I dealt quite a bit with nanowires during my PhD research a decade ago. E-beam lithography can be used for features this small but it's more for one-offs as in

I dealt quite a bit with nanowires during my PhD research a decade ago. E-beam lithography can be used for features this small but it's more for one-offs as in academic research like this. It's not an industrial scalable solution for high-capacity production. Optical lithography works well for industry, but the wavelengths are larger so it's hard to pattern structures this small. Though by pushing wavelengths further into the UV and doing tricks to compensate for diffraction, optical lithography has also come a long way in terms of feature resolution and may be a potential solution. I'm not up-to-date on the current state of the art. In my lab, one technique we used involved carbon nanotubes lying as a bridge over a trench, over which we'd deposit superconducting alloys. This was a way to create nanowires about 10nm wide, allowing us to study superconductivity in one dimension. But you're right in that alignment was difficult. We'd basically put an optimally-dense suspension of carbon nanotubes in a fluid on the substrate and hope that some would cross the trench, find them in a SEM after applying the superconducting alloy. They're also super delicate, so we had to hope we didn't blow them up with tiny static charges while patterning the leads, wire-bonding the silicon substrate to a chip carrier, connecting to our wired cryostat, and evacuate /cool them down to cryogenic temperatures. Felt a bit like Edison and first light bulb attempts! One of my lab colleagues was studying the physics of the carbon nanotube itself, and interesting things that could happen with magnetic leads in magnetic fields (aligning electron spins). Once he had a working sample it stayed in the dewar for over a year as he studied different aspects of it!