In Carbon removal 101 and The chemistry of enhanced weathering you saw the idea: spread crushed silicate rock on soil, let it dissolve, and the released calcium and magnesium end up holding CO₂ as dissolved bicarbonate. The rock that goes on the field is called the feedstock. Choosing it is the first decision in any project, and most of the later numbers (how much CO₂, how fast, at what cost, with what risks) follow from it.

This module builds directly on Minerals & lithology (mafic versus felsic, and the Goldich series) and Weathering, erosion & soil (what controls weathering rate, and why grain size matters).

The chemistry is reported as oxides: geochemists express each element as if it were combined with oxygen (CaO, MgO, Na₂O, K₂O) and give the result in weight per cent (wt%), the share of the rock's mass. "10 wt% CaO" means 100 kg of calcium oxide's worth of calcium in every tonne of rock. The values below are typical and illustrative; real rocks vary from quarry to quarry and even bench to bench.

The vertical axis of the chart above is only qualitative. Chemists put numbers on it by measuring a mineral's dissolution rate: how many moles of it dissolve from each square metre of exposed surface every second. Rates for common minerals span roughly nine powers of ten, so they are drawn on a log scale, where each gridline is ten times faster than the one to its left.

Three things stand out. Calcite dissolves something like ten thousand times faster than olivine, which is why limestone sits at the top of the chart above even though it does not count as removal. Basalt's minerals (glass, pyroxene and calcium-rich plagioclase) dissolve within a factor of ten or so of olivine. And granite's sodium and potassium feldspars are roughly a hundred times slower again, with quartz slower still.

The rate is per square metre, so the other lever is how much surface you expose. Grinding does not change the mineral; it changes the area. Halve a grain's size in every direction and you get eight smaller grains with twice the total surface. A tonne of 10 µm powder has about ten times the surface of a tonne of 100 µm powder. And because a grain dissolves inward from its surface at a steady pace, a grain ten times smaller is also gone about ten times sooner.

Mineral and grain size multiply. A fine powder of a slower mineral can keep pace with a coarse powder of a faster one, which is why labs measure both (see Characterising a feedstock, below). To watch the two combine over months and years, try the dissolution simulator in The chemistry of enhanced weathering.

How much CO₂ could a tonne of rock capture if every bit of its reactive calcium, magnesium, sodium and potassium dissolved? You can work this out from the chemical analysis alone, with nothing more than the bookkeeping of electric charge.

Chemists count atoms in moles: a fixed number of atoms or molecules (about 6 × 10²³), chosen so that one mole of a substance weighs its molecular mass in grams. One mole of CaO weighs 56.1 g; one mole of MgO, 40.3 g; one mole of CO₂, 44 g. Dividing a mass by the molar mass converts "how heavy" into "how many".

Now follow the charges. Each calcium or magnesium ion carries two positive charges (Ca²⁺, Mg²⁺). In water, those must be balanced by negative ions, and in natural waters that mostly means bicarbonate, HCO₃⁻, which carries one negative charge and one carbon atom taken from CO₂. So:

MgO behaves like CaO, and K₂O like Na₂O: every mole of any of these four oxides can, at most, hold two moles of CO₂ as dissolved bicarbonate. That is the bicarbonate pathway.

There is a second, less efficient ending. If the water becomes saturated, calcium and magnesium can precipitate as solid carbonate minerals (CaCO₃, MgCO₃). Each solid carbonate holds only one carbon per calcium or magnesium; the second CO₂ is released again as it forms. Sodium and potassium carbonates are far too soluble to form in soils, so on this solid-carbonate pathway only CaO and MgO count, at one mole of CO₂ each. In practice it is mostly CaCO₃: magnesium carbonates rarely form in soils, so the figure for magnesium is an upper bound.

Put together, the calculation is: for each oxide, divide its mass in a tonne of rock by its molar mass, multiply by the moles of CO₂ it can hold (2 on the bicarbonate pathway; 1 for CaO and MgO on the solid-carbonate pathway), and multiply by 44 kg per kilomole of CO₂. Add up the oxides. This is a widely used formula in the weathering literature, often named after the researchers who popularised it (you may hear it called the Steinour or Renforth approach), and some versions add correction factors.

Before a rock goes on a field, a lab has to answer three questions: what is in it, how fast will it react, and is it safe? Each step depends on careful sampling, because an assay (a lab measurement of what a sample contains) is only as good as the sample behind it. Mineral explorers live by the same rule (see Prospecting & the exploration pipeline).

Two details trip people up. First, chemistry alone cannot tell you reactivity. Two rocks with identical oxide totals can dissolve at very different rates if one holds its calcium in glass and pyroxene and the other in feldspar. That is why XRD sits next to XRF. Second, not all the CaO in an analysis is available. If the rock contains some calcite (calcium carbonate, common in altered basalts), that calcium is already paired with carbon, just as in limestone, and must be subtracted. Labs measure this as inorganic carbon, and it also shows up in the loss on ignition, the weight a sample loses when heated strongly as water and CO₂ are driven off.

Grain size is summarised from a cumulative curve. The d50 is the size that half the mass of the powder is finer than; the d80, the size that 80% is finer than. A human hair is roughly 50–100 µm (micrometres, thousandths of a millimetre) thick, so in the example below, roughly half the powder by mass is finer than a hair is thick.

Safety thresholds are set by regulators and certification standards, not by the rock. As a rough guide, basalts usually carry nickel and chromium at tens to a few hundred parts per million (ppm), grams per tonne, while ultramafic rocks often carry a few thousand. Regulators look both at total concentrations and at how much is in a form that can dissolve and move. The other hazard is dust: respirable crystalline silica, fine quartz particles small enough to reach deep into the lungs, can cause the lung disease silicosis, so quartz content and dust control matter wherever rock powder is handled.