B–N axis as a facilitating agent for the synthesis of 3D structures: the paradigmatic case of BN-[4.4.4]propellane

Abstract

Flat aromatic compounds containing a boron–nitrogen (BN) fragment have gained significant attention in the field of organic optoelectronics. Unsurprisingly, therefore, an increasing number of synthetic methodology groups have devoted efforts and creativity to developing new strategies for accessing diverse structures incorporating a B=N unit, which is isosteric with the olefinic C=C bond. In contrast, the potential of BN isosterism to expand the structural diversity of three-dimensional architectures based on sp³-hybridized atoms remains largely underexplored. In this study, we introduce a strategy to construct an alkane-type quaternary–quaternary Bsp³–Nsp³ molecular axis via double addition of a carbon-based nucleophile/electrophile pair to a readily accessible olefin-type B=N moiety. The approach is showcased through the synthesis of a BN-[4.4.4]-propellane, in which the rapidly assembled tetrahydro-BN-naphthalene intermediate undergoes a polar double allylation of its B=N bond. Despite the unfavorable trans preference in this addition step, efficient [4.4.4]-propellane formation was achieved through a tandem metathesis-based trans-to-cis isomerization and ring-closing reaction. The resulting BN-propellane exhibits a C₃-symmetric helical arrangement in the solid state and shows fluxional behavior in the ¹H NMR spectrum at room temperature due to a helicity flip, for which variable-temperature NMR measurements yielded an activation barrier (ΔG^‡) of approximately 14.6 kcal mol⁻¹.

---

BN-isosterism involves replacing a carbon–carbon (C–C) bond with a boron–nitrogen (B–N) fragment, taking advantage of the fact that both pairs possess eight valence electrons, which allows for analogous Lewis structures and similar molecular geometries. At the same time, the substitution transforms the low-polarity CC unit into a strongly polarized BN bond, thereby altering the compound's orbital energies, reactivity and physical properties (Scheme 1A).^1,2 Owing to its potential for electronic tuning of polycyclic aromatic compounds,³ 1,2-BN isosterism is viewed as a transformative tool in the field of organic optoelectronics,^4,5 prompting significant synthetic efforts toward flat structures containing olefin-type B=N units isosteric with a C=C fragment.⁶ In contrast, very little systematic research exists on the preparation and properties of non-trivial 3D molecules incorporating sp³–sp³ BN moieties mimicking saturated alkane C–C bonds.

Scheme 1 Precedents in the formation of the B=N bicycle 1, and the relevant isosteric analogy between B–N and C–C sp³–sp³ cores. ^aRef. 8b; our unpublished work; ^bref. 10; ^cref. 9.

This disparity between the attention received by sp²-vs.-sp³ type BN compounds stands in sharp contrast to the growing synthetic emphasis – outside of BN isosterism – on highly substituted sp³-hybridized carbon atoms underpinning three-dimensional molecular architectures. Much of the latter interest is fuelled by the escape-from-the-flatland campaign, where a shift away from planar aromatic scaffolds and toward sp³-rich, 3D structures is seen as venturing to the under-populated corners of chemical space and enhancing drug discovery.⁷ In this regard, it would be compelling to readily map diverse carbon-based 3D architectures onto their BN isosteres, anywhere from simpler linear and cyclic molecular cores, to more elaborate spirocycles, cages or propellanes. An interesting possibility that got our attention is to generate saturated 3D structures via di-functionalization of the more synthetically available (vide supra) olefin-like type B=N axes. This idea was inspired by earlier work on the bicyclic aminoborane 1 (Scheme 1B), first prepared by the Ashe and Fang groups on route to BN-naphthalene 2,⁸ and recently revisited by our group through a robust multigram synthesis featuring quantitative condensation of B(allyl)₃ with N,N-diallylamine, followed by ring-closing olefin metathesis (Scheme 1B).⁹ Importantly, while 1 is generally handled under an inert atmosphere, reflecting the general challenge of working with non-aromatic (and, often, even aromatic) B=N derivatives, its B–N hydrate 3^8b proved to be air- and moisture-stable, foreshadowing greater stability of the sp³–sp³ BN derivatives. In terms of isosterism, while 1 maps onto 1,4,5,8-tetrahydronaphthalene, its hydrate 3 is isosteric with the known decalinol 4 (Scheme 1C).¹⁰ Our studies also reveal the potential for generating higher structural complexity, as illustrated by the air- and moisture-stable BN-pentacycle 5, formed via intramolecular addition of a carboxylic acid across a B=N bond in a Diels–Alder-derived tetracyclic B=N precursor; the product is a BN isostere of the hitherto unreported pentacycle 6 (Scheme 1D).⁹ In the present study, the concise synthesis of air- and moisture-stable BN-[4.4.4]-propellane is used to show that B=N double functionalization represents a broader strategy for BN-mapping of 3D structure (Scheme 1E).

