Chiral Jahn-Teller Distortion in Quasi-Planar Boron Clusters

1. Introduction

While they are not experimental observables, molecular orbitals are conceptually useful and elegant tool for elucidating molecular properties [1]. They have long been a significant tool in the arsenal of chemists [2,3,4,5,6], tracing back to the early work of Hückel, Mulliken, and others. One prominent example is the principle of conservation of orbital symmetry [7,8,9], which subsumes the Woodward-Hoffmann rules. Helical frontier molecular orbitals, first introduced by Hendon et al. [10] in 2013, have seeded a surge of interest. Helical frontier molecular orbitals appear in disubstituted allenes and even-n cumulenes. Later, many more types of molecules possessing this interesting property were reported [11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29], and it was discovered that molecules with helical orbitals have interesting physicochemical properties. For example, oligoyne-bridged bifluorenes can induce spin–orbit coupling [29].

Boron forms clusters with unique bonding, aromaticity, and reactivity properties [30,31,32]. Very recently [33], we have observed helical spin densities of anionic boron clusters. In this work, we report that ${\mathrm{B}}_{16}^{-}$ [34], ${\mathrm{B}}_{20}^{-}$ [35], ${\mathrm{B}}_{24}^{-}$ [36], and ${\mathrm{B}}_{28}^{-}$ [37] (see Scheme 1) not only have helical molecular orbitals, but also helical spin densities. This is interesting because, unlike molecular orbitals, spin-densities are experimental observables, and this allows the edifice of (spin)-resolved (conceptual) density-functional theory [38,39,40,41,42,43,44,45,46,47,48,49] to be directly applied to these compounds.

To pin down the origin of spin-density helicity, we forced the quasi-planar boron atoms to be exactly in a plane: helical spin densities are no longer observed. Thus, it is chiral Jahn-Teller distortion that governs the formation of helical spin densities. This seems to be the first observation of this intriguing phenomenon in inorganic boron clusters. We also show that elongation or enlargement of helical molecular orbitals can be achieved by simply adding more structural motifs via a linker. Moreover, we have exhibited that helical-shape molecules have a large propensity to assume helical molecular orbitals as shown in inorganic species ${\mathrm{P}}_9^{+}$ [50], Be₆${\mathrm{B}}_{11}^{-}$ [51], and ${\mathrm{A}\mathrm{s}}_{3-}^{11}$ [52] (vide infra).

2. Results

To characterize the planarity of anionic boron clusters (in both the ground- and excited-state), we used a few parameters [53], including molecular planarity parameter (MPP), span of deviation from plane (SDP), maximum positive/negative deviation (MPD/MND) to the fitted plane, as listed in Table 1. The fitted parameters of a plane are listed in Table S1. One can easily discover that all the systems studied in this work are quasi-planar. Based upon the optimized structures in the ground state, we have observed helical spin densities as exhibited in Figure 1. No similar results are discerned for the excited-state structures (see Figure S1).

To elucidate the origin of helical spin densities, we force the ground-state quasi-planar boron structures to be exactly planar followed by single-point calculations at the PBE0/6-311+G(d) [55,56] level. The spin-densities’ helicity then vanishes (see Figure S2). Accordingly, a chiral Jahn-Teller distortion plays a key role where the right- and left-handed deformations are (quasi)equal in energy, and the planar structure deforms slightly to break symmetry, thus lowering in energy. More intriguingly, these chiral structures [in terms of vibrational circular dichroism (VCD) spectra, see Figure 2 for details] can also have helical frontier molecular orbitals as shown in Figure 3. Of note, excited-state VCD spectra are also observed as shown in Figure S3. In the ground state, except for ${\mathrm{B}}_{24}^{-}, {\mathrm{B}}_{16}^{-}, {\mathrm{B}}_{20}^{-}$, and ${\mathrm{B}}_{28}^{-}$ indeed have helical β–LUMOs (with lower orbital energies than their α counterparts) and ${\mathrm{B}}_{16}^{-}$ and ${\mathrm{B}}_{20}^{-}$ also have β–HOMOs in Figure 3; in the excited state, only ${\mathrm{B}}_{20}^{-}$, and ${\mathrm{B}}_{28}^{-}$ have helical β–HOMOs.

