10.6 GaN Growth Investigations - Hysteresis of 'Dips'

It has been observed that the `Dips' experience growth hysteresis, but not the balance point. There is no hysteresis just to changes in the substrate temperature. Figure 38 demonstrated what is meant with growth hysteresis of the `Dips'. In Figure 38A some material is grown way above the balance point and the substrate temperature is decreased so that the balance point is approached. `Dips' will not be seen until the substrate temperature is very close to the balance point, usually less than 20°. The magnitude of the `Dips' will increase upon approaching the balance point, and behave as shown and discussed earlier below the balance point. Upon rising the substrate temperature above the balance point again the `Dips' will become smaller, but they are visible to a much higher temperature above the balance point than when the temperature was decreased, typically up to about 70° above the balance point. As can be seen, some kind of hysteresis takes place.

The question now is if the hysteresis is due to the change in substrate temperature or due to the growth. Figure 38B shows the same experiment, but no growth is performed while the temperature is decreased below the balance point and increased again through the `Dip' region. If the hysteresis would be due to the substrate temperature manipulation alone a good sized `Dip' would be visible just above the region where `Dips' start coming in upon decrease of the substrate temperature, but none were observed. This indicates that `Dip' hysteresis is growth dependent, but not due to changes in the substrate temperature alone.

To verify that observation the experiment shown in Figure 38C was performed, where material was deposited just below the balance point and the surface recovered again. Then the temperature was raised way above the balance point without growth, and decreased again as shown to the point where the `Dips' would be expected if the hysteresis effect is due to the growth and not just the substrate temperature manipulation. As can be seen the `Dip' is present as in Figure 38A, verifying that the hysteresis effect is due to the growth conducted last at a certain substrate temperature, and not just because the substrate temperature was varied in a certain way.

Since RHEED probes only the surface of the sample this hysteresis must be related to what is happening on the surface, and therefore there is some kind of hysteresis related to that surface. This is further supported since it was also noted that during `Dip' growth there is no reconstruction (1x1), while in no `Dip' regions a faint 2x2 reconstruction is present. It is speculated that the surface might be terminated differently in those regions, and further studies are needed to support that suggestion. A helpful tool will be Auger and one of the planned experiments is the study of surfaces grown in the `Dip' and the non `Dip' regions. Another preliminary experiment was conducted mapping out the hysteresis region as a function of Ga flux as shown in Figure 39. Another more detailed investigations concerning the possibility that the length of growth alters the hysteresis region is planned.

Figure 38: Typical Hysteresis of `Dips' for GaN

Figure 39: `Dip' Hysteresis as Function of Ga Flux

Table of Contents