Modular synthesis of functionalized azetidines through visible-light copper-catalyzed carboamination of 1,3-dienes with azabicyclo[1.1.0]butanes Authors: Yejin Koo, Dongwoo Lee, Jiyong Park, Sungwoo Hong Published: 2026 ABSTRACT The strain-release reactivity of azabicyclo[1.1.0]butanes (ABBs) offers a powerful entry to substituted azetidines, yet modular access to highly functionalized, allylic amine-containing frameworks remains underdeveloped. Here, we report a visible-light copper-catalyzed four-component coupling that effects 1,4-carboamination of aryl-substituted 1,3-dienes, assembling azetidines bearing a quaternary carbon center from dienes, aryl amines, and electrophiles using ABBs. Synergistic copper/photoredox catalysis unites strain-release generation of an azetidinyl radical with allylic C–N bond formation, delivering densely functionalized azetidines bearing an allylic amine in a single, redox-neutral operation. The reaction proceeds with high 1,4-regioselectivity and broad functional-group tolerance across diverse anilines, ABBs, dienes, and electrophiles, and operates under mild conditions. Biorelevant complex-molecule diversification, scalability, and downstream derivatization underscore the synthetic utility of the platform, while DFT calculations trace the 1,4-selectivity to stabilizing noncovalent interactions that favor benzylic C–N coupling and steric repulsion that disfavors the competing pathway. Introduction Structurally rigid nitrogen heterocycles such as azetidines play a pivotal role in medicinal chemistry, where their compact three-dimensional architectures enhance molecular potency and selectivity.1 Among accessible precursors, azabicyclo[1.1.0]butanes (ABBs) serve as strain-activated azetidine surrogates whose electrophilic activation triggers ring opening and installs a quaternary carbon center.2 Advances in radical-mediated catalysis have made it possible to harness strain-release radical intermediates under mild, redox-neutral conditions for downstream C–C and C–heteroatom bond formation. In particular, strain-release of ABBs through radical or polar-radical relay pathways generates open-shell azetidinyl radicals that can be intercepted by π-systems3 or electrophiles,4 establishing ABB-derived radicals as programmable synthetic handles. Representative advances include the radical [2σ + 2π] cycloaddition to access 1-azabicyclo[2.1.1]hexanes (1-aza-BCHs),3c energy-transfer-mediated sulfonimination,4b and strain-release pentafluorosulfanylation.3d,4c,4d Despite this progress, the use of substituted 1,3-dienes as radical acceptors for azetidinyl radicals remains underexplored (Fig. 1A). Huang and co-workers elegantly demonstrated that photoinduced palladium catalysis enables a divergent synthesis of 1-azabicyclo[n.1.1]alkanes, in which radical addition of an azetidinyl radical to an aryl-substituted 1,3-diene is followed by intramolecular N-cyclization onto the resulting allyl fragment.3a This strategy provides streamlined access to structurally complex, saturated bridged bicyclic scaffolds that are otherwise difficult to assemble. By design, however, the intramolecular cyclization incorporates the nitrogen atom into the newly formed bridged ring, thereby precluding engagement of an external nitrogen nucleophile. By contrast, intermolecular trapping of the allyl intermediate with an external amine would offer a modular means of constructing medicinally relevant allylic amines bearing a quaternary azetidine, permit independent variation of the carbon and nitrogen components, and preserve an alkene moiety. Such a convergent transformation remains unrealized, principally because controlling selectivity of amines within the delocalized allyl intermediate is intrinsically challenging. Functionalization of ABBs with dienes. (A) Radical strain-release functionalization of ABBs with aryl substituted 1,3-dienes. (B) Four-component strain-release coupling of ABBs enabled by Cu/photoredox catalysis (this work). To address this gap, we turned to copper catalysis, which, particularly when merged with visible-light photoredox activation, provides a powerful platform for C(sp3)–heteroatom bond formation.5 Through single-electron transfer mediated by Ir photocatalytic cycle, CuII/CuIII manifolds capture and couple carbon-centered radicals, achieving high chemoselectivity even in complex molecular settings.6 This radical-capture reactivity is well suited to conjugated dienes, which, with their delocalized π-systems and wide availability as feedstock chemicals, are ideal partners for such constructions.7 Indeed, transition-metal-catalyzed aminative functionalization of dienes has emerged as an efficient route to allylic amines, forging C–C and C–N bonds with high atom and step economy and without preactivation of the carbon or nitrogen partners.8,9 These precedents suggested that photoredox-enabled copper catalysis could provide a viable means of merging radical generation with aminative diene functionalization. However, achieving reliable 1,4-selectivity in multicomponent reactions of substituted 1,3-dienes with amines remains challenging.9d Guided by these considerations, we developed a metallaphotoredox strategy for the strain-releasing activation of ABBs and regioselective 1,4-aminative difunctionalization of 1,3-dienes (Fig. 1B). Strain-releasing activation of an ABB generates an azetidinyl radical whose steric profile was expected to disfavor direct inner-sphere capture by copper and instead promote addition to the diene. Subsequent copper-mediated capture of the resulting allylic radical would furnish a π-allyl-copper intermediate, enabling C–N bond formation with an amine and sequential assembly of the multiple components.5i In conventional diene aminofunctionalization, amination often occurs at the terminus distal to the arene, preserving alkene–arene conjugation and favoring 1,2-difunctionalization.9c,i We hypothesized that the quaternary carbon center generated upon strain release would reshape the transition state for allylic C–N bond formation, shifting the balance of steric and noncovalent interactions sufficiently to favor the complementary 1,4-pathway.8f,10 Herein, we report a visible-light-driven copper metallaphotoredox four-component coupling that enables the 1,4-carboamination of aryl- and alkyl-substituted 1,3-dienes through strain-releasing activation of ABBs (Fig. 1B). In a single operation, an ABB, a conjugated diene, an arylamine, and an electrophilic activator are combined to furnish azetidines bearing a quaternary carbon center under mild, redox-neutral conditions. The transformation installs a C–C bond and a distal C–N bond across the diene with high 1,4-regioselectivity. The platform tolerates broad functional-group variation and accommodates diverse electrophilic ABB activators, including Boc2O, CbzCl, and sulfonyl chlorides, underscoring its modularity. By intercepting the diene-derived allyl intermediate intermolecularly rather than through intramolecular cyclization, this work delivers acyclic products bearing both an allylic amine and an intact quaternary azetidine, complementary to the bridged bicyclic frameworks of prior diene strain-release chemistry. Results and discussion Guided by this hypothesis, we began optimization with benzoyl ABB (1a), trans-1-phenyl-1,3-butadiene (2a), N-methylaniline (3a) and Boc2O (4a) as model substrates (Table 1). Systematic evaluation of copper sources, ligands, bases, and photocatalysts identified [Ir(dFMeppy)2dtbbpy]PF6 with CuBr/L1 under blue-LED irradiation as the optimal system, affording the allylic amine-containing azetidine 5a in 80% yield with excellent 1,4-regioselectivity (>20 : 1 rr, entry 1), Replacing L1 with other bidentate nitrogen ligands still furnished 5a, albeit in lower yield (entry 2). Although L1 provided the highest yield, it did not induce appreciable enantioselectivity under the standard conditions (<10% ee; see Table S4). Replacing CuBr with CuBr2 proved operative, albeit less efficient, affording 5a in 64% yield (entry 3). Substituting the halide additive (NaI or NaBr for LiBr) or omitting the base likewise eroded efficiency while retaining detectable reactivity (entries 4–5). Notably, omission of the Ir photocatalyst still delivered 5a in 12% yield, demonstrating that a photoexcited Cu(i) species generated in situ can independently drive azetidinyl radical formation, partially