Capacitive Organic Anode Based on Fluorinated‐Contorted Hexabenzocoronene: Applicable to Lithium‐Ion and Sodium‐Ion Storage Cells

Abstract Conducting polymer‐based organic electrochemical capacitor materials have attracted attention because of their highly conductive nature and highly reversible redox reactions on the surface of electrodes. However, owing to their poor stabilities in aprotic electrolytes, alternative organic electrochemical capacitive electrodes are being actively sought. Here, fluorine atoms are introduced into contorted hexabenzocoronene (cHBC) to achieve the first small‐molecule‐based organic capacitive energy‐storage cells that operate at high current rates with satisfactory specific capacities of ≈160 mA h g−1 and superior cycle capabilities (>400) without changing significantly. This high capacitive behavior in the P21/c crystal phase of fluorinated cHBC (F—cHBC) is caused mainly by the fluorine atoms at the end of each peripheral aromatic ring. Combined Monte Carlo simulations and density functional theory (DFT) calculations show that the most electronegative fluorine atoms accelerate ion diffusion on the surface to promote fast Li+ ion uptake and release by an applied current. Moreover, F—cHBC has potential applications as the capacitive anode in Na‐ion storage cells. The fast dynamics of its capacitive behavior allow it to deliver a specific capacity of 65 mA h g−1 at a high current of 4000 mA g−1.


HOMO and LUMO energy levels
To estimate the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) energy levels of cHBC and F-cHBC molecules, spin-polarized density functional theory (DFT) calculations were conducted using the DMol3 module [5,6] of the Materials Studio 2018. [7] The Becke's three-parameter hybrid exchange functional combined with the Lee-Yang-Parr correlation functional (B3LYP) [8,9] was employed for describing the exchange-correlation potential of electrons. The DNP 4.4 basis set was used with a global orbital cutoff of 3.7 Å. The core electrons were explicitly treated as all electrons with relativistic effect. The long-range van der Waals interactions were corrected using the Grimme's method. [10] The self-consistent field calculation was performed with the fixed orbital occupancy, until the convergence criterion of 1.0 × 10 -6 was satisfied. The convergence criteria for geometry optimization were set to 1.0 × 10 -5 Ha for the maximum energy change, 0.002 Ha Å -1 for the maximum force, and 0.005 Å for the maximum displacement, respectively.

Crystal structure prediction
The computational polymorphism study of F-cHBC was carried out using the Polymorph module of the Materials Studio 2018. [7] Using the optimized structure of F-cHBC molecule as input, the ab initio prediction of polymorphs was performed sequentially in six steps (i.e., packing, clustering, geometry optimization, clustering, geometry optimization, and clustering).
In the packing step, the crystal structures belonging to a specific space group were sampled using Monte Carlo simulated annealing. The packing procedure was performed for 10 different space groups: P2 1 /c, ̅ , P2 1 2 1 2 1 , C2/c, P2 1 , Pbca, Pna2 1 , Cc, Pbcn, and C2. To achieve sufficiently wide sampling, we set the maximum temperature to 1.5 × 10 5 K, the minimum temperature to 300 K, the maximum number of steps to 500,000, the number of steps to accept before cooling to 100, the minimum move factor to 1.0 × 10 −50 , and the heating factor to 0.025. In the geometry optimization step, the lattice parameters and atomic positions were relaxed under crystallographic symmetry. The F-cHBC molecule was treated rigid body in the first geometry optimization step, while it was fully relaxed in the second geometry optimization step. The maximum number of steps was set to 10,000 and the convergence criteria were set to 2.0 × 10 −5 kcal mol −1 for the maximum energy change, 0.001 kcal mol −1 Å −1 for the maximum force, 0.001 GPa for the maximum stress, and 1.0 × 10 −5 Å for the maximum displacement. In the clustering step, many similar structures were grouped into clusters, and the lowest energy structure representing each cluster was filtered. The criterion of crystal similarity measure was set to 0.11, which was calculated based on a comparison of radial distribution functions with a cutoff distance of 7 Å and 140 bins. After the final clustering step, the space group symmetry of the predicted crystal structures was reanalyzed and the in-silico screening was carried out on the basis of XRD comparison. The interatomic interactions were described by COMPASS II force field [11] and calculated using the Ewald summation method. [12,13]

Monte Carlo simulation
To figure out the specific adsorption sites of Li-ions in the crystal structure of F-cHBC, Monte Carlo simulated annealing was performed using the Sorption module of Materials Studio 2018. [7] Based on the metropolis algorithm, the Monte Carlo simulated annealing was carried out with the maximum number of loading steps of 1.0 × 10 5 , the maximum number of production steps of 1.0 × 10 8 , and 40 annealing cycles. All simulations were repeated 5 times independently. The interatomic interactions were described by COMPASS II force field [11] with Mulliken [14] charges obtained by DFT calculations.

Density functional theory calculation
DFT calculations were performed using the CASTEP module [15] of the Materials Studio 2018. [7] The generalized gradient approximation with the Perdew-Burke-Ernzerhof (GGA-PBE) functional [16] was used to describe the exchange correlation potential of the electrons.
The interactions between ions and electrons were described by on-the-fly generated normconserving pseudopotentials. The plane-wave basis set with a cutoff energy of 840 eV was employed to expand the wave functions. The van der Waals interactions were corrected using Grimme's method. [10] The convergence criterion for self-consistent field calculation was set to 5.0 × 10 −7 eV atom −1 . Lattice parameters and atomic positions were fully relaxed. The convergence criteria for geometry optimization were set to 5.0 × 10 −6 eV atom −1 for energy, 0.01 eV Å −1 for force, 0.02 GPa for stress, and 5.0 × 10 −4 Å for displacement. The Brillouin zone was integrated using a 1 × 2 × 1 k-point grid with the Monkhorst-Pack scheme [17] for all calculations. The formation energy (E f ) of the inserted structure as a function of Li-ion content was calculated as follows: