113

Received January 29, 2014

Review Accepted for Publication May 20, 2014
©2014 Soc. Mater. Eng. Resour. Japan

Challenges for the Production of Heavy Rare Earth Elements

from Hard Rocks

Yasushi WATANABE＊
＊
Institute for Geo-Resources and Environment, National Institute of Advanced Industrial Science and
Technology, Higashi 1-1-1, Tsukuba, Ibaraki 305-8567, Japan
E-mail : y-watanabe@aist.go.jp
＊
Present address : Faculty of International Resource Sciences, Akita University, Mining Museum of
Akita University, 28-2 Osawa, Tegata 010-8502, Japan
E-mail : y-watanabe@gipc.akita-u.ac.jp

Heavy rare earth elements (HREE) have been produced from ion-adsorption type deposits in southern
China. Despite extremely low REE grades (0.2-0.05wt.%) of the deposits, exploitation of REE from this
deposit type is very competitive because REE are easily extracted from the ores by ammonium sulfate
solution. For the purpose of producing HREE outside China, a variety of sources are examined. They are 1)
ion adsorption deposits outside China, 2) HREE-rich carbonatites, 3) alkaline rock complexes, 4) sediment-
hosted phosphates associated with uranium mineralization, 5) xenotime associated with tin granites, and 6)
iron-oxide apatite deposits. The major challenges for the production of HREE from these hard rock deposits
are mineral beneficiation, REE extraction and treatment of radioactivity.

Key Words : Heavy rare earth elements, alkaline rocks, zirconium silicate, xenotime, apatite

1　INTRODUCTION
　Heavy rare earth elements (Gd to Lu) (HREE) [1], in which often
Y is included, are used to produce permanent magnet, phosphor,
condenser, zirconia stabilizer, etc. Among them, dysprosium (Dy)
is indispensable to improve heat resistance of the NdFeB magnets,
which are applied in modern high-tech materials such as the
driving motors of hybrid vehicles, factory automation machinery,
home electronics and wind turbines [2]. HREE have been mostly
produced from ion-adsorption type deposits in southern China,
except for a very limited amount of production from xenotime in
placer deposits [3].
　Although China has been the main producer of REE in the
world, due to the restriction of the ore production and export of
HREE by the Chinese government, many countries have had
difficulty in the supply of HREE. In particular, Japan, the second

Figure 1　Distribution of major REE exploration and exploitation projects in the world modified from [4].

Int. J. Soc. Mater. Eng. Resour. Vol.20, No.2, (Oct. 2014)

114 Yasushi WATANABE et al.

largest REE magnet producer after China, has been concerned Table 1　Ore minerals and REE grades of representative carbonatite
about the stable supply of HREE to maintain its domestic REE- deposits
applied high-tech industries. However, this Chinese policy has
resulted in the worldwide HREE exploration activities recently
(Figure 1).
　These activities mainly focus on the REE mineralization in
hard rock deposits rather than ion-adsorption type mineralization,
because the formation of ion-adsorption deposits requires several
specific conditions. Ion-adsorption type mineralization is formed in
the following geological and climatic environments ; 1) presence of
HREE enriched host rocks (mostly I-type ilmenite-series granite),
2) tropical to semi tropical climate in stable continents to form a
thick saprolite unit, and 3) presence of soluble HREE minerals Table 2　Representative HREE projects associated with alkaline rock
(carbonates, mostly secondary) in the host rocks. This type of complex
mineralization is found not only in southern China, but also in
neighboring countries such as Vietnam and Thailand. Exploration
is conducted for a similar type of mineralization related to syenite
intrusive complexes in Madagascar and Malawi.
　The prices of rare earth concentrates produced from this deposit
type are low in China (e.g. $25.8/kg for high Y concentrate
in Lognan in June 2013)[5], because of their low production
costs. Thus, the production of HREE from hard rock deposits
is challenging. This paper reviews representative types of
HREE mineral deposits in hard rocks and the challenges for the
development of the deposits.

2　HARD ROCK HREE DEPOSITS

2.1　Carbonatite deposits
　High-grade carbonatite deposits are generally enriched in
LREE because of enrichment of bastnäsite through magmatic
differentiation. However, some carbonatites (e.g. Glenover and
Zandkopsdrift in South Africa) are relatively enriched in HREE
hosted in phosphate minerals, although REE grades are lower
than LREE-enriched ones (Table 1). The Yen Phu deposit in
northern Vietnam and the Lofdal deposit in Namibia are the most
HREE-enriched examples. The Yen Phu deposit consists of two
lensoidal orebodies ; 260 m x 160 m x 60 m (thickness) and 70m x
20-50 m x 13 m (thickness) in size in sericite-quartz schist. The
orebodies occur in the weathered and hydrothermally altered zone
above a dolomitic carbonatite intrusion and include xenotime and
samarskite as ore minerals with abundant iron oxides (Figure 2).
　The Lofdal deposit is a part of a swarm of about 100 carbonatite
Figure 2　Occurrence of xenotime (Xen) in weathered carbonatite at Yen
Phu. Goe : goethite, Qtz : quartz.
and phonolite dikes, mainly 0.5-5 m wide but up to a maximum
of 25 m wide and up to 5 km long in the Huab Basement Complex
rocks. The carbonatite dikes contain xenotime as well as monazite,
synchysite and thorite as REE minerals that are mainly formed
during a hydrothermal stage of mineralization with hematite [6].
Namibia Rare Earths Inc. estimated 0.9Mt of indicated resources
at 0.62% REO and 0.75Mt inferred resources at 0.56% REO (0.3%
cut-off) in this deposit [7] that contain about 1,000t of Dy oxide.
2.2　Deposits associated with alkaline rock complexes
　REE mineralization associated with alkaline rock complexes
has been known from 1980’s, although it has not been regarded
as mineral resources until recently. The only deposit which has
been exploited for REE is Lovozero in Kola Peninsula, where
lopalite-(Ce) [(Na,Ce,Ca,Sr,Th)(Ti,Nb)O 3] in cumulate layers
in the Lovozero peralkaline syenite is mined to produce Nb and
REE [8]. This type of mineralization is explored worldwide for
HREE (Figure 1). Advanced-stage projects are known in Canada
(Nechalacho, Strange Lake, Kipawa), USA (Bokan Mountain),
Greenland (Kvanefjeld), Scandinavia (Norra Kaar), Australia
(Dubbo) and Mongolia (Khaldzan Burgei) (Table 2). These
deposits are associated with peralkaline syenite (Nechalacho,
Kipawa, Kvanefjeld, Norra Kaar, Dubbo) or differentiated alkaline
granites (Strange Lake, Bokan Mountain, Khaldzan Burgei) in
the complexes. The orebodies are variable in form; cumulate
units (Kvanefjeld), dissemination (Kvanefjeld, Kipawa, Dubbo,
Khaldzan Burgei), pegmatite (Strange Lake), and vein (Bokan
Mountain). Some of the deposits (Strange Lake, Kvanefjeld,
Nechalacho) have extremely large reserves, although ore grades

Int. J. Soc. Mater. Eng. Resour. Vol.20, No.2, (Oct. 2014)

 Challenges for the Production of Heavy Rare Earth Elements 115
from Hard Rocks

Table 3　Typical REE composition in the ores at Brown Range (%) [11] Table 4　Recent production of xenotime from Amang in Malaysia [14]

are generally low (2-0.5wt.%). Ore mineralogy is quite variable

and complex in each deposit, including unconventional zirconium
silicates (zircon at Nechalacho and Dubbo, eudialyite at Kipawa,
Kvanefjeld and Norra Kaar, elpidite at Khaldzan Burgei) (Table 2).
　The Kipawa deposit in Canada is hosted by a syenite body that
is part of the Kipawa Alkalic Complex, a concordant folded sheet
of peralkaline syenite and granite. The deposit has been drilled by
Matamec Explorations Inc. over an extent of about 1450 m x 200
m x 100 m depth and contains three mineralized layers (eudialyte,
mosandrite and britholite layers) that, in addition to zirconia,
contain significant REE concentrations. Matamec aims to exploit
the first two layers in which major ore minerals are eudialyite and
mosandrite. At a cut-off grade of 0.3% TREO, mineral resources in
the indicated category show 23.9 Mt@0.41%REO [9].

2.3　Sediment-hosted phosphate deposits associated with

uranium mineralization
　Unconformity-related and/or sediment-hosted uranium Figure 4　Apatite (light color parts)-magnetite (black color parts) ore at
mineralization is occasionally associated with sediment-hosted Chador Malu deposit in Iran.
xenotime mineralization. Representative examples are found in
Canada (Douglass River), Australia (Hastings, Brown Range) and
Kazakhstan [e.g. 10]. Despite relatively small proven reserves of Table 5　REE contents (ppm) in apatite from representative IOA deposits
these deposits so far, ore mineralogy (xenotime) is simple (Table in the world [Watanabe, unpublished data] and apatite from deep-
3, Figure 3) and REE is heavily enriched in HREE. REE is being sea mud around Minami-Torishima Island [16]
recovered from phosphate minerals in the tailing of uranium mine
in Kazakhstan.

