I'm building a mining colony world with large mine-able deposits of rare earth elements (REEs). One source of REEs on Earth is alkaline carbonitite magma. On Earth, I believe, it is thought to be dissolved from the magma and adsorbed onto clay particles. Over a very long time, could slowly cooling magma form a strata of REEs?
I found the below text at http://www.robertbeauford.net/rare_earth_elements/ree_geology. I took away from this and the rest of his monograph that slow cooling magma / fractional crystallization pushes rare earth elements into solution rather than having them crystallize out like copper or something. This process must team up with hydrothermal processes to concentrate rare earth elements, and that does not happen often.
All of the foregoing explanation, though slightly oversimplified, is meant to serve as a foundation to understand the following: As rare earth elements are excluded from silicate melts during fractional crystallization, they become concentrated in late forming peralkaline associated igneous rocks, whether silicates, phophates, or carbonates. As the process of exclusion continues, the rare earth elements, still not fitting nicely into the crystal lattices of many of the forming minerals, get pushed further aside, and wind up in solution in the even lower temperature 'melts' that we refer to as hydrothermal fluids. These water based fluids can either exchange their dissolved elements with surrounding rocks (forming skarns), or precipitate hydrothermal deposits of minerals crystallized directly out of solution. The rocks formed by this type of exchange or crystallization constitute the third major source of primary rare earth element ores.
It should also be pointed out that hydrothermal processes are not limited to shallow regions. They can occur at great depths in crust, where they play an important role in mineral sorting processes (especially in subduction zones).
Hydrothermal fluids don't have to originate during the original cooling of an intrusive pluton. Hot subsurface water can saturate a region of igneous rocks at any point, and add to or deplete the rare earth content of the rocks, either in overall REE content, or in distribution of REE content. There are a few limitations, however. Concentration of REEs to minable levels, by hydrothermal processes, is only going to occur in rock that is already rich in REEs through magmatic or other processes. Hydrothermal processes cannot accumulate mineable concentrations by dissolution of background levels of REE from regions of average igneous or metamorphic rocks. Hydrothermal and igneous processes work together in regions where REE are present, but this also only goes in one direction: Igneous processes can concentrate REE to mineable levels without the help of water, but the corrolary is not true. Magmatic concentration must preceed the action of water.
Including hydrothermal action in your scenario has promise. From http://www.nature.com/news/2011/110703/full/news.2011.393.html
It has long been known that the ocean might provide a wealth of rare earths. Sea-floor hydrothermal vents pump out rare-earth elements dissolved in their hot fluids. And these elements and others accumulate in potato-sized lumps, called manganese nodules, on the sea floor.
These nodules look suspiciously biological to me. I think I have read it posited that their formation is microbially mediated. Nothing better for a science fiction than mysterious xenobiogeochemistry!
One source of REEs on Earth is alkaline carbonitite magma
Carbonatite, not carbonitite. But yes - this is the main source for REE on earth. It's not the only one though!
Here are two ways to form cumulate deposits rich in REE. Note that these are not going to be purely REE because that's pretty much impossible here on earth.
REE-rich carbonatite reacting with other rocks
The REE carrying capacity of carbonatite magmas is dependent on the composition of the magma. If, for example, the carbonatite is phosphorus rich, it can carry large amounts of REE. When this magma flows through the crust, it can react with the regional rocks. If this reaction causes depletion in the element facilitating REE transport, then you deposit the REE. There are actually two examples of something like that that happened here on earth, and there are exploration and mining companies operating them at the moment. One is Hoidas Lake in Canada and the other is Nolans Bore in Australia. This may be a cumulate in the chemical sense, and not the morphological sense of horizontal layers.
True cumulates from alkaline magmas
Alkaline magmas that are not carbonatites can form large magma chambers which undergo fractionation. This means that minerals crystallise within the chamber, and then sink downwards because they're heavier than the magma. In some alkaline intrutions, one of the minerals is eudialyte. Although not strictly a REE mineral, most natural examples are REE-rich enough to make them mineable for a profit. Look at this image from Greenland for example:
The red layers are eudialyte (and therefore REE) rich. Here's another eudialyte layer:
On Earth, I believe, it is thought to be dissolved from the magma and adsorbed onto clay particles
There are some deposits where this happened. One of Mount Weld in Australia, where the REE occur in a carbonatitic laterite. Chinese heavy REE deposits are exactly that - REE that were leached from igneous rocks and concentrated in clays.