Caption for background photo here. Learn more

Search Results

Chemical Diffusion and Crystalline Nucleation...

Chemical Diffusion and Crystalline Nucleation in Fe2+,3+-bearing (Semiconducting) Aluminosilicate Melts and Glasses: Petrological processes involving crystallization and/or phase separation in silicate melts involve the processes of chemical diffusion. Commonly, diffusion problems are addressed by application of the phenomenological models referred to as Fick's Laws.

Very often, taking this approach results in "missing" totally the physics of a dynamics problem. The research we are pursuing in this area emphasizes oxidation/reduction dynamics in multicomponent silicate melts. We have proven that, in most cases, redox dynamics in melts does not involve diffusive motion of an ionic or atomic/molecular oxygen species (see right)! The research has applications in, e.g., magma crystallization and chondrule formation/evolution. As importantly, the reactions studied can form a phenomenological basis for articulating a "point defect" thermodynamics for the amorphous state. To this end, the polymerization model for silicate melts developed and refined by Prof. Paul Hess in our department form a foundation for our "point defect" understanding.

The dynamics of oxidation and reduction reactions in transition-metal-cation-bearing aluminosilicate melts and glasses, at least for anhydrous conditions, involves the coupled diffusion of electron holes (h·; polarons) and component network-modifying cations—a kinetic response orders-of-magnitude faster than any involving the diffusive motion of an oxygen species (ionic or atomic). The redox reaction causes a fundamental change of the amorphous molecular structure, with various results as a function of composition. For example, Fe2+-doped glasses devoid of alkali oxides, oxidation results in the homogeneous nucleation of a crystalline ferrite at an internal front; in melts, e.g., a basaltic magma, the ferrite forms at an internal front as well, in a process well-described as "isothermal undercooling:" the local oxygen fugacity changes liquidus crystalline phase and the liquidus temperature (Figures 4 and 5).

Figure 4: Schematic (A) and transmitted-light micrographs (B,C,D) of a natural basaltic melt droplet processed via AeroAcoustic Levitation (AAL): a small piece of Columbia River flood basalt was levitated on a gas jet, centered and stabilized by ultrasonic transducers and heated to 1400oC with a split-beam CO2 laser. The droplet was first levitated on an argon jet (pO2~10-6atm) for 30s followed by levitation on an air jet (pO2= 0.21atm) for 7s. The micrographs reveal two internal reaction fronts: the nucleation of magnetite occurs at the deepest front (x''') ; the conversion of magnetite to hematite occurs at the shallower front (x''). Also seen are Leisegang bands beyond x''', corresponding to conversion of Fe2+ to Fe3+within the melt as the local oxygen fugacity increases. The dynamics of the oxidation reaction in air are shown in Fig. 5.

Figure 5: Reaction schematic for a basaltic liquid oxidized in air. The physics— revealed unequivocally by ion-backscattering spectroscopy— involves the "outflux" of network-modifying Fe2+ and Ca2+, charge-compensated by a counterflux of electron holes (h·). The flux of Fe2+ dominates the magnetite-formation reaction at x'''; the flux of Ca2+ dominates the reaction turning the magnetite into hematite at x'' (i.e., in the figure, x > y).

Representative Publications:
Cooper, R.F., J.B. Fanselow, J.K.R. Weber, D.R. Merkley and D.B. Poker (1996). Dynamics of oxidation of a Fe2+-bearing aluminosilicate (basaltic) melt. Science 274, 1173-1176.

Cook, G.B. and R.F. Cooper (2000). Iron concentration and the physical processes of dynamic oxidation in an alkaline earth aluminosilicate glass. Am. Mineral. 85, 397-406.

Smith, D.R., and R.F. Cooper (2000). Dynamic oxidation of a Fe 2+ -bearing calcium-magnesium-aluminosilicate glass: The effect of molecular structure on chemical diffusion and reaction-morphology. J. Non-Crys. Solids 278, 145-163.

Reduction processes, e.g., creation of metallic Fe from a FeO-bearing aluminosilicate melt, have been demonstrated to follow a "mirror-image" kinetic process (Fig. 6), at least when the reduction is performed in a carbon-rich, hydrogen-poor environment. In this case, nm-scale, colloidal iron metal is precipitated at an internal interface; the physics involves the "inward" diffusion of divalent network-modifying cations. One ramification of the dynamic reduction process, for example, is enrichment of alkali ions and large alkaline earth cations in the near-surface of the silicate melt, as well as chemical concentration profiles that in no way match the simple profiles associated with chemical diffusion following Fick's 2nd Law. (See, too, the "Technological Applications" page.) Microstructures for experiments at temperatures below the melting point of appropriate iron alloys are particularly comely (Fig. 7).

Figure 6: Reduction dynamic for the FeO-bearing magnesium aluminosilicate melt. For more complicated silicate melt chemistries, the reduction is still dominated by the chemical diffusion of the small, divalent network-modifier cations; consequently, the silicate chemistry as measured from the free surfsce can become quite complicated.

Figure 7: Free surface (left image: BSE/SEM) and internal region (right image, which is between x' and x'' in Fig. 6: TEM) of an FeO-MgO aluminosilicate (FMAS) melt droplet reduced to crystalline Fe metal plus MAS melt at 1355oC and pO2~10-13atm. The surface Fe crystals have experienced vapor-phase-effected coarsening; the {111} face truncated by traces of {100} faces are clearly evident. Internally, 5-20nm crystals of iron are arrayed in a "string-of-pearls" morphology; the morphology is diagnostic of having network-modifying cations in the aluminosilicate melt existing in percolative channels. The reaction proceeds via the dynamic depicted in Fig. 6.

Representative Publication:
Everman, R.L.A. and R.F. Cooper (2003). Dynamic internal reduction in an Fe-doped magnesium-aluminosilicate melt. J. Am. Ceram. Soc. 86, 487-494.

Brown faculty collaborators:

Paul Hess

Other project collaborators:

J.B. Fanselow, J.K.R. Weber, D.R. Merkley, D.B. Poker, Cook, G.B., Smith, D.R., Everman, R.L.A.

Figure 4

Figure 5

Figure 6

Figure 7.a

Figure 7.b

Back to Reid Cooper's Brown Research Profile
Research at Brown: Reid Cooper: Chemical Diffusion and Crystalline Nucleation... Brown Research home page
Contact Us Search