Investigation of olefins as a cocatalyst for halonium catalysis

Abstract

Olefins are rarely used as catalysts in chemical reactions. Here, we explore the use of olefins as cocatalysts to facilitate iodonium catalysis. Our catalytic system employs a simple olefin as a cocatalyst with a simple halogen source N-iodosuccinimide (NIS) in hexafluoroisopropanol (HFIP) to form the active iodonium catalyst. Kinetic studies indicate that the olefin cocatalyst structure impacts the reaction rate significantly. In addition, NMR, kinetics, IR and computational studies were incorporated to understand the role of the olefins in these catalytic reactions, opening doors to other potentially interesting catalytic designs.

Cofactors are widely employed in nature, but the utilization of novel cofactors/cocatalysts in chemical reactions remains an underexplored concept (Scheme 1a).^1–3 Our interest lies in the exploration of olefins as versatile cocatalysts in organic synthesis. One notable example is the use of olefins in palladium/norbornene cooperative catalysis, also known as the Catellani reaction.⁴ The unique structural features of the norbornene facilitate and guide the palladium catalyst in crucial catalytic steps, enabling a wide range of intriguing chemical transformations (Scheme 1b).⁵ Olefins have also recently undergone important advancements in enhancing the Ni(0) catalysts’ stability while broadening the catalyst reactivities (Scheme 1c).^6,7 This concept of olefins as catalysts is similarly evident in shuttle catalysis, where an olefin is masked to transport reagents to another olefin structure, eliminating the need for the direct use of potentially hazardous chemical reagents such as HCN (Scheme 1d).⁸ These innovative catalytic applications of olefins inspired us to probe if classic iodonium species,⁹ generated in situ from olefin and iodine precursors, could deviate from traditional olefin functionalization pathway a, and instead proceed to pathway b to activate Lewis bases for functionalization (Scheme 1e).^10,11

Scheme 1.

Background on olefins as cocatalysts.

Our study commenced with adamantanol 1 and anisole 2 as standard substrates. The chosen catalytic conditions consisted of 1-methylcyclohexene 3 and NIS for iodonium formation.¹² The selection of a sterically demanding tri-substituted aliphatic olefin was aimed to minimize any potential olefin halo-functionalization while facilitating the requisite Lewis base activation.¹³ We were able to quickly identify HFIP as the optimal solvent to yield the desired product 4 in near quantitative yield (Scheme 2a).¹⁴ Control reactions confirmed the essential role of each component in the catalytic system. Having demonstrated the catalytic necessity of the olefin, we proceeded to examine the olefin structural impact on the reaction rate. Based on the standard reaction settings, we examined various olefins labeled A to G with different substitution patterns and electronic properties and summarized the rate of product formation in Scheme 2b. Notably, electron-deficient and -rich styrenes A and C exhibited prolonged induction periods, while electron-neutral olefins B and D displayed similar kinetic profiles. Trans-5-decene E also had a long induction time before minimal product formation. Our standard tri-substituted olefin 3 demonstrated a steady rate of product formation. Interestingly, the tetra-substituted olefin F showed a similar kinetic profile as B and D. Furthermore, no induction period was observed when either an in situ generated iodonium (3-Pregen) from olefin 3 or a pre-synthesized diaryliodonium H was used.¹⁵ The most sterically hindered adamantylideneadamantane G resulted in no product formation, likely due to the blockage of the alcohol binding sites caused by the adamantane’s steric hindrance.

Scheme 2.

(a) Reaction settings for iodonium catalysis. (b) Kinetic studies of olefin structure on the overall reaction rate.

To rationalize the interesting divergent behaviors of these different classes of olefins, we first determined the rate order of the olefin cocatalyst 3 based on the standard assay conditions (Scheme 3a). The reaction was observed to be first order in the olefin cocatalyst, suggesting that an iodonium derived from an olefin is involved in the transition state. Strikingly, the similar kinetic profiles between olefin B, D, F and 3-Pregen seemed surprising at an initial glance. However, when the structure of each iodonium is examined, the variation in the kinetic profiles likely derives from the varying stereo-electronic nature of the resulting iodonium species (Scheme 3b). In these cases, the similar kinetic profiles between olefins B, D, F and 3-Pregen are likely the result of the similarity between the corresponding iodonium structures, in which both binding sites are unaffected by the olefin backbone. The steric environment of the active binding sites of these iodoniums is highly similar with minimal electronic variation of the olefin backbone. The lack of induction time using either the known diaryliodonium H or in situ generated iodonium (3-Pregen) supports the notion that the iodonium formation is likely responsible for the induction time. With that in mind, the observation that different olefin cocatalysts exhibited different induction times is likely a result of how the iodonium catalyst is generated in situ. We also examined the cause of the abnormally long induction time for olefins A, C, and E. For olefin A, the electron-withdrawing group attached to the olefin significantly slowed the iodonium formation, which was confirmed by ¹H-NMR studies where the olefin cocatalyst A remained unreacted after 2 h (see Fig. S23 in SI page 43). Under the same conditions for olefin C, we observed byproducts derived from HFIP or succinimide¹⁶ addition to the iodonium, which essentially terminated the iodonium catalytic activity (see Fig. S22 in SI page 43). The lack of reactivity from olefin E is perhaps the most surprising here, in which our ¹H-NMR study demonstrated the same type of byproduct formation as with olefin C (see Fig. S25 in SI page 46). The lack of catalytic activity from these iodo-byproducts validated two features: (1) the in situ generation of the iodonium intermediate; and (2) these byproducts are stable and not viable as catalytic precursors. The lack of catalytic reactivity from certain alkenes can then be summarized as the result of (1) the lack of iodonium formation (olefin A); or (2) the competing olefin halo-functionalization pathway (olefins C and E); or (3) iodonium binding sites being blocked (olefin G).

Scheme 3.

(a) Olefin rate order studies. (b) Rationale of olefin kinetic studies.

To gain a better understanding of the catalytic system, we conducted an experiment to validate the recoverability of the olefin catalyst. In this case, the low molecular weight of the standard olefin catalyst 3 makes it difficult for recovery, and instead we used a less volatile olefin cocatalyst I in the standard reaction (Scheme 4). We observed 74% yield of product formation, and as mentioned, the control reaction with only NIS did not work. After reaction completion, we were able to recover 84% of the olefin cocatalyst I, confirming the recyclability of the olefin cocatalyst. Importantly, we conducted the Hammett studies based on substituted benzylic alcohols (Scheme 5a). We observed a rho value of −0.89, which suggests a positive charge build up at the benzylic carbon position. The small rho value observed here contrasts with the large rho values often observed in processes involving carbocation formation as the rate limiting step or typical electrophilic aromatic substitutions.^17,18 We have also benchmarked and observed that this catalytic system is comparable, and in some cases, superior to well-known Brønsted or Lewis acid catalysts (see SI page 20).

Scheme 4.

Olefin recovery experiment.

Scheme 5.

(a) Hammett studies. (b) ¹³C-NMR alkene region zoomed in and IR time studies for evidence of iodonium formation. (i) Mixing alkene 3, NIS and HFIP for 2 h. (ii) Alkene 3. (iii) In situ IR studies of mixing 3, NIS and HFIP over time. (c) Calculated potential energy profile of a proposed mechanism. ΔG is defined as the free energy difference between a given state and the immediately preceding state.

