Abstract
Anodic alumina materials exhibiting regular pore structure, microporosity, and extensive surface areas were prepared and characterized. The effects of current density, (J = 12–35 mA/cm2), anodization time (t = 30–150 min), and hydrothermal treatment on pore structure were investigated. Nitrogen sorption hysteresis was simulated using the corrugated pore structure model (CPSM). Pore size distributions, relative specific surface area (SCPSM/Sext = 870–8645), microporosity (max ∼ 33.0%), pore tortuosity (TCPSM = 3.1–5.7), pore connectivity (c = 3.02–4.85, Seaton's model), and nominal pore length values (i.e., Ns = 3–10, from CPSM, and L = 0.91–1.20, Seaton) were evaluated. Pore sizes dpore > 13 nm deduced via CPSM simulation of gas sorption data were also detected by SEM imaging. A minimum external surface pore density of ∼ 7.5 × 1010 pores/cm2 was evaluated from the SEM micrograph. Anodization conditions and the following treatment caused a severe pore structure change. Pore tortuosity (TCPSM) changes inversely proportionally to pore connectivity (c), while the nominal pore length (Ns) varies proportionally to the number of pore length (L). It is concluded that materials possessing microporosity, regular pore architectures, and high surface areas can become potential candidate membranes for gas separation and catalytic reaction applications. They can also be used as templates in electrochemical applications (e.g., solar and fuel cells).
Acknowledgment
The authors thank Professor P. J. Pomonis for offering the use of the SEM facility and A. P. Katsoulidis, Ph.D. candidate, for carrying out the SEM imaging in the Network of Laboratory Units and Centers of the University of Ioannina. Thanks are also due to P. Koukis, who contributed in the anodization experiments.
Notes
a Parameters of the Kelvin and modified Halsey equations are defined in Androutsopoulos and Salmas (Citation2000a).
a Immediate water rinse.
b Water rinse 24 h after anodization.
c Hydrothermal treatment of indicated specimen.
a Use of the DR characteristic curve with the intercept taken at the point A (i.e., the upper end of the linear part of the DR curve).
b An affinity coefficient for nitrogen, β = 0.32 (Equation (10)) was considered (Gauden et al., Citation2004).
c The DA empirical parameter N (Equation (10)) was taken as N = 6 to obtain a linear form of the pertinent adsorption data.
a C: BET equation constant.
b SCPSM/Sexter: reduced surface area, where Sexter (geometric external surface area of pellets) = 5.92583 × 10−4 m2/g (for dAl = 2.7 g/cm3).
a Corresponds to those in Table I.