The CNM solution is based on a metal catalyst deposition method that provides selective area deposition of metallic nanoparticles on silicon surfaces versus oxide surfaces. Before catalyst deposition, the substrates go through several micro-fabrication steps to pattern an oxide layer at the vertical sidewall of supporting silicon microstructures.

To verify the idea, the team used the approach to fabricate well ordered resonator arrays with predefined pitch, which are useful for optical readout. The scientists then made nanomechanical resonators for electrical transduction, where nanowires are precisely positioned with respect to reading and driving electrodes. This second example is particularly representative of the importance of horizontally patterned growth for nanomechanical devices.

Nanowire resonators exhibit a so-called resonant mode splitting effect, by which they can vibrate in two orthogonal directions at slightly different resonance frequencies for each flexural mode. This response had only been observed in nanowires by optical readout at low frequencies (<10 MHz), and it is a crucial effect for the development of nanomechanical sensors. First, because it modifies the expected Lorentzian frequency response; and second, because it provides two different observables (for each mode) to characterize the resonator's response to a given external effect or signal.

In spite of its importance, this effect had never been observed by electrical readout because of the lack of control of nanowire location. The Spanish group has found that horizontal growth with position control allows device makers to exploit the resonant mode splitting effect with electromechanical transduction at frequencies above 100 MHz, which carries important benefits in sensing performance.

Additional details can be found in the journal Nanotechnology.