Coinage Metal-Catalyzed Divergent Heterocyclizations of Underexplored N‑Propargyl Hydrazides

Abstract

Herein, we disclose metal-catalyzed tunable annulations of uncharted alkyne-tethered hydrazides for the divergent preparation of diazacyclic frameworks. Cationic gold facilitates the O-cyclization, providing valuable [1,3,4]­oxadiazine scaffolds with total selectivity, whereas silver catalysis promotes controlled N-cyclization to access N-acyl pyrazoles. Furthermore, density-functional-theory-based theoretical investigations support two different pathways (ionic versus radical) operating in the catalytic heterocyclization reaction of the same N-propargyl hydrazide substrate.

Azaheterocycles are ubiquitous compounds in drugs, agrochemicals, and advanced materials. Among heterocyclic cores, [1,3,4]­oxadiazines can be found in many bioactive molecules and N-acyl pyrazoles are important building blocks in organic synthesis (Scheme , top). , Recent strategies for the preparation of the [1,3,4]­oxadiazine framework include the organocatalyzed cycloaddition of N-acyldiazenes with ketenes or allenoates. Despite hydrazides being of widespread use to prepare several heterocycles, propargyl hydrazides are hitherto unknown, and consequently, its reactivity remains uncharted. Gold and silver catalyses are often utilized for the cyclization of functionalized alkynes due to the π acidity of the metallic salts. , Although the gold-catalyzed reaction of N-propargylcarboxamides to give 5-methyloxazoles is a well-known process (Scheme a and b), similar reactivity for silver or acyl derivatives of prop-2-yn-1-ylhydrazine has not yet been described. New synthetic protocols that provide rapid and selective access to either one heterocyclic motif or a skeleton different from the same starting material are highly desirable. Herein, we report a synthetic strategy based on noble metal-catalyzed cyclization of unexplored N-propargyl hydrazides, to enable divergent access to a series of [1,3,4]­oxadiazines and N-acyl pyrazoles depending on the nature of the metal (Scheme c). Relevant aspects of this approach embrace: (a) ambident nucleophilicity of novel alkyne-tethered hydrazides reacting as either O-nucleophile or N-nucleophile reagents, (b) divergent heterocyclization and convenient synthesis of both [1,3,4]­oxadiazines and N-acyl pyrazoles, (c) high atom economy (leaving or protective groups are not necessary), and (d) metal-controlled chemo- and regioselectivity without noticing competing cyclizations. First of all, we targeted a straightforward and cost-effective protocol for the preparation of alkyne-tethered hydrazides. It was found that the hexafluorophosphate azabenzotriazole tetramethyl uronium (HATU)-promoted coupling reaction of commercially available propargyl hydrazine hydrochloride with a variety of carboxylic acids resulted in the convenient preparation of N-propargyl hydrazides 1 (for details, see the Supporting Information). Our initial investigation focused on the evaluation of various reaction conditions by using N-propargyl hydrazide 1a as a model substrate.

1. Background and Context of the Work.

Different gold-based precatalysts, such as AuCl₃, [AuClIPr] [IPr = 1,3-bis­(2,6-diisopropylphenyl)­imidazol-2-ylidene]/AgOTf, [(XPhos)­AuNTf₂], and Au/TiO₂ resulted in complex reaction mixtures (entries 1–3, Table S1; see the Supporting Information). Fortunately, the use of [(Ph₃P)­AuCl]/AgOTf provided [1,3,4]­oxadiazine 2a in a high 88% yield (entry 4, Table S1), while [(Ph₃P)­AuNTf₂] proved to be the optimal catalyst and procured oxadiazine 2a in a great 96% isolated yield (entry 6, Table S1). The acidic media generated under the gold-catalyzed conditions partially decomposed oxadiazine 2a in the absence of an inorganic base. The addition of stoichiometric amounts of K₂CO₃ granted reproducible results on the formation of heterocycle 2a without impacting either chemo- or regioselectivity. Besides, no isomerization of the exocyclic double bond in 2a toward endocyclic alkene was detected. Unreacted precursor 1a was recovered to a considerable extent (80%) with the introduction of platinum or different Lewis acids as promoters. Screening of different solvents revealed that cyclic ethers are best suited for the directed reaction, while chlorinated solvents, alcohols, and aromatic hydrocarbons resulted in significant deterioration of the unsaturated hydrazide precursor 1a. The heterocyclization reaction can also be conducted in ethyl acetate (entry 9, Table S1), although purification of oxadiazine 2a is problematic because the reaction is more complicated. Noteworthy, silver-based complexes exert opposing activating effects and were able to tune the chemo- and regioselectivities of the process for the specific formation of N-acyl pyrazole 3a (Table S2; see the Supporting Information). Nanometallic AgNO₃·SiO₂ demonstrated superior performance compared to different silver salts under homogeneous conditions and was selected as the catalyst of choice (entry 6, Table S2). With optimized reaction conditions in hand, the scope of the reactions was studied. As previously observed for model substrate 1a, a totally distinct reaction outcome was noticed when gold- or silver-based complexes operated as precatalysts during the transformation of different alkyne-tethered hydrazides 1 (Schemes and ).

