This is a good question, Seb. Much of the available information...

This is a good question, Seb. Much of the available information relating to phosphates, REE and niobium mineralogy and recovery has been helpfully analysed in the past - in particular, from my searching, I found old posts by @halo_3000 and @2ic elucidating - but I’ll attempt to recapitulate with my limited understanding here. I have no geology background, so view the below with discretion.

Phosphates

There is an independent phosphate mine operating at Araxa (ditto Catalao) - Mosaic pit on the attached Araxa map - 130MT @ 12.4% grade, operating capacity 1.3MT phosphate, crudely estimate ~.3b annual revenue? (source). Mosaic acquired it from Vale for, very roughly, A$1b in 2016, including all the processing plants and infra (source). The acquisition included three other mines for a total of $2.5b USD (including Catalao) so I’m pro-rating based on the total deposit size from Table 2.6 in the above MOS 10-K, assuming an equal split between Araxa and Patrocinio, which are pooled and quoted as having a similar surface area.

Crude estimation from WA1’s P2O5 drill % suggest something may be economically viable, but it seems like there are always plenty of material idiosyncrasies to the mineralogy - Mt Weld niobium, discussed below, being an important example of this. Nonetheless, from the WA1 preliminary mineralogy report from 05/06/23.

“The key phosphate-bearing mineral identified was apatite. The apatite was also positively liberated with all three samples reporting over 90% in the ‘well-liberated’ or ‘high-grade middlings’ categories in the -150+38 µm fraction.”

This seems to read positively, in terms of the mineral, liberation, and size under consideration.

REEs

This leads to another question, related to the feasability of the REE at WA1: why is there no REE processing at Araxa? ITAFOS now owns the option on this (source) - ITAFOS REE on the attached Araxa map. The question is well covered in this paper; Araxa has 6MT TREO @ 5% grade, and, as foreshadowed in my earlier paragraph, the issue with processing seems to be subtleties with the mineralogy:

“Considering the large REE reserves—6.37 Mt proved, and additional 21.94 Mt inferred—its 6.8% TREE grade and the already available infra- structure, Araxá certainly is a very interesting prospect. However, complex mineralogy, dominated by gorceixite that accounts for 50% of the mass and 26% of the REE, offsetting the main carrier monazite (73% of the REE), and texture, due to fine crystal and particle size and a strong iron oxide overprint, condition the processing options for the ore. Small particle size and concentration of the REE in the fines (47.4% of the REE in 38.8% of −20 μm size fraction) shall additionally burden to process efficiency.”

So 1. gorceioxite offsets the main carrier, monazite, and 2. poor liberation properties for the monazite, 3. Concentration in the fines.

Similarly to above, there isn’t much information to work with regarding REE mineralogy at WA1 except this snippet from the preliminary mineralogy report:

“The key REE-bearing minerals identified were monazite/rhabdophane and crandallite. The monazite/rhabdophane was observed to have high-REE content (~55-60% TREO1) and all three samples reported over 75% classification in the ‘well-liberated’ to ‘medium-grade middlings’ categories in the -150+38 µm fraction.”

Hopefully this is at least suggestive that 1. No gorceioxite, 2. Liberation properties for monazite are better, at least in the -150+38 range, 3. No indication on what fraction of particles are fine, and I believe there is a general expectation that weathered deposits have a significant fraction of fine particles - again, Mt Weld is an extreme example - but the WA1 release again only mentions the -150+38 µm range.

Niobium

Lastly, a question which has been re-raised is why Mt Weld isn’t pursuing niobium production, despite having an ample deposit. I have been concerned with this question myself and its potential relation to mineralogy, particularly if it may have implications for WA1. Again, this was covered long in the past by @halo_3000 and @2ic, possibly among others, and this paper provides a detailed analysis of the Mt Weld mineralogy, and why it seems difficult to produce a satisfactory niobium concentrate there.

“The Mt Weld niobium deposit occurs in altered carbonatites in Western Australia where Sr pyrochlore and ferroniobate were the major primary Nb–Ta containing minerals. Primary apatites, pyrochlores and niobates were altered to form micrometre to submicrometre size crandallites. The fine crandallite particles were often clustered to form larger aggregates, mixed or cemented with fine niobates (up to 20 mm) and coarse (up to 200 mm) ilmenite. This study showed that the ore was not suitable for gravity concentration, magnetic upgrading and flotation due to fine particle size. Although the fine particle size was advantageous for chemical treatment and ore responded better to caustic and acid leaching, the chemical treatment alone was well short of providing a commercial grade niobium concentrate containing 60% Nb2O5zTa2O5.”

In summary, the weathering process there has altered the mineral composition to such a degree that the resulting niobium-bearing minerals, Sr-pyrochlore and ferroniobiate, are extremely fine, and are suspended in an extremely fine matrix of crandallite particles, which themselves bear a significant fraction of the niobium content (“Phosphate minerals, containing some niobium and to a smaller extent tantalum, were assumed to be the alteration products of primary apatite. It is presumed that during the alteration of primary apatites and pyrochlores of the carbonatite body, calcium was replaced to varying degrees by Al, Sr, Nb, Ta, Ba or REEs giving rise to diverse combinations of these elements with phosphorous. The major alteration phases were Nb–phosphate, Nb–Sr–Al phosphate, Al–REE phosphate, Sr–Al–Ca phosphate and Ca–Sr–Nb–REE– Al phosphate. The niobium content of the crandallites measured by EPMA ranged from 0.7 to 44% Nb2O5 (average 7.4%) and the tantalum content from 0.025 to 0.24% Ta2O5 (average 0.13%), indicating that crandallites were also a major source for niobium mineralisation.”).

Referring again back to the WA1 and the preliminary mineralogy release:

“The key niobium-bearing mineral identified was pyrochlore which is consistent with previous petrographic analyses on the 2022 discovery drillholes, along with columbite being dominant in one of the samples. Importantly, the observed pyrochlore (~70% Nb₂O₅, 0.18% Ta₂O₅) and columbite (~74% Nb2O5, 0.06% Ta₂O₅) primary minerals both have very high-niobium content and low-tantalum levels. These mineral grains were also highly liberated within the sample medium. When sized in the -150+38 µm fraction, all three samples reported over 85% of the dominant niobium bearing mineral as being ‘well-liberated’ or ‘high-grade middlings’ (noting these liberation properties are presented as this represents the size range that flotation would most likely be conducted at)”

I optimistically interpret the above, i.e. the WA1 statement only mentioning pyrochlore and columbite, the usual niobium-bearing minerals found at Araxa, Catalao and Niobec, as well as the WA1 phosphate statement only mentioning apatite (although note that crandellites are mentioned in the REE paragraph..), along with the focus on the -150+38 µm particle size range, as indicative that such an alteration process as the one which has occurred at Mt Weld, distributing the niobium to impractically fine particles, is not a significant concern for Luni.