fulfilling the role of the Ir photocatalyst (entry 6).11 Control experiments confirmed that the copper catalyst and LiBr are both essential (entries 7–8), and no product was detected in the absence of light. Consistent with a photochemically initiated pathway, the reaction failed to proceed in the dark even at 80 °C, excluding thermal initiation (entry 9). Finally, the transformation remained effective under ambient atmosphere, underscoring its operational simplicity and robustness (entry 10). Table 1Optimization for the reaction conditionsa EntryDeviation from standard conditionYield (%)b1None80 (78)2L2, L3, L4 instead of L161/58/523CuBr2 instead of CuBr644NaI/NaBr instead of LiBr62/445Without base656Without photocatalyst127Without LiBrNR8Without Cu or lightNR9Dark condition, 80 °CNR10Ambient atmosphere54aReaction conditions: 1a (0.05 mmol), 2a (0.10 mmol), 3a (0.15 mmol), CuBr (5 mol%), L1 (6 mol%), [Ir(dFMeppy)2dtbbpy]PF6 (1 mol%), 4a (0.10 mmol), LiBr (0.10 mmol), K2CO3 (0.10 mmol) in 1,2-DCE (0.25 mL) at rt under irradiation with blue LEDs (440 nm) for 22 h under argon.bYields determined by 1H NMR spectroscopy with caffeine as an internal standard. Isolated yield in parentheses. Optimization for the reaction conditionsa Reaction conditions: 1a (0.05 mmol), 2a (0.10 mmol), 3a (0.15 mmol), CuBr (5 mol%), L1 (6 mol%), [Ir(dFMeppy)2dtbbpy]PF6 (1 mol%), 4a (0.10 mmol), LiBr (0.10 mmol), K2CO3 (0.10 mmol) in 1,2-DCE (0.25 mL) at rt under irradiation with blue LEDs (440 nm) for 22 h under argon. Yields determined by 1H NMR spectroscopy with caffeine as an internal standard. Isolated yield in parentheses. Having established the optimal conditions, we evaluated the generality of the method across a diverse range of aniline-derived nucleophiles (Table 2). N-Alkylated anilines bearing electron-donating or electron-withdrawing groups at the ortho-, meta-, or para-position uniformly furnished the corresponding azetidine products (5a–5m). The reaction was readily scaled, providing 5a on 1 mmol scale in 65% isolated yield. N-methylanilines with para-methyl (5b), trifluoromethyl (5c), trifluoromethoxy (5d) were compatible with this catalysis. Halogen substituents were well tolerated, with para-chloro (5e), bromo (5f), and iodo (5g), as well as ortho-fluoro (5h) and bromo (5i) groups all retained. Table 2Scope of aryl amines and electrophilesa aReaction conditions: 1a (0.05 mmol), 2a (0.10 mmol), 3 (0.15 mmol), CuBr (5 mol%), L1 (6 mol%), [Ir(dFMeppy)2dtbbpy]PF6 (1 mol%), 4 (0.10 mmol), LiBr (0.10 mmol), K2CO3 (0.10 mmol) in 1,2-DCE (0.25 mL) at rt using blue LEDs (440 nm) for 22 h under Ar atmosphere.bRegioisomeric ratio of 1,4 : 1,2 adducts.c[Ir(dFCF3ppy)2dtbbpy]PF6 (1 mol%).d4 (0.05 mmol), [Ir(dFCF3ppy)2dtbbpy]PF6 (1 mol%).e4 (0.05 mmol) was pre-mixed with 1a and LiBr (0.10 mmol) for 4 h prior to addition of the remaining reagents (see Table S2). Isolated yields. Scope of aryl amines and electrophilesa Reaction conditions: 1a (0.05 mmol), 2a (0.10 mmol), 3 (0.15 mmol), CuBr (5 mol%), L1 (6 mol%), [Ir(dFMeppy)2dtbbpy]PF6 (1 mol%), 4 (0.10 mmol), LiBr (0.10 mmol), K2CO3 (0.10 mmol) in 1,2-DCE (0.25 mL) at rt using blue LEDs (440 nm) for 22 h under Ar atmosphere. Regioisomeric ratio of 1,4 : 1,2 adducts. [Ir(dFCF3ppy)2dtbbpy]PF6 (1 mol%). 4 (0.05 mmol), [Ir(dFCF3ppy)2dtbbpy]PF6 (1 mol%). 4 (0.05 mmol) was pre-mixed with 1a and LiBr (0.10 mmol) for 4 h prior to addition of the remaining reagents (see Table S2). Isolated yields. Anilines substituted with ortho-methyl (5j), meta-phenyl (5k), meta-methyl (5l), and meta-trifluoromethoxy (5m) were efficiently transformed. A 3,4-disubstituted N-methylaniline was likewise competent, delivering 5n. Cyclic secondary anilines, including a substituted tetrahydroquinoline (5o) and a benzomorpholine (5p), engaged smoothly, as did anilines bearing pendant ester (5q), phenyl (5r), and cyano (5s) groups on the N-alkyl chain, demonstrating broad tolerance to diverse substitution patterns. Reactivity diminished for an N-isopropyl aniline (5t), likely owing to steric hindrance from the bulky isopropyl group, which presumably retards efficient ligand exchange. Primary arylamines were also competent coupling partners, albeit with