Indeed, while compounds 3 and 5 shown in Scheme 1 represent a reaction between B=N and a [H⁺/OR⁻] pair, we envisaged that analogous chemistry can potentially also extend to C-based nucleophile/electrophile combinations. Given the lower π-contribution¹¹ and higher polarization in the B=N moiety, disrupting its π component should be considerably easier than in the non-activated C=C moiety. Thus, in light of the appeal of the propellane structure motif, both as a synthetic challenge and as a component of functional materials,¹² this report explores and validates the B=N π-disruption strategy through rapid elaboration of the air- and moisture-stable BN-[4.4.4]-propellane. The process entails a polar (Nu⁻/E⁺) di-allylation of the B=N group followed by endgame ring-closing metathesis of intermediate 7 (Scheme 2).

Scheme 2 The proposed double B=N allylation as a strategy towards a propellane core.

Multigram quantities of bicycle 1 required in this study were readily procured via one-pot solvent-free condensation of B(allyl)₃ with the N,N-diallylamine, followed by ring closing metathesis, as described earlier by our laboratory (see Scheme 1 and SI).⁹

Initial attempts to engage 1 either through N-alkylation with allyl bromide, or through B-quaternization with allyl magnesium reagents, were unsuccessful (Scheme 3, left), reflecting a still considerable N-to-B π-donation. Nevertheless, taking note of recent precedents from the Szymczak group¹³ and the laboratory of García-García and Vaquero¹⁴ on the B-quaternization of related BN-bicycles, we proceeded to test the more nucleophilic organolithium reagents. Indeed, control experiments with MeLi did produce a quaternized B center, as evidenced by the ¹¹B NMR resonance shifting from 41 ppm to −12 ppm (Scheme 3, right). The resulting B–Me anion 9 now readily underwent N-alkylation with allylbromide to give the B_sp³–N_sp³ aminoborane 10 in 45% yield as a white solid.

Scheme 3 Initial exploration of disrupting the B=N π-donation for double functionalization.

Applying this approach to the target B,N-diallylation, a treatment of 1 with freshly prepared allyllithium¹⁵ at −78 °C in THF afforded the B-allyl anion 11, which was quenched with allyl bromide to give the di-allyl target 7 (Scheme 4A). Compound 7 could now be handled without special precautions, and was isolated as a white crystalline solid in 75% yield by conventional column chromatography on silica gel using hexane as eluent. The structure was supported by ¹H, ¹³C and ¹¹B NMR, the lattermost giving a sharp peak at −3.30 ppm for the quaternized boron center.

Scheme 4 Formation and characterization of the bis-allyl structure 7. (A) The double B=N allylation of 1 through B=N π-disruption with allyl-lithium. (B) X-ray crystal structure (ORTEP-type) of trans-7 at 50% thermal ellipsoids; H atoms omitted; (C) ¹³C NMR of 7.

The NMR of 7 showed a mixture of two stereoisomers in an 85/15 ratio, with the X-ray crystal structure of the major component revealing a trans-BN-decalin disposition (Scheme 4B and C). With only the minor cis isomer correctly disposed for ring closing, this stereochemical outcome posed a challenge to our endgame strategy. To test if at least the minor cis fraction is effective under ROM conditions, 7 – used as an isomer mixture – was subjected to the 1st generation Grubbs catalyst (G-I, 5 mol%) in CH₂Cl₂. After 16 h at room temperature, the cis isomer was selectively consumed, leading to the formation of the target BN-[4.4.4]propellane 8, while the trans-7 remained unreacted (Scheme 5A). Compound 8 was identified by its GC-MS molecular ion at m/z = 187, and was isolated as a stable, colorless crystalline solid in 13% yield—consistent with the 15% initial abundance of the reactive cis isomer in the mixture. The C₃ symmetry of 8 is reflected in its four-resonance ¹³C NMR spectrum, where the two olefinic peaks at 133 and 118 ppm are complemented by the two aliphatic resonances at 56 (NCH₂) and 21 ppm (BCH₂).

Scheme 5 (A) Initial test to form propellane 8, (B) ORTEP diagram of the X-ray structure of 8 at 50% thermal ellipsoids. (C) ¹H NMR of 8.

Crystals of 8 suitable for X-ray diffraction were grown by slow evaporation of a hexane solution. The X-ray crystal structure of 8 (Scheme 5B) confirms the expected molecular C₃-symmetry with a helical propeller-like arrangement of the three 6-membered lobes around the saturated B–N axis. While each individual molecule is thus chiral due to its propeller twist, the non-chiral (centrosymmetric) space group of P2_a/c leads to the unit cell containing both enantiomers. Applying the helix chirality descriptors M and P, we can see that the left structure in Scheme 5B possesses a left-handed screw arrangement, thus giving it an M configuration, while the structure on the right is P due to its right-handed screw arrangement. The twist is duly reflected by the ¹H NMR of 8, as shown in Scheme 5C, where the chair-like geometry of each 6-membered blade leads to distinguishable axial and equatorial CH positions in the CH₂ groups. The spectrum features two CH₂ groups, each as a diastereotopic pair of broad singlets: at 4.12 and 2.61 ppm for NCH₂, and at 1.19 and 0.69 ppm for BCH₂. Considering the helicity of the propellane core, this fluxional behaviour, as evidenced by such peak broadening, signals an interconversion (via a ring flip) of the two enantiomeric forms on the NMR timescale, leading to an exchange between axial and equatorial C–H positions, also illustrated in the lower part of Scheme 5B. For the 3D structure, a comparison with the crystal structure of the carbocyclic analogue SI-1 (see the SI), synthesized by Altman and co-workers in 1967 and analyzed by the Dunitz laboratory in 1971,^16,17 revealed subtle structural differences between the two isosteres, most notably in the elongation of the central axis from 1.55 Å for C–C to 1.66 Å for the B–N.

Seeking to overcome the unfavorable stereochemistry produced in the B–N diallylation of 1, we explored the possibility of implementing a metathesis-based cis–trans isomerization of 7. We envisioned that a sufficiently potent metathesis catalyst could both perform the ring closing in cis-7, and also catalyse an intramolecular metathesis between an allyl group and a ring olefin moiety in trans-7 yielding the C_s-symmetric intermediate 12 (Scheme 6),¹⁸ which could then undergo a second metathesis step to form the cis-7. The cis–trans equilibrium could then be driven by the ring closing of the cis component to propellane 8.

Scheme 6 Top: Mechanistic proposal for the trans-to-cis isomerization in 7; Bottom: initial test for the isomerization-ring closing hypothesis. ^aUsing 5 mol% catalyst in CH₂Cl₂ at room temperature.

As discussed above (see Scheme 5), the catalyst G-I proved to be ineffective for such isomerization, only showing activity for ring closing of cis-7. Nevertheless, in a further exploration with isolated trans-7, we became cautiously optimistic upon detecting trace amounts of 8 when applying the Grubbs-II catalyst (G-II), with even higher conversions (2–7%) observed using the 2nd generation Hoveyda–Grubbs (HG-II) or the 3rd generation Grubbs catalysts (G-III, Scheme 6, bottom).^19,20 Returning to the original trans/cis-7 (85/15) mixture, the best-performing HG-II (CH₂Cl₂, room temp., 5–10% catalyst loading) afforded a ∼20% yield of 8, i.e. just ∼5% above the precursor's cis content (Table 1, runs 1 and 2). The resulting product mixture also contained species 13 arising via olefin double bond migration, a phenomenon associated with catalysis by Ru–H species or Ru(0) nanoparticles stemming from the degradation of the Ru carbene catalyst.^21,22 Following literature precedents, the side reaction could be suppressed through the addition of p-benzoquinone,²¹ albeit initially without an increase in the yield of 8 (run 3).

