Rare-Earth Terminology – A Quick Refresher On The Basics

by Gareth Hatch on December 11, 2012 · 15 comments

in News Analysis, Rare Earths, Tools & Metrics

Bookmark and Share

I came across a couple of online articles this week, which make reference to some pretty basic terminology with respect to the rare earths. Unfortunately these articles got much of it wrong, propagating inaccuracies that are particularly egregious when they keep appearing within various natural-resource media channels. I’m therefore going to take a little time here to address some of the key errors that keep cropping up.

First, let’s tackle one of my pet peeves. The term ‘rare-earth mineral’ is NOT synonymous with the terms ‘rare-earth elements’ or ‘rare-earth metals’. The International Mineralogical Association defines a mineral as “an element or chemical compound that is normally crystalline and that has been formed as a result of geological processes” [1]. By this definition the element gold, for example, would be considered a mineral; however the rare earths are never found in elemental form and so in their case, the term ‘mineral’ is never synonymous with ‘element’.

Rare-earth elements (REEs) can be found in many different minerals. Only when the occurrence of such REEs is in significant quantities, or the chemical formula for a particular mineral requires the presence of one or more REEs, does the compound become a rare-earth (or rare-earth-bearing) mineral.

As for what constitutes a rare-earth element or metal; sticking with the definition formulated by the International Union of Pure and Applied Chemistry (IUPAC) will generally keep you out of trouble. IUPAC defines the rare-earth metals as the 15 lanthanoid elements (with atomic numbers of 57 through to 71) in addition to scandium (Sc) and yttrium (Y) [2]. The lanthanoid promethium (Pm) is radioactive, with no stable isotopes; it is thus present in the Earth’s crust in vanishingly small quantities and does not occur with the other REEs. Sc exhibits some properties that are similar to other REEs, but is seldom found in the same minerals as them. It does not selectively combine with the common ore-forming anions and thus it is generally (though not exclusively) confined to trace occurrences [3].

REEs are difficult to separate from one another once they have been liberated from REE-bearing minerals. At the atomic level, the lanthanoid elements have a similar outer electronic structure, with this structure shielding the so-called 4f electrons within the atom. The presence and incremental filling of these electron orbitals are key characteristics of the lanthanoids. This unique structure leads to REE ions that are very similar in size to each other. It is this similarity in ionic radii across the group, not the adjacency of their atomic numbers, which gives rise to the similar chemical properties associated with each of the REEs. This makes them difficult to separate, even with the use of intensive processes such as solvent extraction (SX).

Now we come to one of the most contentious issues with respect to rare earths, namely the definitions used to describe the specific sub-groups of REEs known as light (LREE), medium (MREE) and heavy REEs (HREEs). I freely admit that in the past year or so, I have become less forgiving of the inconsistencies that we see in the industry with respect to the use of these terms. It’s time ‘we’ got our act together on this. One of the articles that triggered my own piece today is particularly egregious in getting this whole topic wrong.

Let’s start with a differentiation with which all chemists and geologists would likely agree, even if the non-techies in the industry don’t, namely the separation of the REEs into two groups by virtue of electronic structure. On this basis alone, the LREEs would be those that have no paired 4f electrons, specifically lanthanum (La), cerium (Ce), praseodymium (Pr), neodymium (Nd), Pm, samarium (Sm), europium (Eu) and gadolinium (Gd). This leaves the HREEs as terbium (Tb), dysprosium (Dy), holmium (Ho), erbium (Er), thulium (Th), ytterbium (Yb) and lutetium (Lu). Because the ionic radius of Y is very similar to that of the HREE Dy, Y is generally included as a HREE, despite not having any 4f electrons, paired or otherwise. In contrast, the ionic radius of Sc is much smaller than any of the other REEs and is therefore generally NOT classified as being either a LREE or HREE.

