05-26 yellow wavy line

[Image above] Credit: Chris Rycroft, Flickr (CC BY 2.0)

In the upcoming June/July issue of the Bulletin—which publishes online in two weeks!—our focus will be on unusual properties of ceramics, specifically dislocation-tuned functional and mechanical properties.

Dislocations are one-dimensional line defects that are the main carriers of plastic deformation in crystalline solids. Since about 2003, researchers have identified and investigated various promising proofs-of-concept for a wide range of properties tuned by dislocations in ceramics.

Electrical properties are one dislocation-based functionality of particular interest to researchers. Unlike metals, dislocations in ceramics may carry charges at their cores, with a surrounding space charge layer for charge compensation.

These charges, together with the high local strain field surrounding dislocations, can significantly influence how current flows through the ceramic. Thus, understanding and harnessing these charged areas can provide a way to create ceramics with improved electrical properties.

Ceramics within the group-III-nitride semiconductor genome, such as gallium nitride and indium nitride, are a focus of considerable attention. Though these materials have considerable potential for electronic and optoelectronic applications, unintended dislocations tend to form in their structure during fabrication.

Currently, group-III-nitride semiconductors are grown on substrates without a suitably matching lattice. This mismatch leads to high mechanical stress at the semiconductor–substrate interface, and the stress is accommodated by the formation of misfit dislocations in the semiconductor. These dislocations can trigger further defects called threading dislocation lines, which extend from the underlying substrate up through the grown crystal.

There is a rich tradition of studies on these threading dislocation lines and how they affect a semiconductor’s electrical properties. Post-doc research fellow Erfan Baghani and professor Stephen K. O’Leary at the University of British Columbia are two researchers who have published on this topic for more than a decade.

In one of their latest publications, they expand on findings Baghani originally presented in his 2012 doctoral thesis, which detailed the striking contrast between the effects of threading dislocation lines in gallium nitride versus indium nitride.

Gallium nitride and indium nitride have very different electronic band structures, with gallium nitride having a wide bandgap (~3.4 eV) and indium nitride having a small bandgap (~0.65 eV).

“Owing to this critical difference, threading dislocation line charge occupancy considerations are expected to be quite different for the two materials,” Baghani and O’Leary write in the recent paper.

Various theoretical and experimental studies have confirmed this expectation. In gallium nitride, threading dislocation lines trap electrons, thereby reducing the free electron concentration. In indium nitride, threading dislocation lines donate electrons, thereby increasing the free electron concentration.

For n-type semiconductors, in which current travels mainly through the movement of electrons, the number of free electrons affects the semiconductor’s conductivity. As such, understanding the implications of the two different charge mechanisms on a semiconductor’s conductivity is an intriguing prospect to explore.

Baghani and O’Leary used two concentric cylinders to model the behavior of the threading dislocation lines: the inner cylinder represented the dislocation’s core charge, while the outer cylinder represented the surrounding space charge layer.

To simplify the model, they assumed two things about their n-type and wurtzite structured semiconductors:

  1. Each threading dislocation line introduces only one energy level into the energy band structure of the semiconductors.
  2. Only one orbital within this single energy level governs the threading dislocation line occupation statistics.

Their simulations using this model showed that, for indium nitride, the concentration of free electrons does not really drop below 1016 cm–3 if there are threading dislocation lines present. In contrast, for gallium nitride, very low free electron concentrations can be achieved by tuning the density of threading dislocation lines.

In other words, tuning the electrical conductivity is easier to do in gallium nitride than indium nitride. The “donating” nature of the indium nitride defects makes it difficult to reduce the free electron concentration if any threading dislocation lines are present. In contrast, careful growth parameters can help reduce the number of threading dislocation lines in gallium nitride, and thereby modulate the strength of the defect’s “trapping” effect.

Baghani and O’Leary conclude their study as a success, as it demonstrated “some of the implications of the different electron acceptor/donor characteristics of the threading dislocation line defect sites.”

The paper, published in Solid State Communications, is “Threading dislocation lines within indium nitride versus gallium nitride: The implications of different dominant dislocation line charge screening mechanisms” (DOI: 10.1016/j.ssc.2022.114833).