Synthesis of 1,4-aminoalcohols from substituted dienes and anilines via molybdooxaziridine catalysis

Abstract

Herein we report the MoODipic(HMPA)ONAr promoted 1,4-aminohydroxylation of dienes using anilines as the quantitative source of N. The transformation involves the formation of the active catalyst Ar–N-molybdooxaziridine 1a, the formation of Ar–NO, a hetero-Diels–Alder reaction, and then reductive cleavage to obtain 1,4-aminoalcohols. Initial kinetic and mechanistic data indicate that Ar–NO formation is the rate-determining step and provides evidence for the proposed catalytic cycle.

---

1,4-Aminoalcohols are ubiquitous structural motifs in many natural products and valuable building blocks in organic synthesis.¹ Their applications extend to the development of functional organic materials and their use as ligands in transition-metal catalysis.² While the synthesis of aminoalcohols has received a lot of attention, methods for the synthesis of 1,4-aminoalcohols are more limited. Moreover, these often rely on strong acidic conditions and oxidizing agents, which limit their synthetic applicability.³ On this front, efficient 1,4-aminooxygenation of 1,3-dienes through Pd catalysis has been reported.⁴ These can also be accessed through the addition of Grignard's reagent to substituted aminotetrahydrofurans,⁵ the Lewis acid-promoted reaction of cyclic N,O-acetals with silyl enol ethers,⁶ the Mannich reactions of α-thiocarbonyls,⁷ and the addition of amines to substituted hetero-dienes.⁸ The sequential nitroso Diels–Alder (NDA), reductive N–O-bond cleavage represents a successful route to the synthesis of 1,4-aminoalcohols.⁹ However, the complex nature of nitroso compound cycloadditions and the lack of efficient access to substituted nitroso compounds due to harsh methods, instability and challenging purification have severely limited this approach.¹⁰ Nitroso compounds are also valuable chemical intermediates in various biological metabolic processes and synthetic chemical transformations.¹¹ Their synthetic applications are often hindered because C-nitroso compounds are unstable and often dimerize to produce azo and/or azoxy arenes through condensation reactions.¹² Therefore, the development of catalytic conditions for the synthesis and cycloaddition of nitroso compounds from simple substrates is of high significance. Metallooxaziridines are organometallic complexes that are derived from the reaction of metal-oxides and N-transfer reagents.¹³ These metalloheterocycles have been reported to function as transfer reagents across different π-systems. Early efforts by Sharpless and Nicholas demonstrated that N–Ar molybdooxaziridines achieve N-transfer across alkyl alkenes to produce N–Ar allylic amines.¹⁴ Recently, we have discovered that metallooxaziridines (Zr, V, W, and Mo) can achieve aziridination, amination, oxyamination, and sulfonation reactions across all types of alkenes.¹⁵ Thus, the potential for novel chemical pathways across metallooxaziridines containing other transition metals is of great significance. These complexes can also serve as R–NO precursors, which under specific chemical environments trigger the formation of highly reactive nitroso compounds.¹⁶ Thus, metallooxaziridines can serve as a catalytic source of highly reactive R–NOs and as such exploit their synthetic value.

Thus, we envisioned creating a reaction that achieves the synthesis of 1,4-aminoalcohols from anilines and dienes by designing a well-defined catalytic system for a highly efficient nitrosoarene Diels–Alder cycloaddition, followed by reductive cleavage in a single operation. We started by identifying catalytic reaction conditions for the tandem nitrosoarene Diels–Alder/reductive cleavage process from dienes and anilines (Table 1). The initial discovery with 10 mol% of MoODipic(HMPA)O₂ 1b, H₂O₂/H₂O, and a 1 : 1 ratio of aniline and cyclohexadiene, followed by treatment with AcOH in H₂O and Zn powder, provided 4a in good yield as a single diastereomer (84% yield, entry 1). Different solvents showed that polar and non-protic environments enhanced the reaction productivity, with 1,4-dioxane affording an 89% yield (entries 2–6). Due to the biphasic nature of the reductive cleavage, we attempted to employ homogenous conditions, but none provided equally high yields (entries 7–9). We were concerned that the high equivalency of water hindered the overall productivity due to competing oxidative pathways, and we tested different terminal oxidants without much improvement (entries 10 and 11). Eventually, we found that H₂O₂·urea elevated the productivity, and we were able to obtain 4a in 94% (entry 12). The product still required a final purification step, but at 5 mol% trace impurities from competing oxidative pathways disappeared, and the yield increased to 99% without a final purification step (entry 13).