There are many in the industry that classify Eu and Gd as HREEs, but there really is no sound basis for doing so, unless you consider the inflation of a project’s HREE numbers as being a “sound basis”…

So now let’s turn our attention to the so-called MREEs. Of the relatively few sources in the industry that acknowledge the existence of this group, most of them get the composition of it wrong as well as its origin. The MREEs specifically refer to Sm, Eu and Gd, and the term arose from the metallurgists, process chemists and others who calculate and design the process flow sheets required to complete the separation of REEs from each other. This definition of the MREEs is the same one that the Chinese authorities use, by the way, when talking about the export quotas allocated to light rare earths, and to medium / heavy rare earths.

The MREE or SEG group as it is also known, arises within the early stages of SX, the main commercial process used to separate and to purify REEs, because of the absence of Pm from the feedstocks used. As we saw earlier, although Pm is a LREE, it does not occur with the other REEs, and so the process chemists take advantage of this ‘blip’ in the sequence of REEs, caused by the resulting ‘gap’ between Nd and Sm, during the separation process.

The figure above shows a summary of the elements of most interest to the REE industry, contained within the LREE, MREE and HREE groups. You can click on it to see an enlarged view.

The MREE group therefore arises for process engineering reasons only and does not require any new definitions, or changes to existing ones. The purists and certain laypeople can argue that these elements should strictly be defined as LREEs; such individuals should simply see the MREEs as a subset of the LREEs and worry themselves no more. I prefer to use all three terms – LREEs, MREEs and HREEs – and have been doing so exclusively for the past 18 months as I get more and more into the processing side of the rare-earth industry. Either way, you now know the origin of the term, and the three REEs to which it correctly refers. Either way, you also now know that Eu and Gd are really NOT HREEs.

Finally, the last term that I use quite a lot is critical REEs (CREEs), in reference to the five REEs that are of critical importance to future demand for sustainable energy sources – specifically Nd, Eu, Tb, Dy and Y.

As I always tell people – just make sure that you understand exactly which definitions a particular company is using, when looking at reported data which use one or more of the group names described above. In the meantime, let’s hope that certain of my fellow commentators on the rare-earth sector start to get the hang of the basic terminology for these materials…


1. E. H. Nickel, The Canadian Mineralogist, 1995, Vol. 33, pp. 689-690.
2. N. G. Connelly, T. Damhus, R. M. Hartshorn and A. T. Hutton, 2005, Nomenclature of Inorganic Chemistry: IUPAC Recommendations 2005, RSC Publishing, Cambridge, p. 366.
3. J. B. Hedrick, 2000, Minerals Yearbook: Volume I – Metals and Minerals, US Geological Survey, Reston, p62.1.

Bookmark and Share
1 Anup Narayan Singh December 12, 2012 at 12:46 AM

Dear Gareth,
I am an exploration geologist with more than 35 years in mineral exploration. based in India. Recently, I got some samples of Ultrabasic rocks analysed and found very high value of Y, Sc, Yb etc. Just wanted to know that does this qualify for rare earth ore and is it worth pursuing the deposit. Its a large deposit.


2 sahharil ahmad December 12, 2012 at 12:55 AM

hi mr Hatch,
Thank you for enroll me as your member. I need advice on clay material that I have contains 0.59% neobium… silica 48%.. Al2O3 35%. Can this clay be market as rare earth.

3 Christer December 12, 2012 at 3:24 AM

Dear Gareth,
Thank you for the very good explanation and clarification on REEs.
Dear Anup,
I’d be interested in knowing where you found the REEs as India in the past have had a great rare earth mining sector (up to the 1950s). You can write to me on christer.hammarlund@ec.europa.eu.

4 Jean-claude Bünzli December 12, 2012 at 4:49 AM

Dear Gareth,
A timely, precise, and welcome article. There is also confusion in several articles between “rare elements” and “rare-earth elements”: in some instance rare-earths elements include indium, lithium, germanium…..
With best personal regards

5 henk mol December 12, 2012 at 7:20 AM

Hi Gareth – can you maybe recommend me scientific articles / books / PhD or MSc dissertations that deal with the substitution of Nd with Dy in magnets based on Nd2Fe14B? It is amongst other this substitution that makes Dy worth the trouble to mine and refine.
Thanks – Henk Mol

6 Gareth Hatch December 12, 2012 at 8:48 AM

@Anup Narayan Singh & @sahharil ahmad: I have sent you messages directly on the queries you made.

