Answers are grounded in this paper and cite their sources.
For nearly three decades, the focused ion beam (FIB) instrument has effectively patterned a variety of different structures and devices from a wide range of materials with micro to nanoscale precision [1-5]. Some examples of this include (but are not limited to) polymer-based microfluidic devices [6], oxide thin films with ordered nanochannels [7], tunnel junctions in superconducting ceramics [8], nanoparticles [9-11], microelectronics [12-13], and two-dimensional (2D) materials [14]. In the case of 2D materials, the use of FIB to mill [15-19], modify [14,20-21], or synthesize [22] graphene into various electrical and nanomechanical devices has been well established, with FIB used very recently to successfully pattern high-quality graphene trampoline nanomechanical resonators [23]. However, the utility of FIB patterning notwithstanding, the task of accurately patterning a precise location on a sample, chip, or device presents a rather unique challenge as the act of FIB imaging (ion-irradiating) the location to be patterned is often necessary for accurate pattern placement while the simple act of doing this (even using the lowest of FIB currents) potentially causes significant undesired structural modification (e.g., point defects, amorphization, etc.) to the material to be patterned [21,24-27]. Furthermore, in the case of a material to be patterned into a functional device, the introduction of structural damage from FIB irradiation will presumably negatively impact device performance. To address this limitation, the dual focused ion beam/scanning electron microscope (FIB/SEM) was developed [28] as the addition of the SEM capability provided a non-destructive means to not only real-time monitor the patterning process, but also locate and image the location to be FIB patterned. Thus, in principle, if there is accurate coincidence between the SEM and FIB columns, the dual FIB/SEM allows FIB patterning of a specific location without the need for prior FIB irradiation of said location (so-called “blind” milling). However, if a very high degree of spatial accuracy is needed for FIB patterning, this blind approach based solely on coincidence between the two beams may produce results that are inconsistent at best (even with careful and proper alignment) or completely inadequate at worst. To address the clear shortcomings in spatial accuracy associated with the standard blind patterning approach, a new blind patterning method is introduced here as tested and developed using an FEI Helios Nanolab 600i dual Ga+ FIB/SEM instrument using graphene flakes suspended over circular cavities with varying diameters in SiO2-coated (001) Si wafers to be patterned into graphene trampoline nanomechanical resonators [23] as test specimens. This new approach builds off the use of coincidence between the SEM and FIB columns while introducing the novel use of “alignment” arrays of 150 nm-diameter circles with 500 nm spacing, which are strategically FIB milled in proximity to the area to be patterned. It will be shown that use of alignment arrays allows the area for FIB patterning to be located with high spatial precision with literally zero need to directly FIB image the area beforehand. Fig. 1 shows top-down ultra-high resolution 2 kV SEM images of part of a graphene flake suspended over a 30 μm-diameter cavity in a SiO2-coated (001) Si wafer; Fig. 1(a) shows the area before FIB patterning with Fig. 2(b) showing the same area after completion of “blind” FIB patterning to produce a trampoline nanomechanical resonator, which clearly demonstrates the effectiveness of this patterning method. It is important to note that literally zero prior FIB imaging of the portion of the graphene flake suspended over the cavity was performed prior to the actual FIB patterning used to produce the final device structure. It will be shown that this novel blind FIB patterning method is spatially accurate to near nanoscale precision for a range of sizes of graphene trampoline nanomechanical resonators. Additionally, while this method is demonstrated here for the specific case of fabricating graphene trampoline nanomechanical resonators [23], it is important to note that this method can be applied to literally any scenario requiring high-precision FIB patterning where prior FIB imaging of the area to be patterned must be avoided. Top-down, ultra-high resolution 2 kV SEM images of part of a graphene flake suspended over a 30 μm-diameter cavity in an SiO2-coated (001) Si wafer: a) prior to “blind” FIB patterning with the cavity outline (yellow dashed circle) and intended FIB path (solid white arcs) schematically indicated and b) after “blind” FIB patterning to produce a trampoline nanomechanical resonator; scale bar is applicable for both a) and b).
No takes yet. Share an insight, caveat, or question.
Nicholas G. Rudawski (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: