Planar defects, such as stacking faults and twins, naturally exist in some as-synthesized one-dimensional (1D) nanostructures . In addition to assisting the growth of nanostructures , these defects can affect the mechanical , electrical , thermal , and optical  properties of 1D nanostructures. Thus, it is crucial to know their nature such as existence, distribution, and orientation within each 1D nanostructure while establishing the structure–property relations. So far, transmission electron microscopy (TEM) has been one major technique commonly used to characterize the structure of individual 1D nanostructures and reveal the nature of planar defects . However, due to the sophistication of the TEM technique, sometimes, experimental artifacts could be erroneously interpreted or lead to controversy [6–10]. To date, most planar defect-related studies have been focused on 1D nanostructures made of silicon, silicon carbide, III-V (e.g., GaAs, InP), or II-IV compounds (e.g., ZnO, CdSe) whose crystal structures are either cubic or hexagonal [8–15].
Boron carbide 1D nanostructures have attracted increasing attention in the last few years because of their potential applications in nanocomposites and thermoelectric energy conversion [16–25]. Most reported boron carbide 1D nanostructures were synthesized by carbothermal reduction or chemical vapor deposition at approximately 1,100°C [16–23]. Field emission [18, 23], photoluminescence , mechanical [21, 23], and thermal conductivity  properties of these 1D nanostructures were reported. However, due to the complicated rhombohedral crystal structure, detailed structural characterization especially on planar defects that could greatly affect the properties of boron carbide 1D nanostructures has not yet gained enough attention, and the structure–property relations have not been established. In our previous study , about one hundred as-synthesized boron carbide nanowires were subjected to TEM study, during which each nanowire was examined throughout the full tilting range allowed by the configuration of our microscope. Approximately 75% examined nanowires were found to have planar defects, while the remaining 25% were planar defect-free-like. The defected nanowires were further categorized into two groups: transverse faults (TF) nanowires with planar defects perpendicular to the preferred growth direction of nanowires and axial faults (AF) nanowires with planar defects parallel to the preferred growth direction of nanowires. The determination of defects’ existence and fault orientations (TF or AF) within each nanowire was based on the characteristic features presented in TEM results, including modulated contrast in high-resolution TEM (HRTEM) images and streaks in diffraction patterns.
In this work, more extensive TEM examination and model simulation were performed to gain a deeper understanding of the nature of planar defects in the aforementioned boron carbide nanowires to answer two questions. (1) Do planar defect-free boron carbide nanowires really exist? Literature review shows that due to its relatively low stacking fault energy (75 mJ/m2) , planar defects have been frequently observed in bulk boron carbides independent of the synthesis methods [27–30]. It has also been reported that the density of planar defects decreases as the synthesis temperature increases . However, the planar defects were still detectable by TEM from bulk samples synthesized at 2,100°C . Considering the common existence of planar defects in bulk boron carbides and the relatively low temperatures researchers used to synthesize boron carbide 1D nanostructures, one may naturally ask ‘Can boron carbide nanowires synthesized at approximately 1,100°C be planar defect-free? Or defects always exist but sometimes are not found by TEM?’ (2) If planar defects exist in all of our as-synthesized boron carbide nanowires, can their orientations be determined from TEM results showing no characteristic features (i.e., results from the off-zone directions as discussed later)? It is expected that different orientations of planar defects could have distinctive effects on the properties of these nanowires, similar to that physical properties of superlattices could be very different along their in-plane and cross-plane directions [31, 32]. Therefore, it is important to know the fault orientation of each boron carbide nanowire when establishing the structure–property relations.
In this paper, a thorough discussion on observing planar defects in boron carbide nanowires by TEM is presented. Results show that planar defects can be easily invisible in boron carbide nanowires even after a full range of tilting examination. Extra attention must be paid and reliable conclusion can only be made based on the results from different viewing directions (i.e., zone axes). Furthermore, a new approach is developed to determine the fault orientations of those boron carbide nanowires whose planar defects are invisible in TEM results. The approach can be extended to other 1D nanostructures whose crystal structure is not rhombohedral.