@Jean-claude Bünzli: thank you for the feedback. And you’re absolutely right – the whole “rare elements / metals” vs. “rare-earth elements / metals” is definitely still an issue.

@henk mol: that’s a great question. There is a lot on this topic in the scientific literature but it can be difficult to figure out where to start with this. Depends how technical you want to get; if you can get hold of a copy, I have found a book called “Rare-Earth Iron Permanent Magnets”, edited by J.M.D. Coey and published by the Oxford University Press to be a useful reference guide on this topic. It was published in the late 1990s (and includes a few diagrams in the chapter on processing by Rex Harris, that I created :-) ).

There are some earlier papers from the late 1980s that also cover this topic in introductory detail; let me ask some of my magnetic materials research pals for some recommendations and I’ll get back to you.

7 Christoph December 12, 2012 at 8:09 PM

It is not so difficult to separate RARE!! Rare-Earth Iron Permanent Magnets can we get this book or part of it? Gareth if you wrote a section can you post a pdf online for us please….

8 Gareth Hatch December 12, 2012 at 8:24 PM

@Christoph: sorry – I only created some of the figures that appear in the book, and that was over 16-17 years ago! Even if I had a PDF copy of part or all of the book (which I do not), the Oxford University Press wouldn’t be too happy with the violation of their copyright…

Here’s a link to it at Amazon –> http://amzn.to/VXObe4. It’s not cheap – but perhaps it can be obtained through your local library…

9 Christoph December 12, 2012 at 9:01 PM

ahh …….the book seems to be very good ,…does anyone else knows where to get Rare-Earth Iron Permanent Magnets book ??

10 Vladimir Seredin December 16, 2012 at 10:58 AM


You well know, as high I high appreciate your work on the blog. However, I am afraid that this post is not the case.
Every REE expert well knows that any classification of REE must base on a single principle.
Meanwhile, REE may be classified according to:
1. Different physical properties (eg, the structure of the electron shells or magnetic qualities, etc.);
2. Different chemical (eg, the ability to change a valence or according to ion parameters, etc.)
3. Different technological properties (eg, the ability to leaching or separation from other REE under the influence of various reagents)
4. Different geochemical properties (eg, capacity of some rare earth abnormally enrich in REE ores of the various geological types etc.).
5. Different industrial demand in various REE (link).
6. Different prices for individual REE.

Your post gives a typical example of such classification that is base only on some of abovementioned and different REE properties. I am afraid that it may only confuse your readers. IMHO, it is unacceptable with the geological and geochemical point of view in any case.


As I know, any fresh ultrabasic rocks don’t contain very high concentrations of REE. (BTW, what the Y content in these rocks, you are considering as very high?). However, if your rocks are fresh alkaline-mafic rocks or mafic, but hydrothermally altered rocks, the REE contents could be very high. For example, REE may be reach up to some wt% in the argillized mafic dykes of the some HREE+Y occurrences of the Russian Far East.

Vladimir Seredin

11 Gareth Hatch December 16, 2012 at 11:18 AM

@Vladimir Seredin: thank you for your comments. I understand your perspective; I know at least a couple of other geologists and geochemists personally who also also very adamant that there should be only the LREE and HREE groups, based on the electronic structure.

The reality though is that the term MREE *IS* being used, most visibly by the Chinese, but it has always been used by the process guys. Right or wrong, my explanation above reflects that reality. I think it’s better to explain what the Chinese mean, for example, when they refer to middle rare earths in their export quota announcements, rather than to say that they shouldn’t be using the term.

Where I am sure that we will agree is on the fact that Eu and Gd should not be referred to as HREEs, since there is no justification at all for it.

