Academia.eduAcademia.edu

Figure 2 – uploaded by Philippe Allongue

See full PDFdownloadDownload figure
ame as in Table 1, but for the complete reaction scheme (reactions 1-4). Note the independence of k, and k_, with the pH of the solut greement with our expectations (see text, Section II). Two hypotheses have been considered for the effect of the doping density in the case of 6]

Table 1 ame as in Table 1, but for the complete reaction scheme (reactions 1-4). Note the independence of k, and k_, with the pH of the solut greement with our expectations (see text, Section II). Two hypotheses have been considered for the effect of the doping density in the case of 6]

Related Figures (11)

Fig. 2. Evolution of the calculated Tafel plots (a) and direct polarization curves (b) as a function of the dissolution kinetic constant k5, (simplified model). (©): experimental data for the GaAs/0.5 M H,SO, junction (Np = 2x 10!6 cm~4) [10]. Curves (1-5): kjg = 1079, 5X 107 !°, 10719, 5x 107"! and 2x 107! cm? s~!. The other parameters are those of Table 1.
Fig. 2. Evolution of the calculated Tafel plots (a) and direct polarization curves (b) as a function of the dissolution kinetic constant k5, (simplified model). (©): experimental data for the GaAs/0.5 M H,SO, junction (Np = 2x 10!6 cm~4) [10]. Curves (1-5): kjg = 1079, 5X 107 !°, 10719, 5x 107"! and 2x 107! cm? s~!. The other parameters are those of Table 1.
Set of parameters derived from simulations according to the simplified corrosion scheme (reactions 1, 5 and 4). For GaAs, two doping levels were available; note the evolution of k39, the corrosion kinetics, with Np in both solutions. For InP, k53 was set to zero for p-InP; it could be neglected for GaAs [10] TABLE 1
Set of parameters derived from simulations according to the simplified corrosion scheme (reactions 1, 5 and 4). For GaAs, two doping levels were available; note the evolution of k39, the corrosion kinetics, with Np in both solutions. For InP, k53 was set to zero for p-InP; it could be neglected for GaAs [10] TABLE 1
Fig. 3. Same as Fig. 2, but for the capture of the first hole, constant k,. (©): experimental data of the GaAs/0.5 M H,SO, junction (Np =2x10!® cm~3) [10]. Curves (1-5): ky = 3X10", 1.2107", 6x107!2, 3x107!2 and 1.2x107!% cm? s~!. The other parameters are those of Table 1. Note the smaller influence of k, upon the polarization curve (b).
Fig. 3. Same as Fig. 2, but for the capture of the first hole, constant k,. (©): experimental data of the GaAs/0.5 M H,SO, junction (Np =2x10!® cm~3) [10]. Curves (1-5): ky = 3X10", 1.2107", 6x107!2, 3x107!2 and 1.2x107!% cm? s~!. The other parameters are those of Table 1. Note the smaller influence of k, upon the polarization curve (b).
Fig. 5. Effect of the introduction of a chemical step, in the reaction scheme, over the calculated J,(/,) variations. (x, ©) are experimental data of the GaAs/K ,Fe(CN), contact. ( ) curve calculated according to parameters of Table 1: £55 = 6X 10-7 cm? s~!; (------ ) curve calculated according to the parameters of Table 2, i.e., with a chemical step in the reaction.
Fig. 5. Effect of the introduction of a chemical step, in the reaction scheme, over the calculated J,(/,) variations. (x, ©) are experimental data of the GaAs/K ,Fe(CN), contact. ( ) curve calculated according to parameters of Table 1: £55 = 6X 10-7 cm? s~!; (------ ) curve calculated according to the parameters of Table 2, i.e., with a chemical step in the reaction.
Fig. 6. Simulations of various functions f(s) vs. c2.5 / Ton OT Crea /J ph for an interface GaAs/acidified solution stabilized by a redox couple which does not adsorb. (a) f(s) = s7/(1— s)+ s /6; (b) f(s) = 57/1 — 5s); (c) f(s)= 57/6 (1— s)+ s3/(1— s)?. The redox transfer is direct (k= 10745 cm* s~!). To cover several decades c,.g /mol 1~' has been changed. The calculated points compare favorably with case (b).
Fig. 6. Simulations of various functions f(s) vs. c2.5 / Ton OT Crea /J ph for an interface GaAs/acidified solution stabilized by a redox couple which does not adsorb. (a) f(s) = s7/(1— s)+ s /6; (b) f(s) = 57/1 — 5s); (c) f(s)= 57/6 (1— s)+ s3/(1— s)?. The redox transfer is direct (k= 10745 cm* s~!). To cover several decades c,.g /mol 1~' has been changed. The calculated points compare favorably with case (b).
Fig. 7. Normalized photocapacitance spectra of the KOH/ (a) and H,SO, /n-GaAs (b) contacts in single beam conditions. The band bending is 0.95 V in both cases [15]. The transitions of interest have been indexed; They correspond to the corrosion states. 5;, the amplitude of a transition, is related to the density of the state. Note the increase of 5, at low pH, although 5, remains almost unchanged. An interesting feature of our model is its relation to the state distribution at the interface. Table 2 shows a significant difference between GaAs and InP with respect to the trapping mechanism of the first hole. For GaAs, using a realistic value for o,=10~'® cm? in eqn. (7), one finds that Ey, =Ey+0.13 eV and E, = E, + 0.5 eV. For InP, only a threshold can be considered for k_, which must be <107'8 cm? s~'. Taking again o, = 10~'° cm’, E,, = Ey + 0.25 eV, Es is found to be at least 0.5-0.7 eV above E,,, the valence band maximum. With respect to the literature, such values for E, are quite compatible with the distribution of defect states generally found at GaAs and InP surfaces in UHV [6.17121 The anreamaent ie avgan hattae when AnAnciddeéringe tha Alctmhingan at
Fig. 7. Normalized photocapacitance spectra of the KOH/ (a) and H,SO, /n-GaAs (b) contacts in single beam conditions. The band bending is 0.95 V in both cases [15]. The transitions of interest have been indexed; They correspond to the corrosion states. 5;, the amplitude of a transition, is related to the density of the state. Note the increase of 5, at low pH, although 5, remains almost unchanged. An interesting feature of our model is its relation to the state distribution at the interface. Table 2 shows a significant difference between GaAs and InP with respect to the trapping mechanism of the first hole. For GaAs, using a realistic value for o,=10~'® cm? in eqn. (7), one finds that Ey, =Ey+0.13 eV and E, = E, + 0.5 eV. For InP, only a threshold can be considered for k_, which must be <107'8 cm? s~'. Taking again o, = 10~'° cm’, E,, = Ey + 0.25 eV, Es is found to be at least 0.5-0.7 eV above E,,, the valence band maximum. With respect to the literature, such values for E, are quite compatible with the distribution of defect states generally found at GaAs and InP surfaces in UHV [6.17121 The anreamaent ie avgan hattae when AnAnciddeéringe tha Alctmhingan at
Fig. 8. Energy diagrams showing the contrasting behavior of GaAs and InP with respect to the corrosion reaction. Dotted arrows indicate improbable charge exchanges. The chemical step (see Fig. 1) has been omitted for clarity.
Fig. 8. Energy diagrams showing the contrasting behavior of GaAs and InP with respect to the corrosion reaction. Dotted arrows indicate improbable charge exchanges. The chemical step (see Fig. 1) has been omitted for clarity.
The constants k, and the ratio k-/k_ ¢ are reflecting the evolution of the dissolution kinetics with the pH of the solution for a given material. At low pH, the kinetics of the corrosion of GaAs is faster than at pH = 14 and the chemical equilibrium is displaced to the right. For InP, these characteristics are little modified. Note that for GaAs, the chemical step becomes almost irreversible in accord with ref. 5 for the GaAs(100)/H,SO, interface (see Table 1 of ref. 5). In this paragraph, we intend to bring more insight into the interpretation of these observations through the identification of surface states. This was possible for GaAs only.
The constants k, and the ratio k-/k_ ¢ are reflecting the evolution of the dissolution kinetics with the pH of the solution for a given material. At low pH, the kinetics of the corrosion of GaAs is faster than at pH = 14 and the chemical equilibrium is displaced to the right. For InP, these characteristics are little modified. Note that for GaAs, the chemical step becomes almost irreversible in accord with ref. 5 for the GaAs(100)/H,SO, interface (see Table 1 of ref. 5). In this paragraph, we intend to bring more insight into the interpretation of these observations through the identification of surface states. This was possible for GaAs only.
Fig. 10. Relative position of GaAs and InP band edges (in darkness) with respect to standard potentials of corrosion reactions (see text). For GaAs the states are E, = Ec — 0.98 eV (As-site) and E, = Ec — 1.15 eV (Ga-site) as derived from Fig. 7 and ref. 15. For InP the states are E, = E, —0.15 eV (P-site) and E, = Ec —0.4 eV (In-site). The potentials Ep(i) refer to standard values of reaction (i) of the text. Note the position of Eo(9) below E, in the case of the GaAs/H,SO, contact which explains that As-rich surfaces are obtained after acidified etching.
Fig. 10. Relative position of GaAs and InP band edges (in darkness) with respect to standard potentials of corrosion reactions (see text). For GaAs the states are E, = Ec — 0.98 eV (As-site) and E, = Ec — 1.15 eV (Ga-site) as derived from Fig. 7 and ref. 15. For InP the states are E, = E, —0.15 eV (P-site) and E, = Ec —0.4 eV (In-site). The potentials Ep(i) refer to standard values of reaction (i) of the text. Note the position of Eo(9) below E, in the case of the GaAs/H,SO, contact which explains that As-rich surfaces are obtained after acidified etching.
Related topics:
Electroanalytical Chemistry
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved