Defect filtering in polycrystalline graphite: Self-selection of pristine interfaces enabling deterministic structural superlubricity

Reliable acquisition of clean, defect-free, single-crystalline van der Waals (vdW) surfaces without relying on post-characterization is crucial for studying intrinsic interface properties and developing high-performance vdW devices, particularly for achieving structural superlubricity (SSL)—a state characterized by wearless and nearly frictionless sliding between two contacting solid surfaces. In this work, we investigate how various nanoscale defects affect friction and lead to probabilistic SSL behavior. We then introduce a defect-filtering method that enables the reliable cleavage of patterned single-crystal graphite surfaces with minimal defects from polycrystalline graphite, transforming SSL from a probabilistic to a deterministic phenomenon in this material for the first time. Using this method, all samples exhibit nearly vanishing friction, and nearly 100% self-retraction motion is observed in samples of different sizes (up to 12 × 12 μm²). This transition is attributed to a spontaneously selective cleavage mechanism that leverages superior SSL when edge static friction is suppressed. Transferred samples with defect-filtering exhibit the lowest friction (1.43 kPa) among previously transferred SSL systems, with markedly reduced sample-to-sample variability. This work provides fundamental insights into the mechanisms governing reliable SSL, and the controlled acquisition of ideal surfaces establishes a foundation for advancing materials and methodologies in interface studies and device applications.