Research
My doctoral thesis was undertaken at the University of Western Australia between 1998 and 2002 and was inspired by a chapter in Secrets of the Soil.
The abstract is presented below. If you would like an electronic copy of the thesis, please email me at andrewharley@ascensionsoil.com.
THE EVALUATION AND IMPROVEMENT OF SILICATE MINERAL FERTILISERS,
Andrew Harley PhD Thesis, UWA 2002
Abstract
The use of silicate mineral fertilisers has been limited by low dissolution rates that release nutrient elements too slowly for adequate crop and pasture nutrition resulting in high application rates that are cost prohibitive. Research on silicate mineral fertilisers has generally focused on solution and plant nutrition chemistry rather than the surface mediated process controlling mineral dissolution and subsequent plant uptake. Consequently, no appropriate methodologies for evaluating silicate mineral fertilisers exist. This thesis is a contribution towards increasing our knowledge in these fields. Three common rock forming minerals, microcline, biotite and hornblende, were subjected to high-intensity attrition milling to produce materials with a wide range of structural, surface and chemical properties which were evaluated in dissolution and plant growth experiments.
Progressive attrition milling of minerals for up to 24 hours resulted in an overall reduction in median particle size and resultant increase in specific surface area in the order hornblende>biotite>microcline. X-ray diffraction analysis indicated that this reduction in particle size was accompanied by an increase in amorphous material as indicated by decreasing peak intensities and increasing background scatter with increasing milling time. Transmission electron microscopy showed the presence of both micron and sub-micron particles in minerals milled for 24 hours as well as the aggregation of sub-micron particles into larger masses. Aggregation was greater for hornblende and biotite than for microcline. The particles of milled minerals produced selected area electron diffraction patterns with both broad rings indicative of amorphous material and sharp, very spotty rings indicating the presence of large numbers of randomly oriented, single crystals. High-resolution field emission source TEM indicated that the presence of amorphous material was mostly limited to rims around particle edges and that rims were more pronounced for microcline than for biotite and hornblende. Energy dispersive spectra obtained during TEM of particles of unmilled and highly milled minerals showed no or minimal chemical changes associated with milling.
Dissolution of minerals in both deionised water and dilute acid increased with milling time by between 10 and 100 fold, with up to 10% of some milled minerals dissolving during a 16 day period. Dissolution was greater in acid solution than in deionised water. Initial dissolution was incongruent with the base cations generally being in excess of Si except for octahedral Mg release from both biotite and hornblende and K release from biotite where it was exchanged from interlayer sites rather than being released by dissolution. Incongruent dissolution was generally greatest for the longest milling times, and dissolution tended towards congruency with dissolution time. Dissolution versus time curves for unmilled minerals were generally between parabolic and linear, tending towards linear with increasing milling time. Unit mass based dissolution rates increased with increased milling time and generally decreased with increased extent of dissolution. In several cases, elemental dissolution rates for highly milled minerals remained constant with increasing dissolution time. Si, K and Ca dissolution were linearly related to specific surface area, while Na and Mg dissolution were not systematically related to surface area.
Milled silicate minerals were less than 30% as effective as soluble fertilisers for ryegrass and clover grown on sand over a six-week period. Ryegrass growth was sustained for up to 268 days where milled microcline and biotite were supplying K and milled hornblende was supplying Ca. Applications of biotite and hornblende to supply Mg were generally too low to supply sufficient concentrations for maximum plant growth. Nutrient uptake increased with increased milling time, and for application rates up to between 10 and 20 (g mineral)(kg soil)-1, beyond which increased application rate did not produce increased nutrient uptake. Exchangeable cations in the soil increased with increased milling time, and plant growth persisted where the exchangeable cation pool was not exhausted. The application of milled mineral increased soil pH with the lime equivalence of silicates being less than 10%. The uptake by ryegrass of nutrients was positively and linearly related to surface area values up to 100 m2 kg-1 for K, 40 m2 kg-1 for Ca and 300m2 kg-1 for Si and at least 60 m2 kg-1 for Mg. It is proposed that surface area may be used as a predictor of plant response to provide a measure for determining optimal milling treatments and application rates of milled silicate minerals.
A simple economic analysis indicates that milling microcline for one hour is optimal when milling costs, application rates, cost of material, transportation and spreading costs are considered. The much greater costs of milled silicate minerals fertiliser relative to soluble K fertilisers indicates that silicate minerals are unlikely to be economically viable replacement for soluble fertilisers for most agricultural settings. Further research is required to determine the benefits of milled silicates in highly leaching, acid soils and to determine non-nutritional benefits such as liming, improved soil charge characteristics and reduced leaching, and the influence of milled silicate rocks on microbial activity and organic matter dynamics in soils.
