Experimental Characterization and Computational Study of Unique C,N-Chelated Lithium Dianions

Abstract

The structure of a unique C,N-chelated dilithio dianion has been established as a solvated monomeric species using a combination of NMR and computational techniques. The highly ordered structure of the dianion may be important in its reactivity in an oxidative C-N bond-forming process.

Structural studies of lithiated anions in solution are important given the ubiquity of these intermediates in organic synthesis.¹ Despite the remarkable advances that have been made in the investigation of C-, O-, and N-lithiated species, dianionic compounds (e.g., C–Li/N–Li) have received relatively little attention.² During a recent total synthesis study of the Lycopodium alkaloid lyconadin A (4), Sarpong et al. discovered that amine 1 was converted to pentacycle 3 by the sequential addition of n-BuLi (2 equiv) and iodine (1 equiv).³ This C–N bond forming reaction was hypothesized to proceed via the intermediacy of a C, N-chelated lithium dianion intermediate (2) of ill-defined structure.

To gain insight into the solution structure of the dianion intermediates, which may enable the generalization of the C–N bond-forming sequence, we have undertaken an in-depth structural study of these unique species. In this communication, we report the synthesis of a ¹⁵N-incorporated iso-topomer related to 1 as well as NMR and computational studies that have resulted in the structural characterization of the affiliated lithium dianion.

Because of its ease of synthesis, desmethyl analog [¹⁵N]8 (Scheme 2), lacking a methyl group on the A-ring, was targeted for synthesis and subsequent NMR spectroscopic studies. Coupling 5⁴ with [¹⁵N]NH₄Cl using standard amidation conditions⁵ gave [¹⁵N]6 in 91% yield.

Scheme 2.

Hofmann rearrangement of [¹⁵N]6, mediated by Pb(OAc)₄, in the presence of benzyl alcohol yielded Cbz-protected amine [¹⁵N]7, which was converted to [¹⁵N]8 in three steps using a previously established protocol.⁶

With [¹⁵N]8 in hand, we performed a series of ⁶Li and ¹⁵N NMR studies. Treatment of [¹⁵N]8 in THF-d₈ at −78 °C with [⁶Li]n-BuLi (2 equiv) recrystallized from pentane⁷ afforded the putative dianion. The ⁶Li NMR spectrum of the dianion generated from [¹⁵N]8—recorded at −90 °C (Figure 1A)—contains two distinct doublets, indicating that each lithium is bound and couples to the isotopically labeled nitrogen. The ¹⁵N decoupled ⁶Li NMR observations (Figure 1B) shows two singlets. The ¹⁵N NMR spectrum, also recorded at −90 °C (Figure 1C), contains a quintet, which confirms the Li–N–Li connectivity.

Figure 1.

NMR spectra of key resonances of the dianion of [¹⁵N]8. (A) ⁶Li NMR spectrum of 0.05 M [¹⁵N]8 and 2.0 equiv [⁶Li]n-BuLi in THF-d₈ at −90 °C. (B) {¹⁵N}⁶Li NMR spectrum of 0.05 M [¹⁵N]8 and 2.0 equiv [⁶Li]n-BuLi in THF-d₈ at −90 °C. (C) ¹⁵N NMR spectrum of 0.05 M [¹⁵N]8 and 2.0 equiv [⁶Li]n-BuLi in THF-d₈ at −90 °C. (D) ¹³C NMR spectrum of 0.025 M 8 and 2.0 equiv [⁶Li]n-BuLi in THF-d₈ at −100 °C, expanded around the benzylic carbon resonance. *Denotes an impurity from [⁶Li]n-BuLi.

The structures of open dianion isomer 9 (Figure 2) and the cyclic isomer 10 are consistent with the ⁶Li and ¹⁵N NMR spectral data and are well-precedented structural motifs.^2,8 Dilithio compound 11 also seems plausible but is inconsistent with the data.

Figure 2.

Possible structures for dilithio anions generated from 8.

¹³C NMR spectra were recorded at −100 °C (Figure 1D) to distinguish open isomer 9 from the chelated dilithio compound 10 by ⁶Li–¹³C coupling. A combination of 2-D NMR techniques including ¹H–¹H gCOSY, ¹H–¹³C gHMBC, and ¹H–¹³C HSQC were used to identify the chemical shift of the benzylic carbon. Unfortunately, the ¹³C spectrum shows no signs of ⁶Li–¹³C scalar coupling. This absence is consistent with previous observations. For example, Fraenkel et al. described difficulties observing an analogous coupling in benzyllithium that could be traced, in part, to small ¹J(⁶Li–¹³C) coupling.⁹

We turned to density functional theory computational studies to provide insight into the structure of the dianion intermediate. The relative free energies (ΔG, kcal/mol) of various forms of the dianion intermediate at −78 °C were calculated at the B3LYP/6-31G(d) level of theory with single-point MP2 correction.^10,11 The relative energies for the computationally most viable dilithio isomers are illustrated in Scheme 3 in monotonically increasing stability. Tetrasolvated open dianion isomer 9c was calculated to be most stable in energy. Additional structures and their associated relative energies are reported in the Supporting Information. On the basis of these results, we conclude that open isomers of type 9 benefit from high solvation numbers.

Scheme 3.

Our studies support the existence of the C, N-chelated dilithio dianion as a monomeric species that is solvated by four molecules of solvent (see 9c) at concentrations that are relevant to oxidative conversion to 3 (see Scheme 1).¹² Furthermore, we have amassed support for the persistence of 9c as a highly organized dianion at low temperature. This structural organization may be critical to the oxidative C–N bond forming process. We expect that ¹⁵N and ⁶Li NMR spectroscopy in conjunction with the computational methods used to characterize dianion 9c could be extended to the characterization of other C,N-chelated dianions. Using this combined approach, the solution structures of a range of such C, N-chelated dianions may be readily determined. These insights will serve as a starting point in our efforts to extend the scope of the oxidative C–N bond-forming reaction.

Scheme 1.