 1.6 Hz, 3H), 7.46 (dd, J = 8.4, 6.8 Hz, 2H), 7.37 (t, J = 7.4 Hz, 1H), 2.98 (t, J = 6.4 Hz, 2H), 2.85 (t, J = 6.3 Hz, 2H), 1.94 (p, J = 6.6 Hz, 2H), 1.86 (q, J = 5.9, 5.4 Hz, 2H). The ¹H NMR exactly matches that of material we previously reported.^41b

3-Phenyl-5-propylpyridine (38).

A solution of 9i^41b (7.9 mg, 0.05 mmol) in HFIP (0.5 mL) was treated with 1-(pent-1-en-1-yl)pyrrolidine (4c, 20.9 mg, 0.15 mmol) and the reaction mixture was stirred at 25 °C for 17 h. The solution was concentrated under vacuum and the residue was purified by PTLC (20% EtOAc–hexanes) to afford 38 (6.9 mg, 70%) as a colorless oil. ¹H NMR (500 MHz, CDCl₃) δ 8.70 (d, J = 2.2 Hz, 1H), 8.45 (d, J = 2.1 Hz, 1H), 7.74 − 7.68 (m, 1H), 7.65 − 7.57 (m, 2H), 7.50 (dd, J = 8.4, 6.8 Hz, 2H), 7.45 − 7.40 (m, 1H), 2.74 − 2.65 (m, 2H), 1.74 (heptet, J = 7.3 Hz, 2H), 1.01 (t, J = 7.3 Hz, 2H). ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 148.5, 145.6, 137.9, 137.8, 134.6, 129.0, 128.0, 127.2, 35.0, 24.3, 13.7. IR (film) ν_max 1455, 1260, 1102, 1068, 1028, 853, 800, 751, 705 cm⁻¹. HRMS (ESI-TOF) m/z: [M + H⁺] calcd for C₁₄H₁₆N, 198.1283; found, 198.1280.

3-Bromo-5-phenylpyridine (39).

A solution of 9k^41b (8.0 mg, 0.05 mmol) in HFIP (0.5 mL) was treated with 4d (26.0 mg, 0.15 mmol) and the reaction mixture was stirred at 25°C for 5 h. The solution was concentrated under vacuum and the residue was purified by PTLC (25% EtOAc–hexanes) to afford 39 (11.7 mg, >99%) as a colorless oil. ¹H NMR (400 MHz, CDCl₃) δ 8.76 (d, J = 2.0 Hz, 1H), 8.65 (d, J = 2.2 Hz, 1H), 8.02 (t, J = 2.1 Hz, 1H), 7.59 − 7.54 (m, 2H), 7.52 − 7.47 (m, 2H), 7.46 − 7.41 (m, 1H). The ¹H NMR exactly matches that of material previously reported.⁶³

2-Methoxylcarbonyl-5-phenylpyridine (40).

A solution of 9l^41b (7.0 mg, 0.05 mmol) in HFIP (0.5 mL) was treated with 4d (26.0 mg, 0.15 mmol) and the reaction mixture was stirred at 25 °C for 20 h. The solution was concentrated under vacuum and the residue was purified by PTLC (25% EtOAc–hexanes) to afford 40 (7.2 mg, 68%) as a white solid. ¹H NMR (400 MHz, CDCl₃) δ 8.97 (dd, J = 2.4, 0.8 Hz, 1H), 8.21 (dd, J = 8.0, 0.9 Hz, 1H), 8.03 (dd, J = 8.1, 2.3 Hz, 1H), 7.68 − 7.60 (m, 2H), 7.58 − 7.49 (m, 2H), 7.49 − 7.43 (m, 1H), 4.04 (s, 3H). The ¹H NMR exactly matches that of material previously reported.⁶⁴

3-Methoxycarbonyl-5-phenylpyridine (41).

A solution of 9i^41b (7.9 mg, 0.05 mmol) in HFIP (0.5 mL) was treated with methyl 3-(pyrrolidin-1-yl)acrylate (4e, 23.3 mg, 0.15 mmol) and the reaction mixture was stirred at 90°C (oil bath) for 13 h. The solution was concentrated under vacuum and the residue was purified by PTLC (33% EtOAc–hexanes) to afford 41 (2.5 mg, 23%) as a white solid. ¹H NMR (400 MHz, CDCl₃) δ 9.20 (d, J = 2.0 Hz, 1H), 9.01 (d, J = 2.3 Hz, 1H), 8.50 (t, J = 2.2 Hz, 1H), 7.66 − 7.60 (m, 2H), 7.55 − 7.48 (m, 2H), 7.47 − 7.41 (m, 1H), 3.99 (s, 3H). The ¹H NMR exactly matches that of material previously reported.⁶⁵

3,5-Bis(methoxycarbonyl)pyridine (42).

A solution of 9j^41b (7.0 mg, 0.05 mmol) in HFIP (0.5 mL) was treated with methyl 3-(pyrrolidin-1-yl)acrylate (4e, 23.3 mg, 0.15 mmol) and the reaction mixture was stirred at 90°C (oil bath) for 13 h. The solution was concentrated under vacuum and the residue was purified by PTLC (33% EtOAc–hexanes) to afford 42 (6.6 mg, 68%) as a white solid. ¹H NMR (400 MHz, CDCl₃) δ 9.37 (d, J = 2.1 Hz, 2H), 8.88 (t, J = 2.1 Hz, 1H), 3.99 (s, 6H). The ¹H NMR exactly matches that of material previously reported.⁶⁶

Methyl 6-Phenylnicotinate (43).

A solution of 9j^41b (7.0 mg, 0.05 mmol) in TFE (0.1 mL) was treated with phenylacetylene (5a, 20.4 mg, 22 μL, 0.2 mmol) and the reaction mixture was stirred at 90°C (oil bath) for 48 h. The solution was concentrated under vacuum and the residue was purified by PTLC (33% EtOAc–hexanes) to afford 43 (3.7 mg, 35%) as a white solid. ¹H NMR (600 MHz, CDCl₃) δ 9.28 (dd, J = 2.2, 0.9 Hz, 1H), 8.35 (dd, J = 8.3, 2.2 Hz, 1H), 8.09 − 8.03 (m, 2H), 7.82 (dd, J = 8.3, 0.9 Hz, 1H), 7.53 − 7.49 (m, 2H), 7.49 − 7.45 (m, 1H), 3.98 (s, 3H). The ¹H NMR exactly matches that of material previously reported.⁶⁷

Cycloadditions of 1,3,5-Triazine (9m).

General method: A solution of 9m⁴⁸ (14.9 mg, 0.05 mmol) in HFIP (0.2 mL) was treated with corresponding aldehyde/ketone (7a–7d, 0.15 mmol, 3 equiv) and the reaction mixture was stirred at the indicated temperature for the time indicated. The solution was concentrated under vacuum and the residue was purified by PTLC to afford the pure pyrimidines 45–48. ¹H NMRs of 45–48 exactly match those of material we previously reported.⁴⁸

Diethyl 6,7-Dihydro-5H-cyclopenta[d]pyrimidine-2,4-dicarboxylate (45).

7.9 mg, 30%. Purified by PTLC (33% EtOAc–hexanes). White solid. ¹H NMR (400 MHz, CDCl₃) δ 4.51 (dq, J = 21.4, 7.1 Hz, 4H), 3.38 (t, J = 7.6 Hz, 2H), 3.20 (t, J = 8.0 Hz, 2H), 2.24 (p, J = 7.9 Hz, 2H), 1.45 (td, J = 7.1, 3.2 Hz, 6H).

Diethyl Pyrimidine-2,4-dicarboxylate (46).

6.9 mg, 31%. Purified by PTLC (33% EtOAc–hexanes). White solid. ¹H NMR (500 MHz, CDCl₃) δ 9.15 (d, J = 5.0 Hz, 1H), 8.13 (d, J = 4.9 Hz, 1H), 4.60 − 4.49 (m, 4H), 1.53 − 1.41 (m, 6H).

Diethyl 5-Phenylpyrimidine-2,4-dicarboxylate (47).

4.5 mg, 15%. Purified by PTLC (33% EtOAc–hexanes). White solid. ¹H NMR (400 MHz, CDCl₃) δ 9.03 (s, 1H), 7.54 − 7.46 (m, 3H), 7.41 (ddt, J = 5.5, 3.1, 1.2 Hz, 2H), 4.57 (q, J = 7.1 Hz, 2H), 4.27 (q, J = 7.2 Hz, 2H), 1.48 (t, J = 7.1 Hz, 3H), 1.14 (t, J = 7.2 Hz, 3H).

Diethyl 5-Propylpyrimidine-2,4-dicarboxylate (48).

4.0 mg, 15%. Purified by PTLC (33% EtOAc–hexanes). White solid. ¹H NMR (500 MHz, CDCl₃) δ 8.87 (s, 1H), 4.54 (q, J = 7.2 Hz, 2H), 4.49 (q, J = 7.2 Hz, 2H), 2.93 − 2.85 (m, 2H), 1.76 − 1.63 (m, 2H), 1.45 (dt, J = 11.5, 7.1 Hz, 6H), 1.00 (t, J = 7.3 Hz, 3H).