Table 1 Optimization of the ring closing on trans/cis-7

Run	Cat.	% Cat.	Solvent	T [°C]	Addit.^a	Yield
a Using 20 mol% of para-benzoquinone when indicated. b The Mo-I catalyst used in entry 8 refers to 2,6-disopropylphenylimidoneophylidene molybdenum(VI) bis(hexafluoro-t-butoxide), CAS 139220-25-0.
1	HG-II	5 mol%	CH2Cl2	RT	—	20%
2	HG-II	10 mol%	CH2Cl2	RT	—	20%
3	HG-II	15 mol%	CH2Cl2	RT	p-BQ	15%
4	HG-II	15 mol%	Toluene	80	p-BQ	40%
5	HG-II	15 mol%	Toluene	110	p-BQ	46%
6	HG-II	2 × 5 mol%	Toluene	110	p-BQ	65%
7	G-III	2 × 5 mol%	Toluene	110	p-BQ	80%
8	Mo-I	15 mol%	Toluene	110	—	90%

---

Nevertheless, up to 40–50% yields were achieved in toluene at elevated temperatures (runs 4 and 5), reaching a 65% yield through catalyst addition in two 5% batches at 110 °C (run 6). To our delight, under the latter conditions the G-III catalyst, originally somewhat less active than HG-II, now led to the BN-[4.4.4]propellane 8 in 80% yield (run 7). A parallel study on other catalyst types also revealed an excellent performance from the Schrock Mo-based catalyst Mo-I,²³ which led to 8 in 90% yield, albeit with the need for rigorous air-free conditions (Table 1, run 8). The BN-[4.4.4]propellane 8 can thus be obtained from the predominantly trans B,N-diallyl precursor 7, either employing the bench-stable HG-II or G-III catalysts, or with the air-sensitive but highly efficient Schrock-type Mo catalyst. Further stability tests for 8, suggested by a referee, revealed that in addition to being air-and moisture stable, solutions of this propellane remain unchanged, as gauged by ¹H and ¹¹B NMR, after heating to 70 °C for 16 h, or after irradiation at 350 nm or 460 nm for 5 h (see the SI).

The tri-olefin 8 could be converted to saturated [4.4.4]BN-propellane 14 in 75% yield by smooth hydrogenation using Pd/C (Scheme 7, top), where this compound formed along with a minor unexpected isomer identified (see the SI) as [4.4.3]-BN-propellane 15, ostensibly due to Pd/C-catalyzed isomerization. As was previously observed for 8, compound 14 exhibits a fluxional ¹H NMR spectrum, with the molecule's helical twist (confirmed through DFT, see Scheme 7 and the SI) resulting in eight broad resonances between +4 and −1 ppm for the four chemically distinct CH₂ groups (see the SI). Preliminary exploration also reveals that the olefinic double bonds in 8 can undergo reactions such as cyclopropanation or epoxidation, albeit leading to a mixture of mono-, bis- and tris-functionalized species. For the sake of illustration, the epoxidation reaction, which can be driven to the double epoxide 16 with moderate selectivity, is shown in Scheme 7, bottom. The product is obtained in three isomer forms, as reflected by three peaks in the GC-MS, with isomers designated using “up” and “down” descriptors to indicate the direction of the epoxide units.

Scheme 7 Top: Transformation of 8 to the saturated BN-propellane 14, along with its DFT-based molecular geometry (B3LYP/6-31+g(d,p)); Bottom: double epoxidation of 8 and the illustration of the three isomeric epoxide products.

As previously noted, the interconversion between the enantiomeric helical forms of propellanes 8 and 14 results in a positional exchange between the equatorial and axial C–H sites in each CH₂ group. For 8, this process was assessed by variable temperature ¹H NMR in toluene-d₈. While the compound's two unique CH₂ groups appeared as four broad resonances at room temperature, cooling to 228 K led to signal sharpening, with each of the four signals displaying well-resolved geminal ²J(HH) coupling constants (16 Hz for the methylene protons α to nitrogen and 18.5 Hz for the methylene protons α to boron). As the temperature was gradually raised to 350 K, the accelerated interconversion led to the coalescence of diastereotopic pairs, initially for the more closely positioned B-CHH resonances (316 K, see the SI), and then for the wider-spaced N-CHH peaks (334 K, see Fig. 1). These coalescence temperatures (T_c) were used with the Eyring equation to calculate ΔG^‡ for the helical flip in 8, yielding internally highly consistent values of 14.64 and 14.73 kcal mol⁻¹ (see the SI).

Fig. 1 Exchange process and coalescence of the N-HH resonances.

For an even more accurate analysis, the experimental ΔG^‡ values were probed through a lineshape analysis using the DNMR software package in TopSpin (v. 4) to fit experimental vs. simulated spectra for each of the variable-temperature ¹H-NMR datapoints recorded. This procedure yielded an interconversion rate constant (k) for the methylene groups at each studied temperature (Fig. 2A). Based on these constants, Eyring plots were constructed by representing ln(k/T) against 1/T. The resulting plots showed a strong linear correlation, described by the equation f(x) = a + bx, where the slope and the intercept represent ΔH^‡ and ΔS^‡, respectively. Using these data and the expression ΔG^‡ = ΔH^‡ − TΔS^‡ thus allowed for the ΔG^‡ values for NHH and BHH to be calculated.

Fig. 2 (A) Simulated ¹H NMR spectrum obtained from DNMR at T = 263 K with a best overlap fitting obtained at 92.26%. vs. the experimental ¹H NMR spectrum at T = 263 K. (B) Eyring plot obtained for NCHH from DNMR extracted k_ex values.

For example, as shown in Fig. 2B, the Eyring plot corresponding to the α protons to N yielded a linear regression with an R² value of 99.9981, from which a ΔG^‡ value of 14.57 kcal mol⁻¹ was calculated. Following an analogous procedure, a ΔG^‡ value of 14.89 kcal mol⁻¹ was determined for the BHH protons (see the SI). We were pleased to observe that these values closely align with those obtained experimentally, further validating our results. We note that analogous dynamic behavior had been documented by Altman et al. for the carbonaceous propellane SI-1, with the ΔG^‡ for this “helical flip” in the BN-propellane 8 being approximately 2 kcal mol⁻¹ higher than that of its all-carbon isostere.²⁴ In contrast, this experimental barrier is almost 3 times higher than the 5–6 kcal mol⁻¹ ring inversion barrier in cyclohexene, a ring representing individual lobes in propellane 8. This is likely the result of the coordination required to flip all three rings within the same movement.²⁵

In summary, this work explores the synthetic potential of the B=N sp²–sp² unit as a platform for constructing more complex 3D architectures via double B/N functionalization. While previous studies demonstrated B=N derivatization through the addition of a protic [H⁺/OR⁻] pair, we now show that two carbon-based fragments can be introduced in a nucleophile/electrophile sequence, enabled by an initial disruption of π-donation using organolithium reagents. This approach allowed the synthesis of BN-[4.4.4]-propellane as a model 3D B–N sp³–sp³ target, where the inherently unfavorable trans-diallylation was overcome through a metathesis-based isomerization and ring-closing strategy. While the 2D BN compound will likely continue to gain importance in organic opto-electronic design, we envisage that this and other synthetic strategies may help launch a new area of BN mapping of both known and unknown 3D cores.

Author contributions

G. S.-L.: methodology, investigation, writing–original draft, review and editing; J. S.: NMR methodology, writing of the section on lineshape analysis; P. N.: variable temperature NMR; A. B. C. and A. S.: project conception, methodology, supervision, management, writing–original manuscript, revisions and final version. A. S. was also responsible for DFT calculations.

Conflicts of interest

There are no conflicts to declare.

Data availability

CCDC 2470569 and 2470570 contain the supplementary crystallographic data for this paper.^26a,b

The datasets supporting this article have been uploaded as part of the supplementary information (SI). Supplementary information: experimental procedures and compound characterization, NMR data (including VT NMR), as well as details on computational modelling by DFT. See DOI: https://doi.org/10.1039/d5sc05132k.

Acknowledgements

This work was supported through MICINN grant PID2020-113661GB-I00, MCIU grants PID2023-146324NB-I00 and PDC2023-145801-I00, and through AGAUR (2021 SGR 00520). We thank the Generalitat de Cataluña (AGAUR) for the Investigo contract to G. S.-L. (2023 INV-2 00014G1). We thank Dr János Rohonczy for his initial assistance in setting up the DNMR analysis.

Notes and references

1. (a) S. J. Blanksby and B. Ellison, Acc. Chem. Res., 2003, 36, 255–263; (b) R. H. Pritchard and C. W. Kern, J. Am. Chem. Soc., 1969, 91, 1631–1635; (c) I. Alkorta and J. Elguero, Struct. Chem., 1998, 9, 59–63.
2. (a) L. R. Thome, R. D. Suenram and F. J. Lovas, J. Chem. Phys., 1983, 78, 167–171; (b) M. Sugie, H. Takeo and C. Matsumara, Chem. Phys. Lett., 1979, 64, 573–575.
3. (a) C. R. McConnell and S.-Y. Liu, Chem. Soc. Rev., 2019, 48, 3436–3453; (b) A. J. Fritsch, J. Chem. Heterocycl. Compd., 1977, 30, 381–440; (c) H. Helten, Chem. Eur. J., 2016, 22, 12972–12982; (d) E. R. Abbey, L. N. Zakharov and S.-Y. Liu, J. Am. Chem. Soc., 2008, 130, 7250–7252.
4. (a) J. Y. Wang and J. Pei, Chin. Chem. Lett., 2016, 27, 1139–1146; (b) J. Huang and Y. Li, Front. Chem., 2018, 6, 341; (c) T. Sakamaki, T. Nakamuro, K. Yamashita, K. Hirata, R. Shang, E. Nakamura, R. Shang and E. Nakamura, Chem. Mater., 2021, 33, 5337–5344; (d) M. Tasior and D. T. Gryko, J. Org. Chem., 2016, 81, 6580–6586.
5. S. Ko, S. Shin, E. Ahn, J. Lyu, J. Lee and J. Han, EP 3940805A1, Samsung Display Co., Ltd, 2022.
6. (a) P. G. Campbell, A. J. V. Marwitz and S.-Y. Liu, Angew. Chem., Int. Ed., 2012, 51, 6074–6092; (b) M. J. D. Bosdet, W. E. Piers, T. S. Sorensen and M. Parvez, Angew. Chem., Int. Ed., 2007, 119, 5028–5031; (c) T. Kaehler, M. Bolte, H. W. Lerner and M. Wagner, Angew. Chem., Int. Ed., 2019, 58, 11379–11384; (d) P. F. Zhang, J. C. Zeng, F. D. Zhuang, K. X. Zhao, Z. H. Sun, Z. F. Yao, Y. Lu, X. Y. Wang, J. Y. Wang and J. Pei, Angew. Chem., Int. Ed., 2021, 60, 23313–23319.
7. (a) F. Lovering, J. Bikker and C. Humblet, Escape from Flatland: Increasing Saturation as an Approach to Improving Clinical Success, J. Med. Chem., 2009, 52, 6752–6756; (b) F. Lovering, Escape from Flatland 2: Complexity and Promiscuity, Med. Chem. Commun., 2013, 4, 515–519.
8. (a) A. D. Rohr, J. W. Kampf and A. J. Ashe, Organometallics, 2014, 33, 1318–1321; (b) F. Sun, L. Lv, M. Huang, Z. Zhou and X. Fang, Org. Lett., 2014, 16, 5024–5027.
9. F. Rulli, G. Sanz-Liarte, P. Roca, N. Martínez, V. Medina, R. Puig de la Bellacasa, A. Shafir and A. B. Cuenca, Chem. Sci., 2024, 15, 5674–5680.
10. B. Halton, R. Boese, D. Bläser and Q. Lu, An Improved Synthesis, Molecular-Structure, and Properties of 1,4-Dihydro-Dicyclopropa[b,g]naphthalene, Aust. J. Chem., 1991, 44, 265–276.
11. As a surrogate for the π bond energy, the rotation barrier in H₂B=NH₂ is ∼30 kcal mol⁻¹, less than half of that (∼65 kcal mol⁻¹) found for ethylene: D. D. J. Grant and D. A. Dixon, σ- and π-Bond Strengths in Main Group 3–5 Compounds, J. Phys. Chem. A, 2006, 110, 12955–12962.
12. A. M. Dilmaç, E. Spuling, A. de Meijere and S. Bräse, Propellanes—From a Chemical Curiosity to “Explosive” Materials and Natural Products, Angew. Chem., Int. Ed., 2017, 56, 5684–5718.