reduced efficiency: aniline, p-toluidine, and p-chloroaniline afforded 5u, 5v, and 5w in 35%, 23%, and 40% yield, respectively. The diminished reactivity may arise from altered copper coordination and catalyst speciation associated with the less sterically encumbered primary arylamines. We next examined ABB activation by electrophiles other than Boc2O (Table 2). Both CbzCl and sulfonyl chlorides proved effective, enabling direct diversification of the azetidine nitrogen substituent while preserving excellent 1,4-regioselectivity throughout. CbzCl furnished the carbamate-protected product 5aa. Sulfonyl chlorides bearing para-methyl (5ab), para-tert-butyl (5ac), and para-methoxy (5ad) groups delivered the corresponding sulfonamide-containing azetidines. The sterically encumbered mesitylenesulfonyl chloride was equally tolerated, affording 5ae, and 2-thiophenesulfonyl chloride engaged as a heteroaryl variant (5af). A celecoxib-derived sulfonyl chloride was likewise competent, delivering 5ag. Subsequently, we applied the carboamination protocol to a range of ABB derivatives (Table 3). Substituents bearing electron-donating or electron-withdrawing groups at various positions of the benzoyl fragment were tolerated throughout. ABBs bearing para-methyl (5ba) and para-tert-butyl (5bb) reacted in good to excellent yields, and the strongly electron-donating para-NMe2 (5bc) and para-OMe (5bd) groups participated efficiently to afford products. Halogenated ABBs were equally well incorporated, delivering fluoro (5be), chloro (5bf), and trifluoromethyl (5bg) products in excellent yields. Positional variation was likewise accommodated, with ortho-methyl (5bh), ortho-chloro (5bi), meta-methyl (5bj), meta-chloro (5bk), and meta-OCF3 (5bl) products obtained. Naphthyl (5bm) and thiophenyl (5bn) variants proceeded smoothly with exclusive 1,4-selectivity. Finally, aliphatic acyl-substituted ABBs, cyclobutyl (5bo), cyclohexyl (5bp), and neopentyl (5bq), were viable. Table 3Scope of ABBs and dienesa aReaction conditions: 1 (0.05 mmol), 2 (0.10 mmol), 3a (0.15 mmol), CuBr (5 mol%), L1 (6 mol%), [Ir(dFMeppy)2dtbbpy]PF6 (1 mol%), 4a (0.10 mmol), LiBr (0.10 mmol), K2CO3 (0.10 mmol) in 1,2-DCE (0.25 mL) at room temperature using blue LEDs (440 nm) for 22 h under Ar atmosphere.bRegioisomeric ratio of 1,4 : 1,2 adducts.c[Ir(dFCF3ppy)2dtbbpy]PF6 (1 mol%). Isolated Yields. Scope of ABBs and dienesa Reaction conditions: 1 (0.05 mmol), 2 (0.10 mmol), 3a (0.15 mmol), CuBr (5 mol%), L1 (6 mol%), [Ir(dFMeppy)2dtbbpy]PF6 (1 mol%), 4a (0.10 mmol), LiBr (0.10 mmol), K2CO3 (0.10 mmol) in 1,2-DCE (0.25 mL) at room temperature using blue LEDs (440 nm) for 22 h under Ar atmosphere. Regioisomeric ratio of 1,4 : 1,2 adducts. [Ir(dFCF3ppy)2dtbbpy]PF6 (1 mol%). Isolated Yields. We next investigated the scope of the 1,3-diene partner. Phenyl-1,3-dienes bearing diverse electronic and positional substituents proved competent, furnishing the desired azetidines in satisfactory yields: para-methyl (5ca), para-bromo (5cb), para-CO2Me (5cc), and para-CF3 (5cd), together with ortho-methyl (5ce) and meta-methyl (5cf), all reacted smoothly. The reaction extended to a sterically hindered 1-naphthyl diene (5cg) and to aliphatic conjugated dienes (5ch, 5ci), delivering the 1,4-adducts with high regiocontrol. To further demonstrate the versatility of this carboamination, we pursued complex molecule diversification using pharmaceutically derived substrates. Probenecid (5br), febuxostat (5bs), flurbiprofen (5cj), and gemfibrozil (5ck) were incorporated smoothly despite their structurally intricate motifs, furnishing the corresponding drug–azetidine conjugates while consistently maintaining high 1,4-regioselectivity. These results establish the methodology as an efficient, modular platform for diversifying bioactive molecules and expanding accessible chemical space in medicinal chemistry. To elucidate the reaction mechanism, we conducted a series of experimental studies (Fig. 2). First, subjecting the independently synthesized 3-brominated azetidine 1aa to the standard conditions furnished the carboamination product 5a, establishing 1aa as a competent intermediate (Fig. 2A). Consistent with this, a radical-clock experiment using (2-(buta-1,3-dien-1-yl)cyclopropyl)benzene (2aa) in place of the diene afforded the ring-opened product 6a, corroborating the involvement of radical species (Fig. 2B). Stern–Volmer luminescence quenching experiments revealed that CuBr efficiently quenches the excited-state Ir photocatalyst, whereas 1aa elicits only negligible quenching, indicating that the photoexcited Ir catalyst oxidizes the Cu(i) species rather than generating the azetidinyl radical directly, consistent with a reductive quenching pathway (Fig. 2C). Addition of the radical scavenger TEMPO (2 equiv) under the standard conditions completely suppressed product formation; instead, the TEMPO-azetidine adduct 6b was isolated in 86% yield, with its exact mass confirmed by HRMS (Fig. S2 in the SI), providing strong evidence for the intermediacy of an azetidinyl radical (Fig. 2D). Of note, in the absence of 2a, the Ir/Cu-mediated reaction under standard conditions afforded no direct coupling product between 1a and 3a (see Scheme S10 in the SI), indicating that copper preferentially captures the diene-derived allylic radical rather than the azetidinyl radical itself. Experimental mechanistic studies. (A) Intermediate experiment with bromoazetidine. (B) Radical-clock experiment with cyclopropyl diene. (C) Stern–Volmer quenching experiments. (D) Radical inhibitor experiment with TEMPO. (E) UV-vis spectroscopy of the copper complexes. (F) Further transformations of 5a. To probe the proposed ligand-exchange step leading to the Cu(ii)–amido complex C,12 we performed UV-vis absorption spectroscopy using CuBr2 as a surrogate for the putative Cu(ii) intermediate (Fig. 2E). A mixture of CuBr2 and L1 exhibited an absorption band at approximately 420 nm, consistent with formation of the Cu(ii) complex B. Subsequent addition of 3a induced a bathochromic shift both in the presence and absence of base (from the red trace to the green and purple traces in Fig. 2E), consistent with ligand exchange to form the Cu(ii)–amido complex C.13 This spectral response is also consistent with previous reports.6c,8e,13 An alternative sequence, in which ligand exchange precedes single-electron transfer, cannot be excluded: replacing CuBr2 with CuBr likewise produced a bathochromic shift upon addition of 3a (Fig. S7 in the SI). With scale-up established, we examined post-synthetic derivatization of the carboamination products to further showcase the synthetic utility of the Ir/Cu-catalyzed platform (Fig. 2F). Boc removal from 5a proceeded cleanly on exposure to TMSOTf and 2,6-lutidine within 1 h, affording the free N–H azetidine 6c as a versatile handle for further N-functionalization. Wittig olefination of the benzoyl ketone in 5a delivered the terminal olefin 6d, and reduction of the same ketone furnished the corresponding alcohol 6e. Collectively, these transformations highlight the azetidine products as versatile intermediates for accessing structurally diverse frameworks. At the outset, we considered whether N-methylaniline might attack a Cu(iii) intermediate as a nucleophile through an outer-sphere pathway to forge the C–N bond.14 To test this, we examined the reaction kinetics by varying the concentration of N-methylaniline under initial-rate conditions. The observed zeroth-order dependence on N-methylaniline indicates that the aniline is not kinetically involved in the rate-determining step,15 making an outer-sphere nucleophilic attack mechanism unlikely (Fig. S9 in the SI). Instead, in light of precedent for C–N bond formation by reductive elimination from Cu(iii) intermediates3a,16 and the well-documented facility of reductive elimination in copper catalysis,17 we propose that the C–N bond is forged through reductive elimination from a Cu(iii) intermediate. A plausible catalytic cycle, consistent with these mechanistic studies and with precedents for copper-catalyzed carboamination,3a,8g,16 is outlined in Fig. 3A. The Cu(i) complex A reductively quenches photoexcited Ir(iii)* (E1/2 = +0.97 V vs. SCE in MeCN),18 generating Ir(ii) and the Cu(ii) complex B. Ligand exchange with the N-nucleophile then furnishes the Cu(ii)–amido species C. Meanwhile, LiBr/Boc2O-promoted, strain-releasing cleavage