Macroporous silicon

Figure 5. SEM image of carbon nanostructures on an internal surface of silicon macropores.

Fang C, Foca E, Xu S, Carstensen J, Foil H (2007) Deep silicon macropores filled with copper by electrodeposition. J Electrochem Soc 154 D45... 

Chiolerio A, Virga A, Pandolfi P et al (2012) Direct patterning of silver particles on porous silicon by inkjet printing of a silver salt via in-situ reduction. Nanoscale Res Lett 7 1-7 Fang C, Foca E, Xu S et al (2007) Deep silicon macropores filled with copper by electrodeposition. J Electrochem Soc 154 D45-D49... 

With electrochemically etched porous silicon structures, quantum size effects are only important in microporous and mesoporous silicon. Macroporous silicon can for many properties be modeled as bulk silicon with voids therein, provided the macropores therein do not contain mesoporosity, for example. 

The etch rate is further increased if H202 is added to the solution, as shown in Fig. 2.5 b. At such low rates the reaction is controlled by the kinetics of the reaction at the interface and not by diffusion in the solution. This etching solution is therefore found to be perfect to remove micro- and mesoporous silicon selectively from a bulk silicon substrate or to increase the diameter of meso- or macropores in an well-controlled, isotropic manner [Sa3],... 

The process described above is expected to produce a random distribution of active and passive spots on the electrode interface. But the electrode surface may also be artificially patterned prior to anodization in order to form nucleation centers for pore growth. This may be a lithographically formed pattern in said passive film or a predetermined pattern of depressions in the electrode material itself, which become pore tips upon subsequent anodization. The latter case applies to silicon electrodes and is discussed in detail in Chapter 9, which is devoted to macropore formation in silicon electrodes. 

An electric field in the semiconductor may also produce passivation, as depicted in Fig. 6.1c. In semiconductors the concentration of free charge carriers is smaller by orders of magnitude than in metals. This permits the existence of extended space charges. The concept of pore formation due to an SCR as a passivating layer is supported by the fact that n-type, as well as p-type, silicon electrodes are under depletion in the pore formation regime [Ro3]. In addition a correlation between SCR width and pore density in the macroporous and the mesoporous regime is observed, as shown in Fig. 6.10 [Thl, Th2, Zh3, Le8]. 

If one studies the growth rate as a function of anodization current density for different PS structures prepared in the same electrolyte, as shown in Fig. 6.5, some inherent laws can be observed. In the regime of stable macropore formation on n-type silicon the growth rate is found to be virtually independent of the applied current density. This is simply a consequence of JPS being present at any pore tip, as described by Eq. (9.5). For the growth rate rPS (in nm s 1) of micro PS in ethanoic... 

If the pore density is plotted versus the doping density of the silicon electrode, it can be seen that the micropore density is independent of doping, while the macropore and mesopore densities increase linearly with doping density, as shown in Fig. 6.10. This is a consequence of the QC formation mechanism being independent of doping, while the SCR-related mechanisms are not, as discussed in Section 6.2. 

Fig. 6.10 Pore density versus silicon electrode doping density for PS layers of different size regimes. The broken line shows the pore density of a triangular pore pattern with a pore pitch equal to twice the SCR width for 3 V applied bias. Note that only macropores on n-type substrates may show a pore spac-...

According to the macropore formation mechanisms, as discussed in Section 9.1, the pore wall thickness of PS films formed on p-type substrates is always less than twice the SCR width. The conductivity of such a macroporous silicon film is therefore sensitive to the width of the surface depletion layer, which itself depends on the type and density of the surface charges present. For n-type substrates the pore spacing may become much more than twice the SCR width. In the latter case and for macro PS films that have been heavily doped after electrochemical formation, the effect of the surface depletion layer becomes negligible and the conductivity is determined by the geometry of the sample only. The conductivity parallel to the pores is then the bulk conductivity of the substrate times 1 -p, where p is the porosity. 

This section is devoted to the formation mechanisms that have been proposed as being responsible for formation of macropores on p-type and n-type silicon electrodes. 

Macropore formation on p-type silicon electrodes was first observed for anodization in water-free mixtures of anhydrous HF and an organic solvent [Pr7, Ril]. Later it was observed that organic HF electrolytes with a certain fraction of water [Pol, We5], or even non-organic, aqueous HF electrolytes [We2, Le21], are also sufficient for the formation of macropores on p-type Si electrodes. This indicates that macropore formation on such electrodes cannot be ascribed to the chemical iden-... 

The growth of a macropore on a p-type substrate can be initiated by artificial etch pits. The growth of predefined pore arrays is observed to be more stable than the growth of random pores on flat electrodes [Chl6, Le21]. If a slit is used for pore initiation the formation of trenches separated by thin walls has been observed on (100) p-type substrates [Oh5]. Note that for slits along the (110) direction the walls become (110) planes, in contrast to trenches produced by alkaline etchants, for which only (111) oriented walls can be formed on (110) oriented silicon substrates. 

A specific feature of macropore formation in n-type silicon is the possibility of controlling the pore tip current by illumination and not by applied bias. This adds another degree of freedom that is not available for mesopore or macropore formation on p-type substrates. The dark current density of moderately doped n-type Si electrodes anodized at low bias is negligible, as shown in Fig. 4.11, therefore all macropore structures discussed below are formed using illumination of the electrode to generate the flux of holes needed for the dissolution process. Illumination, however, is not the only possible source of holes for example, hole injection from a p-doped region is expected to produce similar results. 

Fig. 9.8 SEM micrograph of the surface, cross-section and a 45° bevel of an anodized n-type silicon sample (5 2 cm, (100), 6% HF) showing a random pattern of macropores. Pore initiation was enhanced by applying 10 V bias in the first minute of anodization followed by 149 min at 3 V. The current density was kept constant (10 mA cm-2) by adjusting the backside illumination. After [Le9].

Fig. 9.9 SEM micrograph of an n-type silicon electrode with an etched macropore array (5 2 cm, (100), 3 V, 350 min, 2.5% HF). Pore growth was induced by a square pattern of pits produced by standard lithography and subsequent alkaline etching (inset upper right). In order to measure the depth dependence of the growth rate, the current density was periodically kept at 5 mA cm 2 for 45 min and then reduced to 3.3 mA crrf2 for 5 min. This results in a periodic decrease in the pore diameter, as indicated by the white labels on the left-hand side. After [Le9].

---