13. J. B. Geri, E. Y. Aguilera and N. K. Szymczak, Difluoromethane as a precursor to difluoromethyl borates, Chem. Commun., 2019, 55, 5119–5122.
14. A. Abengózar, M. A. Fernández-González, D. Sucunza, L. M. Frutos, A. Salgado, P. García-García and J. J. Vaquero, C–H Functionalization of BN-Aromatics Promoted by Addition of Organolithium Compounds to the Boron Atom, Org. Lett., 2018, 20, 4902–4906.
15. A. Nakazaki, A. Mori, S. Kobayashi and T. Nishikawa, Diastereoselective synthesis of 3,3-disubstituted oxindoles from atropisomeric N-aryl oxindole derivatives, Tetrahedron Lett., 2012, 53, 7131–7134.
16. J. Altman, E. Babad, J. Pucknat, N. Reshef and D. Ginsburg, Tetrahedron, 1968, 24, 975–998.
17. O. Ermer, R. Gerdil and J. D. Dunitz, Helv. Chim. Acta, 1971, 54, 2476–2487.
18. For prior examples of transannulative ring closing metathesis, see: (a) K. Takao, R. Nanamiya, Y. Fukushima, A. Namba, K. Yoshida and K. Tadano, Org. Lett., 2013, 15, 5582–5585; (b) J. P. A. Harrity, M. S. Visser, J. D. Gleason and A. H. Hoveyda, J. Am. Chem. Soc., 1997, 119, 1488–1489; (c) J. A. Henderson and A. J. Phillips, Angew. Chem., Int. Ed., 2008, 47, 8499–8501.
19. For the development of this 3rd generation Ru catalyst, see: (a) J. A. Love, J. P. Morgan, T. M. Trnka and R. H. Grubbs, Angew. Chem., Int. Ed., 2002, 41, 4035–4037; (b) D. J. Walsh, S. H. Lau, M. G. Hyatt and D. Guironnet, J. Am. Chem. Soc., 2017, 139, 13644–13647.
20. For metathesis-based cis–trans isomerization, we also cannot rule out a sequence operating via ethylenolysis of the ring olefin groups as shown below:.
21. (a) S. H. Hong, M. W. Day and R. H. Grubbs, J. Am. Chem. Soc., 2004, 126, 7414–7415; (b) S. H. Hong, D. P. Sanders, C. W. Lee and R. H. Grubbs, J. Am. Chem. Soc., 2005, 127, 17160–17161.
22. (a) C. S. Higman, A. E. Lanterna, M. L. Marin, J. C. Scaiano and D. E. Fogg, Catalyst Decomposition During Olefin Metathesis Yields Isomerization-Active Ru Nanoparticles, ChemCatChem, 2016, 8, 2446–2449 , For a critical review of isomerization in metathesis,see; (b) B. J. van Lierop, J. A. M. Lummiss and D. E. Fogg, Ring-Closing Metathesis, in Olefin Metathesis – Theory and Practice, ed. K. Grela, Wiley, Hoboken, 2014, pp. 85–152.
23. (a) J. S. Murdzek and R. R. Schrock, Organometallics, 1987, 6, 1373–1374; (b) R. R. Schrock, J. S. Murdzek, G. C. Bazan, J. Robbins, M. DiMare and M. O'Regan, J. Am. Chem. Soc., 1990, 112, 3875–3886; (c) G. C. Bazan, J. H. Oskam, H. N. Cho, L. Y. Park and R. R. Schrock, J. Am. Chem. Soc., 1991, 113, 6899–6907.
24. H. Gilboa, J. Altman and A. Loewenstein, J. Am. Chem. Soc., 1969, 91, 6062–6065.
25. H. Choo and J. Laane, The Barrier to Conformational Interconversion of Cyclohexene, J. Am. Chem. Soc., 1994, 116, 3889–3891.
26. (a) CCDC 2470569: Experimental Crystal Structure Determination, 2025,  DOI:10.5517/ccdc.csd.cc2nxts6; (b) CCDC 2470570: Experimental Crystal Structure Determination, 2025,  DOI:10.5517/ccdc.csd.cc2nxtt7.

---