The dissolution of three synthetic allophanes (A2.0, molar ratio Al/Si = 2.02; A1.6, Al/Si = 1.64, and A1.3, Al/Si = 1.26) and two natural Japanese allophanes (KNP, Al/Si = 1.74 and KYP, Al/Si = 1.70) in the presence of boehmite was studied in 0.01 M NaCl (initial pH 3) at 100-150° C and equilibrium vapor pressures for 4-64 days in order to accelerate the dissolution rate. To estimate the thermodynamic stability of allophane, three assumptions had to be made: (i) that the series of allophane compositions represent stable thermodynamic phases rather than mixtures; (ii) that extrapolation from 100° C to 25° C is valid, and (iii) that conditions close enough to equilibrium can be attained and that extrapolation over time is valid. The apparent equilibrium constant (K) at 25° C and 0.1 MPa was calculated for the equilibrium reaction of allophane with boehmite: Al2O3.xSiO2.yH2O(s) + (2x − y + 1)H2O(1) = 2A1OOH(s) + xH4SiO4(aq), where x depends on the Al/Si molar ratio and y is the structural water, and for the dissolution reaction of allophane alone: Al2O3.SiO2.yH2O(s) + 6H+(aq) = 2Al3+(aq) + xH4SiO4(aq) + (3 − 2x + y)H2O(1). The calculated log K values were: −2.83 ± 0.12 (A2.0), −4.14 ± 0.12 (A1.6), −5.76 ± 0.12 (A1.3), −4.13 ± 0.12 (KNP), −4.32 ± 0.13 (KYP) for allophane dissolution with boehmite, and 14.20 ± 0.93 (A2.0), 12.89 ± 0.98 (A1.6), 11.27 ± 1.07 (A1.3), 12.90 ± 0.96 (KNP) and 12.71 ± 0.97 (KYP) for allophane dissolution alone. Apparent experimental standard Gibbs free energies were slightly less negative than those predicted by a polymer model.
An examination of our experimental results and previous studies indicates that assumptions 1 and 3 present serious problems. Allophane is not a phase of uniform composition and will slowly evolve toward more stable better ordered forms such as halloysite and kaolinite. Extrapolation to infinite time may be invalid because of complex reaction processes. Boehmite did not control Al3+ activity, and allophane dissolution equilibrium was not attained. Consequently, allophane does not seem to have a true thermodynamic stability field in a phase diagram, and it is an unstable solid with respect to halloysite and kaolinite.