Figure 4 shows the GIMIC (gauge-including magnetically induced current) [59,60] distributions of both ground- and excited-state ${\mathrm{B}}_{16}^{-}, {\mathrm{B}}_{20}^{-}, {\mathrm{B}}_{24}^{-}$, and ${\mathrm{B}}_{28}^{-}$. For ²${\mathrm{B}}_{16}^{-}$ and ⁴${\mathrm{B}}_{16}^{-}$, the induced electric currents are running counter-clockwise, which is indicative of aromaticity, as evidenced by the negative NICS (nucleus-independent chemical shift) [61] values in Table 2. Similar results are observed for ${\mathrm{B}}_{24}^{-}$ and ${\mathrm{B}}_{28}^{-}$. However, this is not the case for ${\mathrm{B}}_{20}^{-}$. The overall effect indicates that ${\mathrm{B}}_{20}^{-}$ is antiaromatic while the Z-component of the induced electric current also runs in a counter-clockwise manner as showcased by the NICS_ZZ values in Table 2. Yet, the dominant contributions of the induced electric current lie in the x-y plane which is the source of antiaromaticity. To go a step further, Figure 5 showcases the 3D isotropic shielding surface (ICSS) [62] calculations for both ground- and excited-state boron clusters and it is clearly revealed that there exist strongly shielded chemical environment along the direction perpendicular to the quasi-planar boron clusters.

Hyperfine coupling constants [63,64] provide a direct experimental measure of the distribution of unpaired spin density in paramagnetic molecules. The interactions of unpaired electrons with external magnetic fields arise from the Zeeman effect and from the hyperfine coupling with nuclei having nonzero spins. The latter contribution is related to the chemical environment. For each nucleus $N$ of a molecule located at ${\mathbf{r}}_N$, the isotropic component of the hyperfine interaction tensor, $a\left(N\right)$, is related to the local spin density through [65]

$$a\left(N\right)=\frac{8\pi }{3}{\beta }_{\mathrm{e}}{\beta }_N{g}_N\sum _{\mu \nu }{\mathbf{P}}_{\mu \nu }^{\alpha -\beta }⟨{\phi }_{\mu }|\delta (\mathbf{r}-{\mathbf{r}}_N)|{\phi }_{\nu }⟩$$

(1)

where ${\beta }_e$, ${\beta }_N$, and $g_N$ are the electronic and nuclear magnetons and the nuclear magnetogiric ratio, the indices $\mu$ and $\nu$ run over the basis functions, ${\mathbf{P}}_{\mu \nu }^{\alpha -\beta }$ is the difference between the density matrices of spin $\alpha$ and spin $\beta$ electrons and $\delta (\mathbf{r}-{\mathbf{r}}_N)$ is the Dirac delta function. Therefore, once the density matrices for different spins have been determined the calculation of $a\left(N\right)$ for each nucleus is achieved in a straightforward way. The (isotropic) hyperfine coupling tensor, ${\mathbf{A}}_{\mathrm{i}\mathrm{s}\mathrm{o}}^N$, consists of the Fermi contact term (${\mathbf{A}}_{\mathrm{F}\mathrm{C}}^N$) and a spin orbit correction, the pseudocontact term (${\mathbf{A}}_{\mathrm{P}\mathrm{C}}^N$).

Shown in Table 3 are the isotropic NMR shielding (α_iso) constants and hyperfine coupling (A_iso) constants for both ground- and excited-state ${\mathrm{B}}_{16}^{-}$ at the PBE0/pcJ-2 [66,67] level. It is clearly shown that the 16 boron atoms can be roughly grouped into 5 different atoms in different chemical environments as evidenced both by the α_iso and A_iso data. Among all the boron atoms, one can easily see that atoms 5 and 6 (as shown in Scheme 1), lying at the two ends of the middle line composed of atoms 1 ‒ 6, are the most unique. For example, they have the least positive α_iso values and the most negative A_iso data. In addition, they undergo the largest changes when going from the ground state to the excited state. Specifically, A_iso changes by ~18 MHz while the largest change of the other atoms is ~8 MHz. Similar trends can be observed for ${\mathrm{B}}_{20}^{-}$, ${\mathrm{B}}_{24}^{-}$, and ${\mathrm{B}}_{28}^{-}$ as shown in Table 4, Table 5 and Table 6.