To understand the iodonium intermediate, we conducted ¹³C-NMR analyses of the catalytic reagents as depicted in the overlaid spectra in Scheme 5b. Upon mixing NIS with olefin 3 or HFIP with 3, no discernible new signals emerged. However, when a stoichiometric amount of HFIP reagent was introduced alongside NIS and the olefin and the mixture was stirred for 2 h, the two olefinic ¹³C-NMR signals of 3 completely disappeared with four new sets of ¹³C NMR signals appearing in the olefin region. These new ¹³C NMR signals are likely iodonium intermediates with either succinimide or HFIP anions as the chemical shifts of these intermediates align closely with ¹³C NMR studies conducted by Brown et al. on similar iodonium species.^9c Furthermore, IR studies of the catalyst mixture revealed characteristic C–I stretches indicative of iodonium species. In the NMR and IR studies, the observed rate dependency on the alkene, and the iodo-byproduct formations from olefins C and E, collectively, provide compelling evidence for the iodonium as the key catalytic intermediate in this dehydrative arylation reaction. However, we cannot conclusively rule out Brønsted acid catalysis as the addition of a proton sponge could also inhibit the active halogen source.

The reaction mechanisms based on the iodonium catalytic system were studied computationally via density functional theory (DFT). DFT calculations were performed with Gaussian 16 (Rev. C01) using the M06–2X density functional in combination with the def2-TZVP basis set, with an additional diffuse function (def2-TZVPD) basis set and the corresponding pseudopotential for iodine.^19–21 Additional single point calculations were performed using M06–2X-D3/def2-TZVPD with the same corresponding pseudopotential on the optimized geometries.²² The transition state structure, along with the corresponding Gibbs free energy profile, are depicted in the Scheme 5c, with solid lines denoting equilibrium states (reactants, stable intermediates, and products) and dashed lines denoting transition states. Our computational studies suggest that in the catalytic system (NIS, HFIP and 3), the iodonium formation is an endothermic process (Scheme 5c).²³ The computational studies suggest that the carbocation formation is the rate-limiting step. The transition state involves the olefin transferring the I⁺ onto the tertiary alcohol prior to carbocation formation. The transition state LUMO reveals that the positive charge is mostly localized on the iodine atom with minimal positive charge localized on the carbon atom. This observation is consistent with the small rho value observed in our Hammett studies (Scheme 5a). This phenomenon highlights an interesting mechanistic divergence compared to typical carbocation-based electrophilic aromatic substitutions.^17,18

Finally, we have conducted preliminary substrate scope studies for this reaction (Scheme 6). Different arene nucleophiles including N-tosyl-pyrrole, N-tosyl-indole and benzothiophene participated smoothly in this reaction to afford single constitutional isomeric products 5–7, respectively. Acylaniline also worked efficiently with exceptional site selectivity to generate product 8. For alcohols, tertiary and secondary benzylic alcohols worked well to afford the desired products 9 and 10 in good yields using DCE as a solvent instead. Tertiary aliphatic alcohols also proceeded to generate the coupled products 11 and 12 in good yields. In the case of product 13, a high diastereoselective outcome, guided by the neighboring stereocenter, was obtained. Secondary and primary aliphatic alcohols, however, are not effective substrates under the current catalytic reaction conditions.

Scheme 6.

Preliminary scope.^{a a} Standard conditions. ^b 50 °C. ^c DCE as a solvent, alkene 3 (20 mol%), NIS (20 mol%), 50 °C. ^d Ar is N-Ts-indole. ^e Ar is N-Ts-pyrrole, 50 °C. ^f Ar is N-Ts-pyrrole.

In summary, we have successfully introduced a proof-of-concept for incorporating olefins as cocatalysts in iodonium catalysis. The role of olefins is via the formation of an iodonium catalyst to transfer the active halogen source to activate the alcohol substrate. Our kinetic studies have demonstrated that the olefin structure has a significant impact on the overall reaction rate. These reactions not only validate the feasibility of a novel approach to utilize iodoniums as catalysts but also pave the way to explore new catalytic avenues for chemical reaction engineering.

Data availability

The data underlying this study are available in the published article and its supplementary information (SI). Supplementary information: spectroscopic characterization and mechanistic studies. See DOI: https://doi.org/10.1039/d5cc07020a.