Photoluminescence and the gallium problem for highest-mobility GaAs/AlGaAs-based 2d electron gases

TL;DR

This paper investigates how the purity of gallium affects the mobility and photoluminescence of GaAs/AlGaAs 2D electron gases, highlighting the importance of high-purity Ga for achieving record mobilities.

Contribution

It demonstrates the correlation between gallium purity, photoluminescence, and electron mobility, and compares different MBE systems over several years.

Findings

High-purity Ga leads to strong exciton PL and record mobilities.

Low-purity Ga results in weak exciton PL and limited mobility.

Mobility and PL are correlated across different MBE systems.

Abstract

The quest for extremely high mobilities of 2d electron gases in MBE-grown heterostructures is hampered by the available purity of the starting materials, particularly of the gallium. Here we compare the role of different Ga lots having nominally the highest possible quality on the mobility and the photoluminescence (PL) of modulation doped single interface structures and find significant differences. A weak exciton PL reveals that the purity of the Ga is insufficient. No high mobility can be reached with such a lot with a reasonable effort. On the other hand, a strong exciton PL indicates a high initial Ga purity, allowing to reach mobilities of 15 million (single interface) or 28 million $cm^2/Vsec$ (doped quantum wells) in our MBE systems. We discuss possible origins of the inconsistent Ga quality. Furthermore, we compare samples grown in different MBE systems over a period of several years and find that mobility and PL is correlated if similar structures and growth procedures are used.