Pushing their TLP technique demonstration one notch further, they even grew dual compound mesas with an atomically sharp lateral heterojunction between indium phosphide and tin phosphide, despite the two materials having vastly different crystal structures.
The seemingly simple trick explained in the paper was to heat an array of neighbouring indium and tin mesas to 550°C before introducing precursor gas phosphine. At that temperature, the two metals mixed completely in their liquid phase, and the introduction of a controlled phosphine (PH3) flux triggered the simultaneous, yet separate growth of single indium phosphide and tin phosphide crystals from the unique In/Sn melt.
Because it allows the growth and manufacture of crystalline materials without requiring a nearly lattice-matched substrate, the TLP growth process for which the researchers have filed a provisional patent could enable the design of entirely new crystalline materials (for which traditional vapour-phase epitaxy lacked a suitably matching substrate). This could lead to many new electronic and photonic devices difficult or near impossible to create so far.
The authors conclude that such a TLP growth process will enable device integration at the microscale, way beyond today's coarse 3D integration (physical stacking).