2. Gold-Catalyzed Controlled Preparation of Oxadiazines 2a–2n and 2o-Ac–2q-Ac .

^a Yield of pure, isolated product with correct analytical and spectral data.

3. Silver-Catalyzed Synthesis of Pyrazoles 3a–3q .

^a Yield of pure, isolated product with correct analytical and spectral data.

The Au­(I)-catalyzed cycloisomerization of N-propargyl hydrazides forged the [1,3,4]­oxadiazine skeleton through a 6-exo-dig oxycyclization, while its AgNP-catalyzed counterpart reaction formed the pyrazole scaffold via a 5-endo-dig azacyclization. Our method is not compatible with free OH, NH₂, and COOH groups because the appropriate hydrazide precursors 1 are not achievable through the HATU-promoted coupling reaction. Both a mild electron-donating group (Me) and electron-withdrawing groups (Cl, CF₃, and NO₂) could be present in different positions (ortho, meta, and para) of the aryl ring, leading exclusively to [1,3,4]­oxadiazines 2 in good yields. However, electron-withdrawing groups (Cl, CF₃, and NO₂) gave better yields (72%, quantitative yield) than those with an electron-donating group (Me, 70%). Indeed, under otherwise identical conditions, the gold-catalyzed reaction of N-propargyl hydrazide 1c bearing a more electron-rich 4-MeOC₆H₄ moiety did not occur to form the corresponding oxadiazine. The failure of 1c may be ascribed to the resonance effect of the 4-methoxy substituent in the 4-methoxyphenyl group through the formation of quinone methide, which may be difficult for the nucleophilic attack of amide oxygen. Similarly, the cyclization of 1-N bearing a 4-Me₂NC₆H₄ group was unproductive under gold catalysis. Bromoderivative 2f was generated, but we were unable to isolate it in its pure form. Per contra, when the protocol was applied to internal alkyne-tethered hydrazides 1a-PhOMe and 1a-PhNO ₂ , oxadiazine formation was not attained (Scheme a). Interestingly, heterocycles 2o–2q, which were elusive to chromatographic purification, were isolated as their corresponding acetylated derivatives 2o-Ac, 2p-Ac, and 2q-Ac in a gold-catalyzed/acetylation one-pot process (Scheme b). By contrast, reverse electronic effects were observed for the conversion of alkynes 1 into N-acyl pyrazoles 3, because the electron-donating group furnished five-membered azacycles 3 (e.g., 3c and 3-N versus 3d) in increased yields (Scheme ). To our delight, the indole nucleus was well-tolerated in N-propargyl hydrazides 1, and both O-cyclization products 2j and 2k and N-cyclization products 3j and 3k were efficiently built in the presence of [(Ph₃P)­AuNTf₂] and AgNO₃·SiO₂, respectively. It should be noted that no protection for the NH-free indole was required for either the gold-catalyzed cycloisomerization reaction or the silver-catalyzed process. Besides, other heterocyclic rings, such as furan, thiophene, and quinoline, were also well-accommodated. When alkyl-substituted N-propargyl hydrazides 1p and 1q were used as precursors, both metal-catalyzed reactions delivered the products. However, while oxadiazines 2p and 2q decomposed during chromatographic purification, pyrazoles 3p and 3q were isolated in 64 and 65% yields, respectively. The bulky 2-naphthalene nucleus was tolerated, affording oxadiazine 2n and pyrazole 3n in 59 and 65% yields, respectively. The above results point to a notable dependence of the effectiveness of both sequences related to the electronic nature of the substituent attached to the carbonyl group, while little influence due to steric hindrance is observed (e.g., 2e versus 2h versus 2i). Complete conversion was observed by TLC and ¹H NMR analysis of the crude reaction mixtures of precursors 1. The unwanted side products, the volatility of several pyrazoles 3, and some decomposition perceived during purification by flash chromatography may be the cause of the moderate isolated yields.

Besides unprotected NH-free hydrazides 1, di-tert-butyl 1-(prop-2-yn-1-yl)­hydrazine-1,2-dicarboxylate 1-Boc was also compatible with the above gold catalysis regime. Indeed, the gold-catalyzed heterocyclization was also operative starting from alkyne-tethered bis­(carbamate) 1-Boc, with 2-oxo-[1,3,4]­oxadiazine 4 being smoothly formed in a selective way (Scheme ). The value of the gold-catalyzed reaction of 1-Boc was not just that N-Boc-protected hydrazine can work as a NC­(O)O nucleophile but rather that the gold-catalyzed reaction of 1-Boc proceeded chemoselectively through O-cyclization of the distal carbamate motif, allowing the transfer of the carbonyl group inside the six-membered ring 4. To inspect the behavior of N-homopropargyl hydrazides, we tested N′-(but-3-yn-1-yl)­benzohydrazide (homologue of 1a), but it was not very rewarding because it remained unreacted under gold catalysis and very little conversion was observed after the silver treatment.

4. Gold-Catalyzed Cyclization of Hydrazide 1-Boc .

Post-synthetic transformations of heterocycles 2 under both oxidative conditions (epoxidation using MCPBA and dihydroxylation using OsO₄) and palladium catalysis [4-iodotoluene, catalyzed by either Pd­(OAc)₂ or Pd­(PPh₃)₄ in the presence of SPhos and K₂CO₃] resulted in decomposition. The acetylation and bromobenzoylation of [1,3,4]­oxadiazine 2g were smoothly accomplished to provide acylated derivatives 2g-Ac and 2g-BrBz (Scheme a). Taking into account the utility of vicinal dihalides in pharmacology and material chemistry, we decided to explore the dibromination of selected oxadiazines 2-Ac. The reaction of [1,3,4]­oxadiazine 2g under several dihalogenation conditions resulted in decomposition. Interestingly, the selective formation of 1,2-dibrominated derivatives 4g and 5g was attained after exposure of 2g-Ac and 2g-BrBz to NBS in nitromethane (Scheme b). Adduct 5g was formed along with its rotamer 5g-rot, which were easily separable by flash column chromatography. When a scale-up reaction was run starting from 2 mmol of N-propargyl hydrazide 1a (Scheme c), oxadiazine 2a was obtained in a similar yield (92% versus 96%), which revealed the practicability of our method.

5. Transformations of Oxadiazines 2 and Scale-Up .

^a Yield of pure, isolated product with correct analytical and spectral data.

In order to enlighten the mechanism of the considered reactions, we incorporated the established radical scavenger 2,2,6,6-tetramethyl-piperidinium-1-oxyl radical (TEMPO) into the standard reaction conditions of propargyl hydrazide 1a under both gold and silver catalyses. The addition of TEMPO did not exert influence in the yield of [1,3,4]­oxadiazine 2a but resulted in both a reduction of the yield of pyrazole 3a (from 92 to 58%) and a considerable diminution in the reaction rate. In a similar way, a decrease in the yield (from 75 to 29%) was noticed during the silver-catalyzed formation of target product 3c starting from cyclization precursor 2c in the presence of TEMPO. Likewise, the use of butylated hydroxytoluene (BHT) as a radical-trapping reagent had a considerable effect on the silver-catalyzed reaction of 2c and inhibited it in great extension. Besides, a TEMPO-trapped adduct from 2a was identified by HPLC–MS, displaying a mass-to-charge ratio (m/z) of 330.19 (see the Supporting Information). The above results firmly suggest that the silver-catalyzed sequence progresses via a free radical pathway. A tentative mechanism for the formation of 6-methylene-2-aryl-5,6-dihydro-4H-1,3,4-oxadiazines 2 should start with the η coordination of cationic gold to the triple bond in N′-(prop-2-yn-1-yl)­arenehydrazides 1, which should result in complexes 1-Au. Next, chemoselective 6-exo-dig oxyauration by nucleophilic attack of the carbonyl moiety to form cationic 4H-1,3,4-oxadiazin-3-ium intermediate INT-I followed by deprotonation should construct neutral alkenylgold species INT-II. Finally, protonolysis of the carbon–gold bond should liberate heterocycles 2 with concomitant regeneration of the Au­(I) catalytic species (Scheme ).

6. Tentative Mechanism for the Formation of Oxadiazines 2 .

Based on previous literature, a plausible path for the formation of aryl­(1H-pyrazol-1-yl)­methanones 3 is depicted in Scheme . The reaction should be initiated by molecular-oxygen-promoted oxidation of silver(0) to silver­(I). Next, substitution of the proton in N-propargyl hydrazides 1 with silver­(I) should generate silver amide INT-B, which via homolytic cleavage of the N–[Ag­(I)] bond should form nitrogen-centered hydrazidyl radical INT-C, while oxygen-mediated oxidation of silver(0) to silver­(I) closes the catalytic cycle. An ensuing 5-endo azacyclization should occur, giving rise to 2,3-dihydro-1H-pyrazole radical INT-D. Then, two consecutive hydrogen atom transfers (HATs) assisted by the solvent should build radical INT-F, via neutral pyrazoline INT-E. Oxidation of species INT-F to form pyrazolium intermediate INT-G followed by proton release should liberate final products 3.

7. Plausible Pathway for the Formation of Pyrazoles 3 .

To support the proposed mechanisms for both heterocyclizations, density functional theory (DFT) calculations were conducted (see the Supporting Information). Interestingly, in both Au- and Ag-mediated reactions, the 6-exo-dig ring closure yielding a [1,3,4]­oxadiazine product was found to be energetically more favorable than the 5-endo-dig mode. This implies that, regardless of the kind of noble metal center, the ionic cyclization triggered by π-Lewis acidic activation of the alkyne moiety predominantly gives [1,3,4]­oxadiazines. On the other hand, a selective formation of N-acylpyrazole product 3 was successfully simulated in the cyclization of amidyl radical INT-C. That is, the activation energy of radical 5-endo-dig cyclization (+30.6 kcal mol^–1) was much lower than that of radical 6-exo-dig cyclization (+32.4 kcal mol^–1). The following HAT process between INT-D and a 1,4-dioxane molecule, which was explicitly considered as a solvent molecule, has a very low barrier (+5.5 kcal mol^–1), and the formation of INT-E and a dioxane radical is exergonic. These data support the irreversible formation of INT-E. Although there are several possibilities for the formation of 3 from INT-E, our computation proposes the generation of INT-F through the HAT reaction by the dioxane radical and the following oxidation to yield INT-G, which easily gives 3 through deprotonation.

To summarize, the coinage metal-catalyzed chemo- and regiodivergent cyclization of unexplored N-propargyl hydrazides having multiple reactive sites makes it possible to direct access to structurally diverse heterocycles. The controlled oxycyclization provides [1,3,4]­oxadiazines with the assistance of gold complexes, while silver-catalyzed fine-tuned reactivity delivers N-acyl pyrazoles using the same starting materials. These reactions present high atom and step economies with good functional group tolerance. DFT studies corroborated that two different pathways, namely, ionic (Au) versus radical (Ag), are operating.

The data underlying this study are available in the published article and its Supporting Information.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.orglett.5c03069.

• Experimental procedures, characterization data of new compounds, copies of NMR spectra, and computational details (PDF)

†.

Daniel Diez-Iriepa and Mireia Toledano-Pinedo contributed equally to this work.

The authors declare no competing financial interest.