"Flip chip" approach to nanoelectronics

Most people who aren't experts in the field don't really appreciate how amazing our electronic device capabilities are in integrated circuits.  Every time some lithographic patterning, materials deposition, or etching step is performed on an electrically interesting substrate (e.g., a Si chip), there is some amount of chemical damage or modification to the underlying material.  In the Si industry, we have gotten extremely good over the last five decades at either minimizing that collateral damage, or making sure that we can reverse its effects.  However, other systems have proven more problematic.  Any surface processing on GaAs-based structures tends to reduce the mobility of charge in underlying devices, and increases the apparent disorder in the material.  For more complex oxides like the cuprate or pnictide superconductors, even air exposure under ambient conditions (let alone much lithographic processing) can alter the surface oxygen content, affecting the properties of the underlying material.

However, for both basic science and technological motivations, we sometimes want to apply electrodes on small scales onto materials where damage from traditional patterning methods is unavoidable and can have severe consequences for the resulting measurements.  For example, this work used electrodes patterned onto PDMS, a soft silicone rubber.  The elastomer-supported electrodes were then laminated (reversibly!) onto the surface of a single crystal of rubrene, a small molecule organic semiconductor.  Conventional lithography onto such a fragile van der Waals crystal is basically impossible, but with this approach the investigators were able to make nice transistor devices to study intrinsic charge transport in the material.  

One issue with PDMS as a substrate is that it is very squishy with a large thermal expansion coefficient.  Sometimes that can be useful (read this - it's very clever), but it means that it's very difficult to put truly nanoscale electrodes onto PDMS and have them survive without distortion, wrinkling, cracking of metal layers, etc.  PDMS also really can't be used at temperatures much below ambient.  A more rigid substrate that is really flat would be great, with the idea that one could do sophisticated fab of electrode patterns, and then "flip" the electrode substrate into contact with the material of interest, which could remain untouched or unblemished by lithographic processes.

In this recent preprint, a collaboration between the Gervais group at McGill and the CINT at Sandia, the investigators used a rigid sapphire (Al2O3) substrate to support patterned Au electrodes separated by a sub-micron gap. They then flipped this onto completely unpatterned (except for large Ohmic contacts far away) GaAs/AlGaAs heterostructures.  With this arrangement, cleverly designed to remain in intimate contact even when the device is cooled to sub-Kelvin temperatures, they are able to make a quantum point contact while in principle maintaining the highest possible charge mobility of the underlying semiconductor.  It's very cool, though making truly intimate contact between two rigid substrates over mm-scale areas is very challenging - the surfaces have to be very clean, and very flat!  This configuration, while not implementable for too many device designs, is nonetheless of great potential use for expanding the kinds of materials we can probe with nanoscale electrode arrangements.