[MSA-talk] REE, lanthanides, lanthanoids, Y and Sc

Anton Chakhmouradian Anton.Chakhmouradian at umanitoba.ca
Mon Nov 27 11:13:30 EST 2017

Hello All,

Indeed, chemists include Sc and Y in the REE family, whereas geochemists do not usually consider Sc as part of that family and are split on the Y issue. The difference is: chemists look at elements mostly from the standpoint of pure elements/compounds, whereas geochemists look at how elements ASSOCIATE with one another in nature. For example, ScPO4 has the same zircon-type structure as YPO4, YbPO4 and other heavy-REE orthophosphates, so it would seem natural to a chemist to just consider Sc as part of the REE family. In reality, Sc3+ is >10% smaller than Lu3+, the smallest lanthanide (or lanthanoid, according to the revised IUPAC nomenclature). This difference in ionic radii will drive Sc partitioning in natural processes, effectively forcing it to either form its own minerals or to associate with trivalent or tetravalent cations that are closer to it in ionic radius (IR). For example, zircon associated with the scandium phosphate pretulite (IR [8]Zr = 0.84 vs. IR [8]Sc = 0.87 A) would contain up to 3.2% Sc2O3, whereas xenotime (YPO4) found in the same association (IR [8]Y = 1.02 A) <  0.7% Sc2O3 (Moelo et al., Can. Mineral., 40, 1657).

The differences in radius do also affect the crystal chemistry and metallogeny of Sc. For instance, kolbeckite (ScPO4.2H2O) is isostructural with metavariscite (AlPO4.2H2O), not churchite YPO4.2H2O (Yang et al., Acta Crystal., C63, i91). In laterites, which are probably the most promising future Sc resource (if there is ever any need for much Sc), where this element substitutes in Fe3+ oxides and hydroxides (Chassé et al., Geochem. Persp. Lett., 3, 105).

Yttrium, on the other hand, is a close match for holmium and usually present in heavy-REE minerals at CI-normalized levels comparable to those of Ho. This is the reason why it is often placed between Dy and Ho in chondrite-normalized REE plots. Some minerals (notably, fluorite) exhibit departure of Y from this predictable behavior, which is reflected in Y anomalies in normalized plots. These anomalies are undoubtedly related to differences in Y and Ho speciation in some types of fluids and melts. Despite this intimate GEOCHEMICAL association between the lanthanides and Y, some geoscientists prefer not to include the latter in the REE family, usually without explaining their reasons for doing so.

To those interested in learning more about REE, their applications and politics behind them, as well as the "rare vs. not-so-rare" debate, I can recommend the special issue of Elements Frances and I edited in 2012:

Anton R. Chakhmouradian
University of Manitoba    

-----Original Message-----
From: msa-talk-bounces at minlists.org [mailto:msa-talk-bounces at minlists.org] On Behalf Of Frank Spera
Sent: November 26, 2017 5:51 PM
To: Don Halterman <donhalterman at comcast.net>
Cc: msa-talk at minlists.org
Subject: Re: [MSA-talk] REE...

well, if i recall the REE include Y and Sc which are NOT rare Lanthanides …..but its been a long time since i took an inorganic chemistry class!!

-----Original Message-----
From: msa-talk-bounces at minlists.org [mailto:msa-talk-bounces at minlists.org] On Behalf Of Don Halterman
Sent: November 26, 2017 6:13 PM
To: Frank Spera <spera at ucsb.edu>
Cc: msa-talk at minlists.org
Subject: Re: [MSA-talk] REE...

It has been some time since I read the article deprecating the term, but I recall the phrase "the elements in question are neither rare nor earths," or words to that effect.  Unlike other matters of terminology, this one isn't that critical so I won't pursue the matter further if I seem to be the only one who recalls this.


MSA-talk mailing list
MSA-talk at minlists.org

More information about the MSA-talk mailing list