Table 1 Reaction optimization

Entry	Catalyst (mol%)	Oxidant	Solvent	Reduction	Yield^a^b (%)
a Isolated yields.b ArNH2 (1 equiv.) in 1,4-dioxane was mixed with H2O2·urea (2.2 equiv.), MgSO4 (2.2 equiv.), and MoODipic(HMPA)O2 1b, and then stirred at rt for 5 min. Diene (1 equiv.) was added and the reaction mixture was stirred for 16 h. The crude was filtered through Celite and the solvent was removed and reconstituted in CH2Cl2, then AcOH (10 equiv.), H2O and Zn (40 equiv.) were added and the mixture was vigorously stirred for 2 h. Dipic = dipicolinic acid.c H2O2 30% in H2O.d H2O2 35% in urea.
1	10	H2O2^c	CH3CN	Zn, AcOH, H2O/CH2Cl2	84
2	10	H2O2	CH3CN/H2O	Zn, AcOH, H2O/CH2Cl2	79
3	10	H2O2	Toluene	Zn, AcOH, H2O/CH2Cl2	88
4	10	H2O2	Chlorobenzene	Zn, AcOH, H2O/CH2Cl2	80
5	10	H2O2	DMA	Zn, AcOH, H2O/CH2Cl2	82
6	10	H2O2	1,4-Dioxane	Zn, AcOH, H2O/CH2Cl2	89
7	10	H2O2	1,4-Dioxane	Cu/C, Na2CO3, H2O	64
8	10	H2O2	1,4-Dioxane	Mo(CO)6, CH3CN, 70 °C	75
9	10	H2O2	1,4-Dioxane	In(OTf)3, CH3CN, 80 °C	64
10	10	tBuOOH	1,4-Dioxane	Zn, AcOH, H2O/CH2Cl2	53
11	10	AcOOH	1,4-Dioxane	Zn, AcOH, H2O/CH2Cl2	64
12	10	H2O2·urea^d	1,4-Dioxane	Zn, AcOH, H2O/CH2Cl2	94
13	5	H2O2·urea	1,4-Dioxane	Zn, AcOH, H2O/CH2Cl2	99
14	1	H2O2·urea	1,4-Dioxane	Zn, AcOH, H2O/CH2Cl2	71

---

Based on the results observed for aniline and 1,3-cyclohexadiene, we then focused on investigating the generality across anilines with electron-donating groups (EDGs) and electron-withdrawing groups (EWGs) as a means of activating or deactivating the reaction pathway (Table 2). 4-Methyl-aniline afforded a similar yield at a faster observable rate for the cycloaddition (97%, 6 h, entry 2). Moreover, 3-methyl and 2-methyl provided the corresponding 1,4-aminoalcohols in very good yields, but the yield was slightly slower in the case of 2-methyl (entries 3 and 4). 4-OMe, 3-OMe, and 2-OMe reacted very fast (∼4 h for both) with very good productivities (entries 5–7). Efforts to expand the scope with other EDGs found that 2-phenyl and 3-ethynyl aniline reacted in great yields and at slightly slower reaction rates (10 and 12 h, entries 8 and 9). To our surprise, the EDG ability in 2-phenyl aniline completely overrode any potential steric hindrance on the reaction kinetics. We anticipated that halogenated anilines would react with equal efficiencies but at slower rates and we found that 4-fluoro, 4-chloro, 4-bromo, and 4-iodo reacted with high yields but slower observable rates (24 h, 26 h, 25 h, and 25 h, respectively; entries 10–13). 3-Bromo, 2-bromo, 2-iodo, and 3-chloro-4-fluoro all proceeded with high yields and similar reaction rates (∼32 h, entries 14–17).

