Energy storage is a critical limitation for widespread adoption of renewable energy sources. While various energy harvesters exist, large-scale energy storage remains a key challenge in the 21st century energy system. Carbon nanotubes (CNTs), known for their high strength and modulus, are promising candidates for mechanical energy storage due to their fast charging and discharging capabilities, stability, and reversibility. CNT fibers, composed of densely packed, axially aligned CNTs, are fabricated through techniques like spinning or twisting/rolling, resulting in varied structures and mechanical properties. However, limitations exist; CNT fibers can be metastable, with flattening of constituent CNTs affecting performance, necessitating extensive effort to control their mechanical properties. In 2015, carbon nanothreads, a one-dimensional sp³-bonded carbon structure, were introduced. These ultra-thin structures overcome limitations of CNT fibers, offering consistent fabrication into high-strength bundles and exhibiting excellent mechanical properties comparable to CNTs. This study aims to assess the mechanical energy storage capacity of nanothread bundles using large-scale molecular dynamics (MD) simulations and elasticity theory, focusing on contributions from different deformation modes to determine if they could surpass the energy storage capabilities of advanced Li-ion batteries.

Extensive research explores CNT-based fibers for mechanical energy storage and harvesting. CNT fiber-based systems offer advantages over electrochemical batteries, such as faster and more efficient energy transfer, higher stability, and better reversibility. Various CNT fiber structures have been reported, including knitted, parallel, and twisted structures. The mechanical performance varies significantly due to structural complexities and post-treatments. Recent work highlighted CNT bundles with tensile strength exceeding 80 GPa, utilizing ultralong, defect-free CNTs. However, the flattening of CNTs in bundles with larger diameters, like (10,10) and (18,0) CNTs, negatively impacts mechanical performance. Carbon nanothreads, sp³-bonded carbon structures, offer a potential solution. While experimental characterization is ongoing, theoretical calculations have described potential nanothread configurations. Preliminary studies show nanothreads possess high stiffness and bending rigidity, comparable to CNTs, and exhibit structural-dependent ductility. The synthesis of single-crystalline nanothread packing opens up possibilities for facile exfoliation of nanothread bundles, motivating the investigation of their mechanical energy storage potential.

The research employed a combination of large-scale molecular dynamics (MD) simulations and continuum elasticity theory. Two representative carbon nanothreads, nanothread-A (achiral) and nanothread-C (chiral), were selected for the study based on their low energy and classification. The AIREBO potential was used to describe the C-C and C-H atomic interactions, encompassing short-range interactions, long-range van der Waals interactions, and dihedral terms. A cutoff of 2.0 Å was adopted to address the non-physical high tensile stress under bond stretching inherent in the AIREBO potential. Simulations were conducted at a low temperature of 1 K to minimize thermal fluctuations. The virial stress and atomic Von Mises stress were calculated to analyze stress distribution within the structures. Different bundle configurations (bundle-n, with n representing the filament number from 2 to 19) were constructed using a triangular lattice arrangement based on equilibrium inter-tubular distances determined through relaxation simulations. Four deformation modes—torsion, tension, bending, and radial compression—were investigated through separate MD simulations. For torsion, a constant torsional load was applied; for tension, a constant velocity was applied to one end of the sample; for bending, an initial curvature was introduced; and for radial compression, a close-packed triangular lattice was subjected to lateral pressure. Hooke's law was utilized to describe the linear elastic regime of individual nanothreads, obtaining elastic constants (k) for each deformation mode through quadratic fitting of MD simulation results. A theoretical model, adapted from previous work on CNTs, was developed to quantify the total strain energy in twisted bundles by considering contributions from torsion, tension, bending, and compression. This model enables the assessment of the dominant deformation modes and the calculation of strain energy density for varying filament numbers.

MD simulations revealed that nanothread-A exhibits a higher gravimetric energy density (~884 kJ kg⁻¹) than nanothread-C (~737 kJ kg⁻¹) under torsion. The torsional elastic limit was higher for nanothread-A (~0.71) than nanothread-C (~0.51). In contrast to CNTs, which exhibited flattening at small torsional angles, nanothreads showed no such behavior. Under tensile deformation, nanothreads exhibited significantly higher gravimetric energy density (~2051 kJ kg⁻¹ for nanothread-A and 906 kJ kg⁻¹ for nanothread-C) compared to torsion. Bending simulations yielded lower gravimetric energy densities (468 kJ kg⁻¹ for nanothread-A and 288 kJ kg⁻¹ for nanothread-C). Radial compression showed high elastic limits and gravimetric energy densities (~6051 kJ kg⁻¹ for nanothread-A and ~3063 kJ kg⁻¹ for nanothread-C), but CNTs experienced flattening under compression. Analysis of twisted nanothread bundles showed that gravimetric energy density decreased with increasing filament numbers. Torsion and tension were the dominant contributors to energy storage. CNT bundles, despite individual CNTs having higher Young's modulus, had comparable gravimetric energy density to nanothread bundles. Fracture in nanothread bundles was initiated in the outermost filaments. A theoretical model accurately predicted strain energy components at low strain and, surprisingly, also agreed well with MD simulation results at the elastic limit. For nanothread bundles, tension dominated energy storage in larger bundles. The study found that pure tension of the nanothread bundle showed great potential for achieving high energy storage capacity (1.76 MJ kg⁻¹ for nanothread-A bundle with 19 filaments).

The findings demonstrate the potential of carbon nanothread bundles for high-density mechanical energy storage. The comparable performance to CNT bundles, coupled with the ability to achieve high energy storage through pure tension, makes nanothreads a promising alternative. The absence of flattening in nanothreads, unlike in CNTs, contributes to their superior performance under torsional strain. The theoretical model's accuracy in predicting strain energy components strengthens confidence in the observed behavior. The dominance of tensile energy storage in larger bundles indicates a pathway for maximizing energy storage through bundle design. The high gravimetric energy density achieved under pure tension (up to 1.76 MJ kg⁻¹) significantly exceeds that of steel springs and Li-ion batteries, highlighting the potential for practical applications. The hydrogenated surface of nanothreads, facilitating inter-thread bonding, further enhances their potential for energy storage applications.

This research showcases the potential of carbon nanothread bundles for high-density mechanical energy storage. The ability to achieve a gravimetric energy density of up to 1.76 MJ kg⁻¹ under pure tension, exceeding that of steel springs and Li-ion batteries, highlights their promise. Future research should focus on exploring different nanothread types and bundle configurations to further optimize energy storage capacity. Investigating the effects of temperature and strain rate on the mechanical properties is also crucial for real-world applications. The study's findings open up new avenues for developing advanced mechanical energy storage devices.

The simulations were conducted at a low temperature (1 K) to minimize thermal effects, potentially overestimating the elastic limits and energy storage capacity at room temperature. The high strain rate used in the simulations might also influence the results. The study focused on perfect bundle structures with identical filaments; real-world bundles may exhibit slippage between filaments due to variations in length or defects, affecting the energy storage capacity. Furthermore, the study's findings might not be generalizable to all types of carbon nanothreads, necessitating further investigations.