Studies on the Regioselective Rearrangement of Azanorbornanic Aminyl Radicals into 2,8-Diazabicyclo[3.2.1]oct-2-ene Systems

Abstract

Aminyl radicals are nitrogen-centered radicals of interest in synthetic strategies involving C–N bond formation due to their high reactivity. These intermediate radicals are generated by the reaction of an organic azide with tributyltin hydride (Bu₃SnH) in the presence of substoichiometric amounts of azobisisobutyronitrile (AIBN). In this work, we report the regioselective rearrangement of azanorbornanic ([2.2.1]azabicyclic) aminyl radicals into 2,8-diazabicyclo[3.2.1]oct-2-ene systems. With the aim to establish the structural requirements for this ring expansion, we have studied the effect of different bridgehead atoms of the [2.2.1]bicyclic system and the presence of an alkyl substituent at C4. Attempts to perform this ring expansion on a monocyclic analogue have been also explored to evaluate the influence of the bicyclic skeleton on the rearrangement. A detailed mechanistic proposal supported by computational studies is reported.

Introduction

Nitrogen-containing organic structures are commonly present in natural and non-natural biologically relevant compounds.¹ Thus, the development of novel synthetic methods and strategies toward C–N bond formation is a very active field of research. Aminyl radicals are nitrogen-centered radicals that can be obtained from organic azides and are considered versatile intermediates for C–N bond formation due to their high reactivity.² However, compared to broadly used carbon-centered radicals, aminyl radicals have received much less attention. Kim and co-workers demonstrated that tributyltin hydride (Bu₃SnH) and substoichiometric amounts of azobisisobutyronitrile (AIBN) in refluxing benzene are excellent reagents for the generation of aminyl radicals through the homolytic addition of stannyl radicals to aromatic/aliphatic azides with simultaneous loss of N₂.³ We have previously reported the unexpected formation of ring-expanded bicyclic system 2 as the major product when attempting the desulfonylation of [2.2.1]azabicyclic β-azido sulfones (3-azidoazanorbornanes) 1a and 1b under radical conditions (Scheme 1).⁴ The formation of this unexpected and unknown 2,8-diazabicyclo[3.2.1]oct-2-ene system was postulated to take place through the azanorbornan-3-aminyl radical intermediate A. Together with major compound 2, the products from the radical reduction of the azide group, primary amines 3 and 4, were also obtained as minor compounds. This competing radical reduction of aliphatic azides to amines under Bu₃SnH/AIBN conditions is also known to proceed through an aminyl radical intermediate that is further reduced in the presence of an excess of Bu₃SnH.⁵ Both stereoisomers, 1a and 1b, reacted similarly under these conditions, although 3-exo-azido 1a was slightly more prone to undergo radical expansion than 1b (56% from 1a vs 40% from 1b).

Scheme 1. Ring Expansion of [2.2.1]Azabicyclic β-Azido Sulfones via Aminyl Radicals.

We initially proposed a regiospecific 1,2-shift of the σ (C3–C4) bond of bicyclic aminyl radical A to form the expanded radical intermediate B, which evolves to the final compound 2 (Scheme 2).⁴ However, as radical 1,2-shifts of alkyl carbons were virtually unknown, Spagnolo and co-workers proposed that a ring-opening/ring-closure sequence through intermediate B′ should be more likely involved.^2b Nevertheless, a detailed mechanistic proposal has not been reported to date.

Scheme 2. Proposed Radical Intermediate Structures B and B′.

This radical ring expansion could constitute the key step of a straightforward procedure for the preparation of 2,8-diheterobicyclo[3.2.1]octanes from easily available heteronorbornane systems as starting materials. 2,8-Diheterobicyclo[3.2.1]octanes are present in the skeleton of a wide diversity of biologically active natural products and constitute interesting building blocks in organic synthesis.⁶ The development of synthetic strategies leading to heterofunctionalized bridged bicyclo[3.2.1]octanes justifies our interest in the study of this radical reaction. Therefore, this full account continues our previous publication⁴ and includes the study of the scope, limitations, and mechanism of this transformation. The effect of different bridgehead groups and the presence of an alkyl substituent at C4 of different bicyclic systems have been studied with the aim to establish the structural requirements for this singular ring expansion. Finally, the radical reaction on a monocyclic counterpart was also explored to evaluate the influence of the bicyclic skeleton on the rearrangement. A detailed mechanistic proposal supported by computational studies is also reported.

Results and Discussion

Experimental Studies

We have carried out the synthesis of a variety of 3-azido(hetero)norbornane analogues of 3-exo-azidoazanorbornane 1a (compounds 10a–d) using a shorter and more efficient synthetic strategy than the one used previously.⁷ Moreover, this new strategy affords exclusively the exo-azido derivatives which are slightly more prone to the aminyl radical rearrangement than the corresponding endo-analogues. Thus, Diels–Alder reactions between ethynyl sulfone 5 and commercially available or easily prepared cyclic dienes 6a–d (see the Supporting Information for details) were performed at different temperatures (25–90 °C). Reaction of 5 with pyrrole derivative 6c was performed at higher temperatures than those with furan derivatives (6a and 6d) and cyclopentadiene 6b due to the poor dienic character of pyrrole. All the bicyclic systems (compounds 7a–d) were isolated in moderate to good yield as racemic mixtures (Scheme 3). The selective dihydroxylation of the electron-rich double bond of bicyclic adducts 7a–d afforded the corresponding diols with total exo-face stereoselectivity, as it was previously described for other [2.2.1]heterobicyclic systems.⁸ The resulting diols were not isolated, instead they were directly treated with 2,2-dimethoxypropane (DMP) under catalytic acid conditions to afford protected diol derivatives 8a–d. Attempts to perform the addition of the azide anion to the vinyl sulfone system were unsuccessful (NaN₃ in dimethylformamide, DMF, or TMSN₃ in tetrahydrofuran, THF). Thus, we addressed the incorporation of the azide function in a two-step synthetic strategy. Conjugate addition of ammonia to vinyl sulfones 8a–d showed to be efficient and stereoselective giving the corresponding [2.2.1]bicyclic β-amino sulfones 9a–d. These compounds were directly used in a diazo transfer reaction with nonaflyl azide (NfN₃) under mild conditions⁹ affording the desired [2.2.1]bicyclic β-azido sulfones 10a–d in 22–42% overall yield (from 7a–d, four steps).

Scheme 3. Synthesis of New [2.2.1]Bicyclic β-Azido Sulfones.

With these substrates in hand, we studied the radical reaction in the presence of Bu₃SnH/AIBN under the same reaction conditions previously used with 1a. The results are summarized in Table 1. The substitutions of NBoc in 1a by O or CH₂ (compounds 10a and 10b) were detrimental for the ring expansion and only the non-expanded amino derivatives 9a and 9b were obtained. This result seems to indicate that unlike NBoc, O or CH₂ at the bridgehead position of the bicyclic system do not stabilize the contiguous radical at C4 of the corresponding B′ intermediate (Scheme 2). Heteroatoms containing lone electron pairs can establish stabilizing interactions with radical centers (two-center three-electron interaction), and the magnitude of this interaction is inversely proportional to the electronegativity of the heteroatom.¹⁰ This could explain the fact that oxanorbornane derivative 10a, with a highly electronegative oxygen at the bridgehead position, and norbornane 10b, with no heteroatom adjacent to the radical, do not undergo the radical rearrangement. The stabilizing effect that the NBoc group confers to the contiguous radical can explain the chemoselectivity observed. On the other hand, the radical reaction on azabicycle 10c, which features an additional methyl group at C4 with respect to azabicycle 1a, clearly favored the radical ring expansion with a remarkable increase in the yield of the resulting 2,8-diazabicyclo[3.2.1]oct-2-ene (82% 11c vs 56% 2). A CH₃ substituent at C4 of 10c clearly benefits the radical rearrangement as the new C4 becomes a tertiary radical on the corresponding intermediate B′. However, this extra stabilization was not enough in the case of the oxa-analogue 10d, where the radical rearrangement did not take place and only amine 9d was obtained. Of note, the desulfonylation reaction only took place during the formation of expanded [3.2.1]bicyclic systems (compounds 2 and 11c) but not in the reduced [2.2.1]bicyclic amines (3, 9a, 9b, and 9d). As the order of addition of reagents could influence the outcome of the radical reaction,¹¹ treatment of [2.2.1]bicyclic azides 1a and 10b with Bu₃SnH/AIBN was additionally performed following an alternative experimental procedure (procedure b, see the Experimental Section). However, the inverse order of addition of the reagents did not influence the final result, compounds 2 and 9b being isolated in similar yields. The structure of ring-expanded compound 11c could be confirmed by the appearance of an unambiguous signal at 7.66 ppm in the ¹H NMR spectrum corresponding to imine-type proton H3. Besides, the signals at 162.3 (C3) and 39.3 ppm (C4, coupled to two adjacent protons in DEPT) were observed in the ¹³C NMR spectrum.

Table 1. Results of the Radical Ring-Expansion Assays.

entry	[2.2.1]bicyclic azide	bicyclo[3.2.1]oct-2-enea	[2.2.1]bicyclic aminea
1b	1a (R = H, X = NBoc)	2, 56%	3, 12%
2	10a (R = H, X = O)	c	9a, 90%
3	10b (R = H, X = CH2)	c	9b, 74%
4	10c (R = Me, X = NBoc)	11c, 82%	c
5	10d (R = Me, X = O)	c	9d, 64%

^aYield (%) of isolated compound following experimental procedure a (see the Experimental Section).

^bData from ref (4).

^cNot detected.

To clarify the influence of the dioxolane fused cycle on the radical expansion of azabicyclic derivatives 1 and 10c, we assayed the reaction on diacetylated derivative 14 (Scheme 4). After dihydroxylation of 7c, the resulting diol was acetylated to afford 12. Subsequent conjugate addition of ammonia followed by diazo transfer reaction gave azide 14 in an acceptable yield. Treatment of 14 with Bu₃SnH/AIBN under standard conditions afforded a mixture of compounds where the expanded 2,8-diazabicyclo[3.2.1]oct-2-ene 15 could be isolated in 44% yield. This result indicates that less-strained 3-azidoazanorbornane 14 is less prone to radical ring expansion than acetonide derivative 10c, which could be attributed to the extra strain conferred by the acetonide group to the azabicyclic system and the steric hindrance existing between the NBoc and the isopropylidene group.

Scheme 4. Synthesis and Radical Ring-Expansion of Less-Strained 3-Azidoazanorbornane 14.

Finally, we attempted the radical expansion on pyrrolidine derivative 18 (Scheme 5), a monocyclic analogue of bicyclic compound 14, to fully suppress the influence of the bicyclic skeleton in the reaction. Thus, treatment of 12 with an excess of ammonia in THF/EtOH for 4 days yielded the conjugate addition of NH₃ with concomitant ethanolysis of the ester groups. The resulting bicyclic aminodiol 16 was transformed into azido derivative 17 under standard conditions. Oxidative cleavage of diol 17 with NaIO₄, followed by reduction of the resulting aldehydes with NaBH₄, afforded N-Boc pyrrolidine derivative 18 in good yield. Treatment of this compound with Bu₃SnH/AIBN afforded a complex mixture where no product from a ring radical expansion could be detected. Instead, amine 19 resulting from the radical reduction of the azide function was the major product and could be isolated in 49% yield. This result indicates that the rigid [2.2.1]azabicyclic skeleton clearly favors the C3–C4 bond cleavage on the corresponding aminyl radical intermediate A (Scheme 2) while the radical reduction of the azide function is the only reaction observed for azides embedded in a more conformationally flexible pyrrolidine skeleton (compound 18).

Scheme 5. Attempt to Ring-open the 3-Azidopyrrolidine System 18.