2.4　Xenotime mineralization associated with Sn granites

　Xenotime mineralization is often associated with Sn-bearing
granites [12], and xenotime was recovered from heavy placer sands
in Brazil, Australia and Malaysia [13]. Now, it is recovered as a
byproduct of cassiterite in tin mines in Malaysia. Xenotime is also
recovered or planned to be recovered from Sn tailing materials

(Amang) in Malaysia (Table 4), Indonesia and Brazil.

2.5　Iron oxide apatite deposits

Figure 3　Occurrence of hydrothermal xenotime cementing quartz grains
of quartzite with goyazite at Douglass River.
　Iron oxide apatite (IOA) deposits are called Kiruna-type
deposits and consist of ores dominantly formed by iron oxides
and apatite [15] (Figure 4). IOA deposits are mainly distributed
in North America, Scandinavia, western Eurasia, Middle East,
southern Africa and Australia [15]. IOA deposits are a source of
iron ores (magnetite and/or hematite) (Kirunavaara in Sweden,
Pea Ridge in USA) in which apatite is carefully rejected from the
iron concentrates to avoid the contamination of phosphorus. The

Int. J. Soc. Mater. Eng. Resour. Vol.20, No.2, (Oct. 2014)

116 Yasushi WATANABE et al.
apatite associated with IOA deposits commonly contains up to 0.5-3
wt.% of REE (Table 5), although such apatite has not been utilized
except for the production of fertilizer in a few places.
　In some deposits, apatite is partly or completely replaced by
xenotime and monazite. These apatites are especially enriched in
HREE and thus become targets of HREE production.
2.6　Heavy rare earth mineralization in submarine mud
　It is confirmed that submarine mud in a large part of the
Pacific Ocean contains a few thousands ppm of REE with a
high proportion of HREE [17]. Similar mineralization has been
confirmed off Minami Torishima Island recently [18]. In Minami
Torishima area, several thousand ppm of REE are detected in a
couple of meter thick mud about 3m below the sea floor [18].
　The apatite in the deep-sea mud around Minami-Torishima
Island is enriched in REE, especially HREE (Table 5), and has
characteristics similar to those of fish debris [16]. Thus, the origin
of the deep-sea mud apatite is completely different from those in
IOA deposits, and is characterized by relatively low contents of
radioactive elements (Th: 140 ppm, U : 21 ppm) [16].
3　Challenges for HREE production
Although more than five years have passed since many
REE exploration projects started, no project has started HREE
production except the SARECO project in Kazakhstan, which
planned to produce 1500t rare earth oxides in 2013. Many
challenges exist before the exploitation of HREE from hard-rock
deposits to compete with low-cost HREE production at the ion
adsorption deposits in China.
3.1　Accessibility
　Although some deposits such as Strange Lake and Nechalacho
are promising from viewpoints of ore grades, reserves and
mineralogy, these are located in remote areas and new construction
of infrastructure is required. The HREE-bearing deep sea mud is
also promising from these viewpoints, but it is not easy to access to
the seafloor more than 5000 m from the sea level.
3.2　Ore grades and reserves
　The HREE enriched hard rock deposits mentioned above are
relatively low in REE grades (Tables 1, 2, 5) compared with the
grades of LREE deposits (e.g., about 8 wt.% at carbonatite deposits
at Mountain Pass and Mount Weld). Although HREE is more
valuable than LREE, production of REE from such low grade ores
is challenging.
　Some xenotime deposits (Yen Phu, Lofdal, Bokan Mountain,
Brown Range, Douglass River) consist of the ores with the
most preferable mineral; xenotime, for the production of HREE.
However, the estimated reserves of these deposits are still less than
a few hundred thousand tons.
3.3　Mineralogy
　Simple mineralogy is a key for the production of HREE. Major
target minerals for HREE recovery in the hard-rock deposits
are carbonate [synchysite-(Y)], phosphate (xenotime, apatite)
and zirconium-silicate minerals (zircon, eudialyite, mosandrite,
gittinsite, elipdite), in addition to Nb oxide minerals (fergusonite,
samarskite, euxenite, etc.). The ore minerals listed above are new
for REE extraction except carbonates and phosphates, and therefore
it is required to establish economically feasible beneficiation and
Int. J. Soc. Mater. Eng. Resour.
REE extraction methods for these new minerals [19].
　Ore minerals that contain high amounts of HREE (xenotime,
fergusonite) are preferred because high-grade mineral concentrates
can be produced.
　Simple mineralogy is preferred to conduct simple mineral
dressing and processing, although HREE deposits associated with
alkaline rocks are generally complex in mineralogy. For example,
more than 220 different minerals are recognized at Illimaussac
complex, 34 of which were described from this complex for the
first time in the world, and 34 of which are REE minerals [20].
　Weathered carbonatite ores can be enriched in REE by the
dissolution of carbonates, but contain secondary REE minerals,
which are resistive in a weathering condition, small in size and
hence refractory against acids. It is generally difficult to upgrade
REE concentration because of very small size of target minerals
unless the ores are crushed into very fine fraction. These ores
are usually enriched in iron oxides that are difficult to separate
from REE minerals, and also need special treatment for the REE
extraction, e.g. sulfating roasting.
　Some silicate minerals such as allanite and zirconium silicates
contain high concentration of REE. However, these silicates are
generally resistive against acids and REE extraction methods have
not been well established. An exceptional mineral is eudialyte,
which is soluble with a sulfuric acid solution, and REE extraction
flowsheets are accomplished by the eudialyte projects at Kipawa,
Norra Karr, etc. Zircon is generally very resistive against acids, but
REE extraction from poorly crystallized zircon is possible [21].
3.4　Radioactivity
　REE ores may be radioactive because of concentration of thorium
and uranium. If the ores are highly radioactive, treatment of
radioactivity is indispensable. REE ore is considered as radioactive
if it’s radioactivity exceeds 1 Bq/g [22]. This corresponds to
248 ppm Th 232 or 81 ppm U 238 in the ore. A REE developer
needs to consider whether the deposit has relatively high or low
concentration of radioactive elements from the following points of
view ; 1) U, Th and progeny from the decay chain, 2) occupational
health and safety, 3) environmental issue and 4) product quality
[23].
4　Epilogue
After the soaring of REE prices in 2011 (Dy oxide : >2000 UDS/
kg), REE has been supplied from China with relatively low, stable
prices (e.g., Dy oxide～550 USD/kg) until January 2014. This
price decrease has caused difficulty in developing a new REE mine.
Alternatively, several projects are aiming to produce REE as a
byproduct of other elements such as zirconium (Dubbo, Australia),
niobium (Araxa, Brazil), thorium (Orissa, India), or recover REE
from existing mine tailings (uranium mine in Kazakhstan, bauxite
mine in Jamaica). These projects may have an advantage for the
REE production compared to those that aim to produce REE only.
Despite the relatively decreased demand of REE in recent years
(Figure 5), there is a common perspective that the demand of REE,
in particular, HREE, will steadily increase in future. However,
recent efforts to reduce the amounts of REE used in the products or
during polishing, or to replace REE with other materials have been
successful in Japan.
Dysprosium content in NdFeB magnet has decreased to one third
compared to the content five years ago, and now the production of
Dy free magnet is being commercialized. Recycling of REE also
Vol.20, No.2, (Oct. 2014)
 Challenges for the Production of Heavy Rare Earth Elements
from Hard Rocks
Figure 5　Change of Japanese REE demand from 2008 to 2012.
started, and REE-containing scraps generated during the production
of finished goods are completely recycled in Japan. As the result,
the demand of REE in Japan has significantly decreased in the last
five years (Figure 5), and the demand has not been recovered even
in the present low REE-price market situation. This is because
once the end-users replace REE with other materials, it is not easy
to change the production process again. This means that mining
companies have to supply the raw REE materials, responding
timely to the requests by the end-users. In conclusion, the real
competition in the production of HREE is not among the mining
companies or the countries but with the rapid innovation by the
high technology industry.
References
[1] Henderson, P.; “The rare earth elements ; introduction and
review”, In AP Jones et al. eds., Rare earth minerals Chemistry,
origin and ore deposits, The Mineralogical Society Series 7,
1-19 (1996).
[2] Minowa, T.; “Rare earth magnets ; Conservation of energy
and the environment”, Resource Geology, 58, 414-422 (2008).
[3] Academy of Science Malaysia : “Revitalising the Rare Earth
Mineral Programme in Peninsular Malaysia as a Strategic
Industry”, ASM Study Report 1/2013, 107p (2013).
[4] Watanabe, Y.; “Rare Metal Textbook (1) ; Rare Earth”,
Shigen-Chishitsu, 60, 103-122 (2010) (in Japanese).
[5] Metal Page News ; “Ion-adsorption rare earth concentrate
prices for June published by Ganzhou Rare Earths Association”,
13 June 2013 Metal Page News (2013).
[6] Well, F., Niku-Paavola, V. N., Storey, C., Müller, A. and
Jeffries, T.; Xenotime-(Y) from carbonatite dykes at Lofdal,
Namibia ; Unusually low LREE:HREE ratio in carbonatite,
and the first dating of xenotime overgrowths on zircon. The
Canadian Mineralogist, 46, 861-877 (2008).
[7] Namibia Rare Earths Inc.; The Lofdal project. Website ; http://
www.namibiarareearths.com/ (2013).
[8] Kogarko, L. N.; “Chemical evolution and petrogenic
implications of loparite in the layered, agpaitic Lovozero
complex, Kola Peninsula, Russia”, Mineralogy and Petrology,
Int. J. Soc. Mater. Eng. Resour.
117
74, 1-24 (2002).
[9] Matamec Exploration Inc.; NI 43-101 Report - Feasibility
Study. Matamec Exploration Inc., 768p. (2013).
[10] Cook, N. J., Ciobanu, C. L., O’Relly, D., Wilson, R., Das, K.,
Wade, B.; Mineral chemistry of rare earth element (REE)
mineralization, Browns Ranges, Western Australia. Lithos,
172-173, 192-213 (2013).
[11] AMC Consultants Pty Ltd : “Browns range rare earth project
Mineral Resource Statement AMC Project ; 212088”, AMC
Consultants Pty Ltd, 6p (2012).
[12] Bastos Neto, A. C., Pereira, V. P.; “Fluorine-rich xenotime
from the world-class madeira Nb-Ta-Sn deposit associated
with the albite-enriched granite at Pitinga, Amazonia, Brazil”,
The Canadian Mineralogist, 50, 1453-1466 (2012).
[13] Yamamoto, T., Tokuda, M., Narita, Y., Mita, M.; “Distribution
of Y, Yb and Nb in the tailings of the tin ore from Pitinga
deposit, Brazil”, Bulletin of the Res. Inst. Mineral Dressing
and Metallurgy, Tohoku University, 44, 69-76 (1988) (in
Japanese with English abstract).
[14] Minerals and Geoscience Department Malaysia ; “Malaysian
Minerals Year Book 2011”, Ministry of Natural Resources
and Environment Malaysia, 114p (2012).
[15] Hitzman, M. W.; “Iron oxide-Cu-Au deposits : What,
where, when, and why”, In Porter, T.M., ed., Hydrothermal
Iron Oxide Copper-Gold & Related Deposits ; A general
perspective, PCG Publishing, Adelaide, Volume 1, 9-25
(2000).
[16] Kon, Y., Hoshino, M., Sanematsu, K., Morita, S., Tsunematsu, M.,
Okamoto, N., Yano, N., Tanaka, M. and Takagi, T.; “Geochemical
characteristics of apatite in heavy REE-rich deep-sea mud
from Minami-Torishima area, southeastern Japan”, Resource
Geology, 64, 47-57.
[17] Kato, Y., Fujimoto, K., Nakamura, K., Takaya, Y., Kitamura, K.,
Ohta, J., Toda, R., Nakashima, T., Iwamori, H.; “Deep-sea
mud in the Pacific Ocean as a potential resource for rare-earth
elements”, Nature Geoscience, 4, 535-539 (2011).
[18] School of Engineering, The University of Tokyo ; “Discovery
and distribution of mud containing very high concentrations
of rare earth elements and yttrium around Minami-Torishima
Island (Minami-Torishima Survey Cruise)”, Press Release
http://www.jamstec.go.jp/e/about/press_release/20130321/
(2013).
[19] Castor, S. B. and Hedrick, J. B.; “Rare Earth Elements” In
Industrial Minerals & Rocks, 7th Edition, Society for Mining,
Metallurgy and Exploration, Inc., Littleton, 769-792.
[20] Friis, H.; “Mineralogy of the Kvanefjeld multi-element
deposit”, 12th Biennial SGA Meeting Proceedings, Volume 4,
Uppsala, Sweden, 1713-1716.
[21] Hoshino, M., Watanabe, Y., Sanematsu, K., Kon, Y., Shimizu, M.;
“Characteristics of zircon suitable for REE extraction”,
International Journal of Mineral Processing, 102-103, 130-
135 (2012).
[22] International Atomic Energy Agency ; Application of the
concepts of exclusion, exemption and clearance. Safety Guide
RS-G-1.7, 29p (2004).
[23] Soldenhoff, K.; The impact of radioactivity on rare earth
processing. Power point file at 9th International Rare Earth
Conference, Hong Kong, October 2013, 21p.
Vol.20, No.2, (Oct. 2014)