3. Discussion

Helical frontier molecular orbitals were reported first for hydrocarbon systems, then also for boron-containing molecules [10]. In this work, we have also observed similar results for quasi-planar boron clusters. In addition, as shown in Figure 6, for some other helical (inorganic) motifs, ${\mathrm{P}}_9^{+}$, Be₆${\mathrm{B}}_{11}^{-}$ (a B₁₁ helical structure plus a distorted prism of Be₆), and ${\mathrm{A}\mathrm{s}}_{11}^{3-}$, helical frontier molecular orbitals are also observed. Is this helicity a ubiquitous phenomenon or a special feature of some elements in special molecular topologies? This seems to be an open question, and will be a topic for future research.

While a systematic rule for designing molecular templates with helical spin-densities is unknown to us, we can show how to elongate the helical frontier molecular orbitals from a given template structure. For example, starting from a conjugated hydrocarbon molecule 1 with helical frontier molecular orbitals, combining two monomers of 1 and a linker, such as CH₂(1)₂, [NH₂(1)₂]⁺, and [OH(1)₂]⁺, leads to elongated or enlarged helical frontier molecular orbitals as shown in Figure 7. Yet, when three or four monomers of 1 are grouped together, such as CH(1)₃ or C(1)₄, the helical frontier molecular orbitals are no longer elongated (results not shown). Is it possible to generate an infinite chain of (1)_∞? We do not know; possibly other linkers would work better. For the anionic boron clusters, we failed to even generate a dimer of ${\mathrm{B}}_{16}^{-}$; this is presumably because the repulsion between the anionic monomers prevents electron delocalization between them.

Finally, we have to point out that in a broader sense, dissection of chiral boron clusters and the electron spin should be beneficial to its applications to chiral spintronics and materials [68,69,70].

4. Materials and Methods

For molecular systems, structure optimization was performed at the density functional theory (DFT) [71,72] PBE0/6-311+G(d) level. Stability of molecular wavefunctions was confirmed via keywords of “guess=mix” and “stable=opt” in Gaussian 16 [73]. Vibrational frequency calculations were ensued to make sure that all the structures were true local minima on the potential energy surface. The optimized atomic Cartesian coordinates are supplied in the Supplementary Materials. Multireference (MR) characteristics of all boron clusters were checked via the T1 diagnostics [74] at the coupled cluster theory with single and double substitutions [CCSD/6-311+G(d)] and the frozen core formalism was used for CCSD calculations. The reported values are 0.036, 0.049, 0.034, and 0.041 for the ground state; 0.041, 0.041, 0.040, and 0.037 for the excited state, indicative of substantial multi-reference characteristics (because their T1 > 0.02). To further analyze the aromaticity properties of boron clusters, we employed PBE0/pcJ-2 to calculate the global NICS (nucleus-independent chemical shift) values and GIMIC (gauge-including magnetically induced current) distributions. NMR chemical shielding constants and isotropic hyperfine coupling parameters were obtained at the PBE0/pcJ-2 level with default gauge-including atomic orbitals (GIAOs) [75,76,77,78,79]. All DFT calculations were performed by using the Gaussian 16 package with tight self-consistent field (SCF) convergence criteria and ultrafine integration grids to ensure good accuracy. The Multiwfn [80] software was used to analyze the planarity of boron clusters and prepare the ICSS input files.

5. Summary

We have observed that helical molecular orbitals and helical spin densities can coexist in a chiral quasi-planar boron cluster. We show that this intriguing phenomenon emerges due to chiral Jahn-Teller distortion of planar boron clusters and show how to generate elongated or enhanced helical molecular orbitals by grouping a building blocks together via a linker. Finally, we found that helical inorganic species have a strong propensity to assume helical molecular orbitals. For potential applications in spintronics, it is interesting to study whether the observed helical spin densities would still be observed for boron clusters adsorbed on metal surfaces. Work along these lines is in progress.