Based on the experimental results, we propose a mechanism for the ring-expansion/desulfonylation sequence leading to 2,8-diazabicyclo[3.2.1]oct-2-enes (Scheme 6). Aminyl radical intermediate A would undergo a regioselective ring opening leading to carbon-centered radical B′ which, after ring closing, would yield the expanded system B. A desulfonylation reaction, similar to that reported for allylic sulfones under Bu₃SnH/AIBN conditions,¹² would afford stannyl enamine C that after hydrolysis and tautomerization would originate the 2,8-diazabicyclo[3.2.1]oct-2-ene.

Scheme 6. Mechanism Proposed for the Radical Ring Expansion/Desulfonylation of [2.2.1]Azabicyclic β-Azido Sulfones.

Computational Analysis

The mechanism of the ring expansion for the formation of [3.2.1]bicyclic systems (2, 11a–d) was examined computationally (see the Computational Analysis details). The calculated mechanism starts from norbornan-3-aminyl radical intermediates (Int1) and analyzes a plausible competition between a stepwise ring expansion, to afford compounds 2 and 11a–d, and a radical reduction, i. e., hydrogen atom transfer (HAT),¹³ to give amines 3 and 9a–d (Figures 1a and S1–S4). Abbreviated models were used in all cases by replacing the tosyl and tri-n-butylstannane groups by mesyl and trimethylstannane groups, respectively. Methoxycarbamate (Moc) was also employed instead of the Boc group as a simpler model for the protecting group of azanorbornanes 1a and 10c. Starting from N-centered norbornan-3-aminyl radical intermediates (Int1) formed upon reaction with Bu₃SnH/AIBN, the first step of the ring-expansion reaction is the ring opening by a homolytic cleavage of the C3–C4 bond (TS1, see Table 1 for atom labeling) with activation barriers (ΔG^‡) ranging from 12 to 19 kcal mol^–1. The subsequent ring-closing step takes place from the carbon-centered radical (Int2) to form a bond between C3 and the exocyclic nitrogen (TS2). This step was calculated to be rate-limiting for all considered substrates with activation energies (ΔG^‡) ranging from 17 to 22 kcal mol^–1. Of note, formation of bicyclic carbon-centered radical Int3 upon ring expansion is significantly exergonic with free energies (ΔG) between −12 and −15 kcal mol^–1. A final desulfonylation step through a radical elimination reaction from Int3 was calculated to have very low activation barriers (ΔG^‡ ≈ 2–5 kcal mol^–1) as a result of the high stability of the leaving sulfonyl radical.

Figure 1.

(a) Minimum energy reaction pathway for the model of azanorbornane derivative 1a calculated with PCM(toluene)/M06-2X/6-31G(d,p)+LanL2DZ(Sn). (b) Transition structures (TSs) and relative activation energies for the competing HAT and ring-expansion reactions (ΔΔG^‡_TS2–TSHAT) for all calculated models of bicycles 1a and 10a–d.

On the other hand, the competing HAT reactions from a trimethylstannane hydride molecule to radical intermediates Int1 through transition states TS_HAT, leading to reduced intermediates (Int_HAT), have similar activation barriers (ΔG^‡ ≈ 14–16 kcal mol^–1) to those calculated for the ring-expansion process, although different trends were obtained (Figure 1b). Norbornane (10b) and oxonorbornane (10a and 10d) derivatives showed a preference for the HAT reaction as judged by the difference in the energies of the transition structures for both competing pathways (ΔΔG^‡_TS2–TSHAT ≈ 5–6 kcal mol^–1), in line with the experimental observations. However, these differences in energies were minimal for azanorbornane derivatives 1a and 10c (ΔΔG^‡_TS2–TSHAT ≈ 2 kcal mol^–1), suggesting that both reactions are energetically feasible and, therefore, competitive. The possibility of C-centered radicals Int2 undergoing a HAT reaction—which was never observed experimentally—was discarded considering the relatively high activation energy calculated for this process for compound 1a (ΔG^‡ ≈ 20 kcal mol^–1, see the Supporting Information), due to the large steric hindrance exerted by the trialkyl tin hydride.

These computed trends can be attributed to the relative stability of the N- and C-centered radicals (Int1 and Int2, respectively) for each substrate. The C-centered radicals are thermoneutral with respect to the N-centered radicals for azanorbornane compounds (1a and 10c), while being significantly less stable for oxanorbornene and norbornane. The C-centered radical Int2 for norbornane 10b is unstable due to the lack of an adjacent heteroatom with lone pairs at the bridgehead position, whereas the same C-centered radicals for azanorbornanes 1a and 10c are stabilized by delocalization of the spin density along the adjacent carbamate group (Figure S5). Consequently, and according to Hammond’s postulate, the activation barriers for the ring-closing transition state (TS2) for oxanorbornanes 10a and 10d and, more significantly, for norbornane 10b are higher than those for the competitive TS_HAT, thus exhibiting a preference for the HAT reaction (Figures S1, S2, and S4). In contrast, azanorbornanes 1a and 10c display lower activation energies for the ring expansion, which becomes competitive with HAT. The homolytic cleavage of the C2–C3 bond on Int1 leads to a transition state ca. 4 kcal mol^–1 higher in energy than that calculated for the homolytic cleavage of the C3–C4 bond, which explains the exceptional regioselectivity observed experimentally in the ring expansion.

Conclusions

We have demonstrated that 3-azidoazanorbornanes are excellent substrates for the preparation of 2,3-diazabicyclo[3.2.1]oct-2-enes through a regioselective rearrangement of azanorbornanic aminyl radicals. These systems were previously unknown and can be considered as precursors of the 2,8-diheterobicyclo[3.2.1]octane skeleton, which is present in natural products. Experimental and computational studies allowed us to establish the scope of the reaction and provide a mechanistic proposal to this unusual radical rearrangement. The scope of this reaction is limited to the aza-analogues where the intermediate radicals are adequately stabilized. Moreover, the rigidity of the azabicyclic skeleton showed to be crucial for the rearrangement to proceed efficiently. With the discovery of this new radical rearrangement, new synthetic strategies involving compounds containing a rigid azido-functionalized azacyclic core could be explored.

Experimental Section

General Methods

¹H and ¹³C NMR spectra were recorded with a Bruker AMX300 spectrometer for solutions in CDCl₃ and DMSO-d₆. δ is given in ppm and J in Hz. All of the assignments were confirmed by two-dimensional (2D) spectra (COSY and HSCQ). High-resolution mass spectra were recorded on a Q Exactive quadrupole mass spectrometer. Thin layer chromatography (TLC) was performed on silica gel 60 F₂₅₄ (Merck), with detection by UV light charring with p-anisaldehyde, KMnO₄, ninhydrin, phosphomolybdic acid, or with reagent [(NH₄)₆MoO₄, Ce(SO₄)₂, H₂SO₄, H₂O]. Silica gel 60 (Merck, 40–60 and 63–200 μm) was used for preparative chromatography. Infrared spectra were recorded with a Jasco FTIR-410 spectrophotometer and processed with Jasco Spectra Manager program, using solid and oily compounds in an ATR MIRacle. Maximum absorption wave numbers are indicated.

Synthesis of [2.2.1]Bicyclic β-Azido Sulfones

Synthesis of (7-Hetero)norbornadienic β-Azido Sulfones 10a,b from (7-Hetero)norbornadienes 7a,b. To a solution of (7-hetero)norbornadiene (7a or 7b) (2.2 mmol) in acetone (47 mL) and water (5.1 mL), NMO (3.2 mmol) and OsO₄ (4 wt % in H₂O, 0.079 mmol) were added. The mixture was stirred for 50 min, and then, a saturated solution of NaHSO₃ in water was added at 0 °C. After a few minutes, the reaction mixture was diluted with EtOAc and washed with water. The organic layer was dried over Na₂SO₄ and filtered, and the solvent was removed under reduced pressure to give the dihydroxylated compound that was used without purification in the next step. To a solution of this compound in acetone (26 mL), 2,2-dimethoxypropane (13 mmol) and PTSA (0.2 mmol) were added. The reaction was stirred at room temperature (rt) for 2 h and then diluted with CH₂Cl₂ and washed with an aqueous saturated solution of NaHCO₃. The organic layer was dried over Na₂SO₄ and filtered, and the solvent was removed under reduced pressure to give 8a or 8b. To a solution of 8a or 8b in EtOH (18 mL) and THF (7 mL) at 0 °C, NH₃ was bubbled. The reaction mixture was allowed to warm at rt for 4 h, and then, the solvent was removed under reduced pressure to give the crude protected amine 9a or 9b that was directly dissolved in MeOH (8 mL) and water (3 mL). Then, NaHCO₃ (800 mg, 9.5 mmol), a solution of nonaflyl azide (1.1 g, 3.2 mmol) in Et₂O (5.7 mL), and CuSO₄·5H₂O (57 mg, 0.2 mmol) were added and the reaction mixture was stirred at rt for 3 h. The organic solvents were removed and the resultant aqueous mixture was extracted with dichloromethane (DCM) and washed with an aqueous saturated solution of NaHCO₃. The organic layer was dried over Na₂SO₄ and filtered. The solvent was removed under reduced pressure, and the crude product was purified by column chromatography on silica gel (EtOAc/Cy 1:6). Compound 10a was obtained as a white powder (280.6 mg, 24%, 4 steps); compound 10b was obtained as a white foam (335.8 mg, 42%, 4 steps).

Data for 10a

IR (υ̅, cm^–1) 2975, 2087 (N₃), 1305 1209, 1139, 669. ¹H NMR (300 MHz, CDCl₃) δ 7.80 (d, J = 8.4 Hz, 2H, ArH), 7.42 (m, 2H, ArH), 5.21 (d, J = 5.5 Hz, 1H, H4), 4.65 (dd, J = 5.5, 1.4 Hz, 1H, H1), 4.48–4.42 (m, 2H, H2, H3), 3.88 (d, J = 4.0 Hz, 1H, H5), 3.55–3.45 (m, 1H, H6), 2.47 (s, 3H, CH₃ of Ts), 1.45 (s, 3H, CH₃), 1.33 (s, 3H, CH₃). ¹³C {¹H} NMR (76 MHz, CDCl₃) δ 146.2 (CAr), 136.1 (CAr), 130.7, 128.0 (CHAr), 112.6 (C(CH₃)₂), 86.1, 79.7 (C2, C3), 79.6 (C4), 78.5 (C1), 70.6 (C6), 60.9 (C5), 25.8, 25.3 (CH₃), 21.9 (CH₃ of Ts). HRMS (ESI) m/z: [M + Na]⁺ calcd for C₁₆H₁₉O₅N₃NaS: 388.0938; found: 388.0930.

Data for 10b

IR (υ̅, cm^–1) 2978, 2917, 2104 (N₃), 1595, 1277, 1143, 1041, 865, 663. ¹H NMR (300 MHz, CDCl₃) δ 7.79 (d, J = 8.3 Hz, 2H, ArH), 7.45–7.34 (m, 2H, ArH), 5.10–4.99 (m, 1H, H2 or H3), 4.31–4.19 (m, 1H, H2 or H3), 3.88 (dd, J = 4.2, 1.9 Hz, 1H, H5), 3.18 (t, J = 4.2 Hz, 1H, H6), 2.67 (dt, J = 4.2, 1.3 Hz, 1H, H1), 2.46 (s, 4H, CH₃ of Ts, H4), 1.96 (dq, J = 11.2, 1.7 Hz, 1H, H7a), 1.43 (s, 3H, CH₃), 1.38 (dd, J = 11.2, 1.6 Hz, 1H, H7b), 1.33 (s, 3H, CH₃). ¹³C {¹H} NMR (76 MHz, CDCl₃) δ 145.6, 136.3 (CAr), 130.4, 128.2 (CHAr), 110.1 (C(CH₃)₂), 78.7, 76.4 (C2, C3), 70.8 (C6), 60.2 (C5), 48.1 (C4), 43.3 (C1), 31.1 (C7), 25.3, 24.3 (CH₃), 21.8 (CH₃ of Ts). HRMS (ESI) m/z: [M + Na]⁺ calcd for C₁₇H₂₁O₄N₃NaS: 386.1145; found: 386.1139.

