Domains of Crystallographically Oriented Pores in GaAs

As it was already demonstrated for n-InP and n-GaP, gradual increase of the anodization
current density will result in a switch from crystallographically oriented pores to current line oriented pores.

3.3.3 Domains of Crystallographically Oriented Pores in GaAs

As it was already demonstrated for n-InP and n-GaP, gradual increase of the anodization current density will result in a switch from crystallographically oriented pores to Current-line Oriented Pores. Interestingly, no current-line oriented pores were observed in n-GaAs up to now. Characteristic to n-GaAs at high current densities are domains of crystallographically oriented pores and the so-called tetrahedron like pores.

An overview of an anodized sample containing domains of crystallographically oriented pores obtained on (100)-oriented n-GaAs is presented in Figure 3.20a. The domains are rectangular and a peculiar correlation between rows of individual pores inside the domains can be observed. The rectangular shape of the domain can be divided into four quadrants shown in Figure 3.20b. The quadrants I and II have a higher degree of porosity as compared with the quadrants III and IV. The difference in porosity between the quadrants is more easily distinguished in Figure 3.20c.

By investigating the samples subjected to anodization at high current densities for small periods of time (1 to 5 seconds), it was observed that each domain starts to grow from its central point. The size of the domain depends on the anodization time and current density.

The cross-section of domains (Figure 3.20d) reveals a quite complicated 3D structure, where pores growing downwards into the substrate as well as upwards, i.e., towards the surface of the sample, are observed. A schematic view explaining the principle of domain formation in GaAs is shown in Figure 3.21.



Domains of crystallographically oriented pores in (100)  n-GaAs

Figure 3.43: Domains of crystallographically oriented pores in (100) n-GaAs anodized at j = 80 mA/cm2 in 5% HCl solutions for 1 minute; a) Overview of the domains. The domains have a squared shape and are formed by crystallographically oriented pores; b, c) Each domain can be divided in four quadrants. Two of them (I and II) have a high porosity, while the other two (III and IV) have a lower porosity; d) the structure of a domain in cross section. In this micrograph three direction of pores growth can be seen - one pointing into the substrate (not perpendicular, the arrow presents only the projection of the growth direction) and the other two pointing towards the surface (see the schematic view 'Cross section A2' presented in Figure 3.44.);


The most important feature of a domain is its starting point shown in Figure 3.21 as a gray disk. The dissolution starts here probably due to a defect in the substrate. From the central point normally two pores begin to grow along two <111>B directions pointing into the substrate. During the first few seconds of the experiment the applied current succeeds to be consumed by the two pores, imprinting them correspondingly a high velocity of growth. However, soon enough the electrochemical reaction at the pore tips starts to miss species (reducing and oxide dissolving) from the electrolyte and the amount of current consumed by the two pores decreases.

Nevertheless, the system has to find a way to carry continuously the constant amount of current imposed from exterior (note the galvanostatic conditions). The easiest way to go along with this situation is to nucleate new pores or more specifically - new pore tips. This way increases the diffusion 'freedom' of the electrolyte species and reaction products inside the porous structure.

There are two possibilities for nucleating new pores:


  • On the surface of the sample,
  • Under the surface, e.g. as branches of the two initially nucleated main pores.


Taking into the account that the surface of the sample can be strongly passivated by the electrolyte species, e.g. Cl- or OH-- passivation, it can be assumed that nucleation of new pores on the surface of the sample is less probable than branching of pores somewhere under the surface. The probability of branching is higher also due to the increased number of defects generated at pore tips during the dissolution process. These defects require time in order to be passivated by the electrolyte species and thus can be suitable candidates as nucleation points for new branches.

Pore branching along <111>B directions has a lot of freedom: two <111>B directions point downwards into the substrate and other two point upwards, i.e. towards the initial surface of the sample. In Figure 3.21 the lines illustrating the <111>B directions pointing upwards or downwards have corresponding arrows showing the exact direction.


Schematic representation of  domain formation in (100) n- GaAs obtained at high externally applied current densities;

Figure 3.21: Schematic representation of domain formation in (100) n- GaAs obtained at high externally applied current densities; The plot shows three views of one and the same domain: one from the top, and two reciprocally perpendicular cross sections. The arrows at the end of the lines designate the directions of growth (upwards or downwards) along the <111>B directions. The crystallographic planes (i.e. {011} planes), within which the two different directions (upwards or downwards) are growing, are reciprocally perpendicular.


Similar to the two initial pores stemming from the central point of the domain, the branches at their turn will also tend to branch along <111>B directions. Thus, the process of domain formation is a fractal like branching of pores. On the one hand, branching increases the number of pore tips and thus makes easier the current flow. On the other hand, the increase of the number of pore tips decreases the current density at each pore tip. Consequently, the decrease of the current density at pore tips will cause a gradual retardation of pore branching. It can be supposed that first the upward branching will quench and after that, when the current density will be small enough, the downwards branching will tend to stop as well. Thus, after some time the number of pore tips will reach a fixed value.

Taking into account the anisotropic features exposed by the pores forming the domains, as in the case of 'normal' crystallographically oriented pores, one can canclude that the etching is dominated by direct dissolution of the material rather than by oxide formation. The reason why in GaAs no curro pores have been observed can be attributed to the more pronounced anisotropy exposed by GaAs and also to the fact that in strong acidic solutions the GaAs oxides are much more unstable as compared with those of InP or GaP. This way, even at very high current densities chemical dissolution of the oxide is faster than the oxide formation itself. As a result, even at high current densities no isotropic (amorphous) oxide formation at pore tips, responsible for the formation of the Current-line Oriented Pores, can be generated in GaAs.



Home | Chapter 1 | Chapter 2 | Chapter 3 | Chapter 4 | Chapter 5 | Chapter 6 | Appendixes | Papers | Top
© Copyright www.porous-35.com , all rights reserved.
Quick, Simple, Free SEO CMS website builder and design software