Controllable synthesis of borophene aerogels by utilizing h-BN layers for high-performance next-generation batteries

The borophene aerogels fabrication process is illustrated in Fig. 1A. Prepared similarly to our previous methods, BoAs were synthesized from precursors of hBNAGs (6). hBNAG was immersed into potassium hydroxide and urea propanol solution (1:1, v:v) at 60 °C. The precursors were vacuum dried, at room temperature, followed by an annealing step at 600 °C for 24 h under an Ar environment to form N-defects (Fig. 1B). For surface functionalization, the N-defect hBNAG was submerged into boric acid solution and vacuum dried at 80 °C. This sample was transferred to an induction furnace with boron source (Sigma-Aldrich 99.7% crystalline) and annealed at 1,400 °C for 4h, in an argon atmosphere, to form BoBN. Borophene only grew on N-defected hBNAG samples, indicating N-defects create B-B bridges, as a nucleation site for borophene (Fig. 1C) (7, 8). Thus, borophene can grow between two h-BN sheets, in a self-limiting manner, and form borophene/h-BN monolayers mixed with sp²-sp³ hybridization (Fig. 1 D and E). We believe that the volume expansion (Fig. 1A) is correlated with nucleation growth, but the mechanism is puzzling. Fig. 1F illustrates XPS analysis of BoBN samples. The B-1s spectrum reveals B-N (190.45 eV) and minor B-O (192.85 eV) bonds. The N 1s spectrum shows N−B (398.05 eV) and N−H (398.35 eV) bonds. The B 1s spectrum exposes intricate B-B interactions with peaks at 188.85, 185.45, 187.7, and 186.7 eV, signifying possible coexisting β₁₂ and X₃ structures per literature (ID₅:ID₄ ≈ 1 and IR₅:IR₄ ≈ 1 ratios) (9, 10). However, we also acknowledge the absence of a strong sixfold (R₆) coordination, which could potentially correlate with a different phase, a plausible intermixed phase, currently presents a challenge (11). Thermal annealing in oxygen at 323 K, 353 K, and 373 K confirms durable B-B bonds, evident from consistent B-O to B-B peak ratio ~0.26 (9–11).

BoBN was used for electrode development, adapting generic fabrication techniques for producing high-performance batteries; i) Na-ion (CR2032: BoBN+10wt% CB/NaClO₄+ethylene/PC/ Na-foil, GF separator), ii) Mg-ion (3-electrode cell: MgMn₂O₄ (calcination)+BoBN(1:2wt%)+10wt% CB+10wt% PVDF/graphite/Ag/AgCl in Mg(TFSI)₂ in THF), iii) Ca-ion (3-electrode cell: CaV₂O₅ (solid-state reaction)+BoBN, 10wt% CB+10wt% PVDF/graphite/Ag/AgCl in perchlorate), iv) Li-S (CR2032: BoBN+sulfur powder+10wt% CB+10wt% PVDF in NMP/LiPF6+EC+DMC, Celgard 2320).

The enhanced anodic stability and reversible capacity of BoBN-based batteries with different borophene/h-BN ratios, as exemplified by BoBN(5:1), BoBN(5:2), and BoBN(5:3), are shown in the galvanostatic charge/discharge curves (Fig. 2 A–D). BoBN-modified cells outperform control cells in long-term cycling performance (Fig. 2 E–H). Fig. 2E shows cycling performance of Na-ion batteries with and without BoBN. BoBN(5:1) shows an excellent cyclability over 300 with initial discharge (478 mAh g^–1 at 0.5C) and good CE (>96%), while BoBN(5:2), BoBN(5:3), and samples without BoBN(W/O BoBN) exhibit 325 mAh g^–1, 303 mAh g^–1, and 137 mAh g^–1 respectively. Evidently, cells with BoBN exhibit best performance, attributed to an increase in active sites for sodium ions, due to 3D framework and interstitial space, while lowering diffusion distances and accommodating volume changes during sodiation/desodiation processes. However, only BoBN(5:1) provided over 300 cycles, indicating that optimum loading configuration is critical to sustain efficiency of the charge and discharge process. Similarly, Fig. 2F indicates how important BoBN is for Mg-ion batteries anticipating a similar role as with Na-ion batteries. BoBN(5:1) shows good cyclability over 300 with initial discharge (297 mAh g^–1 at 0.5C) and good CE (>94%), while BoBN(5:2), BoBN(5:3), and samples W/O BoBN exhibit 201 mAh g^–1, 152 mAh g^–1, and 140 mAh g^–1 with significantly less RC but same cyclability, respectively. For Ca-ion batteries (Fig. 2G), only difference is that BoBN(5:3) exhibits the best performance (332 mAh g^–1 at 0.5C) and good CE (>96.5%), while BoBN(5:1), BoBN(5:2), and samples W/O BoBN exhibit 270 mAh g^–1, 264 mAh g^–1, and 189 mAh g^–1 with cyclability over 400, respectively. Samples W/O BoBN exhibit capacity fade and lose ability to store and release charge efficiently over 50 cycles, likely correlated with BoBN’s ability to stabilize the electrolyte and prevent SEI formation. Last, Fig. 2H provides evidence for the importance of BoBN in preserving long-term cycling performance of Li-S batteries. At 0.3C, an initial discharge is shown of 1,559 mAh g⁻¹, 1,393 mAh g⁻¹, 1,045 mAh g⁻¹, and 810 mAh g⁻¹ for BoBN(5:2), BoBN(5:1), BoBN(5:3), and commercial sulfur electrodes (CSE). All BoBN-modified batteries reached 1,000 cycles with good retention while CSE exhibited drastic capacity fade especially around 600 cycles. We speculate because of BoBN’s ability to immobilize sulfur ions and highly conductive pathway. Specifically, cells with BoBN(5:2) and 4 mg/cm² sulfur areal mass loading exhibit best performance with better ionic diffusion and ohmic resistance, as predicted. BoBN-modified batteries show high reversible capacities at >5 C rates due to their high surface area and conductivity (Fig. 2 I–L), with capacities of 316, 168, 158, 1,156 mAh g^–1 (5C) and 273, 101, 87, 901 mAh g^–1 (10C) for Na-ion, Mg-ion, Ca-ion, and Li-S batteries, respectively.

Several noteworthy findings must be highlighted. Na-ion, Mg-ion, Ca-ion, and multivalent ion batteries exhibit unstable behavior with nonstable overshoot capacity, at low discharge rates (<0.5C), attributed to unstable SEI on the electrode surface and limited ion transport. Additionally, the lowest coulombic efficiency for Li-S cells occurs at 1C rates, suggesting that parasitic processes impede total charge extraction. However, at high rates, BoBN aerogel undergoes self-adaptation/arrangement, mitigating these parasitic reactions.

The impedance spectra measurements demonstrate the importance of BoBN reinforcement on cell performances (Fig. 2 M–P). Cells, with BoBN modification, exhibit lower impedance and higher ionic conductivity compared to samples W/O BoBN. Also, semicircles at low frequencies can be associated with the presence of SEI film, apparent in Mg-ion and Ca-ion batteries. However, there is no significant charge transfer resistance at low frequencies, possibly indicating that SEI film is very thin or nonexistent for Na-ion and Li-S batteries. Cells with the proper BoBN ratio perform best, with improved ionic diffusion and ohmic resistance, yielding maximum values of 32%, 21%, 37%, and 25%, respectively. Optimal Li-S cell performance was achieved with high areal sulfur loadings and low E/S ratios. The best yield of 6% was obtained with an E/S ratio of 4.8 μL mg⁻¹, delivering an output of 811 mAh g⁻¹ at 10C. This is a very important finding, being within the ideal E/S ratio <5 μL mg^–1 (12).

In summary, we demonstrated a method to grow stable borophene aerogels utilizing hexagonal boron nitride, as a template. These 3D aerogels, which successfully integrated into next-generation battery chemistries, enhance ionic transport, shorten diffusion routes, and respond to volume fluctuations, enhancing stability. These innovations help us to provide one of the best-performing devices in the literature. We believe that these preliminary findings are key to unlocking many breakthroughs in energy conversion devices to achieve superior batteries in the future.

O.E. acknowledges European-Research-Council (QUEEN: 01043219) and Zettl’s Lab/UC-Berkeley.