Synthesis of (7-Hetero)norbornadienic Vinyl Sulfones 8c,d from (7-Hetero)norbornadienes 7c,d

To a solution of 7c or 7d (0.8 mmol) in acetone (18 mL) and water (2 mL), NMO (1.4 mmol) and OsO₄ (4 wt % in H₂O, 0.004 mmol) were added. The mixture was stirred overnight, and then, a saturated aqueous solution of NaHSO₃ was added at 0 °C. After a few minutes, the crude was diluted with EtOAc and washed with water. The organic layer was dried over Na₂SO₄ and filtered. The solvent was removed under reduced pressure. To the resulting residue dissolved in acetone (10 mL), 2,2-dimethoxypropane (5.5 mmol) and PTSA (0.1 mmol) were added. The reaction was stirred at rt overnight; then diluted with dichloromethane and washed with an aqueous saturated solution of NaHCO₃. The organic layer was dried over Na₂SO₄ and filtered. The solvent was removed under reduced pressure and purified by column chromatography on silica gel (EtOAc/Cy 1:5). Compound 8c was obtained as a white powder (241.0 mg, 67%, two steps); compound 8d was obtained as a white powder (162.6 mg, 58%, two steps).

Data for 8c

IR (υ̅ cm^–1) 2979, 1699, 1596, 1450, 1362, 1266, 1056, 976, 868, 755. ¹H NMR (300 MHz, CDCl₃) δ 7.89–7.61 (m, 2H, ArH), 7.40–7.30 (m, 2H, ArH), 6.83 (s, 1H, H3), 4.29 (d, J = 5.5 Hz, 1H, H5 or H6), 4.12 (d, J = 5.5 Hz, 1H, H5 or H6), 2.46 (s, 3H, CH₃ of Ts), 1.80 (s, 3H, CH₃), 1.68 (s, 3H, CH₃), 1.44 (s, 3H, C(CH₃)₂), 1.35 (s, 9H, C(CH₃)₃), 1.32 (s, 3H, CH₃). ¹³C {¹H} NMR (76 MHz, CDCl₃) δ 153.8 (C=O), 152.7 (C3), 150.6 (C2), 145.2, 136.6 (CAr), 130.1, 128.3 (CHAr), 117.1 (C(CH₃)₂), 83.2, 82.7 (C5, C6), 80.5 (C(CH₃)₃), 72.8, 71.9 (C1, C4), 28.5 (C(CH₃)₃), 26.5, 25.8 (CH₃), 21.8 (CH₃ of Ts), 15.3 (CH₃), 13.4 (CH₃). HRMS (ESI) m/z: [M + Na]⁺ calcd for C₂₃H₃₁NO₆SNa: 472.1770; found: 472.1758.

Data for 8d

IR (υ̅ cm^–1) 2986, 2853, 1598, 1448, 1269, 1208, 1083, 970, 876. ¹H NMR (300 MHz, CDCl₃) δ 7.77 (d, J = 8.3 Hz, 2H, ArH), 7.36 (m, 2H, ArH), 6.84 (d, J = 0.5 Hz, 1H, H3), 4.43 (d, J = 5.3 Hz, 1H, H5 or H6), 4.30 (d, J = 5.2 Hz, 1H, H5 or H6), 2.46 (s, 3H, CH₃ of Ts), 1.55 (s, 3H, CH₃), 1.48 (s, 3H, CH₃), 1.43 (s, 3H, CH₃), 1.35 (s, 3H, CH₃). ¹³C {¹H} NMR (76 MHz, CDCl₃) δ 153.1 (C2), 148.8 (C3), 145.3, 136.6 (CAr), 130.2, 128.3 (CHAr), 117.7 (C(CH₃)₂), 88.7, 88.2 (C1, C4), 83.2, 82.7 (C5, C6), 26.6 (CH₃), 21.8 (CH₃ of Ts), 13.7(CH₃), 12.5 (CH₃). HRMS (ESI) m/z: [M + Na]⁺ calcd For C₁₈H₂₂O₅SNa: 373.1086; found: 373.1074.

Synthesis of (7-Hetero)norbornadienic β-Azido Sulfones 10c,d from (7-Hetero)norbornadienic Vinyl Sulfones 8c,d

Compound 8c or 8d (1.3 mmol) was dissolved in ethanol (3 mL) and THF (6 mL) at 0 °C. Then, NH₃ was bubbled for 5 min. The reaction was allowed to warm at rt for 4 h, and then, the solvent was removed under reduced pressure. The resulting amine 9c or 9d was dissolved in MeOH (5.5 mL) and H₂O(2 mL); NaHCO₃ (5.9 mmol), a solution of nonaflyl azide (2.96 mmol) in Et₂O (4 mL), and CuSO₄·5H₂O (0.2 mmol) were added and the mixture was stirred at rt for 3 h. The organic solvents were removed, and the resulting aqueous mixture was extracted with dichloromethane and washed with an aqueous saturated solution of NaHCO₃. The organic layer was dried over Na₂SO₄ and filtered. The solvent was removed under reduced pressure, and the reaction mixture was purified by chromatography column on silica gel (EtOAc/Cy 1:10 for 10c, dichloromethane/Cy 1:10 for 10d). Compound 10c was obtained as a white foam (269.0 mg, 42%, two steps); compound 10d was obtained as a white powder (189.2 mg, 37%, two steps).