Table 2 Reaction scope

Entry
		Aniline	Product		Yield^a^b (%)
a Conditions: ArNH2 (1 equiv.) in 1,4-dioxane was mixed with H2O2·urea (2.2 equiv.), MgSO4 (2.2 equiv.), and MoODipic((HMPA)O2 1b (5 mol%) and stirred for 5 min. Diene (1 equiv.) was added and the reaction mixture was stirred for 16 h. The crude was filtered through Celite and the solvent was removed and reconstituted in CH2Cl2, then AcOH (10 equiv.), H2O and Zn (40 equiv.) were added and stirred for 2 h.b Isolated yields without silica gel purification.c The crude reaction mixture from NDA was dissolved in CH3CN/H2O (4 : 1) and Mo(CO)6 (1.1 equiv.) was heated at 80 °C for 4 h.
1	3a			4a	99
2	3b		R = 4-methyl	4b	97
3	3c		R = 3-methyl	4c	98
4	3d		R = 2-methyl	4d	96
5	3e		R = 4-MeO	4e	98
6	3f		R = 3-MeO	4f	95
7	3g		R = 2-MeO	4g	91
8	3h		R = 2-Ph	4h	93
9	3i		R = 3-ethynyl	4i	95
10	3j		R = 4-fluoro	4j	98
11	3k		R = 4-fluoro	4k	96
12	3l		R = 4-bromo	4l	97
13	3m		R = 4-iodo^c	4m	93
14	3n		R = 3-bromo	4n	88
15	3o		R = 2-bromo	4o	90
16	3p		R = 2-iodo	4p	89
17	3q		R = 3-chloro-4-fluoro	4q	95
18	3r		R = 4-nitro^c	4r	86
19	3s		R = 3-nitro^c	4s	80
20	3t		R = 2-nitro^c	4t	82

---

The scope was completed with the study of anilines with EWGs: 4-nitro, 3-nitro, and 2-nitro aniline provided the expected 1,4-aminoalcohols in lower yields with a required final purification step (entries 18–20). The initial cycloaddition required heat for 24 h (60 °C) to achieve full conversion and under reductive conditions (Mo(CO)₆ in CH₃CN at 80 °C to avoid NO₂ group reduction) the respective 1,4-aminoalcohols were obtained. These results provide a clear indication that the rate-determining step may involve the initial molybdooxaziridine-mediated nitrosoarene formation. The reaction across different anilines proved to be very successful and kinetic data indicate a clear reaction acceleration with EDGs and reaction deceleration with EWGs. While the reactions with EDG-bound anilines afforded great yields without requiring purification, nitro-anilines provided more complex mixtures and required a purification step.

We then set out to test different dienes, with special emphasis on non-symmetrical dienes (Table 3). NDA reactions have shown different degrees of regioselectivity; while these reactions are known to proceed through a concerted but asynchronous mechanism, the regioselectivity trends are less well-understood.¹⁷ 1-Substituted dienes are expected to have proximal-selectivity, while 2-substituted dienes are expected to provide the distal isomer. 1,3-Cyclooctadiene provided the corresponding 1,4-aminoalcohol in great yield (entry 1). We aimed to investigate the steric biases across polysubstituted dienes and found that α-terpinene and 1,3,5,5-tetramethyl-1,3-cyclohexadiene reacted with complete regioselectivity to produce the proximal-regioisomers with the N–Ph group on the least hindered side (entries 2 and 3).

Table 3 Reaction scope

Entry		Aniline	Product		Yield^a (%)
a Conditions: ArNH2 (1 equiv.) in 1,4-dioxane was mixed with H2O2·urea (2.2 equiv.), MgSO4 (2.2 equiv.), and MoODipic((HMPA)O2 1b (5 mol%), and stirred for 5 min. Diene (1 equiv.) was added and the reaction mixture was stirred for 24 h. The crude was filtered through Celite and the solvent was removed and reconstituted in CH2Cl2, then AcOH (10 equiv.), H2O and Zn (40 equiv.) were added and stirred for 2 h.b Isolated yields without silica gel purification.c The crude reaction mixture from NDA was dissolved in CH3CN/H2O (4 : 1) and Mo(CO)6 (1.1 equiv.) was heated at 80 °C for 4 h.
1	2b			4u	91^b
2	2c			4v	88
3	2d			4w	89
4	2e			4x	79
5	2f			4y	90^b
6	2g			4z	95^b
7	2h			4aa	96^b
8	2i			4ab	91^b
9	2j			4ac	82
10	2k			4ad	84
11	2l			4ae	88
12	2m			4af	71^c
14	2n			4ag	64^c

---

Moreover, optically pure α-phellandrene also reacted with high regioselectivity, but 1,4-aminoalcohols were obtained as a mixture of diastereomers (7 : 1, entry 4). 1,3-Pentadiene was also successful in providing the expected proximal-isomer (entry 5). 1,3-Hexadiene, 2-methyl-1,3-butadiene (selective for the distal-isomer), and 2,3-dimethylbutadiene all reacted in great yields without final purification (entries 6–8). 1,4-Diphenyl-1,3-butadiene reacted with great efficiency, thus proving that highly sterically demanding dienes do not affect NDA reaction kinetics (entry 9). Myrcene also reacted selectively to provide the distal-regioisomer (entry 10), while 1,3-hexadienol provided 1,4-aminoalcohols 4ae and 4ae′ as a 55 : 45 mixture of regioisomers (entry 11). Dienes are also recurrent among some naturally occurring steroids. We found that ergosterol and vitamin D3 reacted with high efficiencies to provide the corresponding 1,4-aminoalcohols as single isomers in very good yields (these two required purifications post reductive cleavage; entries 12 and 13). Ergosterol reacted with surprisingly high selectivity for the proximal-isomer (71% yield), with no other identifiable cycloadduct impurities.¹⁸ Moreover, the vitamin D3 NDA reactivity pattern displays a clear preference for its terminal diene. Although a complex mixture was obtained, no other NDA products were isolated.

Our previous studies on the functionalization of alkenes via metallooxaziridine catalysis has helped design the mechanistic experiments for this new transformation. The proposed catalytic cycle (Fig. 1) starts with fast addition of aniline to molybdodioxirane 1b to form molybdooxaziridine 1a. This step is followed by fast diene–HMPA exchange and then the H₂O₂-mediated rate-determining step nitrosoarene extrusion. The resulting DipicMoO₂(diene) complex undergoes NDA with nitrosoarene and the resulting DipicMoO₂(HMPA) 1c undergoes H₂O₂-mediated oxidation to molybdodioxirane 1b. The oxazine intermediate is then quantitatively reduced to the corresponding 1,4-aminoalcohol 4. Based on the observed conversion rates and selectivities, the reaction exhibits first order kinetics on aniline, zeroth order kinetics for diene and saturation kinetics for H₂O₂ (details in the ESI†). Control experiments show that the reaction has no conversion in the absence of 1b. However, the stoichiometric reaction with 1a provides very low conversion at rt, but the presence of 1 equivalent of H₂O₂ enables full conversion. Moreover, the stoichiometric reaction with 1a and an equivalent of DipicMoO₂(HMPA) 1c at 50 °C also achieves conversion, highlighting the high rate of molybdenum oxide dimerization.¹⁹ The lack of biradical intermediates up to the NDA step was also confirmed when reaction conversion to the oxazine intermediate was not hindered in the presence of radical scavengers.

Fig. 1 Proposed catalytic cycle.

The proposed catalytic cycle is further supported by a Hammett correlation study employing 4-substituted anilines (3a, b, j, k, l). The results show a ρ-value of −1.62 and hence demonstrate enhanced reactivity for anilines with EDGs as a positive charge develops on the nitrosoarene N atom in the transition state (details in the ESI†). Moreover, these results are also in agreement with a nitrosoarene extrusion as the rate-determining step for this catalytic cycle.

In summary, this work reports a novel method to selectively access 1,4-aminoalcohols from substituted anilines and dienes. The reaction works with high efficiency and stereoselectivity for dienes with diverse substitution patterns and for anilines with a variety of functional groups. The proposed catalytic cycle involves the formation of the active catalyst molybdooxaziridine 1a, which releases nitrosoarene for the highly selective NDA to form an oxazine intermediate that then undergoes quantitative reductive cleavage to provide 1,4-aminoalcohol 4. Further experiments to better understand and fully characterize all mechanism intermediates are ongoing, and a follow-up manuscript is in preparation.

This material is based upon work supported by the U. S. National Science Foundation CAREER and MRI awards under grant numbers CHE-1752085 and CHE-1827938.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data (including the experimental protocols, kinetic studies, and spectroscopic characterization of the catalysts and 1,4-aminoalcohols) supporting this article have been included as part of the ESI.†

Notes and references

1. (a) M. Sugiura, H. Hagio, R. Hirabayashi and S. Kobayashi, J. Am. Chem. Soc., 2001, 123, 12510; (b) T. Hudlicky, D. A. Entwistle, K. K. Pitzer and A. J. Thorpe, Chem. Rev., 1996, 96, 1195; (c) T. J. Donohoe, C. K. A. Callens, A. R. Lacy and C. Winter, Eur. J. Org. Chem., 2012, 655; (d) T. Hudlicky, Chem. Rev., 1996, 96, 3.
2. (a) D. J. Diaz, K. G. Hylton and L. McElwee-White, J. Org. Chem., 2006, 71, 734; (b) M. Genov, K. Kostova and V. Dimitrov, Tetrahedron: Asymmetry, 1997, 8, 1869; (c) P. P. Singh, S. Gudup, S. Ambala, U. Singh, S. Dadhwal, B. Singh, S. D. Sawant and R. A. Vishwakarma, Chem. Commun., 2011, 47, 5852; (d) R. Mansueto, V. Mallardo, F. M. Perna, A. Salomone and V. Capriati, Chem. Commun., 2013, 49, 10160; (e) F. C. Sassone, F. M. Perna, A. Salomone, S. Florio and V. Capriati, Chem. Commun., 2015, 51, 9459.
3. (a) D. Scarpi, F. Lo Galbo and A. Guarna, Tetrahedron: Asymmetry, 2006, 17, 1409; (b) M. Knollmuller, M. Ferencic and P. Gartner, Tetrahedron: Asymmetry, 1999, 10, 3969.
4. J.-E. Bäckvall, J.-E. Nystrom and R. E. Nordberg, J. Am. Chem. Soc., 1985, 107, 3676.
5. C. Tejo, Y. F. A. See, M. Mathiew and P. W. H. Chan, Org. Biomol. Chem., 2016, 14, 844.
6. C. Tejo, X. R. Sim, B. R. Lee, B. J. Ayers, C.-H. Leung, D.-L. Ma and P. W. H. Chan, Molecules, 2015, 20, 13336.
7. T. Kano, R. Sakamoto and K. Maruoka, Chem. Commun., 2014, 50, 942.
8. R. Fernandez de la Pradilla, M. Velado, I. Colomer, C. Simal, A. Viso, H. Gornitzka and C. Hemmert, Org. Lett., 2014, 16, 5200.
9. (a) P. Drelich, M. Moczulski and Ł. Albrecht, Org. Lett., 2017, 19, 3143; (b) H. Lin, Y. Tan, X.-W. Sun and G.-Q. Lin, Org. Lett., 2012, 14, 3818; (c) L. Werner, J. R. Hudlicky, M. Wernerova and T. Hudlicky, Tetrahedron, 2010, 66, 3761.
10. (a) M. G. Memeo and P. Quadrelli, Chem. Rev., 2017, 117, 2108; (b) S. M. M. Lopes, A. L. Cardoso, A. Lemos and T. M. V. D. Pinho de Melo, Chem. Rev., 2018, 118, 11324; (c) B. Maji and H. Yamamoto, J. Am. Chem. Soc., 2015, 137, 15957; (d) J. Pous, T. Courant, G. Bernadat, B. I. Iorga, F. Blanchard and G. Masson, J. Am. Chem. Soc., 2015, 137, 11950.
11. (a) H. Song, J. Wei, Z. Wang, Y. Liu, S. Zhao, X. Cai, Y. Xiao, L. Yang, P. Bai, L. Gan, F. Yang, S. Zheng, W. Zhang, J. Pan and C. Xu, ACS Catal., 2024, 14, 12372; (b) S. Fountoulaki, P. L. Gkizis, T. S. Symeonidis, E. Kaminioti, A. Karina, I. Tamiolakis, G. S. Armatas and I. N. Lykakis, Adv. Synth. Catal., 2016, 358, 1500; (c) C. P. Frazier, A. Bugarin, J. R. Engelking and J. Read de Alaniz, Org. Lett., 2012, 14, 3620; (d) C.-N. Chen and R.-S. Liu, Angew. Chem., Int. Ed., 2019, 58, 9831.
12. (a) B. G. Gowenlock and G. B. Richter-Ado, Chem. Rev., 2004, 104, 3315; (b) Y. Gao, S. Yang, W. Xiao, J. Nie and X.-Q. Hu, Chem. Commun., 2020, 56, 13719; (c) A. P. Lightfoot, R. G. Pritchard, H. Wan, J. E. Warren and A. Whitting, Chem. Commun., 2002, 2072.
13. (a) L. S. Liebeskind, K. B. Sharpless, R. D. Wilson and J. A. Ibers, J. Am. Chem. Soc., 1978, 100, 7061; (b) D. A. Muccigrosso, S. E. Jacobson, P. A. Apgar and F. Mares, J. Am. Chem. Soc., 1978, 100, 7063.
14. (a) R. S. Srivastava and K. M. Nicholas, J. Org. Chem., 1994, 59, 5365; (b) E. R. Moller and K. A. Jorgensen, J. Am. Chem. Soc., 1993, 115, 11814.
15. (a) R. Madiu, E. L. Doran, J. M. Doran, A. A. Pinarci, K. Dhillon, D. A. Rivera, A. M. Howard, J. L. Stroud, D. A. Moskovitz, S. J. Finneran, A. N. Singer, M. E. Rossi and G. Moura-Letts, J. Org. Chem., 2024, 89, 4001; (b) R. Madiu, B. Dellosso, E. L. Doran, J. M. Doran, A. A. Pinarci, T. M. TenHoeve, A. M. Howard, J. L. Stroud, D. A. Rivera, D. A. Moskovitz, S. J. Finneran, A. N. Singer, M. E. Rossi and G. Moura-Letts, Org. Biomol. Chem., 2024, 22, 2300; (c) A. A. Pinarci, N. Daniecki, T. M. TenHoeve, B. Dellosso, R. Madiu, L. Mejia, S. E. Bektas and G. Moura-Letts, Chem. Commun., 2022, 58, 4909.
16. E. L. Doran, J. M. Doran, A. A. Pinarci, M. E. Rossi, R. Madiu, A. M. Howard, J. L. Stroud, D. A. Rivera, D. A. Moskovitz, A. N. Singer, S. J. Finneran and G. Moura-Letts, Eur. J. Org. Chem., 2025, e202401335.
17. A. G. Leach and K. N. Houk, J. Org. Chem., 2001, 66, 5192.
18. (a) B. Yang, P. A. Mille, U. Mollmann and M. J. Miller, Org. Lett., 2009, 11, 2828; (b) F. Li, B. Yang, M. J. Miller, J. Zajicek, B. C. Noll, U. Mollmann, H.-M. Dahse and P. A. Miller, Org. Lett., 2007, 9, 2923.
19. The reaction also works very well on a large scale (100 mmol of aniline) to produce 4a in 99% yield (18.711 g).

---

Footnote

† Electronic supplementary information (ESI) available: Experimental protocols and spectroscopic data for all products. See DOI: https://doi.org/10.1039/d5cc02938d

---