of the central C–N bond of the ABB affords the Boc–ABB adduct E. Single-electron reduction of E by Ir(ii) subsequently generates the azetidinyl radical F and regenerates ground-state Ir(iii). Regioselective addition of F to the terminal carbon of the diene furnishes the allyl radical G, which is intercepted by C to deliver a Cu(iii) species D;19 reductive elimination from D forges the C(sp3)–N bond, releasing the 1,4-carboamination product and regenerating Cu(i) A to close the dual cycle. Proposed reaction mechanism and DFT calculations at the M06/def2-TZVP/LANL2DZ/SMD(1,2-dichloroethane)//PBE0-D3(BJ)/def2-SVP/LANL2DZ level of theory. (A) Proposed catalytic cycle. (B) IGMH analysis of the selectivity-determining transition states. (C) Computed free energy profiles. To pinpoint the origin of the 1,4-selectivity, we examined this sequence by density functional theory (DFT); the computed intermediates Int1–Int4, transition states TS1–TS4, and products Pdt1/Pdt2 correspond to the radical-addition and C–N-forming stages of the cycle above (F → G → D → product). Gibbs free energies are referenced to the azetidinyl radical Int1 (0.0 kcal mol−1, corresponding to F). Radical addition to the terminal carbon (C1) of diene 2a proceeds through TS1 with a low barrier (ΔG‡ = 11.6 kcal mol−1) to give the allyl radical Int2 (ΔG = −13.4 kcal mol−1), whereas addition to the internal carbon (C2) viaTS2 is far higher (ΔG‡ = 24.1 kcal mol−1), making terminal addition strongly preferred. Capture of Int2 at C4 by the Cu(ii)–amido complex Cu1 affords the Cu(iii) complex Int3 in a slightly endergonic step (ΔG = +7.0 kcal mol−1),20 from which reductive elimination through the three-membered transition state TS3 (ΔG‡ = 7.3 kcal mol−1) releases the observed 1,4-product Pdt1 and regenerates LnCuBr. The competing 1,2-pathway—capture at C2 to give Int4, then reductive elimination viaTS4 (ΔG‡ = 12.6 kcal mol−1) to Pdt2—is disfavored: TS4 lies 5.3 kcal mol−1 above TS3, accounting for the high 1,4-selectivity (>20 : 1 rr). TS4 is destabilized by steric repulsion between the aniline N-methyl group and C–H hydrogens of the diene and Boc group, manifested by the shorter H⋯H contacts (2.035 and 2.263 Å; Fig. 3C) than in TS3 (2.285 and 2.410 Å; SI Fig. S11)—a more congested arrangement imposed by the azetidine substituent at the diene terminus, consistent with ΔΔG‡ = 5.3 kcal mol−1. Independent gradient model based on Hirshfeld partition (IGMH) analysis (Fig. 3B and S12 in the SI) further revealed stabilizing C–H⋯π and π⋯π contacts in TS3, the latter arising from favorable alignment between the aniline aromatic ring and the allyl π-system at C2; in TS4, only C–H⋯π contacts were found, with the π⋯π interaction absent. The 1,4-selectivity thus arises from the combined effect of stabilizing noncovalent interactions in TS3 and pronounced steric repulsion in TS4. Conclusions In summary, we have developed a visible-light, copper-catalyzed four-component 1,4-carboamination of 1,3-dienes powered by strain-release activation of ABBs. Photogenerated azetidinyl radicals add to conjugated dienes; the resulting allylcopper intermediate then undergoes Cu(iii)-mediated C(sp3)–N bond formation in a single operation that installs both an allylic amine and a quaternary carbon center on a densely functionalized azetidine. The reaction tolerates diverse anilines, ABBs, and dienes under mild, redox-neutral conditions, and interchangeable activators enable programmable N-appendage diversification. DFT identifies noncovalent interactions and steric repulsion in the competing transition states as the origin of the high 1,4-regioselectivity. By intercepting the diene-derived allyl intermediate intermolecularly, this strategy delivers acyclic, sp3-rich architectures pairing an allylic amine with an intact quaternary azetidine, complementary to existing ABB–diene chemistry and provides a versatile route to medicinally relevant nitrogen scaffolds. Author contributions Y. K., D. L. and S. H. conceived the idea of the project. Y. K., D. L. and J. P. performed the experiments, and analyzed the data. All authors wrote the manuscript. Conflicts of interest There are no conflicts to declare. Supplementary Material