Dissolution silicon

Figure 2.5 N2 adsorption isotherms and schematized silicon dissolution (inset) upon alkaline treatment ofZSM-5 zeolites with different framework Si/AI ratios, highlighting the crucial role of framework aluminum.

Thus, no passivating layer is formed due to the fact that silicon dissolves in such media. Therefore, the anodic I-V curve of silicon in HF is quite unique and differs from the I-V curves obtained in acidic (HF-free) and alkaline electrolytes. Less attention has been paid to the electrochemistry of silicon in alkaline electrolytes as compared to the study of the electrochemistry of silicon in acidic electrolytes, probably due to the fact that pore formation is observed only in acidic media. In contrast to acidic solutions, silicon dissolution occurs in alkaline solutions under - open circuit potential (OCP). 

Uhlir (7) found an effective valence of 2.0 + 0.2 for silicon dissolution during thick film formation. He also observed that the him reacted with water, alcohol, and even toluene with gas evolution after being dried and stored in air for as long as one year. The gas evolved has been identified as hydrogen. Turner (29) observed that pieces of the film react with explosive violence when put In contact with a strong oxidizing agent such as concentrated nitric acid. These results all indicate that the silicon in the anode film exists in some reduced form. 

The following series of electrochemical and chemical equations are proposed to explain all the observed facts concerning the thick anode film formed on silicon in HF solutions. The initial process is silicon dissolution ... 

The reaction steps given by Eqs. (13)-(15) represent a general framework for silicon dissolution in fluoride electrolytes. The nature of the X and X intermediates and the chemical reactions are dependent on solution chemistry and dopant type and are discussed in more detail in subsequent sections. 

For the case of pore formation in p-type silicon where the holes are the majority carriers, silicon dissolution occurs at much lower potentials than for n-type silicon. Both hole capture and electron injection have been suggested for the second elec-... 

As for the case of pore formation, although kinetics of silicon dissolution in the electropolishing regime have been studied by a number of groups [149-152], the detailed reaction mechanism is not well understood. The properties of electrochemical-ly formed silicon oxides have been reviewed recently [34, 153] and will not be discussed here. 

As mentioned above, the dispersion stability of the slurry is directly related with CMP performance, removal rate, within-wafer nonuniformity (WIWNU), which is defined as the standard deviation divided by the average of remaining thickness after CMP, microscratching, and the remaining particle on the wafer. To avoid poor CMP performance, the dispersion stability of the slurry must be controlled by preventing silicon ion dissolution. Surface modification of the silica particle was produced by addition of an organic additive. Without surface modification, the amount of silicon dissolution was 1.370 0.002 mol/L, while surfaces modified with poly(vinylpyrrolidone) (PVP) polymer yielded a dissolution of 0.070 0.001 mol/L, almost 20 times less than the unmodified surface. 

FIGURE 5.43. Correlation of silicon dissolution rate with i-V curve in 1M KOH. (Reprinted from Hurd and Hackerman, - 1964, with permission from Elsevier Science.)... 

The mechanisms of the electrochemical reactions of silicon electrodes in alkaline solutions at OCP have been investigated in many smdies due to their importance in the etching processes in micromachining. An important issue involving the reaction mechanisms has been whether the etching process at OCP is of chemical or electrochemical nature, that is, whether charge transfer processes associated with silicon dissolution and hydrogen evolution involve the carriers in the electrode. 

According to Seidel et al. the dissolution at OCP is an electrochemical process with concurrent anodic dissolution of sihcon and reduction of hydrogen ions. The oxidation of silicon gives out electrons which are consumed for the reduction of hydrogen. Both OH" and H2O are the active species in that OH" is involved in silicon dissolution and H2O in hydrogen evolution ... 

The data presented in the following section concern only the hydrogen reaction at cathodic potentials. Those on hydrogen termination are presented in Chapter 2 and on silicon dissolution in Chapter 5. It is to be noted that as a reduction reaction, hydrogen evolution has not been well investigated at cathodic potentials although it has been the subject of numerous studies on the phenomena at anodic or open-circuit potentials. 

Deposition of metals on a silicon surface can be either a conduction band process or a valence band process depending on the redox potential of the metal and solution composition. Deposition of Au on p-Si in alkaline solution occurs only under illumination indicating that it is a conduction band process due to the unfavorable position of the redox couple for hole injection. " On the other hand, deposition of platinum on p-Si can occur in the dark by hole injection into the valence band. For Cu, although the deposition proceeds via the conduction band as shown in Fig. 6.9, it can also proceed via the valence band because a large anodic current of n-Si occurs in the dark in copper-containing HF solution as shown in Fig. 6.10. The reduction of copper under this condition is via hole injection. The holes are consumed by silicon dissolution and the silicon reaction intermediates then inject electrons into the conduction band, resulting in the anodic current on n-Si in the dark. 

Thus, whether a metal can be deposited by electroless deposition onto a silicon surface depends on the redox potential and its relative position to the band edges and on whether the silicon can be dissolved under those conditions. On the other hand, whether the deposition can be sustained to cover the entire surface area depends, on nucleation and growth kinetics of the deposits, the catalytic effect of the deposits on silicon dissolution and hydrogen evolution and the evolution of the morphology of the surface. The formation of a continuous and uniform metal film by electroless deposition is intrinsically difficult because a certain amount of bare silicon surface area is required for silicon dissolution in order to sustain the deposition. 

An anodic plateau current is measured on -Si in the dark thus indicating electron injection into the conduction band. This plateau current is only slightly affected by HF concentration but is proportional to Cr concentration. Illumination of the n-Si causes an increase of the anodic plateau current with a quantum efficiency of about 2.8 due to electron injection into the conduction band by the silicon dissolution intermediates. The quantum efficiency for the anodic photocurrent depends on the HF/CrOs ratio, about 2.8 at a high HF/CrOs ratio (>20) and about 1 at very low ratios. 

The chromium complex may also react with the silicon dissolution intermediate... 

Other Redox Species. Reduction of ferricyanide in KOH solution takes place via hole injection into the valence band. The reaction path depends on whether an oxide film is present on the surface. On an oxide-free p-Si the reduction proceeds by hole injection as shown in Fig. 6.22. On an oxide-covered electrode, which is anodized at 0 V prior to the transient, the drop of current at about 3.5 min is due to the complete dissolution of the oxide film, resulting in the same current as that on the oxide-free surface. The lower current on the oxide-free surface is attributed by Bressers et al. to the reaction of silicon, which consumes a part of the injected holes by the reduction of ferricyanide. On the oxide-covered surface, silicon dissolution does not occur and all of the injected holes flow into the semiconductor. Figure 6.23 shows the dependence... 

The redox couples, which inject holes into the conduction band, induce anodic current on -Si in the dark. According to Gerischer and Lubke, anodic current is due to electron injection by silicon dissolution intermediates for redox couples with one electron transfer per molecule. Depending on the number of electrons involved in the reduction process of the redox couple, the photocurrent may have a quantum yield varying from 1 to 2. The quantum yield is 2 without redox couples. It remains the same when a redox couple with only one oxidation step per molecule such as (IrCU) is present. However, for the oxidizing agents Br2 and MnO, the reduction of which involves two electrons, the quantum yield is smaller than 2 as shown in Fig. 6.25. This is attributed to the adsorption of the reduction intermediates on the surface, which inhibits the electron injection from silicon dissolution intermediates responsible for the quantum yield of 2. 

Although the electrochemical nature of the processes involved in the formation of PS at open-circuit conditions (nonbiased) should be similar to that under anodic bias, there are several major differences in the formation conditions. The first is that at the OCP the driving force is provided by the oxidation agents, the reduction of which provides the anodic polarization of the electrode needed for silicon dissolution. Unlike the externally biased condition, the extent of polarization is limited by the oxidation power of the oxidation agents. The second is that the carrier supply at the open-circuit condition is localized and randomly oriented, while that at anodic potential is perpendicular to the surface. The anodic and cathodic sites in the chemical etching process must be in the vicinity of each other, and continuous alternations must occur between anodic and cathodic reactions on the surface at the pores tips. 

F. Gaspard, A. Bsiesy, M. Ligeon, F. Muller, and R. Herino, Charge exchange mechanism responsible for p-type silicon dissolution during porous silicon formation, J. Electrochem. Soc. 136, 3043, 1989. 

D. J. Blackwood, A. Borazio, R. Greef, L. M. Peter, and J. Stumper, Electrochemical and optical studies of silicon dissolution in ammonium fluoride solutions, Electrochim. Acta 37(5), 889, 1992. 

S. Cattarin, E. Decker, and D. Dini, Anodic silicon dissolution in acidic fluoride electrolyte. A probe beam deflection investigation, J. Phys. Chem. B 102, 4779, 1998. 

S. Cattarin, E. Pantano, and F. Decker, Investigation by electrochemical and deflectometric techniques of silicon dissolution and passivation in alkali, Electrochem. Commun. 1, 483, 1999. 

Since hydrogen evolution occurs at anodic potentials and it is responsible for an apparent dissolution valence smaller than 4, the effective dissolution valence can be used as a measure for the hydrogen efficiency. A silicon dissolution valence of 2 can be used to indicate 100% efficiency for hydrogen evolution, that is, for every dissolved silicon atom one hydrogen molecule is generated. Figure 6 shows the effective dissolution valence and hydrogen evolution in HF solutions under different conditions. 

Hydrogen evolution on silicon may proceed chemically or electrochemically depending on the conditions. Hydrogen evolution near OCP and at anodic potentials can proceed completely chemically, that is, without involving the carriers from the electrode. The chemical nature of hydrogen evolution is responsible for less than 4 of the silicon effective dissolution valence as shown in Fig. 6. A change from a chemical process to an electrochemical process occurs when the potential varies from anodic values to cathodic values as schematically illustrated in Fig. 10. Hydrogen evolution at cathodic potentials is predominantly electrochemical due to the lack of silicon dissolution and abundance of electrons on the surface on ra-Si or illuminated p-Si. 

Mesoporous Fe-MFI zeolites have been successfixlly prepared by treatment of isomorphously substituted Fe-silicalite and ion-exchanged Fe-ZSM-5 in alkaline medium. Iron in framework positions directs the silicon extraction towards mesoporosity development, whereas iron in non-framework positions inhibits silicon dissolution and limits mesoporosity development. 

The relatively high newly developed mesopore surface area of 120 m g is however in contrast to alkaline-treated silicalite-1, in which despite the high degree of unselective silicon dissolution hardly any mesoporosity has been observed. Due to the uncontrolled silicon extraction in the absence of framework aluminum, mostly large macropores were obtained in alkaline-treated silicalite [18]. We have previously reported a correlation between mesopore surface area development and framework Si/Al ratio of the parent... 

---