Data for 10c

IR (υ̅ cm^–1) 2979, 2103 (N₃), 1700, 1597, 1292, 1211, 1082, 849, 707. ¹H NMR (300 MHz, CDCl₃) δ 7.84–7.72 (m, 2H, ArH), 7.45–7.36 (m, 2H, ArH), 4.92 (d, J = 5.7 Hz, 1H, H5 or H6), 4.10 (d, J = 5.7 Hz, 1H, H5 or H6), 3.61 (d, J = 4.1 Hz, 1H, H2), 3.22 (d, J = 4.1 Hz, 1H, H3), 2.47 (s, 3H, CH₃ of Ts), 1.85 (s, 3H, CH₃), 1.64 (s, 3H, CH₃), 1.42 (bs, 12H, C(CH₃)₃, CH₃), 1.34 (s, 3H, CH₃). ¹³C {¹H} NMR (76 MHz, CDCl₃) δ 154.3 (C=O), 146.0, 136.6 (CAr), 130.5, 128.3 (CHAr), 112.0 (C(CH₃)₂), 82.3 (C5 or C6), 80.7 (C(CH₃)₃), 80.2 (C5 or C6), 74.5 (C3), 72.3, 70.8 (C1, C4), 66.9 (C2), 28.5 (C(CH₃)₃), 26.0 (CH₃), 25.4 (CH₃), 21.9 (CH₃ of Ts), 17.3, 14.3 (CH₃). HRMS (ESI) m/z: [M + Na]⁺ calcd for C₂₃H₃₂N₄O₆SNa: 515.1941; found: 515.1931.

Data for 10d

IR (υ̅, cm^–1) 2993, 2937, 2103 (N₃), 1596, 1454, 1256, 1010, 777, 650. ¹H NMR (300 MHz, CDCl₃) δ 7.80 (d, J = 8.3 Hz, 2H, ArH), 7.49–7.33 (m, 2H, ArH), 5.07 (d, J = 5.5 Hz, 1H, H5 or H6), 4.28 (d, J = 5.5 Hz, 1H, H5 or H6), 3.78 (d, J = 4.2 Hz, 1H, H3), 3.21 (d, J = 4.3 Hz, 1H, H2), 2.47 (s, 3H, CH₃ of Ts), 1.54 (s, 3H, CH₃), 1.47 (s, 3H, CH₃), 1.42 (s, 3H, CH₃), 1.35 (s, 3H, CH₃). ¹³C {¹H} NMR (76 MHz, CDCl₃) δ 146.0, 136.5 (CAr), 130.6, 128.2 (CHAr), 112.9 (C(CH₃)₂), 89.0, 86.7 (C1, C4), 83.1, 81.0 (C5, C6), 76.0 (C2), 66.6 (C3), 26.0, 25.8 (CH₃), 21.9 (CH₃ of Ts), 16.1, 12.0 (CH₃). HRMS (ESI) m/z: [M + Na]⁺ calcd for C₁₈H₂₃O₅N₃NaS: 416.1256; found: 416.1244.

Aminyl Radical Ring-Expansion/Azide Radical Reduction

Procedure a

In a two-necked round-bottom flask, AIBN (0.026 mmol, 0.1 equiv) and Bu₃SnH (0.52 mmol, 2 equiv) were dissolved in dry toluene (1.0 mL) under Ar atmosphere. The solution was heated at 110 °C in an oil bath for two min. Then, a solution of the corresponding bicyclic substrate (0.26 mmol, 1 equiv) in dry toluene was added (1.5 mL). The reaction was stirred at 110 °C for 3 h and then quenched with an aqueous solution of NaF (1M). The resultant mixture was extracted twice with AcOEt and washed with brine. The organic layer was dried over Na₂SO₄ and filtered, and the solvent was removed under reduced pressure. The resulting crude mixture was purified by column chromatography.

Procedure b

In a two-necked round-bottom flask, bicyclic substrate (0.23 mmol, 1 equiv) was dissolved in dry toluene (4.0 mL) under Ar atmosphere. The solution was heated at 110 °C in an oil bath and a solution of AIBN (0.12 equiv, 0.029 mmol) and Bu₃SnH (1.25 equiv, 0.29 mmol) in toluene (1.6 mL) under Ar atmosphere was gradually added (0.4 mL each 30 min). The mixture was refluxed for a total time 2.5 h. Then, the solvent was evaporated and the resulting crude mixture was purified by column chromatography

(rac)-N-Boc-6,7-exo-isopropylidendioxy-2,8-diazabicyclo[3.2.1]oct-2-ene (2)

The synthesis of 2 was carried out from 1a (108 mg, 0.23 mmol) according to the general procedure b described above. Compound 2 (33 mg, 51%) was obtained after purification by column chromatography (toluene/acetone 5:1) as a pale yellow syrup. ¹H NMR data for this compound correlates with those previously reported by us⁴ using synthetic procedure a.

(rac)-N-Boc-6,7-exo-isopropylidendioxy-1,5-dimethyl-2,8-diazabicyclo[3.2.1]oct-2-ene (11c)

The synthesis of 11c was carried out from 10c (130.0 mg, 0.3 mmol) according to the general procedure a described above. Compound 11c (66.2 mg, 82%) was obtained after purification by column chromatography (toluene/acetone 10:1 → 5:1) as a yellow syrup. IR (υ̅, cm^–1) 2979, 2934, 1699 (C=O), 1456, 1367, 1283, 1249, 1211, 1158, 1088, 1065, 998, 877, 848, 776, 740, 605. ¹H NMR (300 MHz, CDCl₃) δ 7.66 (m, 1H, H3), 4.15 (d, J = 6.2 Hz, 1H, H6 or H7), 4.05 (d, J = 6.2 Hz, 1H, H6 or H7), 2.98–2.84 (m, 1H, H4a), 2.01–1.88 (m, 1H, H4b), 1.86 (s, 3H, CH₃) 1.53 (s, 3H, CH₃), 1.48 (s, 3H, CH₃), 1.44 (s, 9H, C(CH₃)₃), 1.31 (s, 3H, CH₃). ¹³C {¹H} NMR (76 MHz, CDCl₃) δ 162.3 (C3), 156.1 (C=O), 111.1 (C(CH₃)₂), 85.5, 84.5 (C6, C7), 81.3 (C5 or C1), 80.8 (C(CH₃)₃), 63.2 (C1 or C5), 39.8 (C4), 28.5 (C(CH₃)₃), 26.8, 25.9, 23.5, 21.4,(CH₃). HRMS (ESI) m/z: [M + H]⁺ calcd for C₁₆H₂₆O₄N₂: 311.1965; found: 311.1963.

(rac)-N-Boc-6,7-exo-diacetyl-1,4-dimethyl-2,8-diazabicyclo[3.2.1]oct-2-ene (15)

The synthesis of 15 was carried out from 14 (94.4 mg, 0.2 mmol) according to the general procedure. Compound 15 (25.9 mg, 44%) was obtained after purification by column chromatography (DCM/acetone 10:1 → 5:1) as a yellow oil. IR (υ̅, cm^–1) 2977, 2927, 1748 (C=O), 1699 (C=O), 1456, 1366, 1240, 1160, 1062, 946, 906, 848, 774, 748, 624. ¹H NMR (300 MHz, CDCl₃) δ 7.73–7.72 (m, 1H, H3), 5.05 (d, J = 6.9 Hz, 1H, H6 or H7), 5.00 (d, J = 6.9 Hz, 1H, H6 or H7), 3.04 (dd, 1H, J = 19.4, J = 1.1, H4a), 2.20–2.06 (m, 10H, H4b, 2CH₃ of AcO, CH₃), 1.80 (s, 3H, CH₃), 1.44 (s, 9H, C(CH₃)₃). ¹³C {¹H} NMR (76 MHz, CDCl₃) δ 169.9, 169.8 (COOCH₃), 162.5 (C3), 155.3 (C=O of Boc), 81.4, 80.8, 62.6 (C1, C5, C(CH₃)₃), 41.5 (C4), 28.4 (C(CH₃)₃), 23.7 (CH₃), 20.7, 20.6 (COOCH₃), 20.5 (CH₃). HRMS (ESI) m/z: [M + H]⁺ calcd for C₁₇H₂₇O₆N₂: 355.1864; found: 355.1854.

(2S,3R,4S,5S) and (2R,3S,4R,5R)-N-Boc-3-amino-2,5-bis(hydroxymethyl)-2,5-dimethyl-4-tosylpyrrolidine (19)

The synthesis of 19 was carried out from 18 (93.0 mg, 0.2 mmol) according to the general procedure. Compound 19 (42.4 mg, 49%) was obtained after purification by column chromatography (DCM/methanol 30:1) as a white foam. IR (υ̅, cm^–1) 3388, 2975, 2933, 1688, 1597, 1474, 1351, 1288, 1140, 1062, 846, 817, 771, 734, 706, 671, 621. ¹H NMR (300 MHz, DMSO-d₆, 353 K) δ 7.89 (d, J = 8.3 Hz, 2H, ArH), 7.43–7.40 (m, 2H, ArH), 4.02 (d, J = 11.6 Hz, 1H, CHHOH), 3.95–3.84 (m, 2H, CHHOH), 3.79 (d, J = 10.2 Hz, 1H, CHHOH), 3.57 (dd, J = 12.0, 0.8 Hz, 1H, H4), 3.42 (d, J = 10.2 Hz, 1H, CHHOH), 3.08 (s, 2H, NH₂), 2.42 (s, 3H, CH₃ of Ts), 1.53 (s, 3H, CH₃), 1.42 (s, 9H, C(CH₃)₃), 1.03 (s, 3H, CH₃). ¹³C {¹H} NMR (76 MHz, DMSO-d₆, 353 K) δ 152.0 (C=O), 143.6, 138.3 (CAr), 128.9, 128.2 (CHAr), 79.0 (C(CH₃)₃), 74.6 (C4), 65.6, 64.9 (C2, C5), 63.0 (CH₂OH), 62.2 (CH₂OH), 53.6 (C3), 27.8 (C(CH₃)₃), 24.0 (CH₃), 20.6 (CH₃ of Ts), 16.0 (CH₃). HRMS (ESI) m/z: [M + H]⁺ calcd for C₂₀H₃₃O₆N₂S: 429.2054; found: 429.2047.

(rac)-3-exo-Amino-5,6-exo-isopropylidendioxy-2-endo-tosyl-7-oxabicyclo[2.2.1]heptane (9a)

The synthesis of 9a was carried out from 10a (159.4 mg, 0.4 mmol) according to the general procedure. Compound 9a (133.2 mg, 90%) was obtained after purification by column chromatography (DCM → DCM/methanol 60:1 → 40:1) as a white foam. IR (υ̅, cm^–1) 2980, 2927, 1596, 1457, 1375, 1302, 1288, 1208, 1142, 1058, 999, 930. ¹H NMR (300 MHz, CDCl₃) δ 7.79 (d, J = 8.3 Hz, 2H, ArH), 7.45–7.34 (m, 2H, ArH), 5.17 (d, J = 5.5 Hz, 1H, H5 or H6), 4.49 (dd, J = 5.3, 1.3 Hz, 1H, H1), 4.43 (d, J = 5.6 Hz, 1H, H5 or H6), 4.18 (d, J = 1.3 Hz, 1H, H4), 3.48 (d, J = 4.2 Hz, 1H, H3), 3.17 (dd, J = 5.3, 4.2 Hz, 1H, H2), 2.45 (s, 3H, CH₃ of Ts), 1.44 (s, 5H, C(CH₃)₂, NH₂), 1.32 (s, 3H, C(CH₃)₂). ¹³C {¹H} NMR (76 MHz, CDCl₃) δ 145.6, 136.8 (CAr), 130.5, 127.9 (CHAr), 112.1 (C(CH₃)₂), 88.8 (C4), 80.0 (C5 or C6), 79.8 (C1), 78.3 (C5 or C6), 74.3 (C2), 53.9 (C3), 25.8, 25.2 (C(CH₃)₂), 21.8 (CH₃ of Ts). HRMS (ESI) m/z: [M + H]⁺ calcd for C₁₆H₂₂O₅NS: 340.1213; found: 340.1207.

(rac)-3-exo-Amino-5,6-exo-isopropylidendioxy-2-endo-tosyl-bicyclo[2.2.1]heptane (9b)

General procedure a: The synthesis of 9b was carried out from 10b (132.7 mg, 0.37 mmol). Compound 9b (92.2 mg, 74%) was obtained after purification by column chromatography (DCM/methanol 40:1) as a white foam. General procedure b: The synthesis of 9b was carried out from 10b (84.3 mg, 0.23 mmol). Compound 9b (59.1 mg, 79%) was obtained after purification by column chromatography (DCM/methanol 40:1). IR (υ̅, cm^–1) 2985, 2927, 1596, 1456, 1374, 1281, 1142, 1087, 1036, 917. ¹H NMR (300 MHz, CDCl₃) δ 7.80 (d, J = 8.3 Hz, 2H, ArH), 7.42–7.32 (m, 2H, ArH), 5.08–4.97 (m, 1H, H5 or H6), 4.24 (dt, J = 5.4, 1.2 Hz, 1H, H5 or H6), 3.40 (dd, J = 4.7, 1.7 Hz, 1H, H3), 2.95 (t, J = 4.3 Hz, 1H, H2), 2.46–2.45 (m, 4H, CH₃ of Ts, H1), 2.21–2.16 (m, 1H, H4), 1.91–1.87 (m, 1H, H7a), 1.49–1.48 (m, 1H, H7b), 1.42 (m, 5H, C(CH₃)₂, NH₂), 1.32 (s, 3H, C(CH₃)₂). ¹³C {¹H} NMR (76 MHz, CDCl₃) δ 145.1, 136.7 (CAr), 130.3, 128.1 (CHAr), 109.7 (C(CH₃)₂), 79.5, 76.2 (C5, C6), 73.7 (C2), 51.2 (C3), 50.1 (C4), 43.6 (C1), 30.7 (C7), 25.4, 24.3. (C(CH₃)₂), 21.8 (CH₃ of Ts). HRMS (ESI) m/z: [M + H]⁺ calcd for C₁₇H₂₄O₄NS: 338.1421; found: 338.1421.

(rac)-3-exo-Amino-5,6-exo-isopropylidendioxy-1,4-dimethyl-2-endo-tosyl-7-oxabicyclo[2.2.1]heptane (9d)

The synthesis of 9d was carried out from 10d (100.0 mg, 0.3 mmol) according to the general procedure a described above. Compound 9d (59.7 mg, 64%) was obtained after purification by column chromatography (DCM → DCM/methanol 60:1 → 40:1) as a white powder. IR (υ̅, cm^–1) 2979, 2929, 1599, 1454, 1276, 1068, 930, 884, 709. ¹H NMR (300 MHz, CDCl₃) δ 7.81 (d, J = 8.4 Hz, 2H, ArH), 7.39 (d, J = 7.9 Hz, 2H, ArH), 5.05 (d, J = 5.5 Hz, 1H, H5 or H6), 4.27 (d, J = 5.5 Hz, 1H, H5 or H6), 3.37 (d, J = 4.3 Hz, 1H, H3), 2.95 (d, J = 4.4 Hz, 1H, H2), 2.46 (s, 3H, CH₃ of Ts), 1.52 (bs, 2H, NH₂), 1.47 (s, 3H, CH₃), 1.41 (s, 3H, CH₃), 1.35 (s, 3H, CH₃), 1.33 (s, 3H, CH₃). ¹³C {¹H} NMR (76 MHz, CDCl₃) δ 145.4, 137.1 (CAr), 130.4, 128.3 (CHAr), 112.6 (C(CH₃)₂), 88.5, 86.0 (C1, C4), 83.7, 81.1 (C5, C6), 79.6 (C2), 58.3 (C3), 26.1, 25.8 (CH₃), 21.9 (CH₃ of Ts), 16.3, 11.6 (CH₃). HRMS (ESI) m/z: [M + H]⁺ calcd for C₁₈H₂₅O₅NS: 368.1526; found: 368.1522.

Quantum Mechanical Calculations

Full geometry optimizations and transition structure (TS) searches were carried out with Gaussian 16¹⁴ using the M06-2X hybrid functional,¹⁵ 6-31G(d,p) basis set for C, N, O, S, and H, and LanL2DZ¹⁶ effective core potential for the Sn atoms with ultrafine integration grids. Bulk solvent effects in toluene were considered implicitly through the integral equation formalism polarizable continuum model (IEF-PCM).¹⁷ The possibility of different conformations was considered for all structures. All stationary points were characterized by a frequency analysis performed at the same level used in the geometry optimizations from which thermal corrections were obtained at 383.75 K. The quasi-harmonic approximation reported by Truhlar et al. was used to replace the harmonic oscillator approximation for the calculation of the vibrational contribution to enthalpy and entropy.¹⁸ Scaled frequencies were not considered. Mass-weighted intrinsic reaction coordinate (IRC) calculations were carried out using the Hratchian and Schlegel scheme¹⁹ to ensure that the TSs indeed connected the appropriate reactants and products. For the calculation of bond dissociation energies (BDEs), single-point energy calculations using the correlated ab initio spin-component-scaled SCS-MP2 method, in combination with the cc-pVTZ basis set,²⁰ were performed on the M06-2X/6-31G(d)-optimized geometries. BDEs were defined as the difference in zero-point energies between the neutral species and the sum of the isolated radicals generated upon homolytic C–H cleavage (Table S2). Gibbs free energies (ΔG) were used for the discussion on the relative stabilities of the considered structures. The lowest energy conformer for each calculated stationary point (Figure S6) was considered in the discussion; all the computed structures can be obtained from authors upon request. Electronic energies, entropies, enthalpies, Gibbs free energies, and lowest frequencies of the calculated structures are summarized in Table S1. Cartesian coordinates of the lowest energy structures calculated with PCM(toluene)/M06-2X/6-31G(d,p) are shown in Table S3.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.joc.2c02201.

• Experimental protocols, compound characterization data, computational details, and copy of ¹H and ¹³C NMR spectra for new compounds (PDF)

The authors declare no competing financial interest.