12 Vladimir Seredin December 17, 2012 at 7:20 AM

You misunderstood me. I like the triple REE classification, and also use it in geochemical studies of REE deposits (see the article that I sent you this year: Seredin VV, Dai S. 2012. Coal deposits as potential alternative sources for lanthanides and yttrium / / International Journal of Coal Geology. 94. 67-93). The reason is only using of threefold classification we can adequately describe the distribution of REE in rocks and ores, that actually exist in nature. However, there is the same problem as with dual classification: the vague boundaries between different groups of REE. In our case, it is quite understandable, since sharp boundaries between light, medium and heavy REE are absent in natural processes.
For example, there are three Russian triple REE geochemical classifications.
1. (La-Nd), (Sm-Ho), (Er-Lu) (Mineev, 1974);
2. (La-Nd), (Sm-Dy), (Ho-Lu) (Solodov et al., 1993);
3. (La-Sm), (Eu-Dy + Y), (Ho-Lu) (Seredin VV, Dai S. 2012).

However, I do not quite understand how I can use your classification, and whether to use it at all, if you substantiate the selection of the LREE by some of the REE physical properties, but the selection of the MREE by Chinese export quotas.

13 Gareth Hatch December 17, 2012 at 10:00 AM

@Vladimir Seredin: for a moment, forget about the theoretical, atomic-scale classifications; the classification above is logical from an empirical, processing point of view. When mixed REE precipitates are being separated via solvent extraction (SX), there are different ways of doing so but a common method involves creating intermediate groupings of elements. Because of this, you will frequently see processes (the Lynas Mount Weld processing is a good example) where the result is individually separated La, Ce, Pr and Nd oxides, a mixed Sm-Eu-Gd oxide, and a third mixed oxide containing Tb-Dy-Ho-Er-Tm-Yb-Lu-Y.

These groupings are a “natural” result of the way these elements behave in the SX processes. So one could argue that the above classification is processing-based…

14 Jackie L. Goldsmith January 8, 2013 at 11:41 AM

Love this group discussion…. Since I am an in field application chemist it gives me very good insight to what I am looking for……Have found in a Zinc Depost the following values……. As you can see ZINC is the main collector of all other elements including high end values for Yttrium and Gold…THese values came from GOLDSMITH ZINC PROCESS…..
as you can see these values are high because I used A Excalibur XRF
produced for the Russian Mining kIndustry in Urkanine….. Keiv, Urkanine

Zn…………….K.(Shell)…………3429.(Intensitty)…………..66.03 %(Con)
Fe…………….K…………………… 920……………………….. 17.15%
YTTRIUM…..K…………………… 445……………………….. 5.52%(1609.6oz)
Ni……………..K…………………… 352……………………….. 5.25 %
Cu…………….K……………………. 89……………………….. 1.75 %
GOLD………. L……………………. 42………………………… 1.59 %(463.60z)……..NOTICE HOW HIGHTHE GOLD IS…
Th……………. L……………………… 41………………………… 1.05 %
Zr…………….. K……………………… 91………………………… 0.98 %
V……………… K………………………. 11………………………… 0.69 %
Jackie L. Goldsmith 972 261 8187….emailjackieg1936@yahoo.com and
jackgoldsmith@att.blackberry.net….If you are interested email me back.

15 Enbeng February 11, 2013 at 3:08 AM

Thank you for the detailed explanations. Would you also explain the column headings in the “TMC Advanced Rare-Earth Projects Index”, in particular MR(Mt), wt%, In-Situ TREO, $/t(MR), Basket Price ($/kg). In addition if you could give the words for the acronyms that would be great. TREO wt% sounds like, for every 100 units of stuff that comes out of the ground, the % that is REO so for Araxa for example, the 4.21 units would be REOs, the rest would be other elements. Basket Price sounds like it is obtained by multiplying the % of each REO in 1kg TREO, with the market price of that REO, and adding them up. So for example if 1kg of the TREO from Araxa had 50% La, 30% Ce, 20% Y and prices of each were $10, $5 and $60 respectively, the Basket Price for Araxa would be 10x.40 + 5x.30 + 60x.2 = $17.5. Is that right?

Comments on this entry are closed.

Previous post:

Next post: