In spite of many remarkable prototype device performances reported using semiconductor multilayers that incorporate thin tunnel barriers, it has never been established that the devices are capable of routine manufacture. With reference to the simplest possible tunnel device structure, a single ∼ 3 nm thick AlAs barrier within an asymmetric doping environment in GaAs (the so-called ASPAT microwave detector diode), we show here that is not yet possible to design, grow or qualify the semiconductor multilayers with sufficient accuracy, precision, uniformity or reproducibility that would allow reverse engineering, a prerequisite for manufacture, to be undertaken with confidence. We describe the improvements in design, growth and qualification that will be required to achieve manufacturability, and comment on the feasibility of attaining this goal. Our conclusions for thin tunnel barrier device concepts apply a fortiori to any device ideas that seek to exploit mesoscopic phenomena in semiconductors.

No abstract

No abstract

A second attempt to establish the manufacturability of a simple heterojunction tunnel diode has made some significant progress towards the goal, but has failed to demonstrate a reverse engineering capability, leaving major challenges in design, materials growth and materials qualification.

Maximum operating temperature of the 1.3 μm strained layer multiple quantum well InGaAsP lasers