PHD Thesis - Final

INNOVATIVE TECHNOLOGIES FOR
RARE EARTH ELEMENT RECOVERY

FROM BAUXITE RESIDUE
Rodolfo Andres Marin Rivera
Supervisors:
Prof. Tom Van Gerven
Prof. Koen Binnemans
Dr. Ghania Ounoughene
Members of Examination
Committee:
Prof. Bart Blanpain Dissertation presented in partial
Prof. Georgios Stefanidis fulfilment of the requirements for the
Prof. Yiannis Pontikes degree of Engineering Science (PhD):
Prof. Dimitrios Panias Chemical Engineering
January 2019
© 2019 KU Leuven, Science, Engineering & Technology Uitgegeven in eigen
beheer, Rodolfo Marin Rivera, Celestijnenlaan 200F, 3001 Leuven, België.
Alle rechten voorbehouden. Niets uit deze uitgave mag worden vermenigvuldigd
en/of openbaar gemaakt worden door middel van druk, fotokopie, microfilm,
elektronisch of op welke andere wijze ook zonder voorafgaandelijke schriftelijke
toestemming van de uitgever.
All rights reserved. No part of the publication may be reproduced in any form by
print, photoprint, microfilm, electronic or any other means without written
permission from the publisher
II
ACKNOWLEDGEMENTS
Completing this thesis has been a long but a fruitful journey. I feel deeply indebted
to many people who have greatly inspired and supported me during my PhD study
at KU Leuven.
First and foremost, I would like to thank my promoter, Prof. Tom Van Gerven, for
his invaluable guidance, encouragement, academic stimulus and generous help,
but most of all for giving me the opportunity to learn the rigorous scientific
approach and the dedicating spirit for work. Thanks also to my co-promoters, Prof.
Koen Binnemans and Dr. Ghania Ounoughene for their continuous
encouragement, guidance and suggestions.
I would like to thank all the members of the REDMUD project, in particular Dr.
Ken Evans, Katy Tsesmelis, Gyorgy Banvolgy, Dr. Vicky Vassiliaodou, Dr.
Srecko Stopic, Dr. Alan Tkaczyk, Prof. Dimitris Panias, Prof. Bernd Friedrich,
Prof. Yiannis Pontikes, who influenced my scientific work in a different way.
However, a special thanks to Prof. Risto Harjula (RIP) who still continues
inspiring by his example.
I give special thanks to Michèle Vanroelen, Christine Wouters, Hanne Geunes and
Herman Tollet for their technical support. Special thanks also to Alena Vaes,
Marie-Claude Deflem and Beatrice De Geest. I want to thank my two Master’s
students, Brecht Ulenaers and Jorn Verschelde, for their hard work. Many thanks
to my colleagues from ProcESS, past and present, in particular Dr. Chenna Rao
Borra, Steff Van Loy, Thomas Claes, Senne Fransen, for their valuable assistance
and consultations.
Special thanks to Dr. Annelies Malfliet, Pieter L'hoëst, Tom Van der Donck and
Louis Depré, who provided me assistance with analytical facilities at the
Department of Materials Engineering (MTM).
I would like to acknowledge my colleagues (but moreover friends) Fabio, Marco
and Mohammed with whom I did not only share my lunch time, but also good
times outside the lab. I also want to acknowledge my friends Dzenita, Semir,
Federica, Teresa, Annelies, Francesco, Federico, Carlos, Zoe, for all your support
in all these years. I will always remember and appreciate all the moments of joy
we have had. My heartfelt thanks to Carmen for her constant patience, concern,
support, encouragement and kindness… thank you for being there for me always.
Most of all, I want to thank my parents, Margarita and Rodolfo, but also to my
sister Dennise and my brother Rodrigo. I thank you for always trusting me, for
giving me all your love and knowing when to correct me if necessary. It has been
a very long journey with many sacrifices. Still, without the inspiration, drive, and
support that you have given me, I might not be the person I am today. Thanks for
everything!
III
The research leading to these results has received funding from the European
Community’s Horizon 2020 Programme (H2020/2014–2019) under Grant
Agreement No. 636876 (MSCA- ETN REDMUD). This publication reflects only
the author’s view, exempting the Community from any liability. Project website:
ttp://www.etn.redmud.org. The author thanks Aluminium of Greece for providing
the bauxite residue samples.
IV
ABSTRACT
Bauxite residue is a by-product that originates during the Bayer process

of alumina production from bauxite mineral ores. Every year, approximately 2
tonnes of bauxite residue is generated for every tonne of alumina extracted from
the Bayer process. It is estimated that about 160 million tonnes of bauxite residue
are produced every year, while approximately 4 billion tonnes have already been
stockpiled. Disposal and long-term storage of such waste volumes occupy a lot of
land. In turn, this results in major costs and liabilities for alumina producers.
Some treatments already exist to further utilize this bauxite residue, e.g.,
bioremediation and utilisation in building materials. However, bauxite residue also
contains valuable metals such as rare-earth elements (REEs), in minor but non-
negligible concentration. In the recovery of these valuable metals from bauxite
residue, direct acid leaching is a common applied method. Although, these
methods have demonstrated low extraction yields and/or selectivity, which limit
their application at an industrial scale. Thus, the main objective of this PhD Thesis
was to recover REEs from bauxite residue through the development of innovative
hydrometallurgical methods in order to overcome the main disadvantages of
leaching processes reported in the literature, i.e. consumption of large amounts of
acid during neutralisation of the alkaline bauxite residue, the high co-dissolution
of major metals that affects the efficiency of the separation process (e.g., solvent
extraction or ion exchange), and the decomposition of silicate compounds that
leads to the polymerisation of amorphous silica.
In the first stage of this work, the effect that the high alkalinity of bauxite
residue has on the extraction of REEs during direct acid leaching was studied. The
use of CO2 gas as an alternative neutralisation reagent prior to acid leaching was
investigated. It was found that high CO2 partial pressures and temperatures reduces
the alkalinity of the bauxite residue, but it leads to the formation of large
aggregates as consequence of additional CaCO3 formation. It was observed that
after neutralisation, a large amount of REEs remained unreacted after acid
leaching due to an insufficient amount of acid devoted to leaching due to the
chemical transformation of CaCO3 into CaSO4. The recovery of scandium(III) was
low due to its chemical association with major metals, particularly iron(III). It was
found that cerium(IV) co-exist mostly with the hematite mineral phase.
Presumably, lanthanum(III) and neodymium(III) may follow the same trend as
cerium(IV) as they tend to occur in the same mineralogical phase. Yttrium(III)
may remain associated with aluminium/silicon-compounds.
During the recovery of valuable metals by acid leaching of bauxite
residue, silica gel can be formed due to the dissolution of silicate minerals, which
significantly affects the filtration efficiency of the leach liquor. In the second stage,
therefore, the behaviour of silica during acid leaching of raw bauxite residue was
studied. Silica polymerisation was diminished by using the dry digestion
technique, while REEs were recovered by water leaching of the digested sample.
V
About 40 wt.% of scandium was recovered alongside 25 wt.% of iron dissolution,
due to their simultaneous occurrence in the lattice matrix of iron(III) oxide. The
final concentration of REEs in the leachate was significantly increased by
considering the dry digestion technique in combination with multi-stage leaching.
The high amount of iron dissolved during acid leaching poses problems
in the downstream processes because of the similar behaviour of iron and
scandium in solvent extraction and/or ion exchange processes. Therefore, the
removal of iron from bauxite residue before leaching by smelting reduction was
studied. Thus, iron was successfully separated by using different mixtures of coke,
CaO and SiO2 at a temperature of 1500 °C, while REEs were substantially
enriched in the slag. However, due to the high content of silica in the slag, room-
temperature acid leaching could no longer be considered due to polymerization of
silica. Hence, high-pressure acid leaching (or HPAL) was considered instead.
HPAL with HCl allows an extraction above 90 wt.% of scandium, yttrium,
lanthanum and neodymium at temperatures above 100 ºC. The formation of CaSO4
during HPAL with H2SO4, however, allowed high extraction of scandium (95
wt.%), but also a better selectivity over the other REEs.
Preliminary energy and economic analysis show that H 2SO4-based
leaching processes resulted in higher profit margins than the HCl-based ones due
to the low acid consumption. Furthermore, the treatment of bauxite residue by dry
digestion with multi-stage circulation of the leach solution appears to be
economically the most interesting leaching alternative. The process allows to
reduce significantly the volume of effluents due to the low amount of water
required for the process. Gel formation does not occur and titanium can be
recovered simultaneously with the REEs. High-pressure acid leaching of slag from
bauxite residue smelting also avoids the polymerization of silica gel, but titanium
remains in the solid residue after leaching. These processes must be studied further
as part of an integrated extraction-separation process, so that a comprehensive cost
analysis can be obtained to assess the feasibility and viability of the processes for
the recovery of REEs from bauxite residue.
VI
SAMENVATTING
Bauxietresidu’s zijn afkomstig van het Bayerproces voor de productie van alumina
uit bauxietertsen. Elk jaar wordt er ongeveer twee ton bauxietresidu voortgebracht
per ton geëxtraheerde alumina. Naar schatting wordt er ongeveer 160 miljoen ton
bauxietresidu geproduceerd per jaar en ongeveer 4 miljard ton is opgeslagen
wereldwijd. De verwijdering en de lange termijn opslag van dergelijke
hoeveelheden afval vereist een groot landoppervlak. Dit zorgt voor een grote
additionele kost en bijkomende verplichtingen voor de producenten van alumina.
Bovendien kan de ecologische impact van bauxietresidu’s niet worden onderschat.
Er bestaan al enkele behandelingsmethoden om bauxietresidu’s verder te
gebruiken, e.g. bioremediatie en het gebruik in bouwmaterialen. Echter,
bauxietresidu’s bevatten ook waardevolle metalen zoals zeldzame aardmetalen
(Rare Earth Elements, REEs) in kleine, maar niet verwaarloosbare, concentraties.
De herwinning van deze waardevolle metalen uit bauxietresidu’s gebeurt
voornamelijk door het direct uitlogen in sterke zuren. Echter, deze methode heeft
slechts een kleine opbrengst en/of geringe selectiviteit zodat de toepasbaarheid op
een grotere, industriële schaal beperkt blijft. Het doel van deze doctoraatsthesis is
om de herwinning van REEs te verbeteren door innovatieve hydrometallurgische
methodes te ontwikkelen die de beperkingen van de huidige methodes overkomen.
Dat wil zeggen, het beperken van de grote hoeveelheden zuur die nodig zijn om
het alkalische bauxietresidu te neutraliseren, de uitloging van metalen verhinderen
die de efficiëntie van de verdere scheidingsprocessen (zoals solvent extractie en
ionuitwisseling) verminderen, en de ontbinding van silicaatverbindingen die
leiden tot de polymerisatie van amorf silica voorkomen.
In de eerste fase werd de invloed van de hoge alkaliniteit van het bauxietresidu op
de extractie-efficiëntie bij het uitloging met sterke zuren bestudeerd. Het gebruik
van CO2 als een alternatief neutralisatieagens voor de uitloging met zuren werd
onderzocht en er werd geconcludeerd dat een hoge CO2 partieeldruk en
temperatuur de alkaliniteit van het bauxietresidu vermindert. Echter, dit leidde tot
grotere aggregaten in vergelijking met onbehandelde bauxietresidu’s omwille van
CaCO3-vorming. Ook werd de observatie gemaakt dat silicaatverbindingen
stabiliseerden na neutralisatie met CO2. Bovendien reageerde een groot deel van
de REEs niet na het uitlogen doordat te weinig zuur beschikbaar was door de
transformatie van CaCO3 in CaSO4. De herwinning van scandium(III) uit sterk
geneutraliseerd bauxiet residu was laag ten gevolge van de chemische associatie
met de ijzer(III). Cerium(IV) was vooral terug te vinden in de hematiet
mineraalfase. Vermoedelijk gedragen lanthaan(III) en neodymium(III) zich
hetzelfde als cerium(IV) omdat beiden telkens tezamen voorkomen in dezelfde
mineraalfase. Yttrium(III) blijft geassocieerd met Al/Si-verbindingen.
Tijdens de herwinning van waardevolle metalen door uitloging met sterke zuren
kan silica gel gevormd worden door de oplossing van silicaatmineralen. Dit
beïnvloedt sterk de filtratie-efficiëntie van het extract na het uitlogingsproces.
VII
Daarom werd in de tweede fase het gedrag van silica tijdens de uitloging van het
ruwe bauxietresidu onderzocht. De polymerisatie van silica verminderde door het
gebruik van een “droge digestietechniek” (dry digestion technique), vervolgens
konden de REEs herwonnen worden door uitloging met water. Ongeveer 40 wt.%
van het scandium werd herwonnen tezamen met 25 wt.% ijzer, omdat beiden
voorkomen in het kristalrooster van ijzer(III) oxide. De finale concentratie REEs
werd aanzienlijk verhoogd door gebruik te maken van de droge digestietechniek
in combinatie met meertraps uitloging.
De grote hoeveelheden ijzer die tijdens het uitlogingsproces met sterke zuren
vrijkomen, zijn problematisch voor de volgende scheidingsprocessen. Dit is vooral
omdat ijzer en scandium hetzelfde gedrag vertonen bij solvent-extractie en/of
ionuitwisseling processen. Daarom werd er onderzocht of dat de verwijdering van
ijzer vóór de uitloging kon worden gerealiseerd door het reductieve smelten van
het bauxietresidu. Het werd aangetoond dat ijzer succesvol verwijderd werd door
verschillende mengelingen cokes, CaO en SiO2 bij een temperatuur van 1500 °C.
Tegelijkertijd werd de gevormde slak aanzienlijk verrijkt met REEs. Echter, door
de hoge concentratie van silica in de slak, zal de uitloging met zuren op
kamertemperatuur leiden tot het polymeriseren van silica. Bijgevolg werd hoge-
druk uitloging met zuren (High-Pressure Acid Leaching, HPAL) overwogen als
alternatief. HPAL met HCl zorgt voor een extractie tot 90 wt.% scandium, yttrium,
lanthaan, en neodymium bij temperaturen boven 100 °C. Echter, de vorming van
CaSO4 tijdens HPAS met H2SO4 liet een hogere extractie van scandium (95 wt.%)
toe met een betere selectiviteit tegenover de andere REEs.
Een preliminaire energetische en economische analyse toont aan dat H 2SO4-
gebaseerde uitloging hogere marges biedt omdat het zuurverbruik veel lager ligt
in vergelijking met het HCl-gebaseerde proces. Bovendien is de behandeling van
bauxietresidu’s door droge digestie met een meertraps uitlogingsproces het meest
veelbelovende alternatief in vergelijking met de bestaande processen. Dit proces
laat toe om het debiet van de uitgaande stromen aanzienlijk te verminderen door
de kleine hoeveelheden water die nodig zijn voor het proces. Gel formatie komt
niet voor en titanium kan herwonnen worden in combinatie met de REEs. HPAL
van het gevormde slak na het smelten van het bauxietresidu vermijd ook de
polymerisatie van silicaatverbindingen tot silica gel, maar titanium blijft in het
residu en kan dus niet herwonnen worden. De bovenstaande processen moeten
verder bestudeerd worden als onderdeel van een geïntegreerd extractie-
scheidingsproces zodat een alomvattende kostanalyse kan worden gemaakt. Op
deze manier kan de uitvoerbaarheid en rendabiliteit voor de herwinning van REEs
uit bauxietresidu’s worden geëvalueerd.
VIII
LIST OF ABBREVIATIONS
BR Bauxite residue
BRS Bauxite residue slag
BSE Back-scattered electron
DD 1 Lx Dry digestion with single-stage leaching
DD M Lx Dry digestion with multiple-stage leaching
DLx Direct leaching
DTG Differential thermogravimetry
EPMA Electron microprobe for microanalysis
Eq. Equation
FC Fast cooling
HNBR Highly-neutralised bauxiter residue
HPAL High-pressure acid leaching
HPNBR High-pressure neutralised bauxite residue
HPTNBR High-pressure and high temperature neutralised
bauxite residue
HREEs Heavy rare-earth elements (Gd to Lu and Y)
ICP-MS Inductively Coupled Plasma Mass Spectrometry
ICP-OES Inductively Coupled Plasma Optical Emission
Spectroscopy
L/S Liquid-to-solid
LOI Loss on ignition
LREEs Light rare-earth elements (La to Eu)
Lx HNBR Neutralised-leached bauxite residue
N-Lx Neutralised-leached
NBR Neutralised bauxite residue
NORM Naturally occurring radioactive material
PLS Pregnant leach solution
PTFE Polytetrafluoroethylene
RE Rare-earth
REEs Rare-earth elements
rpm Revolutions per minute
IX
SC Slow cooling
SEM Scanning electron microscopy
TENORM Technologically enhanced naturally occurring
radioactive material
TGA Thermo-gravimetric analysis
USGS United States geological survey
WDS Wavelength dispersive spectroscopy
WDXRF Wavelength dispersive x-ray fluorescence
spectroscopy
XRD X-ray powder diffraction
X
LIST OF PUBLICATIONS
(Doctoral Thesis Context)
Articles in internationally reviewed academic journals
R. M. Rivera, B. Xakalashe, G. Ounoughene, K. Binnemans, B. Friedrich, T. Van
Gerven. High-pressure acid leaching of slag from bauxite residue smelting in view
of selective rare-earth elements recovery (submitted).
R.M. Rivera, G. Ounoughene, A. Malfliet, J. Vind, D. Panias, V. Vassiliadou, K.
Binnemans, T. Van Gerven. A study of the occurrence of selected rare-earth
elements in neutralised-leached bauxite residue and comparison with untreated
bauxite residue (accepted).
R.M. Rivera, B. Ulenaers, G. Ounoughene, K. Binnemans, T. Van Gerven (2018).
Extraction of rare earths from bauxite residue (red mud) by dry digestion. Minerals
Engineering, 119, 82-92.
R.M. Rivera, G. Ounoughene, C.R. Borra, K. Binnemans, T. Van Gerven (2017).
Neutralisation of bauxite residue by carbon dioxide prior to acidic leaching.
Minerals Engineering, 112, 92-102.
Articles in international scientific conferences

R.M. Rivera, B. Xakalashe, G. Ounoughene, K. Binnemans, B. Friedrich, T. Van
Gerven. Recovery of rare-earths from bauxite residue slag by high-pressure acid
leaching. 2nd International Bauxite Residue Valorisation and Best Practices
Conference. Athens (Greece), 7-10 May 2018.
R.M. Rivera, G. Ounoughene, K. Binnemans, T. Van Gerven. Recovery of
selected rare earths from bauxite residue by dry digestion with multi-stage
circulation of leach liquor. 2nd International Bauxite Residue Valorisation and
Best Practices Conference. Athens (Greece), 7-10 May 2018.
R.M. Rivera, B. Ulenaers, G. Ounoughene, K. Binnemans, T. Van Gerven.
Behaviour of Silica during Metal Recovery from Bauxite Residue by Acidic
Leaching. Travaux 46, Proceedings of 35th International ICSOBA Conference,
Hamburg, Germany, 2 – 5 October, 2017. Conference of the International
Committee for Study of Bauxite, Alumina & Aluminium (pp. 547-556).
R.M. Rivera, G. Ounoughene, K. Binnemans, T. Van Gerven. Use of mineral
acids for neutralisation of bauxite residue prior to metal recovery by acidic
leaching. 2nd International Conference on Applied Mineralogy & Advanced
Materials - 13th International Conference on Applied Mineralogy (AMAM-
ICAM2017). Castellaneta Marina - Taranto (Italy), 5-9 June 2017.
XI
Table of contents
1. INTRODUCTION ........................................................................... - 1 -
1.1 TURNING THE BAUXITE RESIDUE INTO AN OPPORTUNITY ............................. - 1 -
1.2 LIMITATIONS FOR RARE-EARTH ELEMENTS RECOVERY FROM BAUXITE RESIDUE - 2 -
1.3 SCOPE OF THE PHD THESIS .................................................................. - 3 -
1.4 OUTLINE ..................................................................................... - 4 -
1.5 REFERENCES ..................................................................................... - 6 -
2. LITERATURE REVIEW ................................................................... - 9 -
2.1 FROM BAUXITE TO BAUXITE RESIDUE ...................................................... - 9 -
2.2 OCCURRENCE OF RARE-EARTH ELEMENTS IN BAUXITE RESIDUE................... - 11 -
2.3 RARE-EARTH ELEMENTS RECOVERY FROM BAUXITE RESIDUE ...................... - 12 -
2.4 TOWARDS AN EFFICIENT RECOVERY OF REES FROM BAUXITE RESIDUE ......... - 15 -
Neutralisation of bauxite residue before acid leaching ... - 15 -
Separation of major metals from bauxite residue ........... - 17 -
Silicon dissolution during acid leaching ........................... - 20 -
2.5 EXISTENCE OF NATURALLY OCCURRING RADIONUCLIDE ACTIVITY IN BAUXITE RESIDUE
................................................................................... - 26 -
2.6 CONCLUSIONS ................................................................................. - 27 -
2.7 REFERENCES ................................................................................... - 28 -
3. NEUTRALISATION OF BAUXITE RESIDUE BY CARBON DIOXIDE PRIOR TO
ACIDIC LEACHING FOR METAL RECOVERY .................................. - 41 -
3.1 INTRODUCTION ................................................................................ - 42 -
3.2 MATERIAL AND METHODS .................................................................. - 43 -
3.3 RESULTS AND DISCUSSION .................................................................. - 44 -
Characterisation of the bauxite residue ........................... - 44 -
Neutralisation - leaching.................................................. - 46 -
Neutralisation of bauxite residue with water .................. - 47 -
Neutralisation of bauxite residue with different CO 2 gas flow
rates ......................................................................................... - 48 -
High-pressure neutralisation ........................................... - 51 -
Leaching of neutralised bauxite residue .......................... - 55 -
3.4 CONCLUSIONS ................................................................................. - 65 -
3.5 REFERENCES ................................................................................... - 66 -
4. OCCURRENCE OF SELECTED RARE-EARTH ELEMENTS IN NEUTRALISED-
LEACHED BAUXITE RESIDUE ....................................................... - 71 -
4.1 INTRODUCTION ................................................................................ - 72 -
4.2 OCCURRENCE OF REES IN BAUXITE AND BAUXITE RESIDUE – A BRIEF REVIEW - 73 -
XIII
4.3 EFFECT OF BAUXITE RESIDUE’S ALKALINITY ON REES RECOVERY BY ACID LEACHING
................................................................................... - 74 -
4.4 MATERIALS AND METHODS ................................................................ - 74 -
4.5 RESULTS AND DISCUSSION.................................................................. - 75 -
Concentrations of major and rare-earth elements in post-
processed bauxite residue ........................................................... - 77 -
Electron probe microanalysis of post-processed bauxite residue
......................................................................................... - 81 -
Reaction mechanism for neutralised-leached bauxite residue
......................................................................................... - 85 -
Comparative analysis for rare earth extraction from bauxite
residue......................................................................................... - 87 -
4.6 CONCLUSIONS ................................................................................. - 89 -
4.7 REFERENCES ................................................................................... - 90 -
5. EXTRACTION OF RARE EARTHS FROM BAUXITE RESIDUE BY DRY
DIGESTION FOLLOWED BY WATER LEACHING ............................ - 95 -
5.1 INTRODUCTION................................................................................ - 96 -
5.2 MATERIAL AND METHODS.................................................................. - 98 -
5.3 RESULTS AND DISCUSSION................................................................ - 100 -
Characterization of the bauxite residue ......................... - 100 -
Silicon dissolution behaviour ......................................... - 101 -
Dry digestion followed by water leaching ..................... - 103 -
Silica dissolution after dry digestion of bauxite residue - 104 -
Metal extraction with dry digestion method ................. - 106 -
Dry digestion method with multi-stage circulation of acid
leaching solution ....................................................................... - 112 -
5.4 CONCLUSIONS ............................................................................... - 119 -
5.5 REFERENCES ................................................................................. - 120 -
6. SELECTIVE RARE EARTH ELEMENT EXTRACTION USING HIGH-PRESSURE
ACID LEACHING OF SLAGS ARISING FROM THE SMELTING OF BAUXITE
RESIDUE ................................................................................... - 125 -
6.1 INTRODUCTION.............................................................................. - 126 -
6.2 MATERIAL AND METHODS................................................................ - 128 -
6.3 RESULTS AND DISCUSSION................................................................ - 132 -
Characterisation of the bauxite residue and slags ......... - 132 -
HPAL of a slag rich in silicon, aluminium and iron ......... - 137 -
REEs recovery by HPAL from other slags of bauxite residue
smelting .................................................................................... - 147 -
Direct treatment of bauxite residue by HPAL versus HPAL of
bauxite residue slags and other acid leaching processes .......... - 151 -
6.4 CONCLUSIONS ............................................................................... - 154 -
6.5 REFERENCES ................................................................................. - 155 -
XIV
7. COMPARATIVE ANALYSIS FOR RARE EARTH EXTRACTION FROM
BAUXITE RESIDUE .....................................................................- 161 -
7.1 METAL EXTRACTION AND SELECTIVITY................................................. - 161 -
7.2 COMPARATIVE PRELIMINARY ECONOMIC ANALYSIS ................................ - 169 -
7.3 SUMMARY ................................................................................. - 176 -
7.4 REFERENCES ................................................................................. - 176 -
8. CONCLUSIONS AND FUTURE PERSPECTIVES .............................- 179 -
8.1 GENERAL CONCLUSIONS AND FINDINGS .............................................. - 179 -
8.2 OUTLOOK ................................................................................. - 182 -
8.3 REFERENCES ................................................................................. - 183 -
APPENDIX .....................................................................................- 185 -
XV
Chapter 1 Introduction
1. INTRODUCTION
Bauxite residue (also called red mud) is a by-product produced in the

alumina industry. More than 95% of the worldwide produced alumina is processed
by the Bayer process [1]. The process is based on contacting aluminium-bearing
ores with a concentrated NaOH solution at temperatures between 150 – 250 ºC in
an autoclave [2]. The process is carried out at high hydroxide concentrations, so
that aluminium hydroxides are selectively dissolved from bauxite minerals, while
most of the other compounds remain insoluble in the bauxite residue. Thus,
aluminium is separated from its accompanying elements. After digestion, the
supersaturated liquor is cooled down, and Al(OH) 3 can be recovered after
crystallization. The bauxite residue is accumulated as a solid waste residue in very
large deposits, which represent major costs and liabilities for alumina producers.
The high alkalinity of bauxite residue (i.e. pH 10 – 13) is a main environmental
concern. Moreover, spills have led to major environmental incidents, as it was the
case of the Ajka disaster in Hungary in 2010 [3]. In 2015, the collapse of a tailing
dam in the state of Minas Gerais in Brazil evidenced as well the serious ecological
and socio-economic impact of storing solid waste residue at large scale [4,5].
Throughout the entire history of alumina production there has been a
desire to utilise the bauxite residue created in the Bayer process, either by
recovering valuable metals from it or by using it completely in bulk applications.
The use of bauxite residue as feedstock in the construction, building or agricultural
industries [1], but also in the area of environment protection by neutralising and
adsorbing toxic components from polluted gas, water and land [6], are the most
representative areas for application. Despite the proposed application areas, there
are incentives for better treatment procedures; such as the processing of bauxite
residue through recovery of major and minor metals to maximize the created value
and to minimise the residual solids produced every year [7].
1.1 Turning the bauxite residue into an opportunity

Bauxite residue is a polymetallic material composed essentially of
compounds that are insoluble in concentrated NaOH solutions such as iron and
titanium minerals, undigested aluminium minerals, sodium aluminium
hydrosilicates, calcium compounds, and significant concentrations of rare-earth
elements (REEs). It has been estimated that the annual global production of
bauxite residue exceeds 150 million tonnes [8,9], and above 4 billion tonnes have
been already accumulated in tailings ponds, lakes or landfills. The material is
composed of particles with sizes in the range of 5 – 10 µm, although particles
larger than 20 µm can also be present. The chemical composition of bauxite
residue depends on the origin of the bauxite ore and the operational conditions
during the Bayer process. However, during the processing of bauxite by the Bayer
process, all the REEs end up in the bauxite residue [10]. Therefore, bauxite residue
-1-
represents an interesting source not only for major elements such as aluminium,
iron and titanium, but also for REEs [11].
Rare earths are the fifteen metallic elements of the lanthanides series,
together with yttrium and scandium. The enrichment factor of the rare earths in
bauxite residue compared to bauxite is approximately a factor of two [12], with
scandium being the most interesting among the REEs because it represents 95%
of the economic value of the REEs present in this solid residue [13]. The REEs are
not rare because their amounts are scarce on the Earth’s crust, but because it is
difficult to find their source minerals in relatively high economic concentrations.
There are approximately 120 million tonnes of proven rare-earth oxide reserves in
the world, spread over countries such as China, Russia, Brazil, Vietnam, India and
the United States [14]. Nowadays, China is the main supplier of REEs with ca.
80% of the world’s production. The REE have a wide range of applications and
they are used in many new electronic devices such as mobile phones, screens,
high-capacity batteries, permanent magnets for wind power stations, ceramics, etc.
[15]. As the world moves towards a cleaner, greener future, demand for REEs
based materials will continue to increase as rare earths remain the material of
choice for many key technologies. However, with China dominating the market of
REEs, prices may not be stable and the access to REEs supplies can be limited. In
fact, REEs have been considered as Critical Raw Materials for the EU economy
[16]. Thus, the control of China on REE market may undermine European
innovation and competitiveness, and slow down the development of green
technologies, such as electric vehicles and offshore wind. The development of new
sources of REEs outside of China and/or recycling from priority waste streams
must therefore remain an urgent priority for Europe. Hence, Europe should reduce
its dependence on China’s REEs, and attempt to re-process waste materials with a
relatively high concentration of REEs.
Nowadays, management of bauxite residue represents a major issue for
the aluminium industry because of its high alkalinity and the large quantities
produced annually (about 150 million tonnes) [9,17]. Despite the potential for
recycling, metal extraction at industrial scale has not been practiced so far, due to
the low concentration of some elements, which makes the recovery of these metals
not economically feasible. Recovery of these and perhaps additional metals
combined with the utilization of the leftover residue could partly solve both the
supply problem of REEs and the storage problem of the bauxite residue.
1.2 Limitations for rare-earth elements recovery from bauxite residue

The combination of pyro- and hydrometallurgical methods allows the
recovery of not only REEs and titanium, but also the recovery of alumina and iron.
However, pyrometallurgical processes are carried out at high temperatures and/or
with the addition of flux, which may lead to a substantial energy consumption.
Hydrometallurgical processes, on the other hand, are in most of the cases
characterised by their high acid consumption due to the presence of multiple
-2-
alkaline solids, remaining after the Bayer process, which render the pH of the
residue very high (pH 10 – 13). Consequently, part of the acid must be used for
the neutralisation of the alkaline products left behind after the alumina production
(i.e. after digestion in the Bayer process). Although high acid concentration can
lead to high recovery of REEs, this can make the process less efficient as other
elements can be dissolved as well. The presence of major metals in the leachate,
in particular iron and/or aluminium, represents a serious drawback during the
further separation process, namely solvent extraction or ion exchange. These
elements are difficult to separate from the REEs because they share the same
trivalent oxidation state, and can be adsorbed simultaneously with the REEs in the
ion-exchange resin of the separation process. In addition, the presence of silicate
compounds (e.g., cancrinite, sodalite) in bauxite residue, and their potential
decomposition during acid leaching, is a major problem in bauxite residue
leaching because silica tends to polymerize once it is in solution. The formation
of this polymer (or silica gel) can significantly reduce the filtration efficiency of
the leach liquor before the separation process.
Bauxite residue can be treated directly by using carbonic acid (H2CO3,
formed during the dissolution of CO2 in water) as leaching reagent [18]. However,
this method allows the recovery of only scandium and not the other REEs due to
its particular amphoteric behaviour [19]. During conventional acid leaching, on
the contrary, most of the REEs can be extracted from bauxite residue, but their
concentrations in the leachate are low, i.e. about one hundred times lower than the
concentration of major elements (e.g., aluminium, iron, titanium, silicon, sodium,
calcium). Various attempts have been reported in the literature to contribute to the
recovery of scandium and other REEs recovery from bauxite residue by
conventional acid leaching. Most of these technologies have demonstrated low
leaching efficiencies (i.e. extraction yields) and/or selectivity, which limit their
application at an industrial scale. Therefore, in order to ensure the cost
effectiveness of the process, it is necessary to reduce the amount of acid needed
during acid leaching and to enhance the concentration of REEs in the leach liquor.
1.3 Scope of the PhD Thesis

This PhD Thesis addresses the development of innovative methods to
overcome occurring drawbacks in the potential industrial extraction process of
REEs from bauxite residue, namely: 1) the cost associated with neutralisation of
bauxite residue by expensive mineral acids (chapter 3), 2) the silica gel formation
(chapter 5) due to excessive acidity leading to difficulties in the subsequent
filtering of the leaching liquor, and 3) the dissolution of major metals and the
enhancement of the leaching efficiency to selectively recover REEs (chapter 6).
The extraction of REEs was studied by considering different options, as
it is depicted in Figure 1.1. In the first stage, direct acid leaching of bauxite residue
was studied to evaluate the yields and selectivity for REEs extraction. The
extraction yield of REEs was low because part of the acid is used for the
-3-
neutralisation of alkaline products remaining after the Bayer process. Although the
extraction yield of REEs can be improved by increasing the acid concentration,
this also leads to a significant increase of the acid consumption. Therefore, in this
first stage, CO2 dissolved in water was used for neutralisation of the bauxite
residue as a pre-treatment before acid leaching. In the second stage, the behaviour
of silica during acid leaching of raw bauxite residue was studied. Silica
polymerisation was diminished by using the dry digestion technique, while REEs
were recovered by water leaching of the digested sample. In the third stage, iron
was removed from bauxite residue through reductive smelting, while the slag rich
in REEs was leached with acid at high temperature in an autoclave.
Solid
charact.
Direct acid
PLS I.1
leaching
I
BAUXITE RESIDUE
Neutralisation Acid leaching PLS I.2
(Multi-stage)
II Dry digestion PLS 2
water leaching
Reductive
III HPAL PLS 3
smelting
PLS: Pregnant leach solution; HPAL: High pressure acid leaching
Figure 1.1: Conceptual flow sheet for bauxite residue processing, with indication of the
three alternative processes that have been studied within the framework of this PhD Thesis
1.4 Outline
This PhD Thesis contains 7 chapters which present the main findings of
this doctoral project followed by one chapter that summarises the major
conclusions and future perspectives. At the end also, an appendix is provided
which includes supplementary material with additional results to support the
information presented in chapters 3 – 6.
Chapter 1 describes the thesis background, scope of this work and the
outline of this PhD Thesis. Chapter 2 gives an overview of the literature on the
recovery of REEs from bauxite residue by conventional acid leaching method, iron
recovery in view of REEs recovery and alumina removal followed by recovery of
iron and REEs. It also provides a detailed introduction of the main drawbacks of
conventional acid leaching methods, i.e. bauxite residue’s alkalinity, the co-
dissolution of major metals and the decomposition of silicate compounds. The
-4-
occurrence of naturally occurring radioactive materials (NORM) in bauxite

residue, and their possible enrichment after processing, is introduced.
Chapter 3 highlights the effect that the high alkalinity of bauxite residue
has on the extraction of REEs during direct acid leaching. It describes the use of
CO2 gas as an alternative neutralisation reagent prior to acid leaching. It also
includes a comparative analysis between CO2 gas and a strong mineral acid (HCl)
as neutralisation reagents. As the chemical and mineralogical composition of the
initial bauxite residue changes after neutralisation with CO2, lower recoveries of
rare-earth elements were obtained during acid leaching of the neutralised sample
compared to that of untreated bauxite residue. Therefore, in Chapter 4, the
distribution of selected REEs in the raw, neutralised and neutralised-leached
bauxite residue is discussed. The occurrence mode was investigated by using an
electron probe micro-analyser (EPMA). The REE-carrier minerals within the
bauxite residue are also discussed. A theoretical base for further processing of
bauxite residue in view of enhanced REEs recovery is proposed.
In Chapter 5, the extraction of REEs from bauxite residue by using the
dry digestion technique with concentrated mineral acids, followed by water
leaching is discussed. Valuable metals can be effectively washed out by using this
method since silica polymerisation does not occur, which significantly can
improve the filterability of the leach liquor. The behaviour of silica was studied
during acid leaching with concentrated mineral acids at different concentrations.
A multi-stage leaching procedure was applied after dry digestion to increase the
REEs concentration in the leachate. The process was compared with the
conventional direct acid leaching method in terms of selected REEs, iron
concentration, and acid consumption.
Chapter 6 presents the leaching behaviour of selected REEs by high-
pressure acid leaching of slags generated during reductive smelting of bauxite
residue. The effects of the different slag’s morphologies on the extraction yield of
REEs is discussed by considering two different mineral acids, at different
concentrations, and different leaching temperatures. High-pressure acid leaching
of bauxite residue slags allows higher recoveries of REEs without silica gel
formation compared to other reported leaching technologies, but with a substantial
co-dissolution of aluminium. The extraction yields were compared with the direct
treatment of bauxite residue by high-pressure acid leaching. A mechanism for
high-pressure acid leaching of slag from bauxite residue smelting is also proposed.
Chapter 7 gives a preliminary comparative economic analysis. It also
provides some information on the best mineral acid that could be used and the
process(es) that could be further developed in pilot scale.
In Chapter 8 the overall conclusions of the thesis are presented. It also
gives an outlook to further studies that can be carried out to improve the process
and take the process to the next level.
-5-
1.5 References
[1] World Aluminium and the European Aluminium Association, Bauxite
Residue Management : Best Practice, London, 2015. http://www.world-
aluminium.org.
[2] G. Power, M. Gräfe, C. Klauber, Bauxite residue issues: I. Current
management, disposal and storage practices, Hydrometallurgy. 108
(2011) 33–45. doi:10.1016/j.hydromet.2011.02.006.
[3] Á.D. Anton, O. Klebercz, Á. Magyar, I.T. Burke, A.P. Jarvis, K. Gruiz,
W.M. Mayes, Geochemical recovery of the Torna-Marcal river system
after the Ajka red mud spill, Hungary, Environ. Sci. Process. Impacts. 16
(2014) 2677–85. doi:10.1039/C4EM00452C.
[4] L.C. Garcia, D.B. Ribeiro, F. De Oliveira Roque, J.M. Ochoa-Quintero,
W.F. Laurance, Brazil’s worst mining disaster: Corporations must be
compelled to pay the actual environmental costs: Corporations, Ecol.
Appl. 27 (2017) 5–9. doi:10.1002/eap.1461.
[5] G.W. Fernandes, F.F. Goulart, B.D. Ranieri, M.S. Coelho, K. Dales, N.
Boesche, M. Bustamante, F.A. Carvalho, D.C. Carvalho, R. Dirzo, S.
Fernandes, P.M. Galetti, V.E.G. Millan, C. Mielke, J.L. Ramirez, A.
Neves, C. Rogass, S.P. Ribeiro, A. Scariot, B. Soares-Filho, Deep into the
mud: ecological and socio-economic impacts of the dam breach in
Mariana, Brazil, Nat. e Conserv. 14 (2016) 35–45.
doi:10.1016/j.ncon.2016.10.003.
[6] M. Gräfe, G. Power, C. Klauber, Bauxite residue issues: III. Alkalinity
and associated chemistry, Hydrometallurgy. 108 (2011) 60–79.
doi:10.1016/j.hydromet.2011.02.004.
[7] M. Schwarz, V. Lalík, Possibilities of Exploitation of Bauxite Residue
from Alumina Production, (2007) 3–23. www.intechopen.com.
[8] K. Evans, The History, Challenges, and New Developments in the
Management and Use of Bauxite Residue, J. Sustain. Met. 2 (2016) 316–
331. doi:10.1007/s40831-016-0060-x.
[9] É. Deady, E. Mouchos, K. Goodenough, B. Williamson, F. Wall, A
review of the potential for rare-earth element resources from European
red muds: examples from Seydişehir, Turkey and Parnassus-Giona,
Greece, Mineral. Mag. 80 (2016) 43–61.
doi:10.1180/minmag.2016.080.052.
[10] J. Vind, V. Vassiliadou, D. Panias, Distribution of Trace Elements
Through the Bayer Process and its By-Products, in: 35th Int. ICSOBA
Conf. Hamburg, Ger. 2 – 5 October, 2017, 2017: pp. 255–267.
[11] K. Binnemans, P.T. Jones, B. Blanpain, T. Van Gerven, Y. Pontikes,
Towards zero-waste valorisation of rare-earth-containing industrial
-6-
process residues: a critical review, J. Clean. Prod. 99 (2015) 17–38.

doi:10.1016/j.jclepro.2015.02.089.
[12] M. Ochsenkuhn-Petropulu, T. Lyberopulu, G. Parissakis, Direct
determination of lanthanides, yttrium and scandium in bauxites and red
mud from alumina production, Anal. Chim. Acta. 296 (1994) 305–313.
doi:10.1016/0003-2670(94)80250-5.
[13] C.R. Borra, Y. Pontikes, K. Binnemans, T. Van Gerven, Leaching of rare
earths from bauxite residue (red mud), Miner. Eng. 76 (2015) 20–27.
doi:10.1016/j.mineng.2015.01.005.
[14] U.S. Geological Survey, Rare earths, in: Miner. Commod. Summ. 2018,
2018: pp. 132–133. doi:https://doi.org/10.3133/70194932.
[15] K. Binnemans, P.T. Jones, B. Blanpain, T. Van Gerven, Y. Yang, A.
Walton, M. Buchert, Recycling of rare earths: A critical review, J. Clean.
Prod. 51 (2013) 1–22. doi:10.1016/j.jclepro.2012.12.037.
[16] European Commission, Study on the review of the list of critical raw
materials, 2017.
[17] É. Ujaczki, V. Feigl, M. Molnár, P. Cusack, T. Curtin, R. Courtney, L.
O’Donoghue, P. Davris, C. Hugi, M.W.H. Evangelou, E. Balomenos, M.
Lenz, Re-using bauxite residues: benefits beyond (critical raw) material
recovery, J. Chem. Technol. Biotechnol. 93 (2018) 2498–2510.
doi:10.1002/jctb.5687.
[18] O. Petrakova, A. Panov, S. Gorbachev, G. Klimentenok, Improved
Efficiency of Red Mud Processing through Scandium Oxide Recovery,
in: Light Met., 2015: pp. 93–96.
[19] E. V. Shkol’nikov, Thermodynamic characterization of the amphoterism
of hydroxides and oxides of scandium subgroup elements in aqueous
media, Russ. J. Appl. Chem. 82 (2009) 2098–2104.
doi:10.1134/S1070427209120040.
-7-
Chapter 2 Literature review
2. LITERATURE REVIEW
2.1 From bauxite to bauxite residue

More than 95% of the worldwide produced alumina is processed by the
so-called Bayer process [1]. The process consists in contacting bauxite mineral
ores with a high concentrated solution of NaOH (100 – 260 g L-1) at temperatures
between 150 – 250 ºC (about 40 atm). Bauxite consists of a mixture of the
aluminium minerals gibbsite (Al(OH)3, boehmite (-AlOOH) and diaspore (-
AlOOH), but minerals such as hematite (Fe2O3), goethite (FeOOH), quartz (SiO2),
rutile/anatase (TiO2) and kaolinite (Al2Si2O5(OH)4) are also part of its
mineralogical matrix. Traces of natural occurring radionuclides are also present in
bauxite residue [2].
The Bayer process is a cyclic process composed of four different steps:
(1) digestion, (2) bauxite residue (red mud) separation, (3) precipitation, and (4)
calcination (Figure 2.1). During digestion, the ground and pre-desilicated bauxite
is dissolved in hot NaOH to produce an aluminium hydroxide solution. The
causticity of the Bayer aluminate caustic solution decreases continuously during
the digestion due to the formation of Na2CO3 as consequence of the decomposition
of organic compounds (e.g., polybasic acids, polyhydroxy acids, alcohols and
phenols, humic and fulvic acids, and other carbohydrates) [3]. Consequently, the
process is carried out deliberately at high pH values, which favours the dissolution
of aluminium(III) from bauxite minerals, while most of the other compounds (e.g.,
iron, titanium and silicates) remain insoluble in the solid residue after leaching,
i.e. the so-called bauxite residue or red mud. Since the leachate is recirculated in
the process, it must be purified from both solids and dissolved impurities. The
supersaturated liquor rich in aluminium (also known as green liquor, mainly
composed of NaAl(OH)4), is separated from the solids, remaining after digestion,
in a series of solid-liquid separation units (i.e. thickeners or settlers). Once the
liquor is separated, the solid fraction is washed in several stages to recover NaOH
and NaAl(OH)4, and treated in a filter or thickener to increase the solid content
before being transported to the bauxite residue disposal area [4]. The solid residue
has a red colour because of its high iron content (ca. 30 wt.%) and, nowadays, is
accumulated in land-based ponds or in dry stacks [5,6]. In 2017, about 126 million
tonnes of alumina were produced globally [7], and it is estimated that for each
tonne of alumina produced, about 1 to 2 tonnes of bauxite residue are generated
[8,9]. Furthermore, it is estimated that the global inventory of bauxite residue
increased from 2.7 to 4.0 billion tonnes in the last decade [1,9,10]. The large
volume of bauxite residue produced each year, and the cost associated with storing
(ca. 4 – 12 US$ per tonne of bauxite residue), are the major concerns for the
alumina producers.
-9-
Figure 2.1: Flow sheet for the Bayer process. Adapted from references [3,4].
Bauxite residue is mainly composed of compounds that have a low

solubility in concentrated NaOH solution: iron and titanium minerals, undigested
alumina minerals, sodium aluminium hydrosilicates and calcium compounds, and
also contains REEs and radioactive elements [5]. The composition can be very
wide as it depends on the characteristics of the bauxite feed and the operational
condition of the Bayer refinery circuit. A composition range of minerals typically
found for bauxite residue is given in Table 2.1.
- 10 -
Table 2.1: Quantitative mineralogical composition range in bauxite residues (adapted from
[6,12]).
Concentration
Mineral Chemical formula
wt.%
Hematite Fe2O3 5 – 30
Goethite FeOOH 0 – 25
Magnetite Fe3O4 0–8
Diaspore -AlOOH 0–5
Boehmite -AlOOH 0 – 20
Gibbsite -Al(OH)3 0–5
Quartz SiO2 2–5
Rutile TiO2 2–5
Anatase TiO2 0 – 11
Ilmenite FeTiO3 0 – 10
Muscovite K2O3Al2O36SiO22H2O 0 – 15
Sodalite 3Na2O3Al2O36SiO2Na2SO4 4 – 40
Cancrinite Na6Ca2(AlSiO4)6(CO3)2 0 – 50
Calcite CaCO3 1 – 20
Perovskite CaTiO3 0 – 22
Imogolite Al2SiO3(OH)4 0 – 32
Amorphous/unidentified - 5 – 50
2.2 Occurrence of rare-earth elements in bauxite residue

The rare-earth elements (REEs) are the fifteen elements of the
lanthanides series, together with yttrium and scandium. They are sub-divided into
light rare-earth elements (LREEs), which comprise lanthanum to europium, and
the heavy rare-earth elements (HREEs), which comprise the lanthanides from
gadolinium to lutetium, plus yttrium. Although scandium is classified as a rare
earth element, it behaves very differently from the rest of its family. The reason
for this is that it has an ionic radius very similar to that of iron (and magnesium)
and, consequently, tends to concentrate easily in major ferromagnesian rock-
forming minerals, notably clinopyroxene [13]. During the production of alumina
from bauxite, all the REEs report to the bauxite residue [14–16]. The ferrotitanate
compound (LREE,Ca,Na)(Ti,Fe)O3, which is mineralogically similar to the
perovskite (CaTiO3) phase, has been identified as the main host for LREEs in
Greek bauxite residue (see Figure 2.2) [16,17]. This compound is formed during
the digestion process with NaOH, and can split up in cerium-rich and
neodymium/lanthanum-rich particles. Such distribution is due to the occurrence of
thorium in perovskite mineral phases, which tends to be mainly associated with
cerium as both elements share the same tetravalent oxidation state in bauxite
residue [18]. All the other REEs occur only in the trivalent oxidation state and,
consequently, thorium is less likely to be incorporated into the mineral structure
of neodymium/lanthanum phases. Minor amounts of LREEs have been found as
carbonates and phosphates, while the HREEs have been found associated with
yttrium phosphate phases. Yttrium is partially associated with other REEs as well
- 11 -
[17]. Scandium is mostly associated with iron-rich mineral phases, i.e. hematite
and goethite, but is also associated with zircon. Goethite is the main host of
scandium with a concentration up to 330 mg kg-1, which is about twice the
concentration found in hematite [19]. This scandium concentration in bauxite
residue is much higher than the average abundance of scandium in the Earth’s
crust (22 mg kg-1). Furthermore, scandium represents more than 95% of the
economic value of the REEs present in bauxite residue [14,20].
Figure 2.2: EPMA-WDS quantitative elemental mapping of a reacted LREE particle of

bauxite residue. The intensely reacted area is also inter-grown with sodium aluminosilicate
phase (indicated with Na-Al-Si) [17].
2.3 Rare-earth elements recovery from bauxite residue

Due to their unique physicochemical properties, the REEs are considered
crucial for the manufacture of many high-tech products [21]. Despite their name,
most REEs are abundant in nature, but they are not often enriched to economic
concentrations in ores. Worldwide there are approximately 120 million tonnes of
proven rare-earth oxide reserves spread in countries such as China, Russia, Brazil,
Vietnam, India and the United States [22]. These resources could sustain the global
REE production at the current pace for more than a hundred years with China
dominating the market [23]. However, a significant increase of the REE demand
is expected for the upcoming years, as consequence of a fast development of clean
- 12 -
technologies (e.g., solar cells, electric vehicles, wind turbines). Therefore, the
control of China over the world mine production of rare-earth oxides is detrimental
for the global market, as small changes in the foreign policy and export criteria
can significantly decrease the imports of REEs. Hence, the development of
comprehensive strategies for sustainable primary mining and recycling are
necessary to avoid the large expenditures of REEs imports. Freshly produced
flows and stocks of bauxite residues can help to overcome the monopolistic supply
situation, as such material can provide major amounts of critical metals like REEs,
but also products for low-carbon building materials [24,25].
In recent years, much effort has been paid to the extraction of valuable
metals from bauxite residue through processes developed at temperatures above
300 ºC (pyrometallurgical methods) and/or processes involving aqueous solutions
at temperatures between 20 – 200 ºC (hydrometallurgical methods). The recovery
of REEs from bauxite residue is commonly performed by hydrometallurgical
treatment of bauxite residue or bauxite residue slags (obtained after reductive
smelting of bauxite residue), which usually consider the use of strong mineral
acids [26–28]. Nonetheless, the acid leaching processes consume large volumes of
reagent due to the high alkalinity of bauxite residue (pH values between 10 – 13).
The chemical dissolution of major metals from bauxite residue starts occurring at
pH values below 3 – 4, while most of the REEs start dissolving only at pH values
below 2 (Borra et al., 2015b). Consequently, in order to ensure a high extraction
yield of REEs from the highly alkaline bauxite residue, a large amount of reagent
must be consumed to acidify the alkaline bauxite residue. Despite the high acid
consumption required (about 500 kg tonnes-1 [29]), direct acid leaching of bauxite
residue has demonstrated high extraction yield of REEs with different mineral
acids (e.g., HCl, HNO3, H2SO4), but with very low selectivity due to the substantial
co-dissolution of iron, aluminium, titanium and silicon (Figure 2.3) [30,31]. A
high extraction yield of REEs can be obtained with HNO 3, especially for yttrium
and scandium, which provides a high selectivity between heavy and light REEs
[31,32]. However, the use of HNO3 involves the formation of toxic vapours that
make its industrial application very difficult to handle. Furthermore, the use of
HNO3 does not fulfil all the EU environmental regulations for solid waste disposal
and HNO3 is expensive [33]. Diluted HCl has also demonstrated high extraction
yield, but with less selectivity, as a high co-dissolution of iron and aluminium takes
place [30,31,34]. The use of diluted H2SO4 has demonstrated relatively high
extraction yields of REEs, although it leads to the transformation of CaCO 3 into
poorly soluble CaSO4. The use of organic acids, such as oxalic acid (C2H2O4) or
citric acid (C6H8O7), in solution, reported comparable results as those achieved
with mineral acids, but only when the temperature is above 90 ºC [35]. Leaching
with alkali (hydrogen) carbonates allows a recovery up to 65 wt.% of scandium in
alkaline media, so that the consumption of acids is avoided, but the recovery of
other REEs is low (< 30 wt%) [36–39].
- 13 -
(a) (b)
(a)
(b)
©
(d)
(c) (d)
Figure 2.3: Effect of (a) HNO3 concentration on the dissolution of major elements in the
bauxite residue sample. The dissolution of selected REEs and iron with (b) HNO 3, (c) HCl
and (d) H2SO4 is also shown (25 °C, L/S: 50, 24 h) (adapted from [30]).
Biometallurgical methods have also been investigated for the recovery of

REEs from bauxite residue. Bioleaching considers the use of microorganism for
recovering valuable metals from low-grade ores, concentrates and solid waste
materials [40,41]. Qu et al. have used a bioleaching method to extract REEs (and
also radioactive elements) from a bauxite residue of Chinalco in Guizhou province
in China [42,43]. Up to 30 wt.% of scandium can be recovered, but the process
requires very long leaching times to let the fungi grow [44]. The use of ionic
liquids for REEs recovery from bauxite residue has been reported as well. High
temperature-leaching with the hydrophilic ionic liquid, 1-ethyl-3-
methylimidazolium hydrogensulfate, [Emim][HSO4], allowed an extraction of
approximately 80 wt.% of scandium from bauxite residue at 200 ºC, but with a
substantial dissolution of major metals [45,46]. However, it was reported that the
selectivity of REEs over major elements could be enhanced at relatively low
leaching temperatures (60 – 90 ºC) [47]. Meanwhile, the hydrophilic ionic liquid
betainium bistriflimide, [Hbet][Tf2N] provided a better selectivity of REEs (up to
70 – 85 wt.%) against the co-dissolution of major metals, but the scandium
recovery was limited to 45 wt.% [48]. Although ionic liquids can enhance the
selectivity of REEs over other elements, they are characterised by their high
viscosity, which requires the use of high-temperature leaching. Moreover, most
ionic liquids are expensive, so that it is very difficult to develop an economically
- 14 -
feasible ionic liquid process for the recovery of metals from a low-grade industrial
process residue, such as bauxite residue.
In general, REEs can be recovered from the leach solution of bauxite
residue by solvent extraction and/or ion exchange processes. It has been reported
that the use of D2EHPA and TBP as extractants allows a high recovery of
scandium from synthetic H2SO4 leachates of Australian bauxite residue over other
extractants [49]. In another study, the recovery of scandium from Greek bauxite
residue leached with HNO3, followed by ion-exchange chromatography with
Dowex 50W-X8 resin and solvent extraction using D2EHPA in n-hexane, has also
been reported [50]. Roosen et al. recovered and purified scandium from a Greek
bauxite residue leachate with a novel biopolymer as sorbent material [51]. The use
of titanium phosphate has also been studied to recover scandium from a synthetic
leach solution of bauxite residue with a high selectivity over iron. Iron was reduced
from its trivalent oxidation state to the divalent state by sodium sulphite [52].
2.4 Towards an efficient recovery of REEs from bauxite residue

Mineral acids are superior lixiviants compared to organic acids for
leaching of REEs from bauxite residue. However, the main disadvantages of acid
leaching are the consumption of large amounts of acid during neutralisation of the
alkaline bauxite residue, the high co-dissolution of major metals (e.g., iron,
aluminium and titanium) that affects the efficiency of the separation process (e.g.,
solvent extraction or ion exchange), and the decomposition of silicate compounds
that leads to the polymerisation of amorphous silica. Polymerisation of silica
represents a common problem during leaching of silicate minerals, due to the fact
that the leach solution can no longer be filtered, reducing significantly the
efficiency of solid-liquid separation [53–55]. In order to develop an efficient
process for recovering REEs from bauxite residue, the use of concentrated
solutions is highly recommended, as it minimizes the amount of reagents needed
for further steps. Furthermore, a highly selective process can allow to significantly
decrease the cost associated to the downstream separation processes [56].
Neutralisation of bauxite residue before acid leaching

In order to reduce the amount of acid needed for neutralisation of bauxite
residue and, consequently, to develop a more sustainable process for metal
recovery, alternative neutralisation routes must be considered. The use of sea water
for neutralisation of bauxite residue, for example, has demonstrated promising
results due to the presence of Ca2+ and Mg2+ ions, which can react with OH- and
Al(OH)4- ions [57–59]. Queensland Alumina Limited (QAL) has successfully
implemented a neutralisation plant that employs sea water to neutralise the
remaining soda from the Bayer process [60]. The neutralised product, however, is
being used for upstream wall construction of the storing dam, so that the capacity
of the dam can be enhanced. The addition of gypsum (CaSO4·2H2O) to weathered
- 15 -
bauxite residue also allows to reduce bauxite residue’s alkalinity due to the release
of Ca2+ ions into the solution [61–64]. Rusal Ltd. at its alumina refinery in
Aughinish, in Ireland, has successfully implemented the use of gypsum as a
neutralisation reagent in their bauxite residue disposal area [8,65]. Both
neutralisation reagents can decrease the pH down to 8 – 9 for a safe disposal.
Hence, seawater and gypsum have been investigated as technologies for
rehabilitation of bauxite residue. However, neutralisation with sea water and/or
gypsum entails the formation of precipitates such as calcite or aragonite (CaCO3),
dawsonite (NaAlCO3(OH)2), dolomite (CaMg(CO3)2), muscovite
(KAl2(AlSi3O10)(F,OH)2), slaked lime (Ca(OH)2) and (hydroxy) tricalcium
aluminate (Ca3Al2(OH)12). The dissolution of these compounds, particularly
carbonates, can again lead to a large acid consumption during acid leaching, which
makes its application unpractical as a pre-treatment for metal recovery. Microbial
neutralisation, on the contrary, can decrease bauxite residue’s alkalinity down to
6 – 7 when bauxite residue is incubated at room temperature for more than 1 month
with alfalfa hay, due to the release of organic acids from the microbes (e.g., lactic,
acetic, propanoic and butyric acid) [66]. This neutralisation method has shown
promising results in terms of bioremediation of the bauxite residue disposal areas,
as several plants were able to grow after this neutralisation treatment.
Nevertheless, the use of microorganisms for bauxite residue neutralisation before
acid leaching has not been studied so far.
Khaitan et al. demonstrated that alkalinity in bauxite residue is mainly
controlled by the dissolution of solids and only for a small portion by the OH -- and
Al(OH)4- ions present in solution [67]. Thus, the dissolution of tricalcium
aluminate buffers the pH at around 11, dissolution of desilication products (DSP,
formed during the Bayer process) buffer the pH at values below 9, and calcite
dissolution buffers the pH at about 6 [67–69]. The use of mineral acids for
neutralisation can allow a substantial reduction of bauxite residue’s pH because
they can dissolve a significant amount of silicate compounds. However, the extent
to which the silicate compounds can decompose, may depend on the type of
mineral acid and concentration. On the other hand, the use of a weak acid such as
acetic acid (CH3COOH) as a neutralisation reagent may contribute substantially
to the depletion of bauxite residue’s alkalinity because silicate compounds are
readily soluble in this media, and the reaction can proceed spontaneously without
consuming large amount of reagent [70]. The contacting period of time for the
reaction between the bauxite residue and acid represents, however, the main
drawback for the utilization of acids for bauxite residue neutralisation.
Furthermore, the amount of acid required is generally large to fully neutralize the
residue and accounts for the relatively high cost (about 200 USD/tonne) [71].
Neutralisation of bauxite residue with CO2 gas represents a potentially
inexpensive and more safe and sustainable technology because CO2 can be
obtained from the gas emissions generated during the alumina refinery or
produced from direct oil burning [72]. A very good example of such application
is Alcoa’s Kwinana refinery, in Western Australia, which uses waste CO 2 from a
nearby ammonia plant reducing the pH of the bauxite residue slurry to a non-
- 16 -
hazardous level (pH < 9.5) and capturing in the process about 35 kg CO 2 per tonne
of bauxite residue by mineral carbonation [73]. It has been reported that during
neutralisation with CO2, the formation of thermodynamically stable carbonates,
such as cancrinite (Na6Ca2[(CO3)2|Al6Si6O24]·2H2O) and chantalite
(CaAl2(SiO4)(OH)4), are responsible for carbonation of bauxite residue [74].
Hence, the pH of bauxite residue can be reduced down to 9 when high CO2-partial
pressures and long-term contacting period are applied [75–78].
Separation of major metals from bauxite residue

During conventional acid leaching, the efficient extraction of REEs from
bauxite residue is highly dependent on the concentrations of iron, aluminium and,
to a minor extent, titanium. The co-dissolution of these elements can significantly
affect the selectivity of REEs during acid leaching. It is known that during acid
leaching with H2SO4, iron reacts with sulfate ions to form a rhomboclase
compound (H5Fe3+O2 (SO4)2·2H2O) that can entrap REEs, in particular scandium
[79]. Nevertheless, during the direct treatment of bauxite residue with mineral
acids, the concentration of REEs in the leachate (ca. 10 mg L-1) is about one
hundred times lower than the iron or aluminium concentration, i.e. ca.
1000 mg L-1. Therefore, the efficiency of the separation process can be
significantly reduced as major metals can also be co-adsorbed on the ion-exchange
resin, for instance, in the solvent extraction and/or ion exchange process, which
hampers the efficient REEs recovery.
The direct magnetic separation of iron from bauxite residue was used to
recover iron oxides from bauxite residue in the 1970s, but the recoveries (up to
55 wt.%) and grade were relatively low due to the fineness and mineralogy
(goethite has a lower magnetic susceptibility than hematite) of the bauxite residue
[28,80–82]. Different pyrometallurgical techniques to separate aluminium and
iron, before REEs recovery by acid leaching, have been reported as well.
Reductive roasting of bauxite residue briquettes, prepared with sodium salts (e.g.,
Na2CO3 or Na2SO4), for example, facilitates the growth of metallic iron particles,
and ca. 95 wt.% of iron recovery can be achieved by an enhancement of the
magnetic separation efficiency [83]. However, the method requires coarse
particles of bauxite residue and a high concentration of iron. The typical particle
size of bauxite residue is < 100 µm [84,85]. The solid fraction remaining after
magnetic separation can be leached with aqueous Na2CO3 at ambient pressures for
alumina recovery, while the REEs can be recovered from the solid residue by an
additional acidic leaching step (full line route in Figure 2.4) [81]. As an alternative
(segmented line route in the same Figure), the non-magnetic fraction can be
leached with diluted H2SO4 for titanium recovery, while the leach liquor ends up
rich in aluminium, sodium and REEs, but also in silicon that can reduce the
efficiency of the solid-liquid separation process due to the silica polymerization
[86].
- 17 -
Figure 2.4: Conceptual process flow sheet of roasting-magnetic separation-leaching for

Fe/Al (full line route) or Fe/Ti recovery (segmented line route). Adapted from references
[81, 83].
Solid-state reduction of bauxite residue by microwave hydrogen plasma

has been proposed as an innovative technology to separate iron from bauxite
residue. Figure 2.5 describes the conceptual process flow sheet of solid-state
reduction of bauxite residue by microwave hydrogen plasma technology. One
option is to pre-treat the bauxite residue at temperatures between 500 – 700 °C for
physical and chemical adsorbed water removal, followed by microwave plasma
with H2 as a reductant reagent. Alumina and silica can be separated as a solid
residue, leaving behind iron and titanium in the leach solution as their chlorides
[87]. Another option is the production of direct reduced iron (DRI) using hydrogen
as reductant at temperatures between 300 – 800 °C without considering a pre-
treatment process [88]. The results have demonstrated a substantial separation of
iron (up to 99 wt.%) from bauxite residue, particularly at high temperatures,
although no information has been reported about REEs distribution between the
solid and liquid phases. The reduction of bauxite residue with a solid reductant
reagent (e.g., lignite coke) has also been investigated by several researchers [89–
93]. Although the resulting product can be used for steelmaking or as a charge to
the blast furnace due to the high concentration of iron [94], it is believed that
scandium may report to the iron-rich phase as iron oxide due to their chemical
association and, consequently, be lost in the final product.
- 18 -
Figure 2.5: Conceptual process flow sheet of solid-state reduction of bauxite residue by
microwave hydrogen plasma technology. Adapted from references [87, 88].
Smelting of bauxite residue (normally performed at temperatures

between 1500 – 1550 ºC) allows to reduce iron-oxides to produce pig iron and a
slag rich in REEs [27,95]. The process strongly depends on a proper separation
between the metal and the slag phase and, consequently, a large amount of flux
(e.g., CaO and SiO2) and reductant are required. The higher the concentration of
aluminium in the bauxite residue is, the higher the flux consumption becomes [96].
The large volume of fluxes may increase not only the energy consumption during
smelting, but also the acid consumption during leaching due to the enrichment of
aluminium and silicon in the slag. Therefore, in order to reduce the overall process
cost and the energy consumption, it is recommended to separate aluminium before
iron. The alkali roasting process has been proposed to separate aluminium prior to
the smelting process (Figure 2.6), so that the consumption of flux can be
diminished [96]. The method allows the recovery of alumina as a valuable product,
but the roasting step must be carried out at temperatures about 500 ºC (when
bauxite residue is mixed with NaOH to convert alumina into sodium aluminate,
which is soluble in water). Meanwhile, the smelting stage is carried out at 1500 ºC
in absence of fluxes for iron removal. Although the total energy consumption of
the process has been estimated to be about 3.5 GJ tonne -1, which is similar to the
energy consumption reported by direct (reductive) smelting of bauxite residue
[29], the volume of slag generated by this process is about 50% lower than the
volume of slag produced by direct smelting process, which limits the amount of
REEs that can be further recovered. Reductive smelting allows to separate iron
from the REEs that can be highly enriched, together with significant amount of
aluminium and silica, but also titanium, in the slag phase. REEs can be extracted
by leaching the slag with mineral acids.
- 19 -
Figure 2.6: Conceptual process flow sheet for recovery of metals from bauxite residue.
Adapted from reference [96].
Silicon dissolution during acid leaching

During the recovery of valuable metals by acid leaching of bauxite
residue, silica gel can be formed due to the leaching of silicate minerals, which
significantly affects the filtration efficiency of the leach liquor, i.e. reduced
filterability [53–55,97]. Therefore, silica gel formation represents a serious
drawback in the extraction of metals from ores and process residues by
hydrometallurgical methods because the gel solutions can no longer be filtered.
Moreover, it is believed that this gelatinous precipitate can blind ore particles and
reduce the leaching efficiency significantly.
The most abundant silica bearing minerals in bauxite residue are quartz,
sodalite, muscovite, kaolinite and cancrinite [10,98]. In Table 2.2 the most
common silicon bearing minerals typically found in bauxite residues are presented
together with their corresponding range of concentration. In particular, quartz is
chemically inert with a high melting point (above 1650 ºC), and it can be dissolved
only by strong acids (i.e. HF), which differentiates it from silicate minerals. Under
standard conditions (1 atm, 25 ºC), the solubility of quartz is only around
6 mg L-1 whereas amorphous silica reaches a solubility of 100 – 120 mg L-1 under
the same conditions [99]. From this point of view, the contribution to soluble silica
by quartz is considered negligible. However, the dissolution of amorphous silica
can lead to an increase of silica concentration in the solution, which can result in
an increase of the silicon supersaturation index, which is considered the driving
- 20 -
force for silica polymerization as it represents the ratio of dissolved silica with
respect to the maximum silica solubility [100,101].
Table 2.2: Silicon bearing minerals and their range of composition in bauxite residue
(adapted from [10, 98]).
Range of
Mineral phase Chemical formula
concentration, wt.%
Quartz SiO2 3-20
Sodalite 3Na2O·3Al2O3·6SiO2·Na2SO4 4-40
Muscovite K2O·3Al2O3·6SiO2·2H2O 0-15
Kaolinite Al2O3·2SiO2·2H2O 0-5
Cancrinite Na6[Al6Si6O24]·2CaCO3 0-50
Carnegieite Si4Al4Na4O16 1-3
Cancrinite H2O Na8(Al,Si)12O24(OH)2·2H2O 4-40
Hydroxycancrinite
(Na,Ca)8(Al,Si)12O24(CO3)·4H2O 0-10
CO3
Cancrinite NO3 Na7.92[Al6Si6O31.56]N1.74 0-45
Cancrinite CO3 Na7.86(AlSiO4)6(CO3)(H2O)3.3 0-5
Lawsonite CaAl2(Si2O7)(OH)2(H2O) 4-13
During acid leaching of bauxite residue, silicate compounds can dissolve

from the bauxite residue by forming H4SiO4 and H3SiO4-, according to Eq. 2.1 and
2.2, which are the precursors for silica gel polymerization [101,102]. Such
products tend to polymerize via dimers, trimers, etc. to cyclic oligomers
(Sin+1Om+2·OH), according to Eq. 2.3, until a gel network is formed via Ostwald
ripening, which finally results in the formation of an acidic silica gel [100]. The
polymerization of silica depends significantly on the pH, temperature and ionic
strength of the solution [103].
M2 SiO4(s) + 4HCl(l) + H2 O(l) → 2MCl2(aq) + Eq. 2.1

H2 O(l) + H4 SiO4 (aq) 
H4 SiO4 (aq) + H2 O → H3 SiO4− + H3 O+ Eq. 2.2
Si𝑛 O𝑚 OH + H4 SiO4 → Si𝑛+1 O𝑚+2 OH + 2H2 O Eq. 2.3
In order to avoid the dissolution of silica and, consequently, the eventual

silica polymerization, several processes have been already reported to enhance the
efficiency of the filtration process and, therefore, the recovery of valuable metals
by acid leaching. The combination of sulfation, roasting and leaching processes
avoids the co-dissolution of silicon and iron, according to the diagram presented
in Figure 2.7 (process route described with full lines) [104]. In the first stage of
the process, bauxite residue is mixed with low amount of water (just to form a
- 21 -
paste, i.e. agglomeration of particles) and concentrated sulfuric acid. During the
second stage, the roasting process of the previously formed mixture is carried out
at temperatures between 600 – 700 ºC, at which sulfates with low thermal stability
such as Fe2(SO4)3 decompose until the corresponding oxide is formed. The REE
sulfates are stable at such temperature and they can be easily dissolved during the
subsequent water leaching stage leaving the iron-oxide in the residue. The SO3
generated during the roasting process can be used to generate H 2SO4, which can
be re-used in the sulfation stage. The dissolution of silicon, iron and aluminium is
significantly reduced when roasting is performed at temperatures between 650 –
700 ºC, but the process requires long contacting period of time (> 1 h). Under those
conditions, about 60 wt.% of scandium and more than 90 wt.% of the other REEs
can be extracted from bauxite residue. It must be noticed that with alkali-roasting,
i.e. roasting with sodium salts (e.g., Na2CO3, Na2SO4, or NaOH), silica
polymerization is highly promoted due to the formation of sodium aluminosilicate
compounds (e.g., Na1.75Al1.75Si0.25O4 and NaAlSiO4), which are highly soluble in
acidic media [83,86]. Therefore, alkaline roasting becomes unpractical in terms of
silica polymerization depletion. The sulfation-roasting-leaching method has been
further enhanced to recover scandium oxide and a mixture of rare-earth oxides as
a side product by considering two consecutives precipitation stages after leaching
with water (process route described with segmented lines in Figure 2.7) [105]. In
the first precipitation stage, NaOH is used to increase the pH of the solution up to
8, so that most of the REEs can precipitate. Scandium, as Sc(OH)3, does not
precipitate due to its amphoteric behaviour, which shows significant solubility at
both low and high pH values [106]; similar amphoteric behaviour is known for
Al(OH)3 [107]. A second precipitation stage with addition of H2SO4 or oxalic acid
allows to precipitate scandium, which can be further recovered by roasting.
Figure 2.7: Conceptual process flow sheet of sulfation-roasting-leaching. Adapted from

reference [20, 56].
- 22 -
The combination of H2O2 and H2SO4 allows to diminish not only the
precipitation of iron(III)-sulfate mineral formed during conventional leaching with
H2SO4, but also to suppress the silica polymerization [108]. The dissolved silicon
from the bauxite residue precipitates as SiO2, which is readily filterable, due to the
highly oxidative conditions created by the addition of H 2O2. At 90 ºC, with an
equal concentration of 2.5 M of H2O2 and H2SO4, and a leaching period of 30 min,
an extraction of approximately 70 wt.% of scandium with a substantial depletion
of silica dissolution can be achieved, but with a significant co-dissolution of iron
(about 35 wt.%). The process also allows to recover 90 wt.% of titanium as [TiO-
O]SO4 complex.
The mixture of different mineral ores with concentrated mineral acids, at
very high solid-liquid ratio (or high pulp density) and prolonged residence times,
followed by water leaching, allows to avoid the hydrolysis of silicon and,
consequently, the eventual polymerization of silica. Hence, the acid pugging (also
reported by some authors as curing, fuming or baking) method has been proposed
to treat uranium mineral ores with high concentration of silicate minerals [109].
The method, depicted in Figure 2.8, considers as a first stage the mixture of ores
with a low amount of concentrated H2SO4 in order to form a paste. The curing
procedure takes place afterwards at a temperature between 100 – 120 ºC, by which
the mixture progressively dries, followed by water leaching. The method allows
to effectively extract uranium and other valuable metals like copper and
manganese from the ores. A similar procedure has been applied to recover REEs
from eudialyte minerals, which are characterized by their high silica content
[110,111]. The authors reported a two-stage treatment considering fuming with
2 mol L-1 concentrated acid (HCl or H2SO4) at temperature between 100 – 110 ºC
followed by water leaching. The leach liquor resulted in a REE extraction above
90 wt.% and low silicon concentrations in the solution. Similar to these methods,
the acid bake process, patented by the Scandium International Mining Corp.
(earlier known as EMC Metals Corporation), allows to recover scandium from
mineral ores rich in scandium and silicon [112]. The process considers (Figure
2.9), in the first stage, the mixture of a scandium-rich feed with concentrated
mineral acids (e.g. H2SO4, HCl, HNO3). During the second stage, the baking
process is carried out at temperatures between 200 – 280 ºC [113]. In the third
stage, the baked material is leached with water from where scandium can be
recovered. The gas (SO3) generated during baking is used for production of acid
that can be used back in the mixing stage, similar to the principle of the
sulfationroastingleaching process described before (Figure 2.7). The pH of the
pregnant leach solution is kept at a pH above 2.5, so that iron does not dissolve.
The redox potential of the solution must be maintained between +0.7 V and
+1.2 V to suppress the formation of jarosite, which can hamper the recovery of
scandium [112]. The scandium extraction can reach up to 75 wt.%, while the
silicon compounds remain in the solid residue produced after water leaching. The
application of the curing method at room temperature, on the other hand, has also
demonstrated high extraction yield of REEs from eudialyte concentrates with HCl
[103].
- 23 -
Figure 2.8: Conceptual process flow sheet of curing method. Adapted from reference [109].
Figure 2.9: Conceptual process flow sheet of acid bake method. Adapted from reference
[109].
High-pressure acid leaching (HPAL) can also limit the silicon

dissolution as water is released as vapor at a high temperature, which inhibits the
generation of silica gel. HPAL has demonstrated to be a suitable technology for
recovering metals from nickel laterites and nickel converter slags, which are also
rich in silicate compounds [114,115]. It must be noticed that bauxite residue and
lateritic ores have similar chemical compositions, as both are rich in scandium,
iron and aluminium [116–119]. Iron phases, such as limonite (FeO(OH)·nH2O) in
the lateritic ore and goethite and hematite in the bauxite residue, react similarly
due to the high temperatures involved in the process. The direct treatment of
bauxite residue by HPAL with H2SO4 allows to selectively dissolve REEs while
leaving iron and titanium in the solid fraction, but the extraction yields are low
(< 45 wt.%) presumably due to the high co-dissolution of sodium and aluminium
that can be present as sodalite and cancrinite [120,121].
Orbite Tehnologies Inc. (Canada) implemented an industrial HPAL
process to extract valuable metals such as aluminum, titanium and REEs from
different feedstocks, among others, different bauxite residues, at a ratio of 100.000
– 200.000 tonnes per year [122]. The process developed by the company had the
capability of producing high value metals with minimum extraction yield of
88 wt.% for aluminium and 96 wt.% for REEs [123]. The conceptual process
applied to bauxite residue is depicted in Figure 2.10. In the first step, bauxite
- 24 -
residue is leached at high temperatures (150 – 170 ºC) in an autoclave using HCl.
All the metals, with exception of titanium, dissolve as chlorides. Specifically,
aluminium and iron dissolve to form AlCl3 and FeCl3. Silicon and titanium remain
insoluble and are removed by filtering. After filtration, the leach liquor is treated
with HCl gas to increase the chloride concentration, which helps in AlCl3·6H2O
precipitation. AlCl3·6H2O is filtered off from the iron-rich liquor. The AlCl3·6H2O
precipitate is transformed to Al2O3 by calcination at 900 – 950 ºC. HCl gas is
produced during the calcination process, which is re-used in either the leaching
and/or the aluminium precipitation stage. The leach liquor after the separation of
aluminium contains, among other elements, FeCl3. The ferric chloride is
hydrolysed at 180 ºC, producing a precipitate of Fe 2O3 while regenerating HCl.
The ferric oxide is very pure and can be sold commercially as a specialty by-
product. After iron removal, the solution is rich in elements such as magnesium,
calcium, sodium, gallium, and REEs. The REEs are separated from the leach
solution by conventional solvent extraction. The process allows to recover a wide
variety of metals and, most of the chlorine is recovered during the conversion of
the chlorides into oxides, and through contact with hydrogen, and is regenerated
into HCl, which is then reused as required throughout the process, thereby
minimizing the need to import additional HCl. Nonetheless, the large volume of
effluents enriched with HCl represents the major concern in the process due to its
high corrosiveness. Although the company reported a positive return on
investment, the use of glass-lined reactor, valves and pipes made with high-
performance chemically-resistance polymers significantly increased the capital
cost, which affected the economy of the process [122].
Figure 2.10: Conceptual process flow sheet of Orbite process. Adapted from reference
[123].
- 25 -
2.5 Existence of naturally occurring radionuclide activity in bauxite

residue
Bauxites residues can contain some levels of naturally occurring
radioactive materials (NORM), also known as technologically enhanced naturally
occurring radioactive material (TENORM), which concentration mainly depends
on the origin of the bauxite ore [124,125]. From these, 238U and its decay products,
232
Th and its decay products, and 40K are the most important ones [10]. During the
Bayer process, uranium dissolves but it subsequently re-precipitates and is
associated with the coarser bauxite residue fraction. Meanwhile, thorium is not
affected by low-temperature extraction process and is most often associated with
the fine bauxite residue fraction [126]. In nature, thorium and uranium tend to
occur in similar mineral phases as those of the REEs [127,128]. In bauxite residue,
however, little is known about their mineralogy and chemical speciation, but it is
estimated that these elements are enriched in the bauxite residue by a factor of
about 2 [5]. In Greek bauxite residue, thorium is associated with titanium phases
(i.e. perovskite) and titanium-iron phases [129].
Their concentration in bauxite residue defines the risk for the safe
utilization of bauxite residue. The IAEA (International Atomic Energy Authority)
Basic Safety Guide for marketable materials sets a limit of 1000 Bq kg-1 in the
case of 238U and 232Th (individually), and of 10000 Bq kg-1 in the case of 40K [130],
which are the recommended limits for exception or clearance. The concentration
of natural radionuclides in bauxite residue produced in different countries is shown
in Table 2.3. According to the limits set by the IAEA and the concentrations shown
in Table 2.3, thorium is only present in significant levels in bauxite residue from
Australia. Nevertheless, bauxite residues can be considered to be unlikely to cause
elevated exposure to ionizing radiation and, therefore, can be exempted from
radiological regulatory control. However, it must be noticed that the radiological
properties of bauxite residues can be altered when the bauxite residue is further
processed to recover valuable metals (e.g., aluminium, iron, titanium and REEs).
It has been reported that about 85% of thorium and polonium ( 210Po, as a decay
product of 238U) remains in the solid residue after leaching with a radioactivity
level < 370 Bq kg-1 [131]. In another study, the radiological properties of post-
processed bauxite residues by sintering, smelting and acid leaching (for the
recovery of aluminium, iron and REE, respectively) were also examined. The
sintered and the leached sample reported a concentration of radionuclides
< 700 Bq kg-1, while the slag produced from the smelting process reported a
concentration higher than 1000 Bq kg-1, which means that this particular sample
(slag from bauxite residue smelting) is classified as NORM and it must be
subjected to regulatory control [132]. Samouhos et al. also reported a substantial
enrichment of NORM when iron was separated from the bauxite residue by
reductive roasting [91]. The separation of iron from bauxite residue, therefore, can
lead to an enrichment of NORM in the corresponding products.
- 26 -
Table 2.3: Natural radioactivity (Bq kg-1) of bauxite residue produced in different countries.
238U 232Th 40K

Country Reference
Greece 230 387 17 [133]
Spain 203 ± 35 598 ± 18 62 ± 13 [134]
China 350 ± 21 414 ± 41 583 ± 18 [126]
Turkey 210 ± 6 539 ± 18 112 ± 7 [135]
Australia 310 ± 20 1350 ± 40 350 ± 20 [136]
Brazil 139 ± 1 350 ± 19 45 ± 2 [125]
Germany 122 183 n.d. [136]
Hungary 299 ± 49 314 ± 75 48 [136,137]
Jamaica 709 ± 479 339 ± 16 300 ± 49 [138]
n.d.: not determined
The EU Radiation Protection Guideline 112 has recommended the range

of 0.3 – 1 mSv y-1 for building materials; the particular limit being determined by
the expected exposure [139]. Doses exceeding these values, i.e. gamma dose
received in the open air, should be considered from a radiation protection point of
view. The direct application of bauxite residue in cement and cementitious
applications has demonstrated a very low radiological impact (< 0.54 mSv y-1),
despite the higher concentration of 226Ra (as a decay product of uranium) and 232Th
with respect to other residues (e.g., metallurgical slags, phosphogypsum) [136].
2.6 Conclusions
The recovery of REEs by direct acid leaching of bauxite residue is limited

by the large acid consumption due to the neutralisation of the alkaline bauxite
residue, by the high co-dissolution of major metals (e.g., iron, aluminium and
titanium), and by the decomposition of silicate compounds that lead to the
polymerization of amorphous silica. Furthermore, direct acid leaching hardly
reduces the volume of solid residue to be stored, because the valuable metals are
present only in small concentrations. Aluminium and iron can be separated from
bauxite residue before acid leaching by considering the combination of roasting-
smelting-magnetic separation methods, although operational conditions, such as
flux consumption and temperatures, must be well optimized to achieve high
recoveries. REEs can be further extracted from the non-magnetic slag. However,
due to the enrichment of silicon on the slag, conventional acid leaching can no
longer be considered due to silica gel formation. The unknown concentration of
NORM in the products generated after removal of valuable elements (aluminium,
- 27 -
iron, titanium and REEs) requires significant research effort, particularly on the
solid residues, to provide the basic information about the speciation of these
constituents for further evaluation.
2.7 References

aluminium.org.
doi:10.1016/j.hydromet.2011.02.004.
[3] A.R. Hind, S.K. Bhargava, S.C. Grocott, Quantitation of
alkyltrimethylammonium bromides in Bayer process liquors by gas
chromatography and gas chromatography-mass spectrometry, J.
Chromatogr. A. 765 (1997) 287–293. doi:10.1016/S0021-
9673(96)00922-3.
(2011) 33–45. doi:10.1016/j.hydromet.2011.02.006.
[5] C. Klauber, M. Gräfe, G. Power, Bauxite residue issues: II. options for
residue utilization, Hydrometallurgy. 108 (2011) 11–32.
doi:10.1016/j.hydromet.2011.02.007.
[6] K. Evans, Successes and Challenges in the Management and Use of
Bauxite Residue, in: Bauxite Residue Valorization Best Pract. Leuven,
(Belgium), 5-7 Ocotber 2015, 2015: pp. 113–127. doi:10.1007/s40831-
016-0060-x.
[7] The International Aluminium Institute, The International Aluminium
Institute, (2018). http://www.world-aluminium.org (accessed June 12,
2018).
[8] A.W. Bray, D.I. Stewart, R. Courtney, S.P. Rout, P.N. Humphreys,
W.M. Mayes, I.T. Burke, Sustained bauxite residue rehabilitation with
gypsum and organic matter 16 years after Initial treatment, Environ. Sci.
Technol. 52 (2018) 152–161. doi:10.1021/acs.est.7b03568.
[9] M. Gräfe, C. Klauber, Bauxite residue issues: IV. Old obstacles and new
pathways for in situ residue bioremediation, Hydrometallurgy. 108
(2011) 46–59. doi:10.1016/j.hydromet.2011.02.005.
331. doi:10.1007/s40831-016-0060-x.
- 28 -
[11] A.R. Hind, S.K. Bhargava, S.C. Grocott, The surface chemistry of Bayer
process solids: A review, Colloids Surfaces A Physicochem. Eng. Asp.
146 (1999) 359–374. doi:10.1016/S0927-7757(98)00798-5.
[12] T.C. Santini, Application of the Rietveld refinement method for
quantification of mineral concentrations in bauxite residues (alumina
refining tailings), Int. J. Miner. Process. 139 (2015) 1–10.
doi:10.1016/j.minpro.2015.04.004.
[13] A.E. Williams-Jones, O. V. Vasyukova, The economic geology of
scandium, the runt of the rare earth element litter, Econ. Geol. 113
(2018) 973–988. doi:10.5382/econgeo.2018.4579.
doi:10.1016/j.jclepro.2015.02.089.
[15] S.H. Patterson, H.F. Kurtz, J.C. Olson, C.L. Neeley, World Bauxite
Resources, in: U.S. Geol. Surv. Prof. Pap., Washington, 1986.
[17] J. Vind, A. Malfliet, B. Blanpain, P.E. Tsakiridis, A.H. Tkaczyk, Rare
Earth Element Phases in Bauxite Residue, Minerals. 8 (2018) 1–32.
doi:10.20944/preprints201801.0288.v1.
[18] P.N. Gamaletsos, A. Godelitsas, T. Kasama, A. Kuzmin, M. Lagos, T.J.
Mertzimekis, J. Göttlicher, R. Steininger, S. Xanthos, Y. Pontikes, G.N.
Angelopoulos, C. Zarkadas, A. Komelkov, E. Tzamos, A. Filippidis, The
role of nano-perovskite in the negligible thorium release in seawater
from Greek bauxite residue (red mud), Sci. Rep. 6 (2016) 21737.
doi:10.1038/srep21737.
[19] J. Vind, A. Mal, C. Bonomi, P. Paiste, I.E. Sajó, B. Blanpain, A.H.
Tkaczyk, V. Vassiliadou, D. Panias, Modes of occurrences of scandium
in Greek bauxite and bauxite residue, Miner. Eng. 123 (2018) 35–48.
doi:10.1016/j.mineng.2018.04.025.
[20] C.R. Borra, B. Blanpain, Y. Pontikes, K. Binnemans, T. Van Gerven,
Recovery of rare earths and other valuable metals from bauxite residue
(red mud): a review, J. Sustain. Met. 2 (2016) 365–386.
doi:10.1007/s40831-016-0068-2.
[21] K. Binnemans, P.T. Jones, B. Blanpain, T. Van Gerven, Y. Yang, A.
Walton, M. Buchert, Recycling of rare earths: A critical review, J. Clean.
Prod. 51 (2013) 1–22. doi:10.1016/j.jclepro.2012.12.037.
[22] U.S. Geological Survey, Rare earths, in: Miner. Commod. Summ. 2018,
- 29 -
2018: pp. 132–133. doi:https://doi.org/10.3133/70194932.

[23] B. Zhou, Z. Li, C. Chen, Global Potential of Rare Earth Resources and
Rare Earth Demand from Clean Technologies, Minerals. 7 (2017) 203.
doi:10.3390/min7110203.
[24] T. Dutta, K.H. Kim, M. Uchimiya, E.E. Kwon, B.H. Jeon, A. Deep, S.T.
Yun, Global demand for rare earth resources and strategies for green
mining, Environ. Res. 150 (2016) 182–190.
doi:10.1016/j.envres.2016.05.052.
[25] S.H. Ali, D. Giurco, N. Arndt, E. Nickless, G. Brown, A. Demetriades,
R. Durrheim, M.A. Enriquez, J. Kinnaird, A. Littleboy, L.D. Meinert, R.
Oberhänsli, J. Salem, R. Schodde, G. Schneider, O. Vidal, N. Yakovleva,
Mineral supply for sustainable development requires resource
governance, Nature. 543 (2017) 367–372. doi:10.1038/nature21359.
[26] A. Akcil, N. Akhmadiyeva, R. Abdulvaliyev, Abhilash, P. Meshram,
Overview On Extraction and Separation of Rare Earth Elements from
Red Mud: Focus on Scandium, Miner. Process. Extr. Metall. Rev. 00
(2017) 1–7. doi:10.1080/08827508.2017.1288116.
Smelting of Bauxite Residue (Red Mud) in View of Iron and Selective
Rare Earths Recovery, J. Sustain. Metall. 2 (2016) 28–37.
doi:10.1007/s40831-015-0026-4.
[28] Z. Liu, H. Li, Metallurgical process for valuable elements recovery from
red mud—A review, Hydrometallurgy. 155 (2015) 29–43.
doi:10.1016/j.hydromet.2015.03.018.
[29] C.R. Borra, B. Blanpain, Y. Pontikes, K. Binnemans, T. om Van Gerven,
Comparative Analysis of Processes for Recovery of Rare Earths from
Bauxite Residue, JOM. 68 (2016) 2958–2962. doi:10.1007/s11837-016-
2111-y.
doi:10.1016/j.mineng.2015.01.005.
[31] M. Ochsenkühn-Petropulu, T. Lyberopulu, K.M. Ochsenkühn, G.
Parissakis, Recovery of lanthanides and yttrium from red mud by
selective leaching, Anal. Chim. Acta. 319 (1996) 249–254.
doi:10.1016/0003-2670(95)00486-6.
[32] M. Ochsenkühn-Petropulu, K.S. Hatzilyberis, L.N. Mendrinos, C.E.
Salmas, Pilot-Plant investigation of the leaching process for the recovery
of scandium from red mud, Ind. Eng. Chem. Res. 41 (2002) 5794–5801.
doi:10.1021/ie011047b.
[33] K. Hatzilyberis, T. Lymperopoulou, L.-A. Tsakanika, K.-M.
- 30 -
Ochsenkühn, P. Georgiou, N. Defteraios, F. Tsopelas, M. Ochsenkühn-

Petropoulou, Process Design Aspects for Scandium-Selective Leaching
of Bauxite Residue with Sulfuric Acid, Miner. 2018, Vol. 8, Page 79. 8
(2018) 79. doi:10.3390/MIN8030079.
[34] R.M. Rivera, B. Ulenaers, G. Ounoughene, K. Binnemans, T. Van
Gerven, Extraction of rare earths from bauxite residue (red mud) by dry
digestion followed by water leaching, Miner. Eng. 119 (2018) 82–92.
doi:10.1016/j.mineng.2018.01.023.
[35] É. Ujaczki, Y.S. Zimmermann, C.A. Gasser, V. Feigl, M. Lenz,
Recovery of rare earth elements from Hungarian red mud with combined
acid leaching and liquid-liquid extraction, Bauxite Residue Valorization
Best Pract. Leuven, (Belgium), 5-7 Ocotber 2015. (2015) 339–345.
in: Light Met., 2015: pp. 93–96.
[37] O. V. Petrakova, A. V. Panov, S.N. Gorbachev, O.N. Milshin, Improved
efficiency of red mud process through scandium oxide recovery, in:
Proc. Bauxite Residue Valoris. Best Pract. Conf., 2015: pp. 355–362.
doi:10.1002/9781119093435.ch17.
[38] A. Suss, N. V Kuznetsova, А. Kozyrev, А. Panov, Specific Features of
Scandium Behavior during Sodium Bicarbonate Digestion of Red Mud,
in: 35th Int. ICSOBA Conf. Hamburg, Ger. 2 – 5 October, 2017, 2017:
pp. 491–504.
[39] S.P. Yatsenko, I.N. Pyagai, Red mud pulp carbonization with scandium
extraction during alumina production, Theor. Found. Chem. Eng. 44
(2010) 563–568. doi:10.1134/S0040579510040366.
[40] J. Willner, J. Kadukova, A. Fornalczyk, M. Saternus,
Biohydrometallurgical Methods, 54 (2015) 255–258.
[41] M. Debaraj, D. jin Kim, J.G. Ahn, Y.H. Rhee, Bioleaching A microbial
process of metal recovery A review, Met. Mater. Int. 11 (2005) 249–256.
doi:10.1007/BF03027450.
[42] Y. Qu, B. Lian, B. Mo, C. Liu, Bioleaching of heavy metals from red
mud using Aspergillus niger, Hydrometallurgy. 136 (2013) 71–77.
doi:10.1016/j.hydromet.2013.03.006.
[43] Y. Qu, B. Lian, Bioleaching of rare earth and radioactive elements from
red mud using Penicillium tricolor RM-10, Bioresour. Technol. 136
(2013) 16–23. doi:10.1016/j.biortech.2013.03.070.
[44] Y. Qu, H. Li, W. Tian, X. Wang, X. Wang, X. Jia, B. Shi, G. Song, Y.
Tang, Leaching of valuable metals from red mud via batch and
continuous processes by using fungi, Miner. Eng. 81 (2015) 1–4.
- 31 -
doi:10.1016/j.mineng.2015.07.022.
[45] C. Bonomi, P. Davris, E. Balomenos, I. Giannopoulou, Ionometallurgical
Leaching Process of Bauxite Residue : a Comparison between
Hydrophilic and Hydrophobic Ionic Liquids, in: 35th Int. ICSOBA Conf.
Hamburg, Ger. 2 – 5 October, 2017, 2017: pp. 557–564.
[46] P. Davris, E. Balomenos, D. Panias, I. Paspaliaris, Leaching Rare Earth
Elements from Bauxite Residue Using Brønsted Acidic Ionic Liquids, in:
Rare Earths Ind. Technol. Econ. Environ. Implic., 2016: pp. 183–197.
[47] P. Davris, E. Balomenos, D. Panias, I. Paspaliaris, The use of ionic
liquids for rare earth element extraction from bauxite residue, in: Bauxite
Residue Valorization Best Pract. Leuven, (Belgium), 5-7 Ocotber 2015,
2015: pp. 323–330.
[48] P. Davris, E. Balomenos, D. Panias, I. Paspaliaris, Hydrometallurgy
Selective leaching of rare earth elements from bauxite residue ( red mud
), using a functionalized hydrophobic ionic liquid, Hydrometallurgy. 164
(2016) 125–135. doi:10.1016/j.hydromet.2016.06.012.
[49] W. Wang, Y. Pranolo, C.Y. Cheng, Recovery of scandium from
synthetic red mud leach solutions by solvent extraction with D2EHPA,
Sep. Purif. Technol. 108 (2013) 96–102.
doi:10.1016/j.seppur.2013.02.001.
[50] M.T. Ochsenkühn-Petropoulou, K.S. Hatzilyberis, L.N. Mendrinos, C.E.
Salmas, Pilot-plant investigation of the leaching process for the recovery
doi:10.1021/ie011047b.
[51] J. Roosen, S. Van Roosendael, C.R. Borra, T. Van Gerven, S. Mullens,
K. Binnemans, Recovery of scandium from leachates of Greek bauxite
residue by adsorption on functionalized chitosan–silica hybrid materials,
Green Chem. 18 (2016) 2005–2013. doi:10.1039/C5GC02225H.
[52] W. Zhang, R. Koivula, E. Wiikinkoski, J. Xu, S. Hietala, J. Lehto, R.
Harjula, Efficient and Selective Recovery of Trace Scandium by
Inorganic Titanium Phosphate Ion-Exchangers from Leachates of Waste
Bauxite Residue, ACS Sustain. Chem. Eng. 5 (2017) 3103–3114.
doi:10.1021/acssuschemeng.6b02870.
[53] E. Abkhoshk, E. Jorjani, M.S. Al-Harahsheh, F. Rashchi, M. Naazeri,
Review of the hydrometallurgical processing of non-sulfide zinc ores,
Hydrometallurgy. 149 (2014) 153–167.
doi:10.1016/j.hydromet.2014.08.001.
[54] L. Shi, S. Ruan, J. Li, A.R. Gerson, Desilication of low alumina to
caustic liquor seeded with sodalite or cancrinite, Hydrometallurgy. 170
(2016) 5–15. doi:10.1016/j.hydromet.2016.06.023.
- 32 -
[55] Y. Zhang, Y. Hua, X. Gao, C. Xu, J. Li, Y. Li, Q. Zhang, L. Xiong, Z.

Su, M. Wang, J. Ru, Recovery of zinc from a low-grade zinc oxide ore
with high silicon by sulfuric acid curing and water leaching,
doi:10.1016/j.hydromet.2016.08.010.
[56] R.P. Narayanan, L.-C. Ma, N.K. Kazantzis, M.H. Emmert, Cost Analysis
as a Tool for the Development of Sc Recovery Processes from Bauxite
Residue (Red Mud), ACS Sustain. Chem. Eng. 6 (2018).
[57] C. Hanahan, D. McConchie, J. Pohl, R. Creelman, M. Clark, C.
Stocksiek, Chemistry of seawater neutralization of bauxite refinery
residues (red mud), Environ. Eng. Sci. 21 (2004) 125–138.
doi:10.1089/109287504773087309.
[58] N.W. Menzies, I.M. Fulton, W.J. Morrell, Seawater neutralization of
alkaline bauxite residue and implications for revegetation., J. Environ.
Qual. 33 (2004) 1877–1884. doi:10.2134/jeq2004.1877.
[59] S. Rai, K.L. Wasewar, D.H. Lataye, J. Mukhopadhyay, C.K. Yoo,
Feasibility of red mud neutralization with seawater using Taguchi’s
methodology, Int. J. Environ. Sci. Technol. 10 (2013) 305–314.
doi:10.1007/s13762-012-0118-7.
[60] B. Cristol, R. Greenhalgh, QAL bauxite residue storage using sea water
neutralisation, in: 2nd Int. Bauxite Residue Valoris. Best Pract. Conf.,
Athens, 2018: pp. 87–97.
[61] P.M. Kopittke, N.W. Menzies, I.M. Fulton, Gypsum solubility in
seawater, and its application to bauxite residue amelioration, Aust. J. Soil
Res. 42 (2004) 953–960. doi:10.1071/SR04034.
[62] R.G. Courtney, J.P. Timpson, Reclamation of fine fraction bauxite
processing residue (red mud) amended with coarse fraction residue and
gypsum, Water. Air. Soil Pollut. 164 (2005) 91–102.
doi:10.1007/s11270-005-2251-0.
[63] A.P. Lehoux, C.L. Lockwood, W.M. Mayes, D.I. Stewart, R.J.G.
Mortimer, K. Gruiz, I.T. Burke, Gypsum addition to soils contaminated
by red mud: Implications for aluminium, arsenic, molybdenum and
vanadium solubility, Environ. Geochem. Health. 35 (2013) 643–656.
doi:10.1007/s10653-013-9547-6.
[64] R. Courtney, G. Mullen, T. Harrington, An evaluation of revegetation
success on bauxite residue, Restor. Ecol. 17 (2009) 350–358.
doi:10.1111/j.1526-100X.2008.00375.x.
[65] R. Courtney, L. Kirwan, Gypsum amendment of alkaline bauxite residue
- Plant available aluminium and implications for grassland restoration,
Ecol. Eng. 42 (2012) 279–282. doi:10.1016/j.ecoleng.2012.02.025.
- 33 -
[66] M.K. Hamdy, F.S. Williams, Bacterial amelioration of bauxite residue

waste of industrial alumina plants, J. Ind. Microbiol. Biotechnol. 27
(2001) 228–233. doi:10.1038/sj.jim.7000181.
[67] S. Khaitan, D.A. Dzombak, G. V. Lowry, Chemistry of the acid
neutralization capacity of bauxite residue, Environ. Eng. Sci. 26 (2009)
873–881.
[68] M. Grafe, G. Power, C. Klauber, Review of Bauxite Residue Alkalinity
and Associated Chemistry, 2009.
[69] L.J. Kirwan, A. Hartshorn, J.B. McMonagle, L. Fleming, D. Funnell,
Chemistry of bauxite residue neutralisation and aspects to
implementation, Int. J. Miner. Process. 119 (2013) 40–50.
doi:10.1016/j.minpro.2013.01.001.
[70] M. Kakizawa, Y. Yanagisawa, Y. Yanagisawa, A new CO2disposal
process via artificial weathering of calcium silicate accelerated by acetic
acid, Energy. 26 (2001) 341–354. doi:10.1016/S0360-5442(01)00005-6.
[71] S. Rai, K. Wasewar, A. Agnihotri, Treatment of alumina refinery waste
(red mud) through neutralization techniques: A review, Waste Manag.
Res. 35 (2017) 563–580. doi:10.1177/0734242X17696147.
[72] L.C. a Venancio, J. Antonio, S. Souza, E.N. Macedo, F.A. Botelho, G.C.
César, Pilot Test of Bauxite Residue Carbonation With Flue Gas ., TMS
Light Met. (2013) 113–118.
doi:https://doi.org/10.1002/9781118663189.ch20.
[73] D.J. Cooling, P.S. Hay, L. Guilfoyle, Carbonation of bauxite residue, in:
6th Int. Alumina Qual. Work., 2002: pp. 185–190.
[74] V.S. Yadav, M. Prasad, J. Khan, S.S. Amritphale, M. Singh, C.B. Raju,
Sequestration of carbon dioxide (CO2) using red mud, J. Hazard. Mater.
176 (2010) 1044–1050. doi:10.1016/j.jhazmat.2009.11.146.
[75] S. Khaitan, D. a. Dzombak, G. V. Lowry, Mechanisms of neutralization
of bauxite residue by carbon dioxide, J. Environ. Eng. 135 (2009) 433–
438. doi:10.1061/(ASCE)EE.1943-7870.0000010.
[76] D. Bonenfant, L. Kharoune, R. Hausler, P. Niquette, CO2 sequestration
potential of steel slags at ambient pressure and temperature, Ind. Eng.
Chem. REs. 47 (2008) 7610–7616.
[77] R.M. Rivera, G. Ounoughene, C.R. Borra, K. Binnemans, T. Van
Gerven, Neutralisation of bauxite residue by carbon dioxide prior to
acidic leaching for metal recovery, Miner. Eng. 112 (2017) 92–102.
doi:10.1016/j.mineng.2017.07.011.
[78] Y.-S. Han, S. Ji, P.-K. Lee, C. Oh, Bauxite residue neutralization with
simultaneous mineral carbonation using atmospheric CO2, J. Hazard.
Mater. 326 (2017) 87–93. doi:10.1016/j.jhazmat.2016.12.020.
- 34 -
[79] G. Alkan, B. Yagmurlu, S. Cakmakoglu, T. Hertel, Ş. Kaya, L. Gronen,

S. Stopic, B. Friedrich, Novel Approach for Enhanced Scandium and
Titanium Leaching Efficiency from Bauxite Residue with Suppressed
Silica Gel Formation, Sci. Rep. 8 (2018). doi:10.1038/s41598-018-
24077-9.
[80] Y. Liu, R. Naidu, Hidden values in bauxite residue ( red mud ):
Recovery of metals, Waste Manag. 34 (2014) 2662–2673.
doi:10.1016/j.wasman.2014.09.003.
[81] T. Chun, D. Zhu, J. Pan, Z. He, Recovery of Alumina from Magnetic
Separation Tailings of Red Mud by Na2CO3 Solution Leaching, Metall.
Mater. Trans. B. 45 (2014) 827–832. doi:10.1007/s11663-014-0023-1.
[82] O.C. Fursman, J.E. Mauser, M.O. Butler, W.A. Stickney, Utilization of
Red Mud Residues From Alumina Production, Washington, 1970.
http://www.scopus.com/inward/record.url?eid=2-s2.0-
0014872360&partnerID=tZOtx3y1.
[83] G. Li, M. Liu, M. Rao, T. Jiang, J. Zhuang, Y. Zhang, Stepwise
extraction of valuable components from red mud based on reductive
roasting with sodium salts, J. Hazard. Mater. 280 (2014) 774–780.
doi:10.1016/j.jhazmat.2014.09.005.
[84] S. Wang, H.M. Ang, M.O. Tadé, Novel applications of red mud as
coagulant, adsorbent and catalyst for environmentally benign processes,
Chemosphere. 72 (2008) 1621–1635.
doi:10.1016/j.chemosphere.2008.05.013.
[85] B.H. Rao, N.G. Reddy, Zeta potential and particle size characteristics of
red mud waste, in: Geoenvironmental Pract. Sustain., Springer Nature
Singapore Pte Ltd., Odisha, India, 2017: pp. 69–89. doi:10.1007/978-
981-10-4077-1.
[86] B. Deng, G. Li, J. Luo, Q. Ye, M. Liu, Z. Peng, T. Jiang, Enrichment of
Sc2O3and TiO2from bauxite ore residues, J. Hazard. Mater. 331 (2017)
71–80. doi:10.1016/j.jhazmat.2017.02.022.
[87] B. Bhoi, P. Rajput, C. Ranjan, PRODUCTION OF GREEN DIRECT
REDUCED IRON ( DRI ) FROM RED MUD OF INDIAN ORIGIN : A
NOVEL CONCEPT Presenter ’ s Bio, in: 35th Int. ICSOBA Conf.,
Hamburg, 2017: pp. 565–574.
[88] B.R. Parhi, S.K. Sahoo, S.C. Mishra, B. Bhoi, R.K. Paramguru, B.K.
Satapathy, Upgradation of bauxite by molecular hydrogen and hydrogen
plasma, Int. J. Miner. Metall. Mater. 23 (2016) 1141–1149.
doi:10.1007/s12613-016-1333-x.
[89] P. Tam, W. Yin, B. Xakalashe, B. Friedrich, D. Panias, Carbothermic
Reduction of Bauxite Residue for Iron Recovery and Subsequent
Aluminium Recovery from Slag Leaching, in: 35th Int. ICSOBA Conf.
- 35 -

[90] M. Samouhos, M. Taxiarchou, G. Pilatos, P.E. Tsakiridis, E. Devlin, M.
Pissas, Controlled reduction of red mud by H2 followed by magnetic
separation, Miner. Eng. 105 (2017) 36–43.
doi:10.1016/j.mineng.2017.01.004.
[91] M. Samouhos, M. Taxiarchou, P.E. Tsakiridis, K. Potiriadis, Greek “red
mud” residue: A study of microwave reductive roasting followed by
magnetic separation for a metallic iron recovery process, J. Hazard.
Mater. 254–255 (2013) 193–205. doi:10.1016/j.jhazmat.2013.03.059.
[92] C. Cardenia, B. Xakalashe, E. Balomenos, D. Panias, Reductive
Roasting Process for the Recovery of Iron Oxides from Bauxite Residue
through Rotary Kiln Furnace and Magnetic Separation, in: 35th Int.
ICSOBA Conf. Hamburg, Ger. 2 – 5 October, 2017, 2017: pp. 595–602.
[93] X. Ni, Q. Li, W. Chen, Dissolution kinetics of Si and Al from
montmorillonite in carbonic acid solution, Int. J. Coal Sci. Technol. 1
(2014) 31–38. doi:10.1007/s40789-014-0005-6.
[94] R.K. Paramguru, P.C. Rath, V.N. Misra, Trends in Red Mud Utilization
– a Review, 2004. doi:10.1080/08827500490477603.
[95] B. Yagmurlu, G. Alkan, B. Xakalashe, B. Friedrich, S. Stopic, Combined
SAF Smelting and Hydrometallurgical Treatment of Bauxite Residue for
Enhanced Valuable Metal Recovery, in: 35th Int. ICSOBA Conf.
Recovery of rare earths and major metals from bauxite residue (red mud)
by alkali roasting, smelting, and leaching, J. Sustain. Met. 3 (2017) 393–
404. doi:10.1007/s11837-016-2111-y.
[97] P.B. Queneau, C.E. Berthold, Silica in hydrometallurgy: an overview,
Can. Metall. Q. 25 (1986) 201–209.
[98] F.R. Feret, Selected applications of Rietveld-XRD analysis for raw
materials of the aluminum industry, Powder Diffr. 28 (2013) 112–123.
doi:10.1017/S088571561300016X.
[99] R.K. Iler, The chemistry of silica: solubility, polymerization, colloid and
surface properties, and biochemistry, New York, 1979.
doi:10.1002/ange.19800920433.
[100] D.J. Tobler, S. Shaw, L.G. Benning, Quantification of initial steps of
nucleation and growth of silica nanoparticles: An in-situ SAXS and DLS
study, Geochim. Cosmochim. Acta. 73 (2009) 5377–5393.
doi:10.1016/j.gca.2009.06.002.
[101] A.A. Hamouda, H.A.A. Amiri, Factors affecting alkaline sodium silicate
gelation for in-depth reservoir profile modification, Energies. 7 (2014)
- 36 -
568–590. doi:10.3390/en7020568.
[102] T.W. Zerda, I. Artaki, J. Jonas, Study of polymerization processes in acid
and base catalyzed silica sol-gels, J. Non. Cryst. Solids. 81 (1986) 365–
379. doi:10.1016/0022-3093(86)90503-X.
[103] D. Voßenkaul, A. Birich, N. Müller, N. Stoltz, B. Friedrich,
Hydrometallurgical Processing of Eudialyte Bearing Concentrates to
Recover Rare Earth Elements Via Low-Temperature Dry Digestion to
Prevent the Silica Gel Formation, J. Sustain. Metall. 3 (2017) 79–89.
doi:10.1007/s40831-016-0084-2.
[104] C.R. Borra, J. Mermans, B. Blanpain, Y. Pontikes, K. Binnemans, T.
Van Gerven, Selective recovery of rare earths from bauxite residue by
combination of sulfation, roasting and leaching, Miner. Eng. 92 (2016)
151–159. doi:10.1016/j.mineng.2016.03.002.
[105] R.P. Narayanan, N.K. Kazantzis, M.H. Emmert, Selective Process Steps
for the Recovery of Scandium from Jamaican Bauxite Residue (Red
Mud), ACS Sustain. Chem. Eng. 6 (2018) 1478–1488.
[106] E. V. Shkol’nikov, Thermodynamic characterization of the amphoterism
of hydroxides and oxides of scandium subgroup elements in aqueous
media, Russ. J. Appl. Chem. 82 (2009) 2098–2104.
doi:10.1134/S1070427209120040.
[107] K.H. Gayer, L.C. Thompson, O.T. Zajicek, The solubility of aluminum
hydroxide in acidic and basic media at 25 °C, Can. J. Chem. 36 (1968)
1268–1271.
[108] G. Alkan, B. Yagmurlu, S. Cakmakoglu, T. Hertel, Ş. Kaya, L. Gronen,
S. Stopic, B. Friedrich, Novel Approach for Enhanced Scandium and
Silica Gel Formation, Sci. Rep. 8 (2018) 5676. doi:10.1038/s41598-018-
24077-9.
[109] T.E. Amer, M.A. Mahdy, N.T. El Hazek, Application of acid pugging
and ferric salts leaching on west central Sinai uraniferous siltstone, in:
M. and P. Canadian Institute of Mining (Ed.), Int. Symp. Process Metall.
Uranium, Saskatoon, Saskatchewan, Canada, 2000: pp. 445–461.
[110] P. Davris, S. Stopic, E. Balomenos, D. Panias, I. Paspaliaris, B.
Friedrich, Leaching of rare earth elements from eudialyte concentrate by
suppressing silica gel formation, Miner. Eng. 108 (2017) 115–122.
doi:10.1016/j.mineng.2016.12.011.
[111] B. Friedrich, M. Hanebuth, S. Kruse, A. Tremel, D. Vossenkaul, Method
for opening a eudialyte mineral, Patent Number EP2995692 A1, 2016.
http://www.google.com/patents/EP2995692A1?cl=en.
- 37 -
[112] W. Duyvesteyn, System and method for recovery of scandium values

from scandium-containing ores, US patent 2012/0207656 A1, 2012.
[113] S.J. Mackowski, R. Raiter, K.H. Soldenhoff, E.M. Ho, Recovery of rare
earth elements, US 2009/0272230 A1, 2009.
[114] F. Huang, Y. Liao, J. Zhou, Y. Wang, H. Li, Selective recovery of
valuable metals from nickel converter slag at elevated temperature with
sulfuric acid solution, Sep. Purif. Technol. 156 (2015) 572–581.
doi:10.1016/j.seppur.2015.10.051.
[115] B.I. Whittington, D. Muir, Pressure Acid Leaching of Nickel Laterites: A
Review , Miner. Process. Extr. Metall. Rev. 21 (2000) 527–599.
doi:10.1080/08827500008914177.
[116] L.E. Mordberg, Patterns of distribution and behaviour of trace elements
in bauxites, Chem. Geol. 107 (1993) 241–244. doi:10.1016/0009-
2541(93)90183-J.
[117] E. Deady, E. Mouchos, K. Goodenough, B. Williamson, F. Wall, Rare
Earth Elements in Karst-Bauxites: a Novel Untapped European
Resource?, in: ERES2014 1st Eur. Rare Earth Resour. Conf. (Milos,
Greece, 4-7 Sept. 2014), 2014: pp. 364–375.
http://eres2014.conferences.gr/.
[118] Z. Li, J. Din, J. Xu, C. Liao, F. Yin, T. Lǚ, L. Cheng, J. Li, Discovery of
the REE minerals in the Wulong-Nanchuan bauxite deposits, Chongqing,
China: Insights on conditions of formation and processes, J.
Geochemical Explor. 133 (2013) 88–102.
doi:10.1016/j.gexplo.2013.06.016.
[119] G.J. Retallack, Lateritization and bauxitization events, Econ. Geol. 105
(2010) 655–667. doi:10.2113/gsecongeo.105.3.655.
[120] K. Sugita, Y. Kobayashi, Y. Taguchi, S. Takeda, Y. Ota, M. Ojiri, K.
Oda, S. Hiroshi, Method of recovering rare-earth elements, US
2015/0086449 A1, 2015. doi:US 2010/0311130 Al.
[121] G.D. Fulford, G. Lever, T. Sato, Recovery of rare earth elements from
Bayer process red mud, US5030424 A, 1991.
[122] Orbite Technologies Inc., Annual information form for the year ended
December 31, 2016., 2016.
http://s2.q4cdn.com/622589029/files/doc_financials/2016/q4/ORT-
AIF_2016-Final_SEDAR.pdf.
[123] R. Boudreault, J. Fournier, D. Primeau, M.-M. Labrecque-Gilbert,
Process for treating red mud, US 9.023.302 B2, 2015. doi:10.1016/j.(73).
[124] M.B. Cooper, Naturally Occurring Radioactive Materials ( NORM ) in
Australian Industries - Review of Current Inventories and Future
Generation, 2005.
- 38 -
[125] V. Cuccia, A.H. De Oliveira, Z. Rocha, Radionuclides in Bayer Process

Residues: Previous Analysis for Radiological Protection, in: Int. Nucl.
Atl. Conf. - Ina., Belo Horizonte, 2011.
[126] H. Gu, N. Wang, Y. Yang, C. Zhao, S. Cui, Features of distribution of
uranium and thorium in red mud, Physicochem. Probl. Miner. Process.
53 (2017) 110–120.
[127] G. Barakos, H. Mischo, J. Gutzmer, Assessments of Boundary
Conditions and Requirements for Rare Earth Underground Mining Due
To Presence of Norms, in: ERES2014 1st Eur. Rare Earth Resour. Conf.,
Milos, 2014: pp. 150–157.
[128] C.K. Gupta, N. Krishnamurthy, Extractive Metallurgy of Rare Earths,
Florida, 2005. doi:10.1007/s13398-014-0173-7.2.
[129] P. Gamaletsos, A. Godelitsas, T.J. Mertzimekis, J. Göttlicher, R.
Steininger, S. Xanthos, J. Berndt, S. Klemme, A. Kuzmin, G. Bárdossy,
Thorium partitioning in Greek industrial bauxite investigated by
synchrotron radiation and laser-ablation techniques, Nucl. Instruments
Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 269
(2011) 3067–3073. doi:10.1016/j.nimb.2011.04.061.
[130] Council of the European Union, Council Directive 2013/59/Euratom of 5
December 2013 laying down basic safety standards for protection against
the dangers arising from exposure to ionising radiation, and repealing
Directives 89/618/Euratom, 90/641/Euratom, 96/29/Euratom,
97/43/Euratom a, 2013.
[131] M. Hegedűs, E. Tóth-Bodrogi, J. Jónás, J. Somlai, T. Kovács, Mobility
of232Th and210Po in red mud, J. Environ. Radioact. 184–185 (2018)
71–76. doi:10.1016/j.jenvrad.2018.01.012.
[132] A. Goronovski, A.H. Tkaczyk, Assessment of NORM in bauxite residue
to facilitate valorization, in: Y. Pontikes (Ed.), 2nd Int. Bauxite Residue
Valoris. Best Pract. Conf., Athens, 2018: pp. 323–329.
[133] P.J. Joyce, A. Goronovski, A.H. Tkaczyk, A. Björklund, A framework
for including enhanced exposure to naturally occurring radioactive
materials (NORM) in LCA, Int. J. Life Cycle Assess. 22 (2017) 1078–
1095. doi:10.1007/s11367-016-1218-2.
[134] D. a Rubinos, M.T. Barral, Fractionation and mobility of metals in
bauxite red mud., Environ. Sci. Pollut. Res. Int. 20 (2013) 7787–802.
doi:10.1007/s11356-013-1477-4.
[135] A. Akinci, R. Artir, Characterization of trace elements and radionuclides
and their risk assessment in red mud, Mater. Charact. 59 (2008) 417–
421. doi:10.1016/j.matchar.2007.02.008.
[136] C. Nuccetelli, Y. Pontikes, F. Leonardi, R. Trevisi, New perspectives and
- 39 -
issues arising from the introduction of (NORM) residues in building

materials: A critical assessment on the radiological behaviour, Constr.
Build. Mater. 82 (2015) 323–331.
doi:10.1016/j.conbuildmat.2015.01.069.
[137] J. Somlai, V. Jobbágy, J. Kovács, S. Tarján, T. Kovács, Radiological
aspects of the usability of red mud as building material additive, J.
Hazard. Mater. 150 (2008) 541–545. doi:10.1016/j.jhazmat.2007.05.004.
[138] W.R. Pinnock, Measurements of radioactivity in jamaican building
materials and ϒ dose equivalents in a prototype red mud house, Health
Phys. 61 (1991) 647–651. doi:10.1097/00004032-199111000-00009.
[139] EC, Radiological protection principles concerning the natural
radioactivity of building materials - Radiation Protection 112, Eur.
Comm. (1999) 1–16.
- 40 -
Chapter 3 Neutralisation of bauxite residue by CO2 prior acid leaching
3. NEUTRALISATION OF BAUXITE
RESIDUE BY CARBON DIOXIDE PRIOR
TO ACIDIC LEACHING FOR METAL
RECOVERY
ABSTRACT - The present work considers the neutralisation of bauxite residue

(red mud) with CO2 as a potential technology for reducing the acid consumption
in the acidic leaching step for metal recovery. The pH of bauxite residue was
reduced during neutralisation by the transformation of hydroxide ions to
(hydrogen) carbonate ions. Neutralisation at high CO2 partial pressures and high
temperatures reduces the alkalinity of the bauxite residue, but it leads to the
stabilisation of silicate compounds such as cancrinite and grossular. After acidic
leaching of the neutralised product with sulfuric acid, a decrease by 20 wt.% in the
dissolution yield of Al, Fe, and Ti was observed, due to an insufficient amount of
acid devoted to leaching as the transformation of calcite into bassanite and the high
concentration of silicate compounds consumed part of the acid. Sc recovery by
leaching of the highly neutralised bauxite residue was about 35 wt.%, which
depends on Fe and Ti recovery. A positive correlation between La/Nd with Sc
indicates that the recovery of La/Nd indirectly depends on the extraction of Fe/Ti,
as Sc is chemically associated to these major metals.
Based on the published paper

R. M. Rivera, G. Ounoughene, C. R. Borra, K. Binnemans, T. Van Gerven,
“Neutralisation of bauxite residue by carbon dioxide prior to acidic leaching for metal
recovery”
Minerals Engineering. 112 (2017) doi:10.1016/j.mineng.2017.07. 011.
Author contributions
R. M. Rivera conceived the research, performed the analytical work, interpreted the results,
and wrote the article. C. R. Borra and G. Ounoughene contributed with part of the analytical
work. K. Binnemans and T. Van Gerven contributed with their guidance and expertise in
this work.
- 41 -
3.1 Introduction
Bauxite residue (red mud) is the waste product generated during alumina
production by the Bayer process, and it is composed essentially of non-soluble
elements such as iron and titanium minerals, un-digested alumina minerals,
sodium aluminium hydrosilicates, calcium minerals and significant amounts of
minor and trace elements such as scandium, yttrium and lanthanides, i.e. the so-
called rare earth elements (REEs) [1]. Bauxite residue is an interesting source of
REEs and especially scandium, but also of iron and titanium, while the residue
after metal recovery can be used for low-carbon building materials [2]. Scandium
concentrations in Greek bauxite residue are about 120 mg/kg [3], but it may reach
up to 260 mg/kg in some bauxite residues [4] depending on the origin and mix of
bauxite minerals (raw material). The scandium concentration in bauxite residue is
much higher than its average abundance in the Earth’s crust (22 mg/kg), and
almost twice that in the original bauxite ores [5]. Although, scandium can be found
in concentration over 100 mg/kg in natural minerals such as thortveitite
((Sc,Y)2Si2O7) and kolbeckite (ScPO4·2H2O), the abundance of these minerals is
very scarce [2,6]. Therefore, bauxite residue represents an interesting source for
the recovery of this metal. In fact, scandium represents more than 90% of the value
of the REEs present in bauxite residue.
Several methods for recovering valuable elements from bauxite residue
have been reported [4]. They are mainly based on hydrometallurgical and
pyrometallurgical processes, or combinations of both. Pyrometallurgical processes
allow a high recovery of scandium and other REEs [7,8], but the energy
consumption is high. Minor elements can be recovered by acidic leaching with
HCl, HNO3, or H2SO4 [5,7,9,10], but co-dissolution of a significant amount of iron
is a serious drawback for REE recovery [3].
The alkalinity of bauxite residue is an important issue for metal recovery
by acidic leaching, because part of the acid must be used for the neutralisation of
the alkaline products left behind after the Bayer process. This leads to a large acid
consumption and makes the recovery of metals from bauxite residue often not
economical. In order to reduce the amount of acid needed for neutralisation of
bauxite residue and, consequently, to develop a more sustainable process for metal
recovery, alternative neutralisation routes must be considered. The use of carbonic
acid (H2CO3), formed during the dissolution of CO2 in water, represents an
inexpensive and safe technology for bauxite residue neutralisation. The use of this
reagent has shown promising results in terms of bauxite residue stabilisation for
further safe disposal [11–15]. Alkaline leaching of bauxite residue with addition
of CO2 for scandium recovery has also being reported for scandium recovery [16].
The process allows the recovery of scandium due to the high solubility of Sc(OH) 3
in a solution of NaHCO3 [17]. However, the process is limited for the recovery of
other metals such as iron, aluminium, titanium and the lanthanides. The use of CO2
as a neutralisation reagent for bauxite residue as part of a flow sheet to reduce acid
consumption during the metal recovery by acidic leaching has not been studied
yet.
- 42 -
The objective of this paper is to evaluate the recovery of valuable metals

after the reduction of bauxite residue’s alkalinity by transformation of hydroxides
into soluble carbonate compounds during the neutralisation by CO 2, at different
conditions. Neutralisation with a mineral acid at similar conditions is compared to
the neutralisation with CO2. The neutralised solid fraction is leached with sulfuric
acid at different concentrations to dissolve major and minor elements, and the
metal extraction efficiency is correlated with the neutralisation conditions. The
acid consumption is also evaluated.
3.2 Material and methods
The bauxite residue studied in this paper was kindly provided by

Mytilineos S.A. - Aluminium of Greece (Agios Nikolaos, Greece). It originates
from a mixture of karst and lateritic bauxites. It was received from the alumina
refinery after dewatering by filter presses and room temperature drying. Upon
arrival in the lab, the sample was further dried at 105 ºC for 24 h (moisture content
20 – 30 wt.%). Chemical analysis of the major elements in bauxite residue was
performed using wavelength dispersive X-ray fluorescence spectroscopy
(WDXRF, Panalytical PW2400). Chemical analysis of minor elements was
performed after complete dissolution of the bauxite residue by alkali fusion and
acid digestion in 3 vol.% HNO3 solution, followed by Inductively Coupled Plasma
Mass Spectrometry (ICP-MS, Thermo Electron X Series) analysis. The alkali
fusion was carried out by mixing 0.5 g of bauxite residue with 1.5 g of sodium
carbonate and 1.5 g of sodium tetraborate decahydrate, followed by heating the
mixture in a platinum crucible at 1100 °C for 30 min. The mineralogy of the
samples was studied by X-ray powder diffraction (XRD, Bruker D2 Phaser). The
obtained data were evaluated with EVA V.3.1 (Bruker AXS) and quantified with
Topas-Academic V.5, using the Rietveld method. Thermo-gravimetric analysis
(TGA, Netzsch STA 409) were carried out in a nitrogen atmosphere from room
temperature to 1000 ºC at a heating rate of 10 ºC min-1 in order to identify phase
transformation and thermal decompositions in the bauxite residue.
Neutralisation at ambient conditions of bauxite residue with CO2 gas was
performed in a three-neck flask reactor (400 mL) with continuous agitation
(agitation at 478 rpm). The agitation was performed by a magnetic stirrer (4 cm
long, 7.5 mm diameter, 6 g). Within the reactor, solid particles were mixed with
water at a liquid-to-solid ratio, L/S, of 5:1. CO2 gas (99.9925%, Air Liquide) was
introduced continuously at defined flow rate, which was controlled with a flow
meter. A diffuser (PTFE, pore size of 10 µm) was used to spread out the gas into
the formed slurry.
High-pressure neutralisation experiments were carried out in a titanium
autoclave (Parr Company, series 4560) varying the partial pressure of CO 2 and
temperature. Bauxite residue and Milli-Q water were mixed during a period of 5
minutes in a beaker at a L/S of 5:1 at ambient conditions. The slurry was poured
into the titanium vessel (200 mL) and mounted in the autoclave. Nitrogen gas at a
- 43 -
pressure of 5 bar was used to remove the excess of oxygen gas inside the reactor,
so that potential oxidative reactions could be prevented.
Neutralisation of bauxite residue with HCl (37%, Fisher Scientific) and
H2SO4 (95 – 97%, Sigma-Aldrich) was performed with a L/S ratio of 5:1 in sealed
polyethylene bottles by constant agitation using a laboratory shaker (Gerhardt
Laboshake) at 200 rpm and 25 °C. The slurries after neutralisation were filtered
using filter paper (pore size 0.45 µm) and diluted with 2 vol.% HNO 3 for
Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES,
PerkinElmer Optima 8300) analysis for major elements (e.g., Al, Fe, Ca, Na, Si,
Ti). Leaching of the non-neutralised and the neutralised bauxite residue was
carried out with sulfuric acid (95 – 97%, Sigma-Aldrich) with a L/S ratio of 10:1.
The experiments were performed in the same laboratory shaker at 200 rpm and
25 ºC. The leach solution (pregnant leach solution, PLS) was filtered using a
syringe filter (pore size of 0.45 µm) and diluted with 2 vol.% HNO 3 for ICP-OES
analysis for major and minor (e.g., Sc, Y, La, Nd) elements. Lutetium was used as
internal standard during the analysis. A synthetic solution with a pre-defined
concentration was used as reference. The distribution of major and minor elements
was detected using an electron microprobe for microanalysis (EPMA, type JXA-
8530F from Jeol Ltd.).
All the experiments, i.e. neutralisation and leaching, were carried out in
triplicate to ensure reproducibility of the results. The errors were determined as
the standard deviations on the results.
3.3 Results and discussion
Characterisation of the bauxite residue

The concentrations of major elements in the form of oxides in our Greek
bauxite residue sample are given in Table 3.1. Iron(III) oxide is the major oxide in
the bauxite residue followed by alumina, calcium oxide, silica, titanium oxide and
sodium oxide. Other oxides may also be present in bauxite residue at low
concentration (e.g. oxides of arsenic, beryllium, cadmium, chromium, copper,
gallium, lead, manganese, mercury, nickel, potassium, thorium, uranium,
vanadium, zinc, zirconium, and rare earth elements), as well as non-metallic
elements such as phosphorus, carbon and sulfur. In Table 3.2, the concentrations
of selected REEs are shown. The concentrations of Sc, Y, La and Nd are shown
because their concentrations in this bauxite residue are relatively high.
- 44 -
Table 3.1: Major chemical components, expressed as oxides, in the bauxite residue.
Compound wt.%
Fe2O3 46.7
Al2O3 18.1
CaO 8.5
SiO2 7.3
TiO2 5.8
Na2O 2.8
Loss on ignition 8.5
Table 3.2: Selected rare-earth elements composition after complete dissolution of the
bauxite residue by alkali fusion and acid digestion in 3 vol.% HNO3 solution [3].
Element Concentration,
g tonne-1
Sc 121 ± 10
Y 76 ± 10
La 114 ± 15
Nd 99 ± 7
Mineralogical analysis allowed the identification of several mineral

phases, based on iron (hematite, goethite), aluminium (gibbsite, diaspore,
bayerite), calcium (calcite, calcium silicates and calcium alumino-silicates),
sodium (sodalite, cancrinite) and titanium (rutile). Perovskite (CaTiO 3) was
detected only in very low concentration (below 1 wt.%). Approximately 95% of
the mineral phases were quantified using the Rietveld method (see Table A-I.1 and
Table A-I.2 in the appendix). Table 3.3 describes the normalised (to 100 wt.%
total) mineralogical composition of the bauxite residue sample. The quantification
of mineral phases was analysed by considering the chemical composition of major
elements obtained by XRF analysis (Table 3.1) with an estimated deviation error
of approximately 10%.
- 45 -
Table 3.3: Normalised mineralogical composition of the bauxite residue sample.
Phase Chemical formula wt.%
Hematite Fe2O3 36
Dicalcium-silicate (C2S) Ca2SiO4 12
Gibbsite Al(OH)3 8
Hydrogrossular Ca3Al2 (SiO4)1.53(OH)5.88 7

Diaspore AlO(OH) 7
Grossular Ca3Al2(SiO4)3 5
Tricalcium-silicate (C3S) Ca3SiO5 5

Goethite FeO(OH) 5
Calcite CaCO3 4
Bayerite -Al(OH)3 4
Sodalite Na6(Al6Si6O24)·xNaOH·(8−2x)H2O 3
Calcium phyllo-dodeca-
Al12CaO27Si4 2
alumotetrasilicate
Rutile TiO2 2
Cancrinite Na6Ca2Al6Si6O24(CO3)2·2H2O 1
Neutralisation - leaching
Neutralisation of bauxite residue by the early removal of hydroxide
compounds may help to decrease the amount of acid during the further recovery
of metals via acidic leaching. The effect of neutralisation of bauxite residue with
water and CO2 gas on the acid consumption and the recovery of selected REEs
from bauxite residue was evaluated in two consecutive experimental processes
according to the diagram presented in Figure 3.1: (1) neutralisation of bauxite
residue slurry with (and without) CO2 gas, and (2) acidic leaching of the (non-)
neutralised bauxite residue.
- 46 -
H2O
Bauxite
CO2
residue
NEUTRALISATION Leachate
Neutralised BR
H2SO4 (solid fraction)
LEACHING Residue
Pregnant leach
solution (PLS)
Figure 3.1: Flow sheet for neutralisation of bauxite residue, followed by metal recovery
from neutralised bauxite residue by leaching with sulfuric acid.
Neutralisation of bauxite residue with water

Bauxite residue was washed with water to study the potential change in
pH of bauxite residue with the removal of sodium hydroxide in absence of CO 2,
i.e. water leaching of bauxite residue. Prior removal of sodium hydroxide by water
would contribute to a reduction of the amount of acid needed to merely neutralise
the high pH. The experiment was conducted in duplicate at a L/S ratio of 5:1. After
a washing period of 24 h, the pH remained stable at 10.7, which corresponded to
the original pH of our bauxite residue sample. This is due to the buffering activity
of alkaline solids that are present in the sample [18]. It was observed that about 8
wt.% of the total sodium was removed from the solid matrix after washing, due to
its chemical association to alumina-silicate compounds, such as sodalite and
cancrinite (see Table 3.3). It has been reported that the solubility of alumino-
silicates decreases with increasing presence of sodium carbonate [19,20].
However, there is a knowledge gap related to the solubility of silicate compounds
remaining after the final separation stages of the Bayer process, so that it is
difficult to assess the conditions at which such compounds become more or less
soluble.
Water leaching of bauxite residue was performed with ultra-pure water
(i.e. Milli-Q, 18.2 MΩ·cm), by which only Na dissolution was favoured. Several
authors have reported promising results by reducing bauxite residue’s alkalinity
with seawater [11,12,21]. Seawater neutralisation allows the removal of alkaline
anions from the solutions as precipitates due to the presence of Ca2+ and Mg2+ ions.
However, this technique has a more useful application for a safe disposal of the
remaining precipitates rather than a further processing for metal recovery, due to
the fact that the entrained liquor after filtration can be disposed of with a much
reduced alkalinity. Furthermore, the accumulation of precipitates in pipes, such as
calcite, dawsonite (NaAlCO3(OH)2), dolomite (CaMg(CO3)2), muscovite
- 47 -
(KAl2(AlSi3O10)(F,OH)2), among others [22] represent a significant drawback for

an in-situ application of seawater neutralisation [21].
Neutralisation of bauxite residue with different CO2 gas flow rates

The studies were carried out in a reactor of 400 mL with gas flow rates
of 0.008, 0.02, 0.1 and 0.25 L/min. The initial pH of the slurry, i.e. 10.7, decreased
over a period of 2 hours with continuous gas blown through the reactor to an
average value of 6.8 for all investigated flow rates (Figure 3.2). The most
remarkable changes were observed in the neutralisation rate, which was evaluated
by considering the slope in the linear portion within the first 5 to 15 minutes of
neutralisation on the corresponding curves in Figure 3.2. High gas flow rates lead
to a high neutralisation rate due to a large amount of CO 2 injection (Figure 3.3),
but not to an improvement of the amount of CO2 sequestration due to the limit
concentration of calcium. It has been reported earlier that CO 2 sequestration is
enhanced by the precipitation of calcite, but is restricted by the calcium
concentration in the bauxite residue [23].
A threshold gas flow rate between 0.1 and 0.25 L/min was observed,
above which a further increase in flow rate did not lead to an increase in
neutralisation rate (the gas flow rate of 0.25 L/min is not shown in Figure 3.2 due
to an overlap with the results for the gas flow rate of 0.1 L/min). In our study the
maximum neutralisation rate was reached at a gas flow rate of 0.1 L/min.
12
0.008 L/min 0.02 L/min 0.1 L/min
10
pH
4
0 20 40 60 80 100 120 140 2880
Time, min
Figure 3.2: Effect of CO2 gas on the pH during the neutralisation of the bauxite residue
sample (T: 25 ºC, L/S: 5).
The neutralisation by CO2 gas in water involves a number of chemical

reactions. The first chemical reaction is the dissolution of CO 2(g) in water to form
CO2(aq) (Eq. 3.1), which reacts with water to form carbonic acid. The concentration
- 48 -
of CO2 in aqueous solution, i.e. CO2(aq), can be determined from Henry’s law (Eq.
3.2).
Eq. 3.1
CO2(g) CO2(aq)
Eq. 3.2
[CO2(aq)] =KCO2 (PCO2(g))
where KCO2 represents Henry’s law constant of CO2, equal to

3.38 × 10-2 mol L-1 atm-1 [24], and PCO2(g) is the partial pressure (in atm) of CO2
gas in the atmosphere. At ambient conditions with a constant PCO2, an increase of
the gas-flow rate of CO2 led to an increase of the mass flow rate, which increased
the kinetics of CO2 dissolution. Thus, larger CO2-dissolution led to higher
neutralisation rate until an equilibrium state was reached at a pH  6.4. The
interaction between CO2 and water leads to the formation of carbonic acid H2CO3,
which is a weak acid (H2CO3/HCO3-, pKa1 = 6.35; HCO3-/CO32-, pKa2 = 10.2 at
25 °C). For instance, the final pH attained after 2 hours of neutralisation with CO2
gas corresponds to the pH associated to the first chemical decomposition of the
carbonic acid, i.e. the formation of the hydrogen carbonate ion, HCO 3-, when the
gas was introduced continuously into the reactor.
The pH of the neutralised sample was measured for 2 days after the end
of the experiment (no CO2 was blown inside the reactor, and the slurry was left at
ambient conditions and in steady state). An increase of the pH to an average value
of 9.3 was observed (Figure 3.2). The pH increase was caused by the release of
CO2 from the slurry into the atmosphere after the neutralization by bubbling was
seized, which decreased the level of acidity in the slurry. Hence, the pH started
increasing until a new equilibrium state was achieved at a pH ~ 9.3, where alkaline
anions in solution such as Al(OH)4- (pKa  10.2) and OH- (pKw = 14) buffer the
pH of the slurry [18, 27].
In bauxite residue, alkalinity is mainly stored in compounds such as
hydrogrossular and sodalite, which have a buffer capacity between pH 11 and 8
[29], respectively. Hence, at pH around 6.8, the decomposition of such compounds
may take place with the partial dissolution of Na and Ca. The dissolution of the
major elements Al, Fe, Ti and Si starts occurring at pH values below 6 [3].
According to Figure 3.3, about 10 to 12 wt.% of the total sodium content was
dissolved. However, as already was described in section 3.3, about 8 wt.% of
sodium is readily soluble in water, therefore only an additional 2 to 4 wt.% of
sodium reacts with the carbonic acid. During the carbonation process (dissolution
of CO2 gas in water), the carbonate and hydrogen carbonate ions may react with
Na+ ions present in solution to form Na2CO3 and/or NaHCO3 according to
reactions 6 and 7. However, these two compounds have a relatively high solubility
in water (30.7 and 8.7 g/100 g water at 25 ºC, respectively). Most of the sodium
remains in the solid matrix mainly associated to alumina-silicates compounds (see
Table 3.3).
- 49 -
2Na+ + HCO3- Na2CO3(aq) + H+ Eq. 3.3
Na+ + HCO3-  NaHCO3(aq) Eq. 3.4
The carbonate and hydrogen carbonate ions may also react with Ca2+ ions
according to Eq. 3.5 and 3.6. CaCO3 tends to precipitate due to its low solubility
(1.4×10-3 g/100 g water at 25 ºC). The solubility of CaHCO 3 is relatively high
(solubility of CaHCO3: 16.6×10-3 g/100 g water at 25 ºC), causing the elements to
remain in solution. However, the chemical association of calcium to silicate
compounds limits the concentration of Ca2+ ions. Consequently, less than 5 wt.%
of Ca was dissolved during neutralisation (Figure 3.3).
Ca2+ + HCO3- CaCO3(s) + H+ Eq. 3.5
Ca2+ + 2HCO3- Ca(HCO3)2(aq) Eq. 3.6
Na dissolution CO2
15 1500
Na and Ca dissolution, wt%
CO2 injected, g CO2/g BR
Ca dissolution
10 1000
5 500
0 0
0.008 0.020 0.100 0.250
Gas flow rate, L/min
Figure 3.3: Effect of gas flow rate on the dissolution of Na and Ca from the neutralised
bauxite residue and CO2 injected. The error bars represent the standard error of the mean
for experiments performed in triplicate (BR: bauxite residue, T: 25 ºC, L/S: 5, t: 2 h).
CO2 neutralization was compared with the use of strong acids for
neutralization in view of a substantial reduction of the material’s alkalinity as
function of the acid concentration [30]. Neutralisation with HCl and H2SO4
demonstrated a significant reduction of bauxite residue’s alkalinity in the pH range
between 6 and 7. However, the amount of calcium dissolution was much lower
during neutralisation with H2SO4 than with HCl, due to the partial transformation
of calcite into less soluble bassanite (CaSO4). HCl allows a high dissolution of Na
and Ca. Figure 3.4 describes the effect of HCl concentration on the pH and on Na
- 50 -
and Ca dissolution in the bauxite residue sample. A higher dissolution of Na and

Ca in comparison to neutralisation with carbonic acid was observed. However, it
is necessary to highlight the significant difference in the corresponding cost
associated to the use of mineral acids or CO2 for neutralisation. The price of the
acid may reach a value of up to 200 USD/tonne (www.sunsirs.com), while the use
of CO2 can be a less expensive reagent, because it can be obtained from the gas
emissions generated during the alumina production (Bayer process). In addition,
the use of CO2 for neutralisation may help to reduce the greenhouse effect due to
its removal from the industrial flue gases.
8 100

Na Ca pH
80
7
60
pH
6
40
5
20
4 0
0 10 20 30 40 50 60
Acid concetration, g/L
Figure 3.4: Neutralisation of bauxite residue with mineral acid and Na and Ca dissolution
at different acid concentration. The error bars represent the standard error of the mean for
experiments performed in triplicate (HCl, T: 25 ºC, L/S: 5, t: 24 h).
High-pressure neutralisation
The effect of CO2 partial pressure on the dissolution of sodium and
calcium and on pH, during neutralisation of the bauxite residue, is shown in Figure
3.5. The increase of CO2 partial pressure leads to an increase of sodium
dissolution, and to a very small increase in the dissolution of calcium due to the
relatively high solubility of CO2(g) in water (0.14 g CO2/100 g water at 25 °C). The
pH of the slurry remained around 9.2. Figure 3.6 describes the effect of
temperature during neutralisation of bauxite residue with 30 bar of CO2 partial
pressure. A significant increase in sodium dissolution was observed when the
temperature increased. The concentration of calcium in solution decreased due to
the precipitation of calcite when the temperature was increased. The solubility of
CO2 in water decreases with an increase in temperature at constant partial pressure
of the CO2 in the gas phase above the bulk liquid, but it does increase when its
partial pressure above the solution increases and, consequently, the concentration
of HCO3-/CO32- also increases. This relationship can be summarised by combining
Le Châtelier’s principle, i.e. the shift of the chemical equilibrium in the direction
- 51 -
at which the pressure is released, and Henry’s law, i.e. the increase in
concentration of the gas in the liquid when the partial pressure increases (see Eq.
3.2). The effect of temperature on the solubility reduction of CO 2 becomes less
significant when the CO2 partial pressure is increased [31]. The decrease in
alkalinity with the increase of temperature could be caused by the reduction of
sodalite concentration (see Figure 3.8)1.
20 11.0
Na
Ca 10.5
15 pH
10.0
pH
10
9.5
5 9.0
8.5
0
0 5 10 15 20 25 30 35
CO2 partial pressure, bar
Figure 3.5: Dissolution of Na and Ca during neutralisation at different CO2 partial pressures.
The error bars represent the standard error of the mean for experiments performed in
triplicate (T: 25 ºC, L/S: 5, t: 2 h).
1
The presence of some mineral phases, such as sodalite, dicalcium silicate and tricalcium
silicate, could not be confirmed by the XDB software (see Chapter 4). This software enables
a more accurate quantification of the mineralogical phases compared to the Rietveld method
because it has a specialized reference database for bauxite and bauxite residue minerals. In
this chapter, the mineral phases were identified by using EVA V.3.1 and quantified with
Topas-Academic V.5 using the Rietveld method. The method is based on the use of scale
factors for refinement as an index of the mineral concentration, which may induce
misinterpretation of the different mineral phases due to overlapping of specific peaks (2 -
position).
- 52 -
40 11.0
Na

10.5
30 Ca
pH 10.0
pH
20
9.5
10 9.0
8.5
0
25 50 75 100 125 150 175
Temperature, ºC
Figure 3.6: Dissolution of Na and Ca during neutralisation at different temperatures. The

error bars represent the standard error of the mean for experiments performed in triplicate
(PCO2: 30 bar, L/S: 5, t: 2 h).
The identification of phase transformation and thermal decomposition in

the non-neutralised and neutralised bauxite residue was confirmed by thermo-
gravimetric analysis (Figure 3.7). The mass loss between 270 and 600 ºC is due to
the loss of chemically bound water, mainly from goethite (between 292 and
355 °C) [32], gibbsite (between 300 and 400 ºC) and diaspore (between 400 and
600 ºC) [33]. The first endothermic peak on the differential thermogravimetry
(DTG) curve describes the dehydration of gibbsite to form boehmite, i.e. between
270 and 320 ºC [34]. The dehydration peak of gibbsite is more intense in the highly
neutralised sample than in the non-neutralised one, presumably due to its chemical
transformation into grossular when Ca and Na were dissolved [35], i.e. during the
carbonation period. The second endothermic reaction is observed at 510 ºC and it
is attributed to the decomposition of diaspore [36]. The mass loss between 600 and
800 ºC was much more pronounced in the neutralised sample compared to the non-
neutralised bauxite residue, due to the release of relatively more CO 2 from the
neutralised sample as a consequence of the decomposition of CaCO3 [3,33] formed
during neutralisation.
- 53 -
100 0.10
BR
98 NBR 150 °C, PCO2: 30 bar
96
94
92
DTG, mg/°C
Mass, wt%
90 0.08
DTG-NBR 150 °C, PCO2: 30 bar 0.05
88
DTG-BR
80 0.04
0.03
0.02
0.01
0 0.00
0 200 400 600 800 1000
Temperature, ºC
Figure 3.7: TGA (left y-axis) and DTG (right y-axis) curves of the non-neutralised (BR)
and neutralised (NBR) bauxite residue sample.
Figure 3.8 describes the change in composition of selected mineralogical

phases. The composition of calcite, C2S, cancrinite, and grossular are shown as
their concentration significantly changed as neutralisation conditions were
intensified (Table A-I.1). The calcite content increases when the CO2 partial
pressure and temperature increased. Nonetheless, the amount of CO 2 consumed
during the formation of calcite at high pressure and temperature registered only
1 wt.%. Formation of calcite may be explained by the decomposition of C 2S, and
also by the partial decomposition of cancrinite [29]. However, strong
neutralisation conditions (temperature and pressure) lead to an increase in the
concentration of cancrinite, presumably as a consequence of the partial sodalite
decomposition [37]. Snars and Gilkes reported that sodium in bauxite residue is
mostly precipitated in sodalite, rather than in cancrinite, while the dissolution of
calcium from calcite and silicon from sodalite is much more significant at pH
values lower than 8 [38]. Silicate compounds such as cancrinite and grossular
(Table A-I.2) became more stable in the neutralised samples rather than in the non-
neutralised bauxite residue, due to the decrease of SiO2 solubility with the increase
of Na2CO3 concentration in the slurry [20].
- 54 -
15
Calcite C2S Grossular
Cancrinite Sodalite
Composition, wt%
10
0
BR NBR HPNBR HPTNBR
Condition
Figure 3.8: Change of composition in selected mineralogical phases. BR: non-neutralised

bauxite residue, NBR: neutralised bauxite residue at ambient conditions
(qCO2: 0.25 L min-1, T: 25 ºC, L/S: 5), HPNBR: high-pressure neutralised bauxite residue
(PCO2: 30 bar, T: 25 ºC, L/S: 5), HPTNBR: high-pressure and high-temperature neutralised
bauxite residue (PCO2: 30 bar, T: 150 ºC, L/S: 5). The error bars represent the standard error
of the mean for experiments performed in triplicate.
Leaching of neutralised bauxite residue

Figure 3.9 shows the effect of acid concentration on the pH of the solution
for both neutralised and non-neutralised bauxite residue. As expected, the pH
decreases and the acid consumption increases when the acid concentration is
increased. However, neutralised samples achieved lower pH values compared to
the non-neutralised material. Furthermore, the sample neutralised with high
pressure and temperature registered the lowest pH values, even with the same acid
consumption as for the non-neutralised samples. This is due to the lower pH of the
sample after neutralisation with high pressure and temperature. As it was already
described in the previous section, during the neutralisation step, high CO 2
concentration enhances the reduction of the sample alkalinity as a consequence of
a high dissolution of NaOH. Figure 3.10 reports the pH values of the sample with
and without neutralisation before acidic leaching.
- 55 -
BR NBR HPNBR HPTNBR

4 Acid consumption 1000
kg H2SO4/tonne BR
800
Acid consumption,
3
600
pH
2
400
1
200
0 0
20 30 40 50 60 70 80 90
Acid concetration, g/L
Figure 3.9: Effect of acid concentration (H2SO4) on the pH in the non-neutralised and
neutralised bauxite residue sample, and acid consumption (T: 25 ºC, L/S: 10, t: 24 h). BR:
non-neutralised bauxite residue, NBR: neutralised bauxite residue at ambient conditions
(qCO2: 0.25 L/min, T: 25 ºC, L/S: 5), HPNBR: high-pressure neutralised bauxite residue
(PCO2: 30 bar, T: 25 ºC, L/S: 5), HPTNBR: high-pressure and high-temperature neutralised
bauxite residue (PCO2: 30BR:bar, T: 150sample
Non-neutralized ºC, L/S: 5).
NBR: Neutralization at ambient conditions (q : 0.25 L/min, T: 25 ºC)
CO2
HPNBR: High pressure neutralization (PCO2: 30 bar , T: 25 ºC)
HPTNBR: High pressure & temperature neutralization (PCO2: 30 bar , T: 150 ºC)
12
10
pH
4
BR NBR HPNBR HPTNBR
Condition
Figure 3.10: Initial pH of the non-neutralised bauxite residue and the resulting pH of the
bauxite residue after neutralisation. BR: non-neutralised bauxite residue, NBR: neutralised
bauxite residue at ambient conditions (qCO2: 0.25 L min-1, T: 25 ºC, L/S: 5), HPNBR: high-
pressure neutralised bauxite residue (PCO2: 30 bar, T: 25 ºC, L/S: 5), HPTNBR: high-
pressure and high-temperature neutralised bauxite residue (PCO2: 30 bar, T: 150 ºC, L/S: 5).
triplicate.
The effect of acid concentration on the leaching of Al, Fe and Ti is shown

in Figure 3.11. Neutralisation at ambient conditions did not result in an
improvement of the extraction efficiencies. Indeed, the major metal recoveries
- 56 -
from the neutralised sample with high calcite and cancrinite concentration, i.e.
samples neutralised with high CO2 partial pressure, were the lowest with higher
acid consumption compared to the non-neutralised sample. This is due to the lack
of acid available for leaching, caused by the transformation of calcite into
bassanite and the partial decomposition of silicate compounds. As it is shown in
Figure 3.12, a further increase in the extraction of Al, Fe or Ti from the non-
neutralised bauxite residue is limited by about 65 wt.% of Si dissolution (Figure
3.12a), while the extraction from the highly neutralised sample is limited by
50 wt.% of Si dissolution (Figure 3.12b), even upon further increasing the acid
concentration. According to the literature, an increase in silica dissolution leads to
a larger supersaturation of monomeric silicic acid, Si(OH) 4, which tends to
polymerise as long it is formed during the dissolution of bauxite residue with
diluted H2SO4 [39]. The difference in the amount of dissolved silica from both
samples, i.e. non- and neutralised bauxite residue, may suggest that precipitation
of silica occurs faster during leaching of highly neutralised sample than during
leaching of non-neutralised bauxite residue, due to the high concentration of
silicate compounds such as cancrinite and grossular, which were enriched during
neutralisation with CO2. It is well known that polymerisation of silica represents a
common problem during leaching of silicate minerals [40–42], which may be
promoted by low pH values and high temperatures [43], resulting in problems of
solid-liquid separation that increase the required acid consumption and affect the
selective recovery of metals. Although, the formation mechanism of silica gel
formation has been studied in detail elsewhere [39,43,44], the polymerization of
silica in acid leach solutions of bauxite residue has not been thoroughly
investigated yet.
40
(a)
30
Al extraction, wt%
20
BR
10 NBR
HPNBR
HPTNBR
0
0 10 20 30 40 50 60 70 80 90
Acid concentration, g/L
- 57 -
5
(b)
4
Fe extraction, wt% 3
2 BR
NBR
1 HPNBR
HPTNBR
0
0 10 20 30 40 50 60 70 80 90
50
(c)
40
Ti extraction, wt%
30
BR
20
NBR
10 HPNBR
HPTNBR
0
0 10 20 30 40 50 60 70 80 90
Figure 3.11: Effect of acid concentration on the extraction of (a) Al, (b) Fe and (c) Ti
(H2SO4, L/S: 10, T: 25 ºC, t: 24 h). BR: non-neutralised bauxite residue, NBR: neutralised
bauxite residue at ambient conditions (qCO2: 0.25 L/min, T: 25 ºC, L/S: 5), HPNBR: high-
pressure neutralised bauxite residue (PCO2: 30 bar, T: 25 ºC, L/S: 5), HPTNBR: high-
pressure and high-temperature neutralised bauxite residue (PCO2: 30 bar, T: 150 ºC, L/S: 5).
triplicate.
- 58 -
40 4
(a)
Al & Ti extraction, wt%

Al Fe Ti
Fe extraction, wt%
30 3
20 2
10 1
0 0
0 20 40 60 80
Si dissolution, wt%
40 3
(b)
Fe extraction, wt%
30 Al Fe Ti
2
20
1
10
0 0
0 10 20 30 40 50 60
Si dissolution, wt%
Figure 3.12: Extraction behaviour of Al, Ti and Fe as function of Si dissolution after acidic
leaching of (a) raw bauxite residue and (b) bauxite residue sample neutralised at 150 °C and
PCO2: 30bar (H2SO4, L/S: 10, T: 25 ºC, t: 24 h). The error bars represent the standard error
Figure 3.13 describes the XRD pattern of non-neutralised and neutralised

bauxite residue samples after acidic leaching. The most notorious effect of silica
gel polymerisation is observed in the large decrease of Al dissolution upon gibbsite
precipitation. Cancrinite was not identified after leaching, presumably because it
is totally decomposed by sulfuric acid.
- 59 -
3000 Leached BR 1
5 1 Hematite 6 Bassanite (Ca2(SO4)2 * H2O)
7 1 2 Diaspore 7 Gibbsite
3 Perovskite (CaTiO3) 8 Goethite
2000 4 C2S 9 C3S
5 Grossular
2
8 1
9 5
5 6 4
1000 6
2 2 6 8
3 2 4
9 7 8 4 4 3
Counts
0
3000 Leached HPTNBR 1
5
1
2000
2 8
5 9 5 1
6
1000 6 4 8
4 6
79 7 8 3 2 2 2
4 3 4
0
20 30 40 50 60 70
Position 2θ (Copper)
Figure 3.13: XRD pattern of non-neutralised and neutralised bauxite residue sample after acidic leaching (1.6 N H2SO4). BR: non-neutralised bauxite
residue, HPTNBR: high-pressure and high-temperature neutralised bauxite residue (PCO2: 30 bar, T: 150 ºC, L/S: 5).
- 60 -
Figure 3.14 shows the effect of acid concentrations on the leaching of

selected REEs. The dissolution of Sc and Y increased with increasing acid
concentration (decrease in pH) and, both elements reached a maximum extraction
yield when the acid concentration was of 35 g/L. A further increase in acid
concentration, does not lead to an improvement in the recovery of Sc and Y. The
lowest recovery of Sc and La were observed in the sample neutralised at high
pressure and temperature. The maximum extraction of Y was about 60 wt.%,
which was independent of neutralisation conditions. At acid concentration lower
than 40 g/L, the extraction of Nd was slightly reduced as the neutralisation
conditions were intensified. The leaching of La and Nd is determined by the
amount of free H+ ions in solution, which are consumed during the dissolution of
major elements and during the decomposition of silicate compounds [9].
Therefore, the extraction of La and Nd could be enhanced by increasing the acid
concentration. All the acid leaching experiments were performed at ambient
conditions, i.e. 25 °C, but the leaching efficiency can be enhanced by increasing
the temperature.
50 (a)
40
Sc extraction, wt%
30
BR
20 NBR
HPNBR
10
HPTNBR
0
0 10 20 30 40 50 60 70 80 90
80
(b)
60
Y extraction, wt%
40 BR
NBR
20 HPNBR
HPTNBR
0
0 10 20 30 40 50 60 70 80 90
- 61 -
40 (c)
La extraction, wt.% 30
20
BR
NBR
10 HPNBR
HPTNBR
0
0 10 20 30 40 50 60 70 80 90
40 (d)
Nd extraction, wt.%
30
20
BR
NBR
10 HPNBR
HPTNBR
0
0 10 20 30 40 50 60 70 80 90
Figure 3.14: Effect of acid concentration on extraction of selected REEs (H2SO4, L/S: 10,
T: 25 ºC, t: 24 h). BR: non-neutralised bauxite residue, NBR: neutralised bauxite residue at
ambient conditions (qCO2: 0.25 L/min, T: 25 ºC, L/S: 5), HPNBR: high-pressure neutralised
bauxite residue (PCO2: 30 bar, T: 25 ºC, L/S: 5), HPTNBR: high-pressure and high-
temperature neutralised bauxite residue (PCO2: 30 bar, T: 150 ºC, L/S: 5). The error bars
represent the standard error of the mean for experiments performed in triplicate.
A chemical correlation between the dissolution of Sc and Y with the

extraction of major metals after leaching of the highly neutralised sample is shown
in Figure 3.15. Although Y recovery was not affected by the neutralisation (Figure
3.14), the extraction of both Sc and Y seems to be limited by the dissolution of
major metals. Urbain and Sarkar (1927) were the first to report on the chemical
association between Sc and Fe, and Borra et al. reported that up to 50% of Sc can
be extracted from bauxite residue without bringing a significant dissolution of Fe
into the solution [3], which is confirmed by our results. Furthermore, it has been
reported that Sc is chemically associated to hematite and goethite minerals [45].
A similar association trend as between Sc and Fe seems to exist between Sc and
Ti [46]. These correlations were confirmed by EPMA analysis as it can be seen in
Figure 3.17. Sc was found to be concentrated in Fe-rich phases, i.e. hematite and
- 62 -
goethite, and also partially associated with Ti-rich phase, i.e. rutile. The
dissolution of Sc resulted to be independent of Al dissolution. Despite the
existence of a chemical correlation between Y and Fe/Ti (Figure 3.15b), such
correlation could not be confirmed by EPMA analysis because Y concentration in
bauxite residue is very low with a heterogeneous distribution in the sample.
Nevertheless, in the association with Fe and Ti dissolution, the extent of Y
leaching is 1.5 times higher than that of Sc leaching. No chemical associations of
Sc and Y with major elements other than Fe have been reported so far in literature.
Up to 40 wt.% of Sc and 60 wt.% of Y is dissolved with negligible
amount of Fe in solution, i.e. iron recovery was less than 3 wt.%. Besides iron
dissolution was low, its concentration is high (about 900 ppm) compared to REEs
concentration (about 5 ppm) due to the high concentration in this particular bauxite
residue.
40 3
(a)
Fe
Al
Fe extraction, wt%
30
Ti 2
20
1
10
0 0
0 5 10 15 20 25 30 35 40
Sc extraction, wt%
40 3
Fe (b)
Al
Fe extraction, wt%
30
Ti
2
20
1
10
0 0
0 10 20 30 40 50 60
Y extraction, wt%
Figure 3.15: Extraction behaviour of (a) Sc and (b) Y as function of Al, Ti and Fe
dissolution after acidic leaching of the bauxite residue sample neutralised at 150 °C and
PCO2: 30 bar (H2SO4, L/S: 10, T: 25 ºC, t: 24 h). The error bars represent the standard error
- 63 -
The extraction behaviour of La and Nd from the highly neutralised

sample is shown in Figure 3.16 as function of Al, Fe and Ti dissolution. The
correlation between the extraction of La or Nd and the dissolution of the major
elements is less clear. Again, above a certain threshold of La or Nd dissolution
(between 0 and 12 wt.% in our results), no relation between La or Nd leaching
exists with Al dissolution. A similar behaviour can be even suggested for Fe and
Ti, where La and Nd leaching is independent from Fe and Ti dissolution from ca.
20 wt.% extraction of La or Nd. However, EPMA analysis demonstrated a close
correlation between La/Nd with Sc (Figure 3.17), which may suggest that the
recovery of La/Nd indirectly depends on the extraction of Fe/Ti, as Sc is
chemically associated to both major metals.
40 3
(a)
Fe extraction, wt%
30
2
20
Fe
1
10 Al
Ti
0 0
0 5 10 15 20 25 30
La extraction, wt%
40 3
(b)
Fe
Al
Fe extraction, wt%
30
Ti 2
20
1
10
0 0
0 5 10 15 20 25 30
Nd extraction, wt.%
Figure 3.16: Extraction behaviour of (a) La and (b) Nd as function of Al, Ti and Fe
dissolution after acidic leaching of the bauxite residue sample neutralised at 150 °C and
PCO2: 30bar (H2SO4, L/S: 10, T: 25 ºC, t: 24 h). The error bars represent the standard error
- 64 -
Fe 5 mm Ti 5 mm Sc 5 mm
Y 5 mm Nd 5 mm La 5 mm
BSE 20 mm
Figure 3.17: Elemental distribution of major elements (Fe and Ti) and minor elements (Sc,
Y, La, Nd), and a back-scattered electron (BSE) image of the neutralised-leached bauxite
residue.
3.4 Conclusions
The two-step processing of bauxite residue via neutralisation and

subsequent leaching establishes the effect of CO2 gas as neutralisation reagent on
the extraction efficiencies of metal recovery via acidic leaching. The minimum pH
attained after neutralisation in our bauxite residue with CO2 was 8.6. Further pH
decrease is limited, because of the chemical association of sodium to the sparsely
soluble silicate compounds. The increase of temperature leads to a strong
decomposition of silicate minerals, which allows the reduction of alkalinity upon
carbonation of sodium and calcium. However, the intensification of neutralisation
at elevated temperatures and high CO2 partial pressures leads to the formation of
calcite and cancrinite. Silicate compounds became more stable in the neutralised
- 65 -
samples rather than in the non-neutralised bauxite residue dissolution with the
increase of CO2 partial pressure and temperature. The extraction of Al, Fe and Ti
was reduced due to the insufficient availability of acid caused by acid consumption
for the chemical transformation of calcite into bassanite and silicate compounds.
The recovery of Sc from highly neutralised bauxite residue is lower compared to
that from the less neutralised, or even non-neutralised, presumably due to its
chemical association with Fe and Ti. The recovery of La and Nd is determined by
the acid concentration upon a limit recovery of major elements, although a positive
correlation with Sc was found.
3.5 References

331. doi:10.1007/s40831-016-0060-x.
doi:10.1016/j.jclepro.2015.02.089.
doi:10.1016/j.mineng.2015.01.005.
doi:10.1007/s40831-016-0068-2.
doi:10.1016/0003-2670(95)00486-6.
[6] W. Wang, Y. Pranolo, C.Y. Cheng, Metallurgical processes for scandium
recovery from various resources: A review, Hydrometallurgy. 108
(2011) 100–108. doi:10.1016/j.hydromet.2011.03.001.
[7] C.R. Borra, J. Mermans, B. Blanpain, Y. Pontikes, T. Van Gerven,
Selective leaching of rare earths from bauxite residue after sulphation
roasting, in: Bauxite Residue Valorization Best Pract. Leuven,
(Belgium), 5-7 Ocotber 2015, Pp. 301-307, 2015.
[8] I.N. Tanutrov, M.N. Sviridova, A.N. Savenya, A new technology for
coprocessing man-made wastes, J. Non-Ferrous Met. 54 (2013) 136–142.
doi:10.3103/S1067821213020132.
- 66 -
[9] S. Abhilash, M. Kumar, B. Dhar, Extraction of lanthanum and cerium

from Indian red mud, Int. J. Miner. Process. 127 (2014) 70–73.
doi:10.1016/j.minpro.2013.12.009.
doi:10.1021/ie011047b.
[11] C. Hanahan, D. McConchie, J. Pohl, R. Creelman, M. Clark, C.
Stocksiek, Chemistry of seawater neutralization of bauxite refinery
residues (red mud), Environ. Eng. Sci. 21 (2004) 125–138.
doi:10.1089/109287504773087309.
[12] S. Rai, K.L. Wasewar, D.H. Lataye, J. Mukhopadhyay, C.K. Yoo,
Feasibility of red mud neutralization with seawater using Taguchi’s
methodology, Int. J. Environ. Sci. Technol. 10 (2013) 305–314.
doi:10.1007/s13762-012-0118-7.
176 (2010) 1044–1050. doi:10.1016/j.jhazmat.2009.11.146.
[14] D. Bonenfant, L. Kharoune, R. Hausler, P. Niquette, CO2 Sequestration
by aqueous red mud carbonation at ambient pressure and temperature,
Ind. Eng. Chem. Res. 47 (2008) 7610–7616.
[15] N.A. Lutpi, J. Zhu, Carbonation of bauxite residue : a solution for carbon
dioxide capture in alumina industry, in: ICSTIE 2010, 2010.
in: Light Met., 2015: pp. 93–96.
[17] L.A. Pasechnik, A.G. Shirokova, O. V Koryakova, N.A. Sabirzyanov,
S.P. Yatsenko, Complexing Properties of Scandium(III) in Alkaline
Medium, Russ. J. Appl. Chem. 77 (2004) 1070–1073.
doi:10.1016/j.hydromet.2011.02.004.
[19] K. Zheng, A.R. Gerson, J. Addai-Mensah, R.S.C. Smart, The influence
of sodium carbonate on sodium aluminosilicate crystallisation and
solubility in sodium aluminate solutions, J. Cryst. Growth. 171 (1997)
197–208. doi:10.1016/S0022-0248(96)00480-0.
[20] K. Zheng, R.S.C. Smart, J. Addai-Mensah, A. Gerson, Solubility of
Sodium Aluminosilicates in Synthetic Bayer Liquor, J. Chem. Eng. Data.
43 (1998) 312–317. doi:10.1021/je970187i.
- 67 -
[21] N.W. Menzies, I.M. Fulton, W.J. Morrell, Seawater neutralization of

alkaline bauxite residue and implications for revegetation., J. Environ.
Qual. 33 (2004) 1877–1884. doi:10.2134/jeq2004.1877.
[22] R. Dilmore, P. Lu, D. Allen, Y. Soong, S. Hedges, J.K. Fu, C.L. Dobbs,
A. Degalbo, C. Zhu, Sequestration of CO2 in mixtures of bauxite residue
and saline wastewater, Energy and Fuels. 22 (2008) 343–353.
doi:10.1021/ef7003943.
[23] Y.-S. Han, S. Ji, P.-K. Lee, C. Oh, Bauxite residue neutralization with
simultaneous mineral carbonation using atmospheric CO2, J. Hazard.
[24] S.K. Reddy, S. Balasubramanian, Carbonic acid: molecule, crystal and
aqueous solution, Chem. Commun. 50 (2014) 503–514.
doi:10.1039/C3CC45174G.
[25] D.J. Cooling, P.S. Hay, L. Guilfoyle, Carbonation of bauxite residue, in:
6th Int. Alumina Qual. Work., 2002: pp. 185–190.
[26] S. Khaitan, D. a. Dzombak, G. V. Lowry, Mechanisms of neutralization
of bauxite residue by carbon dioxide, J. Environ. Eng. 135 (2009) 433–
438. doi:10.1061/(ASCE)EE.1943-7870.0000010.
[27] R.C. Sahu, R.K. Patel, B.C. Ray, Neutralization of red mud using CO2
sequestration cycle, J. Hazard. Mater. 179 (2010) 28–34.
doi:10.1016/j.jhazmat.2010.02.052.
[28] S. Rai, K.L. Wasewar, J. Mukhopadhyay, C.K. Yoo, H. Uslu,
Neutralization and utilization of red mud for its better waste
management, Arch. Environ. Sci. 6 (2012) 13–33.
[29] L.J. Kirwan, A. Hartshorn, J.B. McMonagle, L. Fleming, D. Funnell,
Chemistry of bauxite residue neutralisation and aspects to
implementation, Int. J. Miner. Process. 119 (2013) 40–50.
doi:10.1016/j.minpro.2013.01.001.
[30] S. Khaitan, D.A. Dzombak, G. V. Lowry, Chemistry of the acid
neutralization capacity of bauxite residue, Environ. Eng. Sci. 26 (2009)
873–881.
[31] R.M. Santos, J. Van Bouwel, E. Vandevelde, G. Mertens, J. Elsen, T.
Van Gerven, Accelerated mineral carbonation of stainless steel slags for
CO2 storage and waste valorization : Effect of process parameters on
geochemical properties, Int. J. Greenh. Gas Control. 17 (2013) 32–45.
doi:10.1016/j.ijggc.2013.04.004.
[32] B. Rizov, Phase transformations from goethite to hematite and thermal
decomposition in various nickeliferous laterite ores, J. Univ. Chem.
Technol. Metall. 47 (2012) 207–210.
http://www.scopus.com/inward/record.url?eid=2-s2.0-
- 68 -
84872850941&partnerID=tZOtx3y1.
[33] S. Agatzini-Leonardou, P. Oustadakis, P.E. Tsakiridis, C. Markopoulos,
Titanium leaching from red mud by diluted sulfuric acid at atmospheric
pressure, J. Hazard. Mater. 157 (2008) 579–586.
doi:10.1016/j.jhazmat.2008.01.054.
[34] J.T. Kloprogge, H.D. Ruan, R.L. Frost, Thermal decomposition of
bauxite minerals : Infrared emission spectroscopy of gibbsite, boehmite
and diaspore, J. Mater. Sci. 37 (2002) 1121–1129.
[35] B.I. Whittington, C.M. Cardile, The chemistry of tricalcium aluminate
hexahydrate relating to the Bayer industry, Int. J. Miner. Process. 48
(1996) 21–38. doi:10.1016/S0301-7516(96)00011-7.
[36] A. Atasoy, An investigation on characterization and thermal analysis of
the Aughinish red mud, J. Therm. Anal. Calorim. 81 (2005) 357–361.
doi:10.1007/s10973-005-0792-5.
[37] M.C. Barnes, J. Addai-Mensah, A.R. Gerson, The solubility of sodalite
and cancrinite in synthetic spent Bayer liquor, Colloids Surfaces A
Physicochem. Eng. Asp. 157 (1999) 101–116. doi:10.1016/S0927-
7757(99)00058-8.
[38] K. Snars, R.J. Gilkes, Evaluation of bauxite residues (red muds) of
different origins for environmental applications, Appl. Clay Sci. 46
(2009) 13–20. doi:10.1016/j.clay.2009.06.014.
[39] S. Wilhelm, M. Kind, Influence of pH, temperature and sample size on
natural and enforced syneresis of precipitated silica, Polymers (Basel). 7
(2015) 2504–2521. doi:10.3390/polym7121528.
doi:10.1016/j.hydromet.2016.08.010.
doi:10.1016/j.hydromet.2014.08.001.
(2016) 5–15. doi:10.1016/j.hydromet.2016.06.023.
Hydrometallurgical processing of eudialyte bearing concentrates to
recover rare earth elements via low-temperature dry digestion to prevent
the silica gel formation, J. Sustain. Metall. 3 (2017) 79–89.
- 69 -
doi:10.1007/s40831-016-0084-2.
[44] J. Schlomach, M. Kind, Investigations on the semi-batch precipitation of
silica, J. Colloid Interface Sci. 277 (2004) 316–326.
doi:10.1016/j.jcis.2004.04.051.
[45] N. Zhang, H.-X. Li, H.-J. Cheng, X.-M. Liu, Electron probe
microanalysis for revealing occurrence mode of scandium in Bayer red
mud, Rare Met. 36 (2017) 295–303. doi:10.1007/s12598-017-0893-x.
[46] R.D. Shannon, Revised effective ionic radii and systematic studies of
interatomic distances in halides and chalcogenides, Acta Crystallogr.
Sect. A. 32 (1976) 751–767. doi:10.1107/S0567739476001551.
- 70 -
Chapter 4 Occurrence of REEs in neutralised-leached bauxite residue
4. OCCURRENCE OF SELECTED RARE-

EARTH ELEMENTS IN NEUTRALISED-
LEACHED BAUXITE RESIDUE
ABSTRACT - This work investigates the chemical association of selected rare-

earth elements (Sc, Y, Ce, La and Nd) with major element phases in post-
processed bauxite residue and compares it with untreated bauxite residue. The
treatment of bauxite residue considers our previous published process which
involved the neutralisation with CO2, followed by leaching with H2SO4.
Neutralised bauxite residue resulted with larger aggregates than the untreated
bauxite residue after contacting it with CO2 as consequence of additional CaCO3
formation. Neutralisation with CO2, however, has a negligible effect on the
distribution of the rare-earth elements (REEs) with respect to the untreated bauxite
residue, but a large amount of rare earths remained unreacted after acid leaching.
Electron probe microanalysis (EPMA) confirmed the chemical association of
Sc(III) and Ce(IV) with Fe(III) and Al(III)-containing minerals in post-processed
bauxite residues, i.e. bauxite residue subjected to CO2-neutralisation and
neutralisation-acid leaching processes. The occurrence of Nd(III) is positively
correlated to that of La(III) in the untreated bauxite residue, but both of them may
be associated with the same mineralogical phase as Ce(IV) after processing. Y(III)
may remain associated with the Al/Si-minerals cancrinite and chamosite. Ergo, the
extractability of Sc, Y, Ce, La and Nd from neutralised bauxite residue is more
difficult in H2SO4 media due to the presence of coarser particles compared to
untreated bauxite residue, but also due to the formation of a solid product layer
(i.e. CaSO4) that is presumably absorbed on the surface of Fe(III)-rich phases
(hematite and goethite) and Al(III)-containing minerals (diaspore, gibbsite,
boehmite and chamosite).
Based on the accepted paper

R. M. Rivera, G. Ounoughene, A. Malfliet, J. Vind, D. Panias, V. Vassiliadou, K.
Binnemans, T. Van Gerven, “A study of the occurrence of selected rare-earth elements in
neutralised-leached bauxite residue and comparison with untreated bauxite residue”
Journal of Sustainable Metallurgy.
R. M. Rivera conceived the research, performed part of the analytical work, interpreted the
results, and wrote the article. G. Ounoughene, A. Malfliet, J. Vind contributed with part of
the analytical work. D. Panias, V. Vassiliadou, K. Binnemans and T. Van Gerven
contributed with their guidance and expertise in this work.
- 71 -
4.1 Introduction
The name bauxite is widely used to describe Al-ores containing high

amounts of Al-(oxy) hydroxides, such as gibbsite (-Al(OH)3), boehmite
(-AlO(OH)) and diaspore (-AlO(OH)). The deposits are typically formed by
the weathering of Al-bearing rocks [1, 2]. There are three main categories of
bauxite deposits, i.e. lateritic, karst and Tikhvin-type. The concentration of rare-
earth elements (REE), i.e. the group of the lanthanides plus Y and Sc, is much
higher in karst bauxites than in lateritic or Tikhvin-type bauxites, due to
crystallisation and accumulation of authigenic REEs-bearing minerals, and
adsorption of ions on clays, gibbsite and diaspore minerals [3–6]. Lateritic bauxite
deposits are characterised by a high Fe-content (more than 20 wt.%) [7], and Sc is
widely distributed in minerals such as goethite, clay minerals, or manganese
oxides due to the relative similar ionic radii of Sc and Fe [8, 9]. REEs have a strong
affinity for oxygen, so that they are commonly found in oxide, carbonate,
phosphate and silicate compounds, but they can also be found adsorbed on clay
minerals.
Bauxite residue (red mud) is the waste product generated during the
alumina production from bauxite ore by the Bayer process and is composed
essentially of compounds that are insoluble in concentrated NaOH solutions such
as Fe- and Ti-rich minerals, undigested Al-rich minerals, Na/Al-hydrosilicates, Ca
compounds and significant amounts of REEs. Hence, bauxite residue represents
an interesting source for not only major elements such as Al, Fe and Ti, but also
for REEs. Among the REEs, Sc represents about 95% of the economic value of
the REEs present in bauxite residue [3, 10]. The Sc concentration depends on the
origin of the bauxite ore and process parameters of the Bayer process, but can be
as high as 260 mg kg-1 [11]. In Greek bauxite residue, Sc concentrations are about
120 ± 10 mg kg-1 [12]. These Sc concentrations are much higher than the average
abundance of Sc in the Earth’s crust (22 mg kg-1) [13], which renders bauxite
residue a suitable source of Sc and other rare earths (e.g., Y, Ce, Nd, La) to meet
future demands.
During the production of alumina from bauxite, the REEs, in particular
Sc, end up in the bauxite residue due to their association with Fe- and Ti-minerals,
which react in the course of the high temperature digestion process [10, 14–18].
The transfer of Sc from the bauxite ore into the bauxite residue may occur
by: (1) formation of an independent insoluble phase (e.g., Sc(OH) 3 or ScOOH)
that is adsorbed on some mineral phases (e.g., hematite), and/or (2) associated with
some newly formed mineral phases, such as Na/Al-hydrosilicates or Ca-
hydrogarnet compounds, during the Bayer process [19]. Sc can also remain
associated to undissolved minerals from bauxites (e.g., hematite, goethite and
zircon [20]). Bauxite residues tend to be characterised by their high concentration
of Al2O3 and Fe2O3, although other oxides such as SiO2 and TiO2 may also be
present [3, 21]. Urbain and Sarkar (1927) were the first to report a chemical
interaction between Sc and Fe. Borra et al. reported that Sc cannot be extracted
- 72 -
from bauxite residue without bringing Fe into the solution [12]. Rivera et al. have
confirmed this association trend between Sc and Fe and reported a similar trend
between Sc and Ti [22]. Ochsenkühn-Petropulu et al. found Sc in Ti minerals in
mixtures of different bauxite deposits from Greece [23]. Vind et al. have reported
a detailed study about the occurrence of rare-earth-containing minerals in karst
and lateritic bauxite, and in dewatered bauxite residue [24]. According to
experimental results obtained after acid leaching of bauxite residue, the
lanthanides seem not to be associated with major elements in bauxite residue [22],
but a linear relationship between adjacent lanthanides has been observed due to
the common trivalent oxidation state in aqueous solution and similarities in ionic
radii [12].
4.2 Occurrence of REEs in bauxite and bauxite residue – a brief review
In karst-type bauxites, the REEs occur mostly as authigenic and/or

diagenetic light REE (LREEs) fluorocarbonate minerals, although Ce can also be
found in its tetravalent oxidation state (Ce4+) associated either with
LREE/actinides oxides and/or sorbed in Fe-rich oxides [25]. Bastnäsite is the most
common host for rare-earth oxides in karst bauxites [26, 27]. However, REEs have
also been reported in other fluorocarbonate minerals such as parisite
(REE2Ca(CO3)3F2), röntgenite (REE3Ca2(CO3)5F3), synchysite (REECa(CO3)2F),
and also in cerianite (CeO2) and churchite-(Y) (Y(PO4)·2(H2O)) [6, 26]. Wagh and
Pinnock reported that REEs tend to concentrate in secondary aluminium
phosphates such as crandallite (CaAl3(PO4)2(OH)5·H2O) and variscite
(AlPO4·2H2O) [11]. Sc, in particular, tends to be associated with Fe-mineral
phases such as goethite and/or hematite [20, 28], which is attributed to the
substitution and charge compensation of Fe(II) by Sc(III) during the geological
formation of Fe(III)-rich oxidic minerals [29], due to the similarity in ionic radii
of Sc(III) (88.5 pm) and Fe(II) (78.5 pm), and their coordination numbers. An
association to Ti(IV) mineral phases has also been reported, although the
substitution of Ti(IV) by Sc(III) is not of the same nature as the substitution
occurring in Fe(III)-rich mineral phases, despite their similar ionic radii (Ti(IV):
74.5 pm) [30, 31]. The substitution of Sc(III) into the Ti(IV) phase requires a
charge compensation to maintain the electrostatic balance [30]. A comparable
behaviour between Sc and Fe (Sc/Fe) and between Sc and Ti (Sc/Ti) has been
reported by several authors in post-processed samples of bauxite residue [12, 19,
22, 23, 28, 32, 33]. Sc has also been found associated with Al(III) mineral phase
such as boehmite (-AlO(OH)), although this substitution is less common to occur
due to the relatively high difference between the ionic radii of Sc(III) and Al(III)
(ionic radius Al(III): 67.5 pm) [19].
During the Bayer process, most of the trace elements end up in the
bauxite residue, which distribution is mainly controlled by mass flows during the
Bayer process [15]. The most abundant phase containing REEs in Greek bauxite
residue is a LREE ferrotitanate (REE,Ca,Na)(Ti,Fe)O 3 compound, which is
- 73 -
formed during the digestion process of bauxite minerals in the Bayer process. This
host compound can be sub-divided into Ce-predominant and Nd/La- predominant
particles. Minor amounts of LREEs have been found as carbonates and
phosphates, while heavy REEs were found associated with Y-phosphate phases.
Y was reported to be partially associated with other REEs as well [24].
4.3 Effect of bauxite residue’s alkalinity on REEs recovery by acid

leaching
The alkalinity of bauxite residue represents an important factor during
acid leaching for metal recovery from bauxite residue as it may lead to a large acid
consumption and makes the recovery of metals from bauxite residue economically
not feasible. The use of CO2 gas as an alternative neutralisation reagent before
metal extraction by acid leaching was previously investigated to diminish the acid
consumption [22]. However, the change of the chemical and mineralogical
composition of the bauxite residue after neutralisation led to lower recovery of
REEs during acid leaching compared to that from non-neutralised bauxite residue.
At present, there is no systematic study reported to explain the differences in
leaching mechanisms of non-neutralised and neutralised bauxite residue. Such
knowledge requires a detailed phase and mineralogical investigation on the
corresponding leach residues, with the aim to enhance the further recovery of
REEs with an optimal consumption of acid.
In this paper, the mode of occurrence of selected REEs in neutralised-
leached bauxite residue is investigated by electron probe microanalysis (EPMA).
The REE-carrier minerals within the bauxite residue are discussed. A comparative
analysis with the occurrence mode of the same elements in the untreated bauxite
residue is presented. A theoretical framework for further processing of bauxite
residue in view of enhanced REEs recovery is proposed. A comparative analysis
for rare earth extraction from non- and neutralised bauxite residue has been
developed to assess the technical viability of the proposed method.
4.4 Materials and methods
The bauxite residue used in this study was kindly provided by Mytilineos
S.A. - Aluminium of Greece (Agios Nikolaos, Greece). It originates from a
mixture of 80 wt.% karst and 20 wt.% lateritic bauxites. It was received from the
alumina refinery after dewatering by filter pressing and drying at ambient
temperature. Upon arrival in the laboratory, the sample was further dried at 105
ºC for 24 h. Chemical analysis of the major elements in the pre-processed bauxite
residue (BR, also termed as untreated material) was performed by wavelength
dispersive X-ray fluorescence spectroscopy (WDXRF, Panalytical PW2400).
Chemical analysis of the minor elements was performed after complete dissolution
of the untreated bauxite residue by alkali fusion and acid digestion in 3 vol.%
- 74 -
HNO3 solution, followed by Inductively Coupled Plasma Mass Spectrometry

(ICP-MS, Thermo Electron X Series) analysis. The sample was further processed
via CO2 neutralisation in a titanium autoclave (Parr Company, series 4560,
capacity of 400 mL) with a CO2-pressure of 30 bar at 25 and 150 °C, a liquid-to-
solid ratio, L/S, of 5 during 2 h, followed by acid leaching for metal recovery. The
sample subjected to CO2 neutralisation at 25 ºC is called neutralised bauxite
residue, NBR, and the sample neutralised at 150 ºC is called highly-neutralised
bauxite residue, HNBR. Leaching of the non- and neutralised bauxite residue was
carried out with 1.6 N H2SO4 (95–97%, Sigma-Aldrich) with a L/S ratio of 10:1 to
ensure a substantial leaching concentration of REEs in the leach liquor with a
reasonable consumption of bauxite residue. The experiments were performed in
sealed polyethylene bottles by constant agitation using a laboratory shaker
(Gerhardt Laboshake) at 200 rpm and 25 °C. The sample after CO2 neutralisation
(at 150 ºC) and leaching is called neutralised-leached bauxite residue, Lx HNBR.
The procedure, as well as the results of the neutralisation by CO 2 and subsequent
leaching step, have been reported elsewhere [22]. The slurries after neutralisation
and neutralisation-leaching were filtered using filter paper (pore size of 0.45 µm)
and diluted with 2 vol.% HNO3 for Inductively Coupled Plasma Optical Emission
Spectroscopy (ICP-OES, PerkinElmer Optima 8300) analysis for major (Al, Fe,
Ca, Na, Si, Ti) and minor (Sc, Y, Ce, La, Nd) elements. The chemical composition
of the major and minor elements in the solid residues obtained from the post-
processed bauxite residue samples, i.e. after neutralisation and neutralisation-
leaching processes, were determined from the corresponding slurry by mass
balance. The mineralogy of the samples was studied by X-ray powder diffraction
(XRD, Bruker D2 Phaser). The data obtained were analysed and quantified by
XDB Powder Diffraction Phase Analytical System [34]. The particle size of the
untreated and neutralised bauxite residue was measured by laser particle size
analysis (Malvern Mastersizer 3000). The distribution of major and minor
elements was identified using an electron probe for microanalysis (EPMA, type
JXA-8530F from Jeol Ltd.). The accelerating voltage was 15 kV, and the beam
current was 100 nA. The counting time on peak for REEs was 200 s and on the
background 50 s. EPMA-WDS quantifications of these elements were carried out
using the standards Sc metal, monazite (for Ce and La) and the CAL-STD from
the instrument (for Y).
The effect of neutralisation of bauxite residue with CO 2 gas on the

recovery of selected REEs from bauxite residue was previously investigated in
two consecutive processes according to the diagram in Figure 4.1: (1)
neutralisation of bauxite residue slurry with CO2 gas, and (2) acid leaching of the
neutralised bauxite residue with H2SO4 [22]. H2SO4 was selected as a leaching
reagent because is commonly used in hydrometallurgical plants due to its
relatively lower cost (about 200 USD/tonne), effectiveness, availability and less
corrosiveness than other mineral acids (e.g., HCl, HNO3). The results on process
- 75 -
optimisation allowed to decrease the initial pH of the bauxite residue (ca. 10.3) to
a value < 9 during neutralisation with CO2, so that an enhancement of metal
extraction during leaching could be expected under the same acid consumption
used to leach the bauxite residue directly (without CO2-neutralisation). However,
as it is depicted in Figure 4.2, the extraction yields were slightly lower in
comparison to acid leaching of non-neutralised bauxite residue under similar
leaching conditions. It was reported that such decrease in metal recovery was
caused by an enrichment of CaCO3 after neutralisation with CO2, which later on
was transformed into CaSO4 during leaching with H2SO4. The decomposition of
silicate compounds can also influence the extractability of metals from the
neutralised bauxite residue, as silicate compounds were enriched after
neutralisation [22]. During neutralisation with CO2, the dissolution of major
metals did not take place because the pH during neutralisation was kept around 7,
while the dissolution of major elements (e.g., Al, Fe, Ti and Si) start occurring
only at pH < 6 [12]. However, only the dissolution of Na (about 27 wt.%) and
calcium (about 5 wt.%) took place, but most of the calcium tends to precipitate as
CaCO3. Hence, the leachate obtained after neutralisation was rich in Na, and could
be recirculated back in the Bayer process.
H2O
Bauxite
CO2
residue
NEUTRALISATION Leachate
Neutralised BR
H2SO4 (solid fraction)
LEACHING Residue
Pregnant leach
solution (PLS)
Figure 4.1: Flow sheet for neutralisation of bauxite residue by CO2, followed by metal
recovery by leaching with H2SO4 (reprinted from [19] with permission of Elsevier; license
number: 4212401452947).
- 76 -
70
BR
60
Lx HNBR
Extraction, wt.%
50
40
30
20
10
0
Al Fe Ti Sc Y La Nd Ce
Element
Figure 4.2: Extraction yields of Al, Fe, Ti and selected REEs (Sc, Y, La, Nd, Ce) after acid
leaching of the bauxite residue (BR) and neutralised bauxite residue (Lx HNBR) (adapted
from [19]).
Concentrations of major and rare-earth elements in post-processed

bauxite residue
The concentrations of the major elements in the pre- and post-processed
Greek bauxite residue samples are given in Table 4.1. Fe is the major element in
the bauxite residue, also after neutralisation and acid leaching. The concentrations
of Al, Ca, Na, Ti and Si decreased after processing the bauxite residue by acid
leaching. The concentration of selected trace elements, i.e. Sc, Y, Ce, La and Nd,
is shown in Table 4.2. These trace elements were selected because their
concentrations in this particular bauxite residue are relatively high. The REEs
concentrations remain unchanged during neutralisation with CO 2 gas, but slightly
decreased during leaching with H2SO4, although a significant amount of REEs still
remained unreacted in the sample. The concentration of elements in the sample
subjected to room-temperature, i.e. at 25 ºC, neutralisation was similar to that of
the original sample.
- 77 -
Table 4.1: Major chemical elements, expressed as oxides, in the bauxite residue samples
(in wt.%).
Element BRa NBRb HNBRb Lx HNBRb

Al2O3 18.1 18.3 18.3 14.4
CaO 8.5 8.3 8.5 7.6
Fe2O3 46.8 47.2 47.0 45.5
NaO 3.6 3.1 2.5 0.8
SiO2 7.3 7.3 7.3 3.9
TiO2 5.8 5.8 5.8 4.0
aXRF analysis, bmass balance
BR: untreated bauxite residue, NBR: neutralised bauxite residue (at 25 ºC),
HNBR: highly-neutralised bauxite residue (at 150 ºC), Lx HNBR: neutralised-
leached bauxite residue
Table 4.2: Minor elements, expressed as oxide, in the bauxite residue samples
(in mg kg-1).
Element BRa NBRb HNBRb Lx HNBRb

Sc2O3 186 186 186 118
Y2O3 97 97 97 39
CeO2 452 452 452 356
La2O3 134 135 135 96
Nd2O3 115 115 115 80
aICP-MS analysis, bmass balance
The particle size distribution of the untreated and (highly) neutralised

bauxite residue samples is shown in Figure 4.3. In the bauxite residue sample,
90 vol.% of the particles are smaller than 10 µm and 50 vol.% of the particles are
smaller than 0.05 µm. Such particle size distribution is comparable to the typical
values for common bauxite residues, i.e. 90 vol.% below 75 µm [32,33]. In our
sample, 77 vol.% of the particles is concentrated in clays, i.e. particle size below
2 µm [34,35]. Therefore, our bauxite residue is essentially composed of very fine
particles that tend to form aggregates after neutralisation. Hence, after
neutralisation, coarser particles were obtained as 90 vol.% of the particles were
< 115 µm, and 50 vol.% of the particles were < 0.1 µm. The morphological
structure of both samples (Figure 4.4), i.e. untreated and neutralised bauxite
residue, indicates the existence of scattered particles smaller than 1 µm, although
neutralised bauxite residue can exhibit slightly larger aggregates. It has been
reported that larger agglomerates can be achieved during neutralisation with
seawater or other Ca- and Mg-rich brines due to the precipitation of hydroxide,
carbonate or hydroxycarbonate [36,37]. In our study, neutralisation with CO2
- 78 -
changed the aggregates size due to the precipitation of CaCO3, despite an increase
of only 4 wt.% with respect to the initial content of calcite in the untreated bauxite
residue [19]. However, this increase of particle size caused a significant reduction
on the extraction of REEs from the neutralised bauxite residue, due to a substantial
reduction of the diffusion leaching rate [38,39]. It must be noticed that the
dissolution of REEs from bauxite residue significantly depends on the acid
concentration. Hence, the extraction of REEs from neutralised bauxite residue
demands much more acid than the non-neutralised bauxite residue.
Clay Silt Sand

d < 2 mm 2 < d < 63 mm d < 2000 mm
100
Volumetric distribution, %
80
60
40 BR
HNBR
20
0
0.01 0.1 1 10 100 1000
Particle size, mm
Figure 4.3: Cumulative particle size distribution of the untreated and (highly) neutralised
bauxite residue samples determined by laser diffraction (particle size scale reference from
[43].
Untreated bauxite residue
- 79 -
Neutralised bauxite residue
Figure 4.4: SEM pictures of the untreated bauxite residue (previous page) and neutralised
bauxite residue (on this page).
The normalised mineralogical composition of untreated and processed

bauxite residue is given in Table 4.3. This quantification led to a significantly
different result compared to the one reported in section 3.3.1, as in the current
research the samples were analysed and quantified by the – presumably more
correct – XDB software. With this software, the presence of some mineral phases,
such as sodalite, dicalcium silicate and tricalcium silicate, could not be confirmed.
The XDB software enables a more accurate quantification of the mineralogical
phases than other softwares, as it has a specialized reference database for bauxite
and bauxite residue minerals, and it uses the bulk concentration and the content of
lost-on-ignition (LOI) elements as an additional input to fit the profile to be
conform with the mass balance. It must be noticed that in Rivera et al. (2017),
identification of phases was done by EVA V.3.1 and quantified with Topas-
Academic V.5 using the Rietveld method, which is based on the use of scale
factors for refinement as an index of the mineral concentration [37]. The XDB
approach takes advantage of the fact that scale factors and phase weights can be
determined without the knowledge and refinement of structural parameters, and it
is widely used by alumina producers [34]. Consequently, the following mineral
phases and compounds were identified in the untreated and/or CO 2-neutralised
bauxite residue: hematite, goethite (Fe-rich); gibbsite, diaspore, boehmite (Al-
rich); calcite, portlandite, bassanite (Ca-rich); perovskite, rutile, anatase (Ti-rich);
cancrinite and chamosite. An increase in calcite content (from 4 to 8 wt.%) was
observed after neutralisation with CO2. The concentrations of cancrinite,
perovskite and the Ca/Fe/Al-silicate compound Ca3FeAl(SiO4)((OH)4)2 were
significantly reduced after leaching, presumably due to the dissolution of CaO that
reacts with H2SO4 to form CaSO4. Hence, a detectable amount of bassanite was
identified only after acid leaching.
- 80 -
Table 4.3: Normalised mineralogical composition of the untreated and treated bauxite
residue (in wt.%).
Lx
HN
Phase Chemical formula BR NBR HN
BR
BR
Hematite Fe2O3 37 37 36 38
Ca-Al-Fe-Si- Ca3FeAl(SiO4)((OH)4)2 16 16 11 0
hydroxide
Diaspore AlO(OH) 11 11 10 11
Cancrinite Na8(Al,Si)12O24(OH)2·3H2O 11 11 9 3
Goethite FeO(OH) 7 7 9 7
Calcite CaCO3 4 4 8 0
Chamosite Fe1.8Mg0.2Al0.8(Si1.3Al0.7)O5(OH)4 3 3 5 3
(berthierine)
Perovskite CaTiO3 5 5 4 2
Gibbsite Al(OH)3 2 2 4 3
Boehmite AlO(OH) 2 2 2 3
Anatase TiO2 0.5 0.5 0.5 0.8
Rutile TiO2 0.5 0.5 0.5 2
Quartz SiO2 0.5 0.5 0.5 1
Portlandite Ca(OH)2 0.7 0.7 0 0
Bassanite CaSO4 0.5H2O 0 0 0 22
Electron probe microanalysis of post-processed bauxite residue

Bauxite residue samples subjected to CO2-neutralisation and
neutralisation-leaching were analysed by electron probe microanalysis (EPMA).
Fe(III)-rich phases appear in the backscattered imaging mode as bright particles,
while Al(III)-rich phases appear slightly duller than Fe(III)-rich phases, but
brighter than the bulk phase (Figure 4.5 and 4.6). Both mineral phases were easily
distinguishable from each other in EPMA analysis at the same brightness level.
Also, the Fe(III)- and Al(III)-rich phases were found simultaneously as a solid
solution, i.e. Fe(III)/Al(III) solid solution. Ti(IV) and REEs were found together
with Fe(III)- and Al(III)-mineral phases, but not as single particles. The presence
of Sc(III)-, Y(III)- and Ce(IV)- containing phases were revealed because their
concentrations are relatively high in this particular bauxite residue. La(III)- and
Nd(III)-containing phases, on the contrary, could not be detected by EPMA
because their concentrations were lower than the detection limits [44]. New data
reported elsewhere, based on analysis by wavelength-dispersive X-ray
- 81 -
fluorescence and transmission electron microscopy, have demonstrated that

Nd(III) tends to occur simultaneously with La(III) and Ce(IV) [24]. Therefore,
Ce(IV) can serve as a good approximation to understand the behaviour of La(III)
and Nd(III) in bauxite residue samples after processing.
Figure 4.5: BSE image of the highly neutralised bauxite residue (HNBR, 15 kV, 1.004e-7
A, dwell: 25000 ms).
Table 4.4: EPMA (average) quantitative analysis results (expressed as oxide) of marked
areas in Figure 4.5 (from Fe2O3 until ZrO2 in wt.%, from Sc2O3 until Y2O3 in mg kg-1, total
in wt.%).
Phase Fe2O3 Al2O3 TiO2 SiO2 CaO ZrO2 Sc2O3 CeO2 Y2O3 Total
10 83 1 0.2 0.1 0.0 5 28 25 94

Al(III) ±6 ± 0.1 ± 0.1 ± 0.1 ± 0.0 ± 0.0 ±7 ± 39 ± 21 ±6
82 2 3 0.5 0.5 0.1 160 153 123 89

FeO(OH) ±1 ± 0.1 ± 0.2 ± 0.0 ± 0.0 ± 0.0 ± 46 ± 21 ± 21 ± 1.3
94 1 3 0.3 1 0.0 157 334 112 100

Fe2O3 ±4 ± 0.6 ±1 ± 0.2 ±1 ± 0.0 ± 40 ± 74 ± 61 ± 0.2
- 82 -
Figure 4.6: BSE images of the neutralised-leached bauxite residue (Lx HNBR, 15 kV,
1.004e-7 A, dwell: 25000 ms).
Table 4.5: EPMA (average) quantitative analysis results (expressed as oxide) of marked
areas in Figure 4.6 (from Fe2O3 until ZrO2 in wt.%, from Sc2O3 until Y2O3 in mg kg-1, total
in wt.%).
Phase Fe2O3 Al2O3 TiO2 SiO2 CaO ZrO2 Sc2O3 CeO2 Y2O3 Total
87 2 2 2 0.4 19 8 47 92
FeO(OH) 0
±1 ± 0.5 ± 0.6 ± 0.6 ± 0.1 ± 15 ± 25 ± 32 ±2
94 1 3 0.4 0.1 123 394 79 100

Fe2O3 0
±1 ± 0.1 ±1 ±0 ±0 ± 14 ±115 ± 29 ±0.3
Fe(III)/
Al(III) 28 62 2 0.4 0.1 0.03 52 73 68 92
solid ± 16 ± 13 ±1 ± 0.1 ± 0.1 ± 0.02 ± 32 ± 60 ± 26 ±4
solution
In the untreated and neutralised bauxite residues, Sc(III) and Y(III) were
found to be associated to Fe(III)- and Al(III)-rich phases (Table 4.4). In both the
neutralised bauxite residues, i.e. NBR and HNBR, the highest concentrations of
Sc(III) were observed in Fe(III)-rich mineral phases, namely hematite and goethite
(160 ± 40 and 157 ± 40 mg kg-1, respectively). Such concentration is in agreement
with the Sc(III)-concentration in the untreated bauxite residue, as CO2-
neutralisation did not alter the concentration of REEs in the treated bauxite residue
(see Table 4.2). Figure 4.7 describes the distribution of Sc(III) (in mg kg-1, based
on EPMA analysis) in goethite, hematite and Al(III)-rich mineral phases. The
quantitative analysis based on EPMA results after leaching is shown in Table 4.5.
The lowest concentration of Sc(III) was observed in goethite (about
19 ± 15 mg kg-1, in Table 4.5). The concentration of Sc(III) in the Al(III)-rich
- 83 -
mineral phases after leaching was about 52 ± 32 mg kg-1. This may suggest that
Sc(III) was mainly extracted from the goethite mineral phase, though the
concentration of goethite remained the same as in the untreated sample, while a
significant content (about 123 ± 14 mg kg-1) remained associated to hematite after
leaching. In both the neutralised bauxite residues, Ce was also found in the same
mineral phases as Sc, particularly associated to hematite (about 334 ± 74 mg kg-1)
and goethite (about 153 ± 21 mg kg-1), because both phases can incorporate into
its lattice matrix the Ce(IV) ion, as this is the predominant oxidation state of Ce in
bauxite residue [41]. However, after leaching, Ce was found widely distributed
among different mineral phases, which makes it difficult to estimate the
distribution of Ce(IV) among Fe(III)- and Al(III)-rich mineral phases. Ce(IV) can
also be partially associated with Ti(IV)-rich mineral such as perovskite [21,42],
but such phase was not detected by EPMA. It is believed that La(III) and Nd(III)
may also present similar associations as Ce(IV) due to the fact that La(III) and
Nd(III) tend to be positively correlated due to their similarities in ionic radii, but
also because they tend to have a similar mineralogical distribution in bauxite
residue as both occur only in the trivalent oxidation state [43]. However, it is not
very plausible that either Nd(III) or La(III) can be associated with the crystalline
structure of Fe(III)- or Al(III)-rich mineral phases, as Sc(III) or Ce(IV) does, due
to the large difference in ionic radii between the trivalent lanthanide ions and
Fe(III) or Al(III). Nevertheless, due to the existence of the light REEs (LREEs,
group of the lanthanides from La to Eu) inside of a calcium ferrotitanate shell, the
extraction of LREEs is more difficult in H 2SO4 media due to the formation of a
solid product layer (i.e. CaSO4) and/or due to the precipitation as a sodium-double
sulfate, which hinders the dissolution of the LREEs [21,44,45]. The presence of a
sodium-double sulfate, however, was not detected in our samples. The formation
of a double sulfate with heavy REEs (HREEs, group of the lanthanides from Gd
to Lu plus Y), on the other hand, is easily leachable due to higher solubility
compared with LREEs [46]. Hence, it is believe that the dissolution of Y(III) was
mostly limited by its association to Fe(III)- and Al(III)-mineral phases, although
it was also found in very small particles (< 1 µm) associated to Zr and/or Si,
presumably associated with a Zr/Si-mineral phase. The association of Y(III) to a
Zr/Si-rich mineral phase was found in the neutralised and neutralised-leached
bauxite residue, but only in very localised spots. It is reported that heavy REEs
(HREEs, group of the lanthanides from Gd to Lu plus Y) tend to be associated
with Zr/Si compounds, but also can be found associated to phosphate minerals
[21,47–49]. Therefore, it is believed that in the bauxite residue obtained after
neutralisation-leaching, Y(III) can remain mostly associated with Al/Si-
compounds, such as cancrinite and/or chamosite, as they were not completely
dissolved by H2SO4, while no phosphate minerals were found in the samples
investigated in this work. Due to the low concentration of Ti(IV) compared to
Fe(III)- and Al(III)-rich phases, the correlation between the REEs and Ti(IV)
could not be confirmed by EPMA.
- 84 -
250 Goethite Hematite Al-phases
Sc concentration, mg L-1
200
150
100
50
0
NBR HNBR Lx HNBR
Sample
Figure 4.7: Concentration of Sc in goethite, hematite and Al-rich phases (concentration

based on EPMA quantified-analysis).
Reaction mechanism for neutralised-leached bauxite residue

According to Figure 4.4, bauxite residue is composed of particles < 1 mm
that tend to form aggregates after neutralisation with CO 2, which finally results in
an increase of the average particle size (P50 of neutralised-leached bauxite residue
was  0.1 mm, while P50 of untreated bauxite residue was  0.06 mm, according to
Figure 4.3). The particle shape of the non- and neutralised bauxite residue was not
spherically symmetric, as the shrinking core model suggest. However, it is
believed that the formation of large aggregates, with an irregular morphology after
CO2-neutralisation, can lead to a reduction in the surface-to-volume ratio [41].
Particle size and shape are the two main factors that determine the surface area of
the particles. It is stated that the rate at which particles are leached out from the
solid matrix, i.e. leaching kinetics, are affected by the particle size of the material
that is subjected to leaching [42]. Thus, smaller particles size gives faster leaching
kinetics because they have a larger surface area compared to bigger particles. In
addition, as it has been described in the previous section, the REE-host minerals
can be surrounded by other minerals located inside the particles. Hence, during
leaching of untreated bauxite residue with H2SO4, the solution can only access into
the particles via pores and/or cracks [54]. During leaching of the neutralised
bauxite residue, however, the diffusion of reagent and dissolved elements, for
instance REEs, is further limited by the formation of large aggregates during
neutralisation with CO2, as consequence of an enrichment of the CaCO3
concentration (see Figure 4.3). Furthermore, SEM-EDX analysis of the
neutralised-leached bauxite residue (Figure 4.8) revealed that in these aggregates,
Ca(II) can co-exist simultaneously with Al(III), Si(IV) and Ti(IV)-rich phases, but
it can also enclose partially the Fe(III)-rich phases (hematite and/or goethite).
Although Ca2+ ions precipitate as CaSO4 in the presence of H2SO4, calcium
dissolution is low due to the low solubility in water of calcium sulfate, i.e. 2 g L -1
- 85 -
at 20 ºC [55]. Therefore, it is most likely that at the beginning, leaching can be

surface chemical reaction controlled due to the small surface area of the particles,
but as leaching advance, the presence of other minerals reduces the rate by
blocking access of the leach solution to most of the particle’s surface. Thus,
diffusion control may become dominant. Furthermore, it is believed that CaSO 4
may be adsorbed on the surface of the particles and acts as barrier for the H 2SO4
solution, so that Fe(III)- and Al(III)- rich compounds are less dissolved [56].
Consequently, the decomposition of REE-host minerals is not fully achieved, and
the dissolution of REEs becomes less efficient. The formation of CaSO4 can be
avoided by using an alternative mineral acid as a leaching reagent (e.g., HCl,
HNO3), but the leaching can still be surface-chemical reaction controlled due to
the formation of large aggregates after neutralisation with CO2.
Figure 4.8: Back-scattered electron (BSE) image of the neutralised-leached bauxite residue
with H2SO4, and elemental distribution of Al, Ti, Fe, Ca, Si, S and O (15 kV, 3500x).
- 86 -
Table 4.6: SEM-EDX (average) semi-quantitative analysis results of elements in Figure 4.8.
Element Concentration, wt.%

Al 7.2 ± 6.3
Ti 5.1 ± 4.5
Fe 43.2 ± 3.0
Ca 6.9 ± 3.7
Si 4.2 ± 5.6
S 3.9 ± 4.7
O 25.2 ± 7.9
Comparative analysis for rare earth extraction from bauxite residue

The flow sheet and mass balance of the conventional direct acid leaching
and the neutralisation-leaching methods are shown in Figure 4.9 and 4.10,
respectively. The low extraction yields of Sc, La and Nd obtained from the
neutralised bauxite residue were caused by an insufficient amount of acid devoted
to leaching, as consequence of the chemical transformation of CaCO3 into CaSO4,
and the dissolution of silicate compounds. In addition, an increase of the average
particle size after neutralisation led to a reduction of the diffusion leaching rate.
The extraction yield of Y, on the contrary, resulted to be the highest one because
was not affected by the CO2-neutralisation and, therefore, ca. 60 wt.% of Y was
recovered from the neutralised material (Figure 4.2). Also, under the CO2-
neutralisation conditions studied in this work, about 9 kg of NaOH (based on
27 wt.% of Na in solution) can be recovered per tonne of bauxite residue, which
can be sent back into the Bayer process. However, the water consumption of the
neutralisation-leaching method, i.e. 15 kg H2O per tonne of bauxite residue, was
three times higher than the consumption required by the conventional leaching
method, i.e. 5 kg H2O per tonne of bauxite residue, because CO2 must be dissolved
in water to form H2CO3 during the neutralisation stage.
- 87 -
Figure 4.9: Flow sheet and mass balance of the conventional direct acid leaching process
with H2SO4 as leaching reagents (PLS: pregnant leach solution, chemical composition of
Ti and REEs in the PLS are based on results depicted in Figure 4.2).
Figure 4.10: Flow sheet and mass balance of the neutralisation-leaching process with H2SO4
as leaching reagent (PLS: pregnant leach solution, chemical composition of Ti and REEs
in the PLS is based on results depicted in Figure 4.2).
The technical feasibility of the neutralisation-leaching method can be

further improved by recirculating back the leachate of the neutralisation stage
(solution rich in Na) to the Bayer process, while the consumption of bauxite
residue could be substantially increase by using the dry digestion technique in
combination with multi-stage leaching (see Chapter 5). As it has been reported,
the dry digestion technique does not only avoid the silica gel formation during
leaching, but also allows to process a larger quantity of bauxite residue, with a
- 88 -
substantial reduction in water consumption, compared to conventional direct

leaching [33]. The consumption of large amount of acid during neutralisation of
the alkaline products left behind after the Bayer process, still remains as one of the
main drawbacks for REEs recovery. The use of H2CO3 (formed during the direct
dissolution of CO2 gas in water) did not improve the efficiency of the leaching
process, but the residue obtained after acid leaching can be used for building
materials. The remaining sodium is totally dissolved into the acid leach solution,
whereby its final concentration is very low. Hence, the residue can be used in the
clinker production industry. In addition, the formation of large aggregates after
CO2-neutralisation, as consequence of the formation of extra CaCO3 and its
subsequence chemical transformation into CaSO4, can represent a benefit because
CaSO4 can replace the addition of gypsum in the ordinary Portland cement clinker
[57].
4.6 Conclusions
The distribution mode of selected REEs in untreated, CO 2-neutralised

and neutralised-leached bauxite residue was investigated. Neutralisation of
bauxite residue with CO2 led to an enrichment of CaCO3, which also led to an
increase of the average particle size and, consequently, to a reduction of the
leaching efficiency.
Electron probe microanalysis confirmed the chemical association of
Sc(III) with Fe(III)- mineral phases, but also of Sc(III) with Al(III)-mineral
phases, in the CO2-neutralised sample, which has also been observed in a similar
bauxite residue reported in the literature. The same association was observed in
the residual solid after acid leaching. Sc(III) and Ce(IV) co-exist mostly with the
Fe(III)-rich mineral hematite. It is believed that La(III) and Nd(III) may follow the
same trend as Ce(IV) as they tend to occur simultaneously in the same
mineralogical phase. However, their extraction from the neutralised bauxite
residue was limited by the transformation in H2SO4 media of calcite into bassanite
or double sulfates. It is believed that CaSO4 can be adsorbed on the surface of the
particles and acts as a resistant for the acid. On the other hand, Y(III) may remain
associated with cancrinite and/or chamosite compounds after leaching, due to the
association of Y(III) with Al/Si-compounds and because these minerals were not
completely dissolved by H2SO4. The technical and eventually also the economic
viability of the neutralisation-leaching method could be improved by considering
the recirculation of the leachate obtained after CO2-neutralisation into the Bayer
process, and by considering the dry digestion of the neutralised solid product to
recover REEs. The solid residue obtained after leaching can be considered for
building materials as it has low Na-content but it can also be rich in CaSO 4. The
selectivity of REEs during leaching can be enhanced, however, after the separation
of Al and Fe by pre-treatment process (e.g., roasting, smelting).
- 89 -
4.7 References
[1] B.K. Gan, Z. Taylor, B. Xu, A. Van Riessen, R.D. Hart, X. Wang, P.
Smith, Quantitative phase analysis of bauxites and their dissolution
products, Int. J. Miner. Process. 123 (2013) 64–72.
doi:10.1016/j.minpro.2013.05.005.
[2] G. Bárdossy, Berthier, les Baux et l’histoire de la bauxite, Comptes
Rendus l’Academie Sci. - Ser. IIa Sci. La Terre Des Planetes. 324 (1997)
1031–1040. doi:10.1016/S1251-8050(97)83989-4.
doi:10.1007/s40831-016-0068-2.
[4] L.E. Mordberg, Patterns of distribution and behaviour of trace elements in
bauxites, Chem. Geol. 107 (1993) 241–244. doi:10.1016/0009-
2541(93)90183-J.
[5] E. Deady, E. Mouchos, K. Goodenough, B. Williamson, F. Wall, Rare
Earth Elements in Karst-Bauxites: a Novel Untapped European
Resource?, in: ERES2014 1st Eur. Rare Earth Resour. Conf. (Milos,
Greece, 4-7 Sept. 2014), 2014: pp. 364–375.
http://eres2014.conferences.gr/.
[6] Z. Li, J. Din, J. Xu, C. Liao, F. Yin, T. Lǚ, L. Cheng, J. Li, Discovery of
the REE minerals in the Wulong-Nanchuan bauxite deposits, Chongqing,
China: Insights on conditions of formation and processes, J. Geochemical
Explor. 133 (2013) 88–102. doi:10.1016/j.gexplo.2013.06.016.
[7] G.J. Retallack, Lateritization and bauxitization events, Econ. Geol. 105
(2010) 655–667. doi:10.2113/gsecongeo.105.3.655.
[8] S. Kaya, Y.A. Topkaya, Extraction Behavior of Scandium From a
Refractory Nickel Laterite Ore During the Pressure Acid Leaching
Process, in: Rare Earths Ind. Technol. Econ. Environ. Implic., 2015: pp.
177–188. doi:10.1016/B978-0-12-802328-0.00011-5.
[9] O. Birgul, Sacndium-iron correlation in clay minerals, Earth Planeta~ Sci.
Lett. 55 (1981) 450–452.
doi:10.1016/j.jclepro.2015.02.089.
[11] A.S. Wagh, W.R. Pinnock, Occurrence of scandium and rare earth
elements in Jamaican bauxite waste., Econ. Geol. 82 (1987) 757–761.
doi:10.2113/gsecongeo.82.3.757.
- 90 -

doi:10.1016/j.mineng.2015.01.005.
[13] R. Rudnick, S.G.-T. on Geochemistry, U. 2003, Composition of the
continental crust, Pergamon, Oxford, 2003.
http://adsabs.harvard.edu/abs/2003TrGeo...3....1R�Ü.
[16] K.E. Laintz, C.M. Wai, C.R. Yonker, R.D. Smith, Extraction of metal
ions from liquid and solid materials by supercricitical carbon dioxide,
Anal. Chem. 64 (1992) 2875–2878.
[17] P. Smith, Reactions of lime under high temperature Bayer digestion
conditions, Hydrometallurgy. (2015).
doi:10.1016/j.hydromet.2016.02.011.
[18] G.G. Bánvölgyi, Scale Formation in Alumina Refineries, in: Trav. 45,
Proc. 34th Int. ICSOBA Conf., Quebec, 2016: pp. 1–14.
[19] A. Suss, N. V Kuznetsova, А. Kozyrev, А. Panov, Specific Features of
Scandium Behavior during Sodium Bicarbonate Digestion of Red Mud,
pp. 491–504.
Tkaczyk, V. Vassiliadou, D. Panias, Modes of occurrences of scandium in
Greek bauxite and bauxite residue, Miner. Eng. 123 (2018) 35–48.
doi:10.1016/j.mineng.2018.04.025.
331. doi:10.1007/s40831-016-0060-x.
[22] R.M. Rivera, G. Ounoughene, C.R. Borra, K. Binnemans, T. Van Gerven,
Neutralisation of bauxite residue by carbon dioxide prior to acidic
leaching for metal recovery, Miner. Eng. 112 (2017) 92–102.
doi:10.1016/j.mineng.2017.07.011.
doi:10.1016/0003-2670(94)80250-5.
- 91 -

doi:10.20944/preprints201801.0288.v1.
[25] P.N. Gamaletsos, A. Godelitsas, A. Filippidis, Y. Pontikes, The Rare
Earth Elements Potential of Greek Bauxite Active Mines in the Light of a
Sustainable REE Demand, 2018. doi:10.1007/s40831-018-0192-2.
[26] G. Mongelli, Ce-anomalies in the textural components of Upper
Cretaceous karst bauxites from the Apulian carbonate platform (southern
Italy), Chem. Geol. 140 (1997) 69–79. doi:10.1016/S0009-
2541(97)00042-9.
[27] Z. Maksimovic, G. Panto, Contribution to the geochemistry of the rare
earth elements in the karst-bauxite deposits of Yugoslavia and Greece,
Geoderma. 51 (1991) 93–109. doi:10.1016/0016-7061(91)90067-4.
[29] H.A. Das, J. Zonderhuis, Scandium in rocks, minerals and sediments and
its relations to iron and aluminium, Contrib. Mineral. Petrol. 32 (1971)
231–244.
[30] Z. Liu, Y. Zong, H. Li, Z. Zhao, Characterization of scandium and
gallium in red mud with Time of Flight- Secondary Ion Mass
Spectrometry ( ToF-SIMS ) and Electron Probe Micro- Analysis ( EPMA
), Miner. Eng. 119 (2018) 263–273.
[31] R.D. Shannon, Revised effective ionic radii and systematic studies of
interatomic distances in halides and chalcogenides, Acta Crystallogr.
Sect. A. 32 (1976) 751–767. doi:10.1107/S0567739476001551.
[32] G. Urbain, P.B. Sarkar, Sur les analogies du scandium avec les éléments
des terres rares et avec les éléments trivalents de la famille du fer, Compt.
Rend. 185 (1927) 593–596.
doi:10.1016/j.mineng.2018.01.023.
[34] I.E. Sajó, X-Ray Diffraction Quantitative Phase Analysis of Bayer
Process Solids, in: Int. ICSOBA Conf. Bhubaneswar, India, 2008: pp. 71–
76.
[35] S. Wang, H.M. Ang, M.O. Tadé, Novel applications of red mud as
coagulant, adsorbent and catalyst for environmentally benign processes,
Chemosphere. 72 (2008) 1621–1635.
doi:10.1016/j.chemosphere.2008.05.013.
- 92 -
[36] B.H. Rao, N.G. Reddy, Zeta potential and particle size characteristics of
red mud waste, in: Geoenvironmental Pract. Sustain., Springer Nature
Singapore Pte Ltd., Odisha, India, 2017: pp. 69–89. doi:10.1007/978-981-
10-4077-1.
[37] T.C. Santini, Application of the Rietveld refinement method for
quantification of mineral concentrations in bauxite residues (alumina
refining tailings), Int. J. Miner. Process. 139 (2015) 1–10.
doi:10.1016/j.minpro.2015.04.004.
doi:10.1016/j.hydromet.2011.02.004.
[39] S. Rai, K. Wasewar, A. Agnihotri, Treatment of alumina refinery waste
(red mud) through neutralization techniques: A review, Waste Manag.
Res. 35 (2017) 563–580. doi:10.1177/0734242X17696147.
[40] S.J. Palmer, R.L. Frost, Characterisation of bauxite and seawater
neutralised bauxite residue using XRD and vibrational spectroscopic
techniques, J. Mater. Sci. 44 (2009) 55–63. doi:10.1007/s10853-008-
3123-y.
[41] Y. Ghorbani, M. Becker, J. Petersen, A.N. Mainza, J.P. Franzidis,
Investigation of the effect of mineralogy as rate-limiting factors in large
particle leaching, Miner. Eng. 52 (2013) 38–51.
doi:10.1016/j.mineng.2013.03.006.
[42] Y. Ghorbani, M. Becker, A. Mainza, J.P. Franzidis, J. Petersen, Large
particle effects in chemical/biochemical heap leach processes - A review,
Miner. Eng. 24 (2011) 1172–1184. doi:10.1016/j.mineng.2011.04.002.
[43] S.J. Blott, K. Pye, Particle size scales and classification of sediment types
based on particle size distributions: Review and recommended
procedures, Sedimentology. 59 (2012) 2071–2096. doi:10.1111/j.1365-
3091.2012.01335.x.
doi:10.1016/0003-2670(95)00486-6.
[45] F. Mohd, H. Abdul, S. Stoll, Hematite (α-Fe2O3) – A potential Ce4+
carrier in redmud RalphM., Sci. Total Environ. 622–623 (2018) 849–860.
doi:https://doi.org/10.1016/j.scitotenv.2017.12.043.
[46] S. Reid, J. Tam, M. Yang, G. Azimi, Technospheric Mining of Rare Earth
Elements from Bauxite Residue (Red Mud): Process Optimization,
Kinetic Investigation, and Microwave Pretreatment, Sci. Rep. 7 (2017)
15252. doi:10.1038/s41598-017-15457-8.
- 93 -
[47] W. Lei, P. Linsalata, E.P. Franca, Distribution and mobilization of

cerium, lanthanum and neodymium in The Morro do Ferro Basin, Brazil,
Chem. Geol. 55 (1986) 313–322.
doi:10.1007/s40831-015-0026-4.
[49] M. Kul, Y. Topkaya, I. Karakaya, Rare earth double sulfates from pre-
concentrated bastnasite, Hydrometallurgy. 93 (2008) 129–135.
doi:10.1016/j.hydromet.2007.11.008.
[50] R.D. Abreu, C.A. Morais, Purification of rare earth elements from
monazite sulphuric acid leach liquor and the production of high-purity
ceric oxide, Miner. Eng. 23 (2010) 536–540.
doi:10.1016/j.mineng.2010.03.010.
[51] V.G. Caccia, F.J. Millero, Distribution of yttrium and rare earths in
Florida Bay sediments, Mar. Chem. 104 (2007) 171–185.
doi:10.1016/j.marchem.2006.11.001.
[52] J.L. Graf, Rare earth elements, iron formations and sea water, Geochim.
Cosmochim. Acta. 42 (1978) 1845–1850. doi:10.1016/0016-
7037(78)90239-9.
[53] V. Johan, Z. Johan, Accessory minerals of the Cínovec (Zinnwald) granite
cupola, Czech Republic: indicators of petrogenetic evolution, Mineral.
Petrol. 83 (2005) 113–150. doi:10.1007/BF01159735.
[54] W. Liu, J. Yang, B. Xiao, Review on treatment and utilization of bauxite
residues in China, Int. J. Miner. Process. 93 (2009) 220–231.
doi:10.1016/j.minpro.2009.08.005.
[55] A. Myerson, Handbook of Industrial Crystallization, second, Boston,
2002. https://www.google.com/#cns=1.
[56] C.R. Borra, J. Mermans, B. Blanpain, Y. Pontikes, K. Binnemans, T. Van
Gerven, Selective recovery of rare earths from bauxite residue by
151–159. doi:10.1016/j.mineng.2016.03.002.
[57] Y. Pontikes, G.N. Angelopoulos, Bauxite residue in cement and
cementitious applications: Current status and a possible way forward,
Resour. Conserv. Recycl. 73 (2013) 53–63.
doi:10.1016/j.resconrec.2013.01.005.
- 94 -
Chapter 5 Extraction of REEs from bauxite residue by dry digestion
5. EXTRACTION OF RARE EARTHS FROM

BAUXITE RESIDUE BY DRY DIGESTION
FOLLOWED BY WATER LEACHING
ABSTRACT - In this work, the extraction of selected rare earth elements from
bauxite residue by dry digestion method followed by water leaching was
investigated. Kinetic studies performed with HCl and H2SO4 demonstrated that, at
ambient temperatures, silica dissolution increases with increasing acid
concentration, which leads to the formation of silica gel. Dissolution of silica is
limited to less than 5 wt.% by applying a two-step process: dry digestion of bauxite
residue with HCl or H2SO4, followed by water leaching. The extraction of
aluminium was low because of the low solubility of aluminosilicate compounds.
The extraction of iron and titanium increased with increasing acid concentrations.
High extraction of the rare-earth elements (REEs) were achieved with the HCl-
based dry digestion method, but the concentration in the leachate was limited to
approximately 6 to 8 mg L-1. About 40 wt.% of scandium was recovered with a
high co-dissolution of iron, due to the occurrence of scandium(III) ions in the
lattice matrix of iron(III) oxide. Dry digestion method with multi-stage circulation
of the acid leaching solution significantly increased the REEs concentration up to
20 mg L-1, while achieving an acid consumption of 788 g of HCl per kilogram of
bauxite residue, and a significant reduction of water consumption (60%) relative
to the single-stage acidic leaching method. The low water consumption allows to
increase the filtration efficiency of the leach liquor due to the avoidance of silica
gel formation.
Based on the published paper

R. M. Rivera, B. Ulenaers, G. Ounoughene, K. Binnemans, T. Van Gerven,
“Extraction of rare earths from bauxite residue (red mud) by dry digestion followed by
water leaching”
Minerals Engineering. 119 (2018) doi: 10.1016/j.mineng.2018.01.023.
R. M. Rivera conceived the research, supervised Master’s thesis student B. Ulenaers,
performed part of the analytical work, interpreted part of the results, and wrote the article.
G. Ounoughene contributed with part of the analytical work. K. Binnemans and T. Van
Gerven contributed with their guidance and expertise in this work.
- 95 -
5.1 Introduction
Bauxite residue (BR, also called red mud) is the waste product generated
during alumina production from bauxite by the Bayer process [1]. It is composed
essentially of compounds that are insoluble in concentrated sodium hydroxide
solutions: iron and titanium minerals, undigested alumina minerals, sodium
aluminium hydrosilicates and calcium compounds. It has been estimated that the
annual global production of bauxite residue exceeds 150 million tonnes [2,3] and,
according to numbers from the year 2007, about 2.7109 tonnes have been already
accumulated in tailing ponds, dry stacking and other dry disposal methods [4]. It
is believed that today, this amount has increased to approximately 4109 tonnes
[5].
Bauxite residue represents an interesting source for major elements such
as aluminium, iron and titanium, but also for rare-earth elements (REEs) [6]. The
chemical composition of bauxite residue shows a very wide variability, as it
depends on the origin of the bauxite ore and the operational conditions during the
Bayer process. A range on composition of minerals typically found on bauxite
residue has been reported by Evans [2]. REEs are normally associated to phosphate
(e.g., monazite (REEPO4)) and fluorocarbonate (e.g., bastnäsite (REECO3F))
compounds [7]. Bastnäsite is the most common host for rare-earth oxides in karst
bauxites [7,8]. The enrichment factor of the rare earths in bauxite residue
compared to bauxite is about a factor of two [9]. The concentration of rare earths
in bauxite residue may vary between 500 to 1700 mg kg-1 [10].
Scandium represents 95% of the economic value of the REEs present in
bauxite residue [11]. Ores with a scandium content range between 20 –
50 mg kg-1 are considered as resources and are worthy of exploitation [12]. The
highest concentration of scandium is found in Jamaica bauxite residue, with
concentration as high as 260 mg kg-1 [13]. This concentration is ten times higher
than the average concentration of scandium in the Earth’s crust. Several methods
based on direct leaching by acids have been reported for recovering valuable
metals from bauxite residue [13–16]. The extraction efficiencies depend on the
acid concentration with more extraction achieved at higher acid concentrations,
but the amount of iron dissolution increases with acid concentration as well and
this limits the REE concentration in the leachate [9,11,17,18]. Although
conventional direct acidic leaching of bauxite residue allows a high extraction of
REEs with a low iron dissolution [18], the concentration of REEs in the leachate
(5 – 9 mg L-1) is still about one hundred times lower than the iron concentration
(about 900 mg L-1), which hampers the extraction of REEs in further processing,
for instance in the solvent extraction step [19]. However, the concentration of
REEs in the leachate may be increased by a multi-stage treatment. The method has
demonstrated promising results in terms of acid consumption reduction and
enhancement of leaching efficiency, by which the concentration of metals in the
leach liquor is increased [19–21]. The application of the multi-stage leaching
method to bauxite residue may increase the REE concentration in the leachate,
although a high silica dissolution may also be expected, as well as an accumulation
- 96 -
of major metals in the final leachate. A high silica dissolution leads to a greater
silicon supersaturation index (SSI), which represents the ratio of dissolved silica
with respect to the maximum silica solubility [22]. The increase of SSI is
considered to be the driving force for silica polymerization [23,24]. In HCl or
H2SO4 media, at relatively low pH values, i.e. below the isoelectric point for silica
in the solution (pHiso between 1.7 and 2.2) [25], the hydrolysis of silica occurs very
fast to produce H4SiO4 and H3SiO4- (Eq. 5.1 and 5.2), which are the precursors
for silica gel polymerization [24,26]. Initially, silica monomers (e.g., H4SiO4,
H3SiO4-) polymerize via dimers, trimers, etc. to cyclic oligomers (Si n+1Om+2·OH),
according to Eq. 5.3 and 5.4. These oligomers continue to react until a gel network
is formed via Ostwald ripening, i.e. dissolution of smaller particles and
precipitation on larger particles, which finally results in the formation of an acidic
silica gel [23].
M2 SiO4(s) +4HCl(l) +H2 O(l) → 2MCl2(aq) +H2 O(l) +H4 SiO4 (aq) Eq. 5.1
H4 SiO4 (aq) +H2 O → H3 SiO4− +H3 O+  Eq. 5.2
H3 SiO4− + H4 SiO4 → OH − +H6 Si2 O7 Eq. 5.3
Si𝑛 O𝑚 OH + H4 SiO4 → Si𝑛+1 O𝑚+2 OH + 2H2 O Eq. 5.4
Among others, quartz, sodalite and cancrinite are the most abundant silica
bearing minerals in bauxite residue, although silica gel is produced by the
acidification of amorphous silicate minerals. Such compounds have a higher
solubility in comparison to quartz, which makes these minerals of special interest
for further research [27].
Silica gel formation represents a serious drawback in the extraction of
metals from ores and process residues by hydrometallurgical methods because the
gel solutions can no longer be filtered [28–31]. Additionally, this gelatinous
precipitate may blind ore particles from further dissolution and reduce the leaching
kinetics significantly. According to the literature, silica (and also iron) can remain
undissolved when the bauxite residue is processed by the consecutive combination
of sulfation, roasting and leaching processes, which also allows to achieve a high
selectivity for REEs [19,32]. However, this method is limited by the high energy
consumption due to the decomposition of sulfates during roasting and evaporation
of water during sulfation [33]. Another method, dry digestion, is an effective way
to avoid the dissolution of amorphous silica from silicate minerals mainly because
it may take place at ambient conditions. The method consists in contacting silicate
minerals with strong acids in a water-deficient system, i.e. a high solid
concentration, by which silica is effectively rejected from the respective minerals
[34]. It is stated that in a water-deficient system, the interaction between the metal
and the acid (i.e. formation of MSO4 or MCl2 with concentrated H2SO4 or HCl,
respectively, where M represents the valuable metal) scavenges the available water
- 97 -
of the system, so that no hydration of the silica is possible. This way, the silica
polymerization is avoided, and the dehydrated silica (i.e. SiO2) is readily filterable.
The same principle applies in the so-called acid pugging and curing method, but
the technique requires temperatures in the range of 100 – 200 °C [35]. The dry
digestion method is applied at ambient temperature, which represents its most
remarkable advantage.
Although there exist no literature reports on the application of dry
digestion to bauxite residue, researchers have reported promising results on the
use of this technique for extraction of metals from eudialyte [27,36–38], but also
from different silicate minerals for zinc and manganese extraction [28,34,39,40].
The objective of this paper is to evaluate the extraction of REEs from
bauxite residue by dry digestion with concentrated mineral acids (i.e. HCl, H2SO4),
followed by water leaching. Valuable metals can be effectively washed out by
using this method since silica polymerization does not occur, which significantly
can improve the filterability of the leach liquor. Direct leaching experiments of
bauxite residue are performed with HCl and H2SO4 at different concentrations in
order to study the behaviour of silicon during acid leaching over time. A multi-
stage leaching was applied after dry digestion to increase the REEs concentration
in the leachate. The process has been compared with the conventional direct acidic
leaching method in terms of selected REEs and iron concentration, and acid
consumption.

Mytilineos S.A. - Aluminium of Greece (Agios Nikolaos, Greece). It originates
from a mixture of karst and lateritic bauxites. It was received from the alumina
refinery after dewatering by filter pressing and drying at room temperature. Upon
arrival in the lab, the sample was further dried at 105 ºC for 24 h. Chemical
analysis of the major elements in bauxite residue was performed by wavelength
dispersive X-ray fluorescence spectroscopy (WDXRF, Panalytical PW2400).
Chemical analysis of the minor elements was performed after complete dissolution
of the bauxite residue by alkali fusion and acid digestion in 3 vol.% HNO 3
solution, followed by Inductively Coupled Plasma Mass Spectrometry (ICP-MS,
Thermo Electron X Series) analysis. The alkali fusion was carried out by mixing
0.5 g of bauxite residue with 1.5 g of sodium carbonate and 1.5 g of sodium
tetraborate decahydrate, followed by heating the mixture in a platinum crucible at
1100 °C for 30 min. The mineralogy of the samples was studied by X-ray powder
diffraction (XRD, Bruker D2 Phaser). The obtained data were evaluated with EVA
V.3.1 (Bruker AXS) and quantified with Topas-Academic V.5, using the Rietveld
method.
Leaching experiments of bauxite residue over time were performed in a
150 mL glass reactor with HCl (37 vol.%, Fisher Scientific) and H 2SO4 (95 –
- 98 -
97 vol.%, Sigma-Aldrich) solutions of a fixed concentration (0.5, 1.0 and 1.5 N)

at ambient conditions, i.e. 25 °C. The experimental set-up consists of a hot-plate
magnetic-stirrer device, a pH-electrode (Hamilton, VWR) and a thermocouple
(Pt100, VWR). Within the reactor solid particles were mixed with 100 mL of the
corresponding solution at a liquid-to-solid ratio (L/S) of 5:1. The entire experiment
lasted for 60 min, in which aliquots (about 2 mL) of slurry were extracted from
the glass reactor at specific time intervals. The mixtures were continuously
agitated during the whole experiment to ensure a homogenous suspension. Each
aliquot was filtered using a syringe and a 0.45 µm filter (CHROMAFIL PET-45/25
Polyester), and diluted with 2 vol.% HNO3 for Inductively Coupled Plasma
Optical Emission Spectroscopy (ICP-OES, PerkinElmer Optima 8300) analysis
for silicon. Lutetium was used as an internal standard. A pre-prepared liquid
solution was used as a reference sample during each chemical analysis by ICP-
OES. The chemical composition of the reference sample considers: 30 mg L -1 of
Al and Fe, 18 mg L-1 of Si and Ti, 40 mg L-1 of Na and 0.12 mg L-1 of Sc. The
experiments were conducted at ambient temperature, i.e. 25 °C. The pH of each
sample was measured with the same pH probe. At the end of each experiment, the
remaining reactor content was filtered through a Buchner filter (VWR® Grade 413
Filter Paper, Qualitative, 5 µm retention), and the solid filter cake was dried
(24 h, 105 °C) and stored for further XRD analysis.
Dry digestion experiments were initiated by mixing 2 g of sample with
2 mL of concentrated acid. The desired water-to-acid ratio, W/A, was reached
through dilution with Milli-Q water (18.2 MΩ·cm). The time of digestion and
leaching was evaluated during 24 h. After digestion, the paste was washed with
10 mL of Milli-Q water. The experiments were performed at 25 °C in sealed
polyethylene bottles with constant agitation (200 rpm) using a laboratory shaker
(Gerhardt Laboshake). The pregnant leach solution (PLS) was filtered using a
syringe filter (pore size of 0.45 µm) and diluted with 2 vol.% HNO3 for ICP-OES
analysis of major (Al, Fe, Ti) and minor elements (Sc, Y, La, Nd).
Multi-stage leaching was performed by repeating the dry digestion
process five times, i.e. five stages, each time on a virgin red mud sample, but
instead of leaching with water during the second and following cycles, the leachate
of the previous cycle was used as the liquid phase. Since a small volume of the
leachate is lost during filtration, the volume of the leachate after filtration is less
than 10 mL. Therefore, extra water was added in each cycle to maintain a 5:1 L/S
ratio. Each cycle, about 8.5 mL of the leachate of the previous cycle was mixed
with approximately 1.5 mL of ultrapure water to achieve exactly 10 mL of leach
solution and used as the liquid phase for leaching. Aliquots of PLS (1 mL) were
extracted at the end of each cycle, which were diluted with 2 vol.% HNO3 for ICP-
OES analysis. All other steps were performed in the same way as described above.
All the experiments were carried out in triplicate to ensure reproducibility
of the results. The errors were determined as the standard deviations on the results.
- 99 -
Characterization of the bauxite residue

The chemical analysis of the major elements present in bauxite residue is
shown in Table 5.1. Bauxite residue has a high content of iron and aluminium. In
Table 5.2, the concentrations of scandium, yttrium, lanthanum, and neodymium
are shown because their concentrations are among the highest for the REEs in
bauxite residue. Mineralogical analysis allowed the identification of several
mineral phases (Table 5.3), rich in iron (hematite, goethite), aluminium (gibbsite,
diaspore, bayerite), calcium (calcite, calcium silicates and calcium
aluminosilicates), sodium (sodalite, cancrinite) and titanium (rutile).
Table 5.1: Major chemical components in the bauxite residue.
Element Concentration, wt%

Fe 33
Al 10
Ca 6
Si 3
Ti 3
Na 2
Loss on ignition 8.5
Table 5.2: Selected rare-earth elements composition of the bauxite residue sample [11].
Concentration,
Element
mg kg-1
Sc 121 ± 10
Y 76 ± 10
La 114 ± 15
Nd 99 ± 7
- 100 -
Table 5.3: Mineralogical composition of the bauxite residue sample (estimated relative
error 10%).
Concentration
Phase Chemical formula
wt.%
Hematite Fe2O3 36
Grossular Ca3Al2(SiO4)3 13
Dicalcium-silicate
Ca2SiO4 12
(C2S)
Gibbsite Al(OH)3 8
Diaspore AlO(OH) 7
Tricalcium-silicate
Ca3SiO5 5
(C3S)
Goethite FeO(OH) 5
Calcite CaCO3 4
Bayerite -Al(OH)3 4
Sodalite Na6(Al6Si6O24)·xNaOH·(8−2x)H2O 3
Calcium-aluminium-
Al12CaO27Si4 2
silicate (CAS)
Rutile TiO2 2
Cancrinite Na6Ca2Al6Si6O24(CO3)2·2H2O 1
Silicon dissolution behaviour

The behaviour of silicon dissolution over time was studied during the
direct acid leaching of bauxite residue. The experiments were conducted with
mineral acids at different concentrations. Figure 5.1 describes the effect of HCl
and H2SO4 on the dissolution of silicon as a function of time at different
concentrations. At ambient temperature, the dissolution of silicon in the leach
solution increases with increasing acid concentration, and the maximum
dissolution was obtained between the first 5 to 10 minutes of chemical reaction.
In hydrochloric (or sulfuric) acid media, silicates dissolve from the bauxite residue
by forming H4SiO4 and H3SiO4- according to Eq. 5.1 and 5.2. These monomers
polymerize to cyclic oligomers (Sin+1Om+2·OH), according to Eq. 5.3 and 5.4,
which react until a gel network is formed via Ostwal ripening, i.e. formation of an
acidic silica gel [23].
The lowest silicon dissolution was observed with the lowest acid
concentration due to the reduction of the monomer’s solubility as consequence of
a fast reactivity of the acid with solid particles [25]. The initial concentration of
silicon in the solution was about 2.0 – 2.3 g L-1 when the bauxite residue started
being leached with 0.5 N of HCl and H2SO4, i.e. within the first 3 – 4 minutes,
which represents the minimal concentration of silicon in the leachate for silica gel
formation.
- 101 -
80
0.5 N (a)
1.0 N
1.5 N
60
Si dissolution, wt%
40
20
0
0 10 20 30 40 50 60
Time, min
80
(b)
Si dissolution (wt%)
60
0.5 N
40
1.5 N
20
0
0 10 20 30 40 50 60
Time (min)
Figure 5.1: Effect of (a) HCl and (b) H2SO4 concentration on the dissolution of silicon (Si)
during leaching of bauxite residue sample (L/S: 5, 200 rpm).
The highest silicon dissolution was observed at 1.5 N H2SO4. This

suggests that silicate compounds, such as grossular, C2S and C3S (Figure 5.2), are
much more soluble in H2SO4 than in HCl. The low concentration of these
compounds in the leached samples indicates a favoured dissolution of these
compounds in H2SO4 media rather than in HCl.
- 102 -
15
HCl H2SO4 Raw BR
Composition, wt%
10
0
Grossular C2S C3S
Mineral
Figure 5.2: Concentration of selected mineralogical phases in the solid after acidic leaching
with HCl and H2SO4 (acid concentration: 1.5 N, t: 1 h, T: 25 °C, L/S: 5).
Dry digestion followed by water leaching

Direct acid leaching experiments revealed the importance of mineral acid
concentration during silica gel formation. As it was discussed in the previous
section, silica polymerization can be limited by avoiding the hydrolysis of silica.
The dry digestion method allows to treat silicate minerals with highly concentrated
mineral acids in a water-deficient system at ambient conditions [34]. The effect of
dry digestion of bauxite residue on silica depletion, acid consumption and the
extraction of selected REEs from bauxite residue was evaluated at ambient
conditions in two consecutive experimental processes, according to the diagram
depicted in Figure 5.3: (1) dry digestion of bauxite residue with highly
concentrated mineral acid, and (2) subsequent water leaching and filtration of the
digested bauxite residue sample.
Bauxite
0 Mineral
residue acid
Dry digestion
H2O
Water leaching
Solid
Filtration Residue
Liquid
Pregnant leach
solution
(PLS)
Figure 5.3: Flow sheet for dry digestion of bauxite residue, followed by metal extraction
from the digested bauxite residue by water leaching.
- 103 -
Silica dissolution after dry digestion of bauxite residue

Figure 5.4 shows the effect of acid consumption on the pH and silicon
dissolution after water leaching. The acid consumption increases due to an
increase in the acid concentration, which leads to a dramatic reduction of the pH.
The lowest silicon dissolutions were observed with H2SO4 rather than with HCl.
This was due to the preferential dissolution of silicate compounds in H 2SO4 (see
Figure 5.2). In HCl media, silicon dissolution was reduced from 6 wt.% when the
acid consumption was 400 g HCl/kg BR to about 2 wt.% with an acid consumption
of 1200 g HCl/kg BR. However, dry digestion with sulfuric acid caused only
1.5 wt.% of silicon dissolution with a much lower acid consumption
(620 g H2SO4/kg BR). This is due to the fact that sulfuric acid releases two protons
per molecule, whereas this is only one proton in the case of hydrochloric acid. The
acid consumptions were defined as function of pH by using different mineral acids.
It must be noticed that the dissolution yields achieved by this method were much
lower than the one previously described in section 5.3.2 by direct leaching (Figure
5.1), and even much lower than the values reported in the literature [11,18]. Only
about 5 wt.% of silicon was soluble over the whole acid consumption range. The
standard error of silicon was less than 0.05% with HCl, while with H2SO4 was less
than 0.1%.
- 104 -
4 8
(a)
3 6
Si dissolution, wt%
pH Si
pH 2 4
1 2
0 0
250 500 750 1000 1250 1500
Acid consumption, g HCl/kg BR
4 8
(b)
3 6
Si dissolution, wt%
pH Si
pH
2 4
1 2
0 0
150 300 450 600 750
Acid consumption, g H2SO4/kg BR
Figure 5.4: Effect of (a) HCl and (b) H2SO4 consumption on the pH and silicon (Si)
dissolution after dry digestion-water leaching processing. (BR: bauxite residue, L/S: 5,
tdigestion: 24 h, tleaching: 24 h, T: 25 °C, 200 rpm).
The lowest amount of silicon dissolution achieved with this method was
due to the low amount of water used during the process, which avoids the
hydrolysis of silica. When a silicate compound is in contact with an excess of water
in acidic media, i.e. conventional direct acid leaching of bauxite residue, the
silicon liberation and the subsequent generation of silicic acid (Eq. 5.5) results in
the silica gel formation (Eq. 5.4). However, when the amount of water is restricted,
the reaction proceeds according to Eq. 5.6. As this is an intermediate step, the
partially hydrated metal ion reacts further with silicic acid, diminishing the
condensation of monomeric silicic acid according to Eq. 5.7.
excess H2 O
M2 SiO4(s) +4HCl(l) +H2 O(l) → 2MCl2(aq) + Eq. 5.5
H2 O(l) +H4 SiO4 (aq) 

Eq. 5.6
M2 SiO4 +4HCl +𝑛H2 O → 2MCl2 ∙𝑛H2 O+ H4 SiO4
- 105 -
Eq. 5.7
2MCl2 ∙𝑛H2 O+ H4 SiO4 → 2MCl2 ∙ (𝑛 + 2)H2 O+SiO2 (filterable)
The overall reaction is represented in Eq. 5.8. It is postulated that in a

water-deficient system, the produced chloride scavenges the available water. Thus,
hydration of silica may be restricted and so is the silica polymerization. The
dehydrated silica (i.e., SiO2) can be readily filterable [34,40]. The same principle
is applied in H2SO4 media (Eq. 5.9).
Eq. 5.8
M2 SiO4 +2HCl+∙𝑛H2 O → 2MCl2 ∙ (𝑛 + 2)H2 O+SiO2 (filterable)
M2 SiO4 +2H2 SO4 +∙𝑛H2 O Eq. 5.9

→ 2MSO4 ∙ (𝑛 + 2)H2 O+SiO2 (filterable)
Metal extraction with dry digestion method

The extraction of major and minor elements was evaluated after applying
the dry digestion method. The effect of acid concentration on the extraction of
aluminium, iron and titanium is depicted in Figure 5.5. At lower acid consumption,
aluminium was easily leachable in both acidic media compared to iron and
titanium. The extraction of aluminium increased with higher acid concentrations
until a maximum of about 30 wt.% was reached. As expected, iron and titanium
dissolution gradually increased with the increase in acid consumption. A further
increase in acid concentration did not lead to an improvement of aluminium
extraction, due to the low dissolution of aluminium-silicate compounds and
gibbsite. As it is described in Figure 5.6, these compounds were identified not only
in the raw bauxite residue, but also after dry digestion with HCl and H2SO4.
Grossular demonstrated a higher decomposition after dry digestion with HCl than
with H2SO4, while cancrinite was not identified after leaching, presumably
because it is totally decomposed by dry digestion. The XRD pattern also revealed
that the intensity of gibbsite was increased prominently after dry digestion with
HCl, presumably due to the low solubilty of gibbsite at ambient temperature [41].
- 106 -
40 (a)
Al
Ti
Al, Fe & Ti extraction, wt%

30 Fe
20
10
0
250 500 750 1000 1250
40 (b)
Al
Ti
30
Fe
20
10
0
100 200 300 400 500 600 700
Acid consumption, g H2SO 4/kg BR
Figure 5.5: Effect of acid consumption on the extraction of aluminium (Al), iron (Fe) and
titanium (Ti) after water leaching. Dry digestion with (a) HCl and (b) H2SO4 (L/S: 5, tdigestion:
24 h, tleaching: 24 h, 200 rpm). The standard error of Fe in H2SO4 was below 0.1%.
- 107 -
(a) 1
1
1500
1000 2 1
9 5
7 5 4 8
10 4 3 2 1 8
500 5 9 8 2 4
2 3
4000 (b) 7
1
3000
2 1
Counts
2000 1
9
2 4 2 8
5 9 4 3 2 1 8
1000 8 4
3
3000 1
(c)
1
2000
5 2 5 1
6 8
9 6
7 4 2 2
1000 3 2 4 1 8
9 8 4 3
0
20 30 40 50 60 70
Position 2Ө (Copper)
Figure 5.6: XRD pattern of (a) raw bauxite residue and, water-leached samples after dry digestion with (b) 1000 g HCl/kg BR and (c)
517 g H2SO4/kg BR (BR: bauxite residue, L/S: 5, tdigestion: 24 h, tleaching: 24 h, 200 rpm). 1: Hematite; 2: diaspore; 3: perovskite (CaTiO2);
4: C2S; 5: grossular; 6: bassanite (CaSO4 0.5H2O); 7: gibbsite; 8: goethite; 9: C3S; 10: cancrinite.
- 108 -
Figure 5.7 describes the effect of acid concentration on the leaching of

selected REEs. The extraction of REEs gradually increased with increasing acid
concentration. However, the lowest extraction yields were achieved by using
H2SO4 during the dry digestion stage (Figure 5.7b), which confirms the results
already reported in the literature [11,17,18,42]. The difference in extraction yield
of scandium between dry digestion with HCl and H2SO4 can be explained by the
co-precipitation of scandium with calcium during the chemical transformation of
calcite into bassanite in H2SO4 media (Figure 5.6) [36]. Additionally,
scandium(III) ions are present in the iron(III) oxide lattice [43], which limits its
complete dissolution. Iron dissolution was much higher when dry digestion was
performed with HCl (Figure 5.5) compared to H2SO4 dry digestion, which explains
the higher extraction of scandium in a HCl-based dry digestion procedure upon a
significant amount of iron dissolution (about 25 wt.% in HCl) (Figure 5.8).
Although the elemental correlation between scandium and the major elements is
less clear with H2SO4 in comparison to HCl dry digestion, scandium dissolution
does not depend on the extraction of aluminium, but it may have some correlation
with the dissolution of titanium. It has been reported that REEs, with the exception
of scandium, tend to be associated to perovskite phases, which makes the
dissolution in acidic solutions very difficult [44,45]. No correlation between
yttrium and the other metals was found, which may indicate that yttrium is not
bonded to other metals in this particular bauxite residue, but the extraction depends
only on acid concentration.
- 109 -
60
(a)
Sc Y La Nd
50
Extraction, wt% 40
30
20
10
0
250 500 750 1000 1250
40
(b)
Sc Y La Nd
30
Extraction, wt%
20
10
0
100 200 300 400 500 600 700
Acid consumption, g H2SO 4/kg BR
Figure 5.7: Effect of acid consumption on the extraction of scandium (Sc), yttrium (Y),
lanthanum (La) and neodymium (Nd) after water leaching. Dry digestion with (a) HCl and
(b) H2SO4 (L/S: 5, tdigestion: 24 h, tleaching: 24 h, 200 rpm). The error bars represent the standard
error of the mean for experiments performed in triplicate
- 110 -
40 (a) 8
Al Fe Ti Si

30 6
Si dissolution, wt%
20 4
10 2
0 0
0 10 20 30 40
Sc extraction, wt%
40 (b) 4
Al Fe Ti Si
Fe & Ti extraction, wt%
30 3
Si dissolution, wt%
20 2
10 1
0 0
0 5 10 15 20 25
Sc extraction, wt%
Figure 5.8: Extraction behaviour of scandium (Sc) as function of Al, Fe, Ti and Si
dissolution after water leaching. Dry digestion with (a) HCl and (b) H2SO4 (L/S: 5, tdigestion:
24 h, tleaching: 24 h, 200 rpm). The error bars represent the standard error of the mean for
experiments performed in triplicate. Trend lines do not represent measurement data, but are
added just to aid in detecting trends.
The extraction of neodymium and lanthanum was higher with the HCl-
based dry digestion (about 37 wt.%) in comparison to H 2SO4-based dry digestion
(about 20 wt.%). Leaching of both elements tends to be positively correlated due
to their similarities in atomic and ionic radii (Figure 5.9) [46,47]. However, in HCl
media, the extraction depends on scandium dissolution (Figure 5.10). No trend
between the selected lanthanides and scandium was found in the sample subjected
to dry digestion with H2SO4. (Figure A-II. 1 in the appendix), presumably due to
the occurrence of a double sulfate precipitation [48].
- 111 -
40
H2SO4
30 HCl
La extraction, wt.%
R2 = 0.977
20
R2 = 0.992
10
0
5 10 15 20 25 30 35 40
Nd extraction, wt.%
Figure 5.9: Correlation between extraction of lanthanum (La) and neodymium (Nd). Dry
digestion with HCl and H2SO4 (L/S: 5, tdigestion: 24 h, tleaching: 24 h, 200 rpm). The error bars
represent the standard error of the mean for experiments performed in triplicate.
40 40
R2 - Nd: 0.903
Nd extraction, wt.%
30 30
La extraction, wt.%
20 20
R2 - La: 0.950
La
10 10
Nd
0 0
0 10 20 30 40 50
Sc extraction, wt%
Figure 5.10: Correlation between extraction of lanthanum (La) and neodymium (Nd) with
scandium (Sc) in HCl-based dry-digested bauxite residue sample with (L/S: 5, tdigestion:
experiments performed in triplicate.
Dry digestion method with multi-stage circulation of acid leaching

solution
In order to decrease the water consumption and enhance the leaching
efficiency of REEs, the multi-stage leaching method was studied by considering a
constant acid consumption. According to Figure 5.5 and 5.7, the lowest Al/Fe
leaching ratio (i.e. 2:1), given the ratio between the dissolved amount of Al with
respect to the dissolved amount of Fe, was obtained under the acid consumptions
of 788 g HCl/kg BR and 412 g H2SO4/kg BR. The same acid consumptions allow
- 112 -
a relatively high REEs extraction without bringing too much iron into the solution
(about 4 wt.%). In this leaching technique, the obtained leachate is filtered and
repeatedly contacted again with dry-digested solid samples in multiple leaching
stages (Figure 5.11). Consequently, a more concentrated leachate is obtained and
less water is consumed. The corresponding dry-digested samples were prepared in
each stage by mixing the same mentioned amount of acid. The water leaching step
was always performed in a 5:1 L/S ratio.
Mineral
acid
Bauxite
residue
Dry digestion Dry digestion Dry digestion Dry digestion Dry digestion
H2O
Water leaching Water leaching Water leaching Water leaching Water leaching
Liq. Liq. Liq. Liq. Final

Filtration Filtration Filtration Filtration Filtration
residue
Residue Residue Residue Residue Final PLS
Stage 1 Stage 2 Stage 3 Stage 4 Stage 5
Figure 5.11: Conceptual process flow sheet for dry digestion of bauxite residue with a multi-
stage leaching.
The concentrations of the selected REEs and iron in the leachate after
each stage are summarized in Figure 5.12. A single-stage dry digestion-water
leaching limits the final concentration of REEs in the leachate to approximately 6
to 8 mg L-1. However, the concentration of REEs increased during the following
leaching stage. The treatment of fresh bauxite residue with a pre-defined quantity
of acid allows to obtain the same extraction yields described in section 5.3.5
(Figure 5.7). Moreover, the enrichment was much more significant for the HCl-
based method than for the H2SO4-based method. During multi-stage leaching with
HCl, the scandium concentration increased from 7.3 mg L-1 in stage 1 to
19.3 mg L-1 in stage 5, while the concentration of yttrium, lanthanum and
neodymium increased from an average value of 6.8 mg L-1 in stage 1 to an average
of 18 mg L-1 in stage 5, more than twice the concentration observed by a single-
stage method. During multi-stage leaching with H2SO4, the scandium
concentration increased from 5.4 mg L-1 in stage 1 to 14.2 mg L-1 in stage 5, while
the concentration of yttrium, lanthanum and neodymium increased from an
average value of 3 mg L-1 in stage 1 to a concentration about 8 mg L-1 in stage 5.
Note that the concentration of silicon in both systems, i.e. HCl and H 2SO4,
remained below 0.5 g L-1 (Figure A-II. 2 in the appendix), so that no gel formation
took place during the different stages. Therefore, the multi-stage leaching allows
- 113 -
to successfully increase the concentration of REEs, but also the concentration of

major metals. When HCl was used, a higher concentration of iron was obtained in
comparison to the H2SO4-based method. This was caused by the lower pH values
achieved with HCl in the different stages, but also due to the low solubility of
silicate compounds (Figure 5.2). The concentration of aluminium was similar with
both acids (about 12 g L-1 in stage 5), but the concentration of titanium was slightly
higher with H2SO4 (4 g L-1 in stage 5) in comparison to the concentration achieved
with HCl (0.5 g L-1 in stage 5) (Figure A-II. 2 in the appendix). Purging part of the
leach liquor during the multi-stage treatment of bauxite residue can help to reduce
the concentration of major metals, in particular of iron.
25 (a) 15
REEs concentration, mg L-1
20 Sc Y
Fe concentration, g L-1
La Nd
10
15 Fe
10
5
0 0
1 2 3 4 5
Stage
25 (b) 15
Sc Y
REEs concentration, mg L-1
20
Fe concentration, g L-1
La Nd
Fe 10
15
10
5
0 0
1 2 3 4 5
Stage
Figure 5.12: Concentration of selected rare earths (expressed in mg L-1) and iron (expressed
in g L-1) in the leachate after water leaching of each stage. Dry digestion was performed
with (a) 788 g HCl/kg BR and (b) 412 g H2SO4/kg BR (L/S: 5, tdigestion: 24 h, tleaching: 24 h,
200 rpm).
- 114 -
1.5
HCl
H2SO4
1.0
pH
0.5
0.0
1 2 3 4 5
Stage
Figure 5.13: Change of the pH of the leachate in the different stages of multi-stage leaching
with 788 g HCl/kg BR and 412 g H2SO4/kg BR (L/S: 5, tdigestion: 24 h, tleaching: 24 h,
200 rpm).
The present method allows to process much more bauxite residue with a
significant decrease in water consumption (Figure 5.14). The water consumption
was reduced from 5.5 cm3 H2O/g BR in stage 1 to about 2.3 cm3 H2O/g BR in
stage 5. This was caused by the treatment of fresh bauxite residue and the low
amount of water added during the different stages. An increase in bauxite residue
processing may help to reduce the inventories and costs associated with storing,
which are major concerns for the alumina producers. Nevertheless, this method
still needs more investigation at a larger scale, as the obtained reduction in water
consumption may vary as a function of the chemical and physical properties of the
different bauxite residue samples.
8
Water consumption, cm 3 H2O/g BR
HCl
H2SO4
6
0
1 2 3 4 5
Stage
Figure 5.14: Water consumption in the different stages of multi-stage leaching with 788 g
HCl/kg BR and 412 g H2SO4/kg BR (L/S: 5, tdigestion: 24 h, tleaching: 24 h, 200 rpm).
- 115 -
Table 5.4 compares the results of the direct (single-stage) acidic leaching
and results obtained with the dry digestion method with multi-stage circulation of
the acid leaching solution. Dry digestion method with multi-stage circulation of
solution leads to a reduction in water consumption of about 60 wt.% and 48 wt.%
when HCl and H2SO4 are considered, respectively. During multi-stage leaching,
the enrichment of REEs is three to four times higher than the concentration
achieved with a single-stage leaching method, particularly with HCl, while the
iron concentration increases significantly. Therefore, a severe reduction in the
REE/Fe leaching ratio is obtained, i.e. less selective leaching of REEs. However,
as it was discussed in section 5.3.3, hydrolysis of silica does not occur and
polymerization can be avoided with the dry digestion technique, which may lead
to a significant improvement of the filtration efficiency for further processing of
the leach liquor.
Table 5.4: Comparison between single-stage direct acidic leaching and dry digestion with
a five-stage circulation of acid solution in terms of selected REEs, iron and aluminium
concentration.
Direct Multi- Multi-stage

leaching stage leaching
Parameters
with leaching with H2SO4
H2SO4 with HCl
Acid consumption, g acid/kg BR 438 788 412
Water consumption, cm3 H2O/g BR 5 2 2.6
Scandium concentration, mg L-1 5 19 14
Yttrium concentration, mg L-1 4 19 2.5
Lanthanum concentration, mg L-1 3 18 7.6
Neodymium concentration, mg L-1 2 17 5.3
Iron concentration, g L-1 0.9 7 10
Aluminium concentration, g L-1 2.5 12.5 12
Leaching ratio Al/Fe, g g-1 2.8 1.8 1.2
Leaching ratio selected REE/Fe, mg g-1 15.6 10.4 3.0
The residue generated after leaching is rich in silica and low in sodium
content. It can also be rich in CaSO4 when H2SO4 is considered in the process. For
instance, the residue can be further studied for their applicability in building
materials or cementitious binder.
A preliminary economic analysis of the processes was developed in order
to assess the economic feasibility of pilot-scale test (Table 5.5). The analysis
considered the extraction of titanium and selected rare-earth elements, as oxides,
by the methods described in Table 5.4. The results are given for 1 tonne of bauxite
residue processed by a single unit, i.e. single-stage process. During multi-stage
- 116 -
leaching, 5 stages were considered. It has been reported that the operating cost of
bauxite residue may vary between 4 – 12 USD/tonne, which mainly depends on
the site remediation [2]. However, the consumption of bauxite residue was
considered as a benefit in this preliminary evaluation, as the material is a liability
to the company [49]. The price of acid used in the calculation was taken from
www.alibaba.com, while the price of titanium dioxide was taken from USGS
Minerals [50]. REE prices were taken from mineralprices.com. For the scandium
oxide price two values were considered: 1) market price (accessed on 13/12/2017)
and 2) half of the market price. This is to check the decrease in margin if there is
a price drop for scandium oxide due to increase in production from the proposed
processes. Hence, two margins were calculated. The cost associated to electrical
energy and further recovery processes (e.g. solvent extraction, ion exchange), as
well as labor and equipment was not considered in these preliminary calculations.
Table 5.5: Preliminary comparative economic analysis of the processes studied in this
investigation.
Direct leaching with H2SO4

Mass Unit price Value
(tonne) (USD/tonne) (USD)
Costs
Acid 0.412 -200 -82.4
Water 5 -1 -5
Benefits
Bauxite residue 1 10 10.0
TiO2 1.68E-02 1.5E+03 25.2
Sc2O3 7.42E-05 4.2E+06 311.8
Y2O3 5.58E-05 6.0E+03 0.3
La2O3 3.08E-05 2.0E+03 0.1
Nd2O3 2.42E-05 4.2E+04 1.0
Margin 1 - - 261.0
Margin 2 - - 105.1
Multi-stage leaching with H2SO4
Costs
Acid 2.06 -200 -412
Water 2.6 -1 -2.6
Benefits
TiO2 6.23E-02 1.5E+03 93.5
Sc2O3 1.98E-04 4.2E+06 830.8
Y2O3 6.35E-05 6.0E+03 0.4
La2O3 8.07E-05 2.0E+03 0.2
- 117 -
Nd2O3 5.66E-05 4.2E+04 2.4

Margin 1 - - 562.7
Margin 2 - - 147.3
Multi-stage leaching with HCl
Costs
Acid 3.94 -200 -788
Water 2.6 -1 -2.6
Benefits
TiO2 9.24E-03 1.5E+03 13.9
Sc2O3 2.97E-04 4.2E+06 1247.5
Y2O3 2.42E-04 6.0E+03 1.5
La2O3 2.13E-04 2.0E+03 0.4
Nd2O3 1.95E-04 4.2E+04 8.2
Margin 1 - - 530.9
Margin 2 - - -92.9
The profit margins of these methods are shown in Figure 5.15. It is

evident from this Figure that recovery of scandium is very important. The results
show that the margins are not going below zero at the current price of scandium
oxide. Instead, the margins are significantly improved by considering multi-stage
leaching, in particular with H2SO4. This is mainly because of the high consumption
of bauxite residue and the low consumption of water with multi-stage leaching.
Furthermore, multi-stage leaching allows to decrease the large volumes of
effluents generated during direct leaching, which also leads to an improvement of
the profit margin. The margin described by multi-stage leaching with H2SO4 is
slightly higher than the margin depicted by multi-stage leaching with HCl, because
of the lower acid consumption and higher titanium extraction. However, the profit
margin does not show a significant improvement when half of the scandium price
is considered. Indeed, the margin goes below zero when multi-stage leaching with
HCl is applied. This is due to the low amount of titanium extracted with HCl
media, despite the high amount of extracted scandium. Therefore, the
simultaneous recovery of REEs and titanium oxide is the key factor to ensure a
positive profit margin.
- 118 -
600
Margin 1
400 Margin 2
USD/tonne BR
200
-200
Direct leaching Multi-stage Multi-stage
with H2SO4 leaching leaching
with H2SO4 with HCl
Figure 5.15: Profit margin for three leaching methods at current REEs price (source:
http://mineralprices.com, accessed on13th of December 2017).
5.4 Conclusions
Kinetic studies demonstrated that, at constant temperatures, the

dissolution of silicon in the leachate increases with increasing acid concentration,
due to the high solubility in water of the orthosilicic acid at decreasing pH. High
silicon dissolution leads to an increase of the silica saturation index and, therefore,
to the formation of silica gel. Reducing the hydrolysis of silica is key to avoid
silica polymerization during acidic leaching of bauxite residue.
The two-step processing of bauxite residue by dry digestion and
subsequent water leaching demonstrated a high extraction of REE while at the
same time preventing silica polymerization. The extraction of aluminium, iron and
titanium increased with increasing acid concentration. However, only up to
30 wt.% of aluminium can be recovered by this method due to the poor solubility
of alumino-silicate compounds in water. The HCl-based dry digestion method
allowed the extraction of much more REEs, particularly scandium, compared to
H2SO4-based dry digestion. However, about 40 wt.% of scandium was recovered
alongside 25 wt.% of iron dissolution, due to their simultaneous occurrence in the
lattice matrix of iron(III) oxide. A single-stage dry digestion-water leaching limits
the final concentration of REEs in the leachate to approximately 6 to 8 mg L-1, but
it was significantly increased up to 20 mg L-1 by considering the same acid
consumption in a multi-stage leaching, particularly with HCl. By considering the
dry digestion and water leaching methods as processes integrated in a multi-stage
treatment, much more bauxite residue can be processed for REEs extraction
without demanding too much water for the process. The low water consumption
allows to increase the filtration efficiency of the leach liquor due to the avoidance
of silica gel formation.
The preliminary comparative economic analysis shows that dry digestion
with H2SO4 and multi-stage leaching is the most interesting method for the
- 119 -
extraction of scandium and titanium from bauxite residue. However, this method
needs to be studied in pilot-scale. The method looks promising, but further studies
are required to increase the extraction of scandium.
5.5 References
331. doi:10.1007/s40831-016-0060-x.
doi:10.1180/minmag.2016.080.052.
doi:10.1016/j.hydromet.2011.02.007.
aluminium.org.
doi:10.1016/j.jclepro.2015.02.089.
[7] G. Mongelli, Ce-anomalies in the textural components of Upper
Cretaceous karst bauxites from the Apulian carbonate platform (southern
Italy), Chem. Geol. 140 (1997) 69–79. doi:10.1016/S0009-
2541(97)00042-9.
[8] Z. Maksimovic, G. Panto, Contribution to the geochemistry of the rare
earth elements in the karst-bauxite deposits of Yugoslavia and Greece,
Geoderma. 51 (1991) 93–109. doi:10.1016/0016-7061(91)90067-4.
doi:10.1016/0003-2670(94)80250-5.
[10] A. Akcil, N. Akhmadiyeva, R. Abdulvaliyev, A. Meshram, P. Meshram,
Overview On Extraction and Separation of Rare Earth Elements from
Red Mud : Focus on Scandium Overview On Extraction and Separation
- 120 -
of Rare Earth Elements from Red Mud : Focus on Scandium, Miner.

Process. Extr. Metall. Rev. (2017) 1–7.
doi:10.1080/08827508.2017.1288116.
doi:10.1016/j.mineng.2015.01.005.
[12] X. Shaoquan, L. Suqing, Review of the extractive metallurgy of
scandium in China (1978-1991), Hydrometallurgy. 42 (1996) 337–343.
doi:10.1016/0304-386X(95)00086-V.
doi:10.1007/s40831-016-0068-2.
[14] N. Zhang, H.-X. Li, X.-M. Liu, Recovery of scandium from bauxite
residue—red mud: a review, Rare Met. 35 (2016) 887–900.
doi:10.1007/s12598-016-0805-5.
recovery from various resources: A review, Hydrometallurgy. 108
(2011) 100–108. doi:10.1016/j.hydromet.2011.03.001.
[16] Z. Liu, H. Li, Metallurgical process for valuable elements recovery from
red mud—A review, Hydrometallurgy. 155 (2015) 29–43.
doi:10.1016/j.hydromet.2015.03.018.
doi:10.1016/0003-2670(95)00486-6.
[18] R.M. Rivera, G. Ounoughene, C.R. Borra, K. Binnemans, T. Van
Gerven, Neutralisation of bauxite residue by carbon dioxide prior to
acidic leaching for metal recovery, Miner. Eng. 112 (2017) 92–102.
doi:10.1016/j.mineng.2017.07.011.
[19] B. Onghena, C.R. Borra, T. Van Gerven, K. Binnemans, Recovery of
scandium from sulfation-roasted leachates of bauxite residue by solvent
extraction with the ionic liquid betainium
bis(trifluoromethylsulfonyl)imide, Sep. Purif. Technol. 176 (2017) 208–
219. doi:10.1016/j.seppur.2016.12.009.
[20] X. Zhu, Y. Zhang, J. Huang, T. Liu, Y. Wang, A kinetics study of multi-
stage counter-current circulation acid leaching of vanadium from stone
coal, Int. J. Miner. Process. 114–117 (2012) 1–6.
doi:10.1016/j.minpro.2012.07.001.
[21] Z.M. Xia, M.T. Tang, S.H. Yang, Z.M. Xia, M.T. Tang, S.H. Yang,
- 121 -
Materials balance of pilot-scale circulation leaching of low-grade zinc

oxide ore to produce cathode zinc, Can. J. Metall. Mater. Sci. 54 (2015)
439–445. doi:10.1179/1879139515Y.0000000016.
[22] P. Kokhanenko, K. Brown, M. Jermy, Silica aquasols of incipient
instability: Synthesis, growth kinetics and long term stability, Colloids
Surfaces A Physicochem. Eng. Asp. 493 (2016) 18–31.
doi:10.1016/j.colsurfa.2015.10.026.
doi:10.1016/j.gca.2009.06.002.
568–590. doi:10.3390/en7020568.
(2015) 2504–2521. doi:10.3390/polym7121528.
379. doi:10.1016/0022-3093(86)90503-X.
[27] B. Friedrich, M. Hanebuth, S. Kruse, A. Tremel, D. Vossenkaul, Method
for opening a eudialyte mineral, Patent Number EP2995692 A1, 2016.
http://www.google.com/patents/EP2995692A1?cl=en.
doi:10.1016/j.hydromet.2016.08.010.
doi:10.1016/j.hydromet.2014.08.001.
(2016) 5–15. doi:10.1016/j.hydromet.2016.06.023.
Can. Metall. Q. 25 (1986) 201–209.
[32] C.R. Borra, J. Mermans, B. Blanpain, Y. Pontikes, K. Binnemans, T.
Van Gerven, Selective recovery of rare earths from bauxite residue by
- 122 -
151–159. doi:10.1016/j.mineng.2016.03.002.
2111-y.
[34] R. Dufresne, Quick leach of siliceous zinc ores, JOM. 28 (1976) 8–12.
[35] T.E. Amer, M.A. Mahdy, N.T. El Hazek, Application of acid pugging
and ferric salts leaching on west central Sinai uraniferous siltstone, in:
M. and P. Canadian Institute of Mining (Ed.), Int. Symp. Process Metall.
Uranium, Saskatoon, Saskatchewan, Canada, 2000: pp. 445–461.
Hydrometallurgical processing of eudialyte bearing concentrates to
recover rare earth elements via low-temperature dry digestion to prevent
the silica gel formation, J. Sustain. Metall. 3 (2017) 79–89.
doi:10.1007/s40831-016-0084-2.
[37] B. Friedrich, D. Voßenkaul, S. Stopic, Leaching of high silica containing
ores - preventing gel-formation by dry digestion - a case study on
Eudialyte ores, in: ERES2017 2nd Eur. Rare Earth Resour. Conf., 2017:
pp. 79–89. doi:10.1007/s40831-016-0084-2.
[38] P. Davris, S. Stopic, E. Balomenos, D. Panias, I. Paspaliaris, B.
Friedrich, Leaching of rare earth elements from eudialyte concentrate by
doi:10.1016/j.mineng.2016.12.011.
[39] D.R. Groot, D.M. Kazad, H. Pollmann, J.P. Villiers, T. Redtmann, J.
Steenkamp, The recovery of manganese and generation of a valuable
residue from ferromanganese slags by hydrometallurgical route, in:
Thirteen. Int. Ferroalloys Congr. - Effic. Technol. Ferroalloy Ind., 2013:
pp. 1051–1060.
[40] D.M. Kazadi, D.R. Groot, J.D. Steenkamp, H. Pöllmann, Control of
silica polymerisation during ferromanganese slag sulphuric acid
digestion and water leaching, Hydrometallurgy. 166 (2016) 214–221.
doi:10.1016/j.hydromet.2016.06.024.
[41] A.C. Zhao, Y. Liu, T.A. Zhang, G.Z. Lü, Z.H. Dou, Thermodynamics
study on leaching process of gibbsitic bauxite by hydrochloric acid,
Trans. Nonferrous Met. Soc. China (English Ed. 23 (2013) 266–270.
doi:10.1016/S1003-6326(13)62455-3.
[42] R. Boudreault, J. Fournier, D. Primeau, M.-M. Labrecque-Gilbert,
Process for treating red mud, 2015. doi:10.1016/j.(73).
- 123 -

[44] F. Zheng, F. Chen, Y. Guo, T. Jiang, A.Y. Travyanov, G. Qiu, Kinetics
of Hydrochloric Acid Leaching of Titanium from Titanium-Bearing
Electric Furnace Slag, JOM. 68 (2016) 1476–1484. doi:10.1007/s11837-
015-1808-7.
[45] O.P. Shrivastava, N. Kumar, I.B. Sharma, Solid state synthesis and
structural refinement of polycrystalline La x Ca 1 – x TiO 3 ceramic
powder, Bull. Mater. Sci. 27 (2004) 121–126. doi:10.1007/BF02708493.
[46] C.R. Borra, J. Mermans, B. Blanpain, Y. Pontikes, T. Van Gerven,
Selective leaching of rare earths from bauxite residue after sulphation
roasting, in: Bauxite Residue Valorization Best Pract. Leuven,
(Belgium), 5-7 Ocotber 2015, Pp. 301-307, 2015.
Chem. Geol. 55 (1986) 313–322.
016-0060-x.
[50] M. Bedinger, Titanium and titanium dioxide, 2017. Commod. Summ. 1,
172–173
- 124 -
Chapter 6 REEs extraction using HPAL of bauxite residue slags
6. SELECTIVE RARE EARTH ELEMENT

EXTRACTION USING HIGH-PRESSURE
ACID LEACHING OF SLAGS ARISING
FROM THE SMELTING OF BAUXITE
RESIDUE
ABSTRACT - During acid leaching of bauxite residue (red mud), the

increase in dissolution of rare-earth elements (REEs) is associated with a
substantial co-dissolution of iron; this poses problems in the downstream
processing (i.e. solvent extraction or ion exchange). Six different slags generated
by reductive smelting of the same bauxite residue sample were treated by high-
pressure acid leaching (HPAL) with HCl and H2SO4 to selectively extract REEs.
Thus, up to 90 wt.% of scandium was extracted from the slags using H 2SO4 at
150 ºC, while with HCl the extraction of scandium reached up to 80 wt.% at
120 ºC. The extraction of yttrium, lanthanum and neodymium was above 95 wt.%
when HCl was used as a reagent, but it was much lower (< 20 wt.%) with H 2SO4,
presumably due to the formation of a double sulfate (NaLn(SO 4)2nH2O) and/or
due to the adsorption on the surface of silicon/aluminium-oxides compounds. In
addition, HPAL of bauxite residue slags led to a significant co-dissolution of
aluminium (> 90 wt.%, 18 g L-1), while the concentration of the remaining iron
was of 3 g L-1 in the leachate. The co-dissolution of silicon and titanium was lower
than 5 wt.%.
Based on the submitted paper

R. M. Rivera, B. Xakalashe, G. Ounoughene, K. Binnemans, B. Friedrich, T. Van Gerven,
“Selective rare earth element extraction using high-pressure acid leaching of slags arising
from the smelting of bauxite residue”
Hydrometallurgy.
R. M. Rivera conceived the research, performed the analytical work, interpreted the results,
and wrote the article. B. Xakalashe contributed with part of the analytical work. G.
Ounoughene, K. Binnemans, B. Friedrich and T. Van Gerven contributed with their
guidance and expertise in this work.
- 125 -
6.1 Introduction
Over 95% of the worldwide produced alumina is processed by the Bayer

process [1]. The process consists in contacting aluminium-bearing ores with
NaOH at temperatures between 150  250 ºC, in autoclaves or tubular reactors, at
pressures of up to about 4 MPa [2]. As the process is carried out at high pH values
(about 14), aluminium hydroxides are selectively dissolved from bauxite minerals,
while most of the other compounds remain insoluble in the bauxite residue.
Annually, about 1.35 tonnes of bauxite residue are generated per tonne of alumina
produced [3,4]. It is believed that the inventories of bauxite residue have increased
up to 4 billion tonnes, with an annual growth rate of approximately 160 million
tonnes [5].
Bauxite residue is mainly composed of metallic oxides of aluminium,
iron and titanium, but also of rare-earth elements (REEs) [6]. Soda and silica are
also present. The chemical composition is very wide, as it depends on the origin
of the mineral ore and also on the operational conditions used during the
production of purified aluminium hydroxide from bauxite minerals. A typical
range of component in bauxite residue has already been reported elsewhere [3].
Rare-earths are contained in minor extent, but their concentration can be as high
as 1300 mg kg-1 [7]. Scandium represents about 95% of the economic value of the
REEs present in bauxite residue, and its concentration can be as high as
260 mg kg-1 [8,9]. This concentration is higher than the average abundance in the
Earth’s crust (22 mg kg-1), and significantly enriched compared to that in the
original bauxite ore [10].
Although conventional acid leaching with different mineral acids (e.g.,
HCl, H2SO4, HNO3) can partially or fully extract REEs from bauxite residue, the
leaching process is controlled by intra-particle diffusion as metals are transported
from the solid phase to the solution through the cracks and pores present in the
particles [11,12]. However, the co-dissolution of silicon and iron represents a
drawback in the downstream processing (e.g., solvent extraction or ion exchange).
A high concentration of silicon in the leach solution can lead to silica gel
formation, which significantly affects the leaching efficiency as the gel solution
can no longer be filtered. It is believed that these aggregates may act as a resistance
for the acid solution as they are partially adsorbed on the surface of bauxite residue
minerals [13]. Meanwhile, the co-dissolution of iron is detrimental as it is difficult
to separate it from the rare-earths, particularly from scandium, requiring a large
quantity of reagents during the downstream processing. Scandium(III) ions are
present in the iron(III)-rich oxide lattice [14,15], which limits its complete
dissolution [16,17]. Therefore, iron must be removed in advance in order to
improve the extraction and selectivity of REEs.
Smelting of bauxite residue leads to the reduction of iron oxides to
produce pig iron, i.e. a metallic product generated from a smelting furnace with an
iron content usually above 90 wt.% [18,19], and a REE-rich slag [20–22]. The
process strongly depends on flux consumption and/or high operating temperature
- 126 -
due to the high concentration of alumina. The relatively large volume of fluxes
may increase not only the energy consumption during smelting, but also the acid
consumption during leaching. The removal of alumina by alkali roasting before
the smelting process was proposed by Borra et al. to diminish the flux consumption
[22,23]. Although the method allows for the recovery of alumina as a valuable by-
product, the roasting step must be carried out at temperatures of about 500 ºC
(when NaOH is used as a reducing agent), while the smelting stage is carried out
at 1500 ºC without using fluxes for iron removal. Although the energy
consumption of the process is about 3.5 GJ tonne-1, i.e. similar to the energy
consumption reported by direct (reductive) smelting of bauxite residue [24], the
volume of slag generated by this process is about 50% lower than the volume of
slag produced by the direct smelting process, which limits the amount of REEs
that can be further recovered. Reductive smelting allows the separation of iron
from the rare-earths which can be highly enriched, together with a significant
amount of aluminium and silicon, in the slag phase. REEs can then be extracted
by leaching the slag with mineral acids. The extraction efficiency of REEs from
the slag during leaching, however, is determined by their mineralogical association
during solidification of the slag after smelting. It has been reported that perovskite
(CaTiO3) phase, formed during solidification, has the capability of converting the
lanthanides into titanate solid solutions, CaxLn(1-x)TiO3, which is very difficult to
leach, but its formation can be avoided with a rapid cooling rate (or rapid
solidification) of the slag [23,25]. Nonetheless, little is known about the
mineralogy and chemical speciation of REEs in slags arising from bauxite residue
smelting. Hence, the association of REEs to other major elements present in the
slag, such as aluminium and/or silicon, cannot be underestimated during acid
leaching. During the recovery of valuable metals by acid leaching of bauxite
residue, silica gel can be formed due to the leaching of silicate minerals, which
significantly affects the filtration efficiency of the leach liquor, i.e. reduced
filterability. It is stated that under acidic conditions, below the isoelectric point for
silica in the solution (i.e. pHiso between 1.7 and 2.2) [26], the hydrolysis of silica
occurs very fast to produce silica monomers, H4SiO4 and H3SiO4- according to
chemical reactions Re 6.1 – 6.3, which tend to form cyclic oligomers
(Si(𝑛+1) O(𝑚+2) OH) via Ostwald ripening (Re 6.4) until a gel network is formed
[27–29].
M2 SiO4(s) + 4HCl(l) + H2 O(l) → 2MCl2(aq) + H2 O(l) + H4 SiO4(aq) Re 6.1
M2 SiO4(s) + 2H2 SO4(l) + H2 O(l) Re 6.2

→ 2MSO4(aq) + H2 O(l) + H4 SiO4(aq)
H4 SiO4(aq) + H2 O(l) → H3 SiO4− + H3 O+ Re 6.3
Si𝑛 O𝑚 OH + H4 SiO4(aq) → Si(𝑛+1) O(𝑚+2) OH + 2H2 O Re 6.4
- 127 -
It has been reported that during conventional acid leaching of bauxite

residue at room temperature, a substantial decomposition of silicate compounds is
achieved, which promotes the silicon dissolution and, consequently, the leach
solution can no longer be filtered due to the polymerization of silica monomers
(see Chapter 5) [13,30]. Due to the high concentration of silica in the slag, room-
temperature acid leaching can no longer be considered because of silica
polymerization. Hence, high-pressure acid leaching (or HPAL), i.e. acid leaching
performed at high temperature in an autoclave, cannot only avoid problems
associated to silica gel formation, but also to the depletion of other major metal
such as titanium due to its hydrolysis at high temperature [34]. Therefore, the acid
consumption can be reduced and the extraction of REEs can be done selectively.
This process has demonstrated promising results in the recovery of scandium from
refractory lateritic nickel ore [31,32], but also in the recovery of other valuable
metals from nickel converter and smelter slags [33,34].
In this work, the bauxite residue sample was mixed with different
proportions of lignite coke, silica and lime, and smelted in an electric arc furnace
for iron separation, so that different slags with low iron content were generated.
The slags were cooled down at two different cooling rates. Subsequently, the
leaching behaviour of selected rare earth elements (Sc, Y, La, Nd) by high-
pressure acid leaching of the slag was analysed. Firstly, the REEs recovery by acid
leaching of a slag with the highest content of aluminium, silicon and iron was
studied. The acid concentration of two different mineral acids (HCl and H2SO4)
and the leaching temperature were investigated. Secondly, the optimal conditions
for REEs recovery from bauxite residue slags were established in terms of
maximizing the extraction yield of REEs with the lowest dissolution of major
elements (aluminium, iron, silicon, titanium). Thirdly, the untreated bauxite
residue and the remaining slags (also generated by reductive smelting) were
studied under optimal conditions to evaluate the effect of different chemical
composition, mineralogy and morphology on the recovery of REEs. The treatment
of bauxite residue and bauxite residue slags by HPAL was compared in terms of
selected REEs, iron, aluminium and titanium concentration.
The bauxite residue used in this work was provided by Mytilineos S.A. -
Aluminium of Greece (Agios Nikolaos, Greece). It was generated predominantly
from Greek (karst) bauxite ore. Chemical analysis of the major elements (Al, Ti,
Fe, Ca, Si, Na) was performed using wavelength-dispersive X-ray fluorescence
spectroscopy (WDXRF, Panalytical PW2400). The concentration of selected rare
earth elements (Sc, Y, La and Nd) were obtained by lithium metaborate (LiBO 2)
fusion with nitric acid (3 vol.% HNO3) digestion followed by Inductively Coupled
Plasma Mass Spectrometry (ICP-MS, Thermo Electron X Series) analysis.
The reductive smelting of bauxite residue was performed with three
different mixtures of fluxes according to the distribution presented in Table 6.1.
- 128 -
The corresponding amount of fluxes, i.e. CaO and SiO2, were determined by
FactSage 7.0 software [35] based on three different slag melting points. The major
slag forming oxides (Al2O3, CaO, SiO2 and TiO2) were considered in the
calculations. Na2O was not considered in the calculations because of its low
amount in the sample and its volatile behaviour during smelting. FeO x was not
used in the calculations as it is reduced to metallic iron during smelting. Each flux
was prepared on the base of 1000 g of bauxite residue. The smelting reduction
experiments were carried out in an electrical arc furnace (100 kVA direct current)
at a temperature of 1500 ± 50 °C during 1 h. With each mixture of bauxite residue
+ flux + coke, three different slags were generated: slag I, II and III, respectively.
After heating, the molten material was cooled down at two different cooling rates:
(1) quenching with cold water (pouring the slag in water at room temperature,
cooling rate about 1400 ºC min-1, fast cooling) and (2) room temperature cooling
by keeping the slag in the crucible (cooling rate about 30 ºC min-1, slow cooling).
Hence, the slag I subjected to fast and slow cooling are described as “slag I.FC”
and “slag I.SC”, respectively. The same denomination was considered for slags II
and III. The iron metal fraction produced in the smelting experiment was separated
from the slag and kept for further analysis. The slag was crushed into small pieces
(< 2 cm) with a hammer and was subsequently crushed in a laboratory jaw crusher
(Retsch BB 51) to produce material 100% finer than 1 mm. Later on, all the
material was ground in a planetary ball mill (tungsten carbide grinding balls of
2 cm, 4 min at 510 rpm) to reduce the particle size to 100% < 400 μm (P 80 = 135
± 30 mm, P60 = 77 ± 22 mm). Small iron particles in the slag sample were removed
by a NdFeB magnet after grinding. The magnet was covered with a plastic layer,
and it was passed through the surface of the milled slag before sieving. These
metallic pieces were subsequently removed and weighed. This procedure was
repeated several times until no more magnetic particles were found attached to the
magnet. The chemical analysis procedure for the slag is the same as the one
described for the bauxite residue. Figure 6.1 depicts the flow sheet for reductive
smelting of bauxite residue, followed by high-pressure acid leaching of the slag
arising during the smelting process.
High-pressure acid leaching experiments were carried out in a titanium
autoclave (Parr Company, series 4560, 400 mL capacity) varying the temperature
between 60 and 180 ºC, in separate experiments, that led to an increase of the
pressure by increasing the temperature. Consequently, the pressure increased from
0.3 bar at 60 °C until 15 bar at 180 °C when H 2SO4 was used, while with HCl the
pressure increased from 0.3 bar at 60 °C until 10 bar when the temperature was
180 °C. The experiments were performed with a liquid-to-solid ratio, L/S, of 10:1
to ensure a substantial leaching concentration of REEs in the leach liquor with a
reasonable consumption of bauxite residue. It must be noted that the utilization of
low L/S-ratios does not enhance the recovery of REEs, with a concentration of less
than 1 wt.% in bauxite residue, due to the limited mass transfer as consequence of
the high pulp density [16,34]. Moreover, a ratio of 10:1 has been proposed as an
optimal L/S-ratio for bauxite residue leaching [36]. Analytical reagent grade
H2SO4 (95 – 97 vol.%, Sigma–Aldrich) and HCl (37 vol.%, Fisher Scientific) were
- 129 -
used in the present study as leaching reagents. The slurries after leaching were
filtered using filter paper (pore size 0.45 µm) and diluted with 2 vol.% HNO 3
(65 vol.%, Chem-lab) for Inductively Coupled Plasma Optical Emission
Spectroscopy (ICP-OES, PerkinElmer Optima 8300) analysis of major (Al, Fe, Ti,
Si) and minor elements (Sc, Y, La, Nd). The corresponding extraction yield of
metal was calculated according to Eq 6.1:
𝐴𝑚𝑜𝑢𝑛𝑡 𝑜𝑓 𝑚𝑒𝑡𝑎𝑙 𝑖𝑛 𝑡ℎ𝑒 𝑙𝑒𝑎𝑐ℎ𝑎𝑡𝑒 Eq 6.1

%=
𝑇𝑜𝑡𝑎𝑙 𝑎𝑚𝑜𝑢𝑛𝑡 𝑜𝑓 𝑚𝑒𝑡𝑎𝑙 𝑖𝑛 𝑡ℎ𝑒 𝑠𝑎𝑚𝑝𝑙𝑒
The samples before and after leaching were embedded in epoxy resin and
polished with SiC abrasive paper down to 1200 grit size followed by polishing
with diamond paste (6, 3, and 1 µm) on a cloth disk. Then the samples were coated
with platinum and analysed with a Scanning Electron Microscope (SEM-EDX,
Philips XL30). The mineralogy of the samples before and after leaching was
studied by X-ray powder diffraction (XRD, Bruker D2 Phaser). The obtained data
were evaluated with EVA V.3.1 (Bruker AXS).
- 130 -
Table 6.1: Flux distribution, melting temperature and cooling rates for reductive smelting of bauxite residue.
Flux, g Melting
Bauxite Cooling rate Slag
Mixture Lignite temperature,
residue, g CaO SiO2 of slag (b), name (b)
coke(a) °C
I 1000 200 - 100 1350 Fast Slag I.FC

I 1000 200 - 100 1350 Slow Slag I.SC
II 1000 100 150 100 1390 Fast Slag II.FC
II 1000 100 150 100 1390 Slow Slag II.SC
III 1000 80 100 100 1485 Fast Slag III.FC
III 1000 80 100 100 1485 Slow Slag III.SC
(a) Lignite coke was composed by 87 wt.% of carbon, 10 wt.% ash content
(b) Fast cooling (FC) = 1400 °C min-1; Slow cooling (SC) = 30 °C min-1
- 131 -
Bauxite Flux
residue (CaO, SiO2, coke)
Slag
Electrical arc furnace
(1450 ± 50 ºC; 15 kVA, 1 h)
Fast cooling Slow cooling

(quenching with cold water, (room temperature,
20 ºC min-1) 30 ºC h-1)
Crushing and grinding
Metallic Magnetic separation

Fe (NdFeB magnet)
Diluted mineral acid
(1 - 3 N H2SO4 and HCl)
High pressure
leaching
(60 - 180 ºC; L/S: 10; 1 h)
Solid
Filtration Residue
Liquid
Pregnant leach
solution
(REEs)
Figure 6.1: Flow sheet for reductive smelting of bauxite residue, followed by high-pressure
acid leaching of the slag arising during the smelting process.
Characterisation of the bauxite residue and slags

The cumulative particle size distributions of the bauxite residue and the
slags samples produced after reductive smelting are shown in Figure 6.2. In the
bauxite residue, 90 vol.% of the particles are smaller than 21 µm and 50 vol.% of
the particles are smaller than 0.06 µm. All the slags generated after smelting were
crushed and milled under the same conditions. Coarser particles were obtained
with the slag subjected to fast cooling in comparison to the slag subjected to slow
cooling. Thus, the slags subjected to fast cooling resulted in 90 vol.% of the
particles < 230 µm and 50 vol.% of the particles < 75 µm. The slags subjected to
slow cooling, on the other hand, resulted in 90 vol.% of the particles < 180 µm
and 50 vol.% < 45 µm. A fast cooling led to a slag with an amorphous morphology
(Figure 6.3), of which the particles easily tend to agglomerate when the slags are
ground in a planetary ball mill [37,38]. The same grinding conditions were applied
to the other slags. However, due to the crystallinity of the slags formed after slow
- 132 -
cooling (Figure 6.4), agglomeration does not occur, but just a particle size
reduction.
Figure 6.2: Cumulative particle size distribution of the bauxite residue and the slags
generated after melting (FC: fast cooling; SC: slow cooling; Slag I, II and III produced with
mixture I, II and III, respectively).
The composition of major elements in the bauxite residue used in this

study and in the slag generated after reductive smelting experiments is shown in
Table 6.2. Iron is the main constituent in the bauxite residue sample (about
33 wt.%). Iron was significantly removed during the smelting experiment, by
which its concentration in the slag is low (< 4 wt.%). The highest silicon
concentration was obtained in slags II and III due to the addition of an excess of
SiO2 during smelting. The slag I resulted in the highest content of calcium due to
the largest addition of CaO. In average, aluminium and titanium were enriched by
a factor of 2 in the slag with respect to the original bauxite residue.
Table 6.2: Major chemical components (in wt.%) in the bauxite residue and in slag samples
generated after smelting experiments.
Bauxite Slag Slag Slag Slag Slag Slag

Element
residue I.FC I.SC II.FC II.SC III.FC III.SC
Al 9.6 20.2 20.0 16.9 17.8 19.3 19.1
Ca 6.1 27.2 27.1 17.7 15.9 16.9 16.3
Si 3.4 4.6 4.7 11.8 12.3 10.4 10.1
Ti 3.5 5.2 5.5 6.1 5.4 5.5 5.2
Na 2.1 1.8 0.9 2.2 2.0 2.1 2.3
Fe 32.7 1.5 2.5 2.0 2.4 2.5 3.7
FC: Fast cooling; SC: Slow cooling; Slag I, II and III produced with mixture I, II and
III, respectively.
- 133 -
In Table 6.3, the concentrations of selected REEs (i.e. Sc, Y, La and Nd)
are shown. The concentrations of all the REEs in the bauxite residue and the slags
are shown in Table A-III.1, in the appendix. The concentration of these selected
REEs are among the highest in the bauxite residue and in the slags. The
concentration of REEs in the slags was increased by a factor of about 1.4 compared
to the concentration of REEs in the bauxite residue sample.
Table 6.3: Selected rare earth elements (in mg kg-1) in the bauxite residue and in the slag
samples generated after smelting experiments.
Elemen Bauxite Slag Slag Slag Slag Slag Slag

t residue I.FC I.SC II.FC II.SC III.FC III.SC
Sc 97.8 125.5 122.5 117.4 128.3 128.1 124.0
± 1.7 ± 2.0 ± 4.9 ± 1.4 ± 1.7 ± 1.5 ± 1.6
Y 65.7 117.4 118.2 92.4 105.2 107.1 105.6
± 1.6 ± 1.1 ± 5.2 ± 0.7 ± 0.8 ± 1.6 ± 1.3
La 143.9 211.9 206.9 159.2 167.9 174.8 168.7
± 3.0 ± 5.2 ±18.0 ± 1.3 ± 2.7 ± 1.1 ± 1.5
Nd 64.8 126.1 126.9 103.7 111.2 116.3 111.8
± 2.6 ±2.1 ± 5.2 ± 1.0 ± 3.9 ± 2.2 ± 0.7
FC: Fast cooling; SC: Slow cooling; Slag I, II and III produced with mixture I, II and
III, respectively.
Figure 6.3 and 6.4 exhibit the XRD pattern of the slags generated after
reductive smelting. It is stated that large cooling rates lead to high glass transition
temperature (Tg), i.e. temperature range at which the molten material freezes into
an amorphous solid without a discontinuity in the volume and the heat capacity
and, consequently, allows the system a short period of time to relax. With low
cooling rates, however, the molten material freezes at the melting temperature into
a crystalline solid, with an abrupt discontinuity in both the volume and the heat
capacity [39]. The figures show that fast-cooled slags (Figure 6.3), i.e. cooling rate
of about 20 °C min-1, resulted in less crystalline mineral phases compared to the
slags that were cooled down slowly at room temperature (Figure 6.4), i.e. cooling
rate of about 30 °C h-1. However, the particles of the slag subjected to fast cooling
easily tend to agglomerate when the slags were ground in a planetary ball mill
(Figure 6.2) [37,38]. The same grinding conditions were applied to the other slags,
but due to the crystallinity of the slags formed after slow cooling (Figure 6.4),
agglomeration did not occur, but only a particle size reduction was obtained
(Figure 6.2). Thus, from the slags that were cooled down slowly, slag I, which was
produced with addition of 20 wt.% CaO as flux, was mainly characterised by the
presence of gehlenite and perovskite, although calcium magnesium silicate was
also identified. Meanwhile, in slags II and III, similar mineralogical phases were
recognized due to the use of CaO and SiO2 as flux in slightly different proportions.
Consequently, yoshiokaite, nepheline and tricalcium aluminate were the main
mineral phases formed after cooling. The slag II subjected to fast cooling,
- 134 -
however, resulted in an almost completely amorphous structure. The mineral

phases identified in the slags differ significantly from the minerals and compounds
identified in the original sample of bauxite residue: hematite and goethite (iron
minerals); gibbsite and diaspore (aluminium minerals); calcite (calcium mineral);
and cancrinite. SEM analysis on the smelted slag III.SC indicated that small
particles of iron were locked in the slag phase, particularly in the Ca-Al-Si-Ti
matrix phase (Figure 6.5). These iron particles could not be liberated from the slag
matrix due to the very small size (< 50 µm).
5000 Slag I
Counts
G
2500
G P
G P G
0
10 20 30 40 50 60 70
5000
Slag II
Counts
2500
Fe
0
10 20 30 40 50 60 70
5000 Slag III

Counts
2500 NY
Y Y T Fe T T
0
10 20 30 40 50 60 70
Position 2q (Copper)
Figure 6.3: XRD pattern of fast-cooled (FC) slags (G: gehlenite (Al2Ca2O7); P: perovskite
(CaTiO3); Fe: metallic iron; Y: yoshiokaite (Al5.4Ca2.7O16Si2.7); N: nepheline
(AlNa0.98O4Si); T: tricalcium aluminate (Ca3Al2O6)).
- 135 -
10000 C Slag I
7500
Counts
5000
G ZC P G G
2500 C C
P PG
0
10 20 30 40 50 60 70
10000
Slag II
7500 Y
Counts
Y N P
5000 T
Y N Ti Y T
2500 Ti P P
0
10 20 30 40 50 60 70
10000 Slag III

7500
Counts
5000 Y T
Y P T T
2500 Y N N Y Ti P P
0
10 20 30 40 50 60 70
Figure 6.4: XRD pattern of slow-cooled (SC) slags (G: gehlenite (Al2Ca2O7); C: calcium
magnesium silicate (Ca2MgO7Si2); P: perovskite (CaTiO3); Z: zirconium oxysulfide
(ZrOS); Ti: calcium-silico-titanate (CaO5SiTi); Y: yoshiokaite (Ca7.5Al15SiO32); N:
nepheline (AlNa0.98O4Si); T: tricalcium aluminate (Ca3Al2O6)).
Ca-Al-Si-Ti
matrix
Fe
Resin
50 µm
Figure 6.5: SEM image of slag III.SC (10 kV, 200x, 10 WD).
In view of these results, slag III subjected to slow cooling, i.e. slag III.SC,
represents the most unfavourable material to recover rare-earths by acid leaching,
due to the high content of silicon, aluminium and iron. It must be noted that during
- 136 -
conventional acid leaching, at room temperature, an over-saturation of silica in

acid solution can lead to silica polymerization due to the hydrolysis of the silica
monomers, i.e. H4SiO4 and H3SiO4-, formed during the decomposition of silicate
compounds [27,29,40]. Silica gel formation represents a serious drawback in
hydrometallurgy because the gel solutions can no longer be filtered [36,41–44].
Meanwhile, a high co-dissolution of aluminium and iron can significantly decrease
the efficiency of the further separation process, e.g., solvent extraction and/or ion
exchange, to recover selectively the REEs due to their co-adsorption to the ion-
exchange resins [45]. Therefore, slag III.SC will be considered from now on as the
main subject of this research. The remaining slags as well as bauxite residue were
then studied afterwards under the experimental conditions that turned out to be the
best for slag III.SC.
HPAL of a slag rich in silicon, aluminium and iron

The slag III.SC was the one with the highest content of silicon,
aluminium and iron. This slag was leached with two different mineral acids (HCl
and H2SO4) to evaluate the selectivity of the different elements. The effect of the
temperature on the extraction of the major elements (Al, Fe, Ti and Si) with H2SO4
is shown in Figure 6.6. As it has been reported in the literature, the dissolution of
silicon in the leach solution increases with increasing acid concentration [18,30].
Thus, the highest dissolution of silicon was about 80 wt.% at 60 °C with 3 N
H2SO4. Further increase in temperature led to a significant drop in silicon
dissolution, particularly at temperatures above 130 °C (< 0.5 wt.%), with the same
acid concentration. This is because at high temperatures, water is released as
vapour and, therefore, hydration of silica, i.e. formation of silicic acid, is
substantially reduced. Hence, the partially hydrated metal ion reacts further with
silicic acid, diminishing the condensation of monomeric silicic acid according to
Re 6.5. The dehydrated silica precipitates and is readily filterable [34,46,47].
M2 SiO4(s) + 2H2 SO4(l) + (4𝑛 − 2)H2 O(l) Re 6.5

→ 2MSO4 ∙ 𝑛H2 O(l) + SiO2(filterable)
As a result, the concentration of silicon in solution was less than

0.5 g L-1 and no traces of silica gel formation was observed. The leaching process
with low acid concentration resulted in a similar behaviour. It must be noted that
less than 1 wt.% of silicon dissolution was obtained at 100 °C with 1 N H2SO4 (pH
 2.5). However, with 3 N H 2SO4 (pH  0.7), the same level of dissolution was
observed just at 150 °C. This is because, with low acid concentration, the
monomer’s solubility is reduced as consequence of a fast reactivity of the acid
with the solid particles. Moreover, high acid concentration and temperature
enhance the decomposition of silicate compounds that lead to an increase of silicon
dissolution [26,48]. High acid concentrations also led to high co-dissolution of
aluminium and iron, which decrease the selectivity of the leaching process in terms
of REEs recovery. Between 60 to 100 °C, about 80 – 95 wt.% of aluminium was
- 137 -
leached (ca. 18 g L-1), while between 65 – 80 wt.% of iron was transferred into the
solution (ca. 3 g L-1). A further increase in the leaching temperature, however,
reduces the dissolution of aluminium, most probably caused by the lack of acid
available for leaching due to the chemical precipitation of Ca2+ ions as CaSO4 and
the partial decomposition of silicate compounds, which phenomena have been
reported earlier to occur in the presence of H2SO4, but not with HCl [17,49].
Although aluminium sulfate can also precipitate at high temperatures due to the
evaporation of water [50]. The slight reduction of iron dissolution, in the range of
temperature between 150 – 180 °C, can be explained by the formation of sodium-
jarosite that was detected with low intensity only in the solid residue obtained after
leaching at 180 °C. Sodium was totally dissolved when the leaching temperature
was 100 °C, with a maximum concentration of 3 g L-1 in solution. A further
increase in the leaching temperature caused sodium to precipitate, as its
concentration at 180 °C was reduced till 0.9 g L-1 (about 40 wt.%). The average
iron concentration was 2.8 ± 0.3 g L-1. The pH of the solution during leaching with
3 N H2SO4 was about 0.7 ± 0.1. Therefore, the precipitation of sodium and the
acid conditions could explain the formation of jarosite at high temperatures [51].
100 1 N H2SO4
Extraction, wt.%
80
60
40
20
0
40 60 80 100 120 140 160 180 200
100
3 N H2SO4
Extraction, wt.%
80
60
40
20
0
40 60 80 100 120 140 160 180 200
Temperature, ºC
Al Fe Si Ti Ca Na
Figure 6.6: Effect of the temperature and H2SO4 concentration on the dissolution
of Al, Fe, Ti and Si from slag III.SC (L/S: 10, t: 1 h).
The extraction of titanium was < 5 wt.%, particularly at high

temperatures. This low dissolution of titanium has been attributed by different
- 138 -
authors to the formation of TiOSO4, which tends to have a low solubility [52,53],
but this phase was not detected during this investigation. However, the XRD
analysis of the solid residue obtained after HPAL with H 2SO4 (Figure 6.7)
confirmed that high acid concentration led to a large decomposition of silicate
compounds, but also to the formation of CaSO 4. The formation of CaSO4 can
hinder the dissolution of titanium-containing mineral phases, such as perovskite
(CaTiO3) and/or calcium-silico-titanate (CaO5SiTi), but also Ca/Al/Si-compounds
such as yoshiokaite (Ca7.5Al15SiO32) and tricalcium aluminate (Ca3Al2O6) (Figure
6.4), as it was confirmed by SEM analysis. Figure 6.8 describes the elemental
distribution of the residue obtained after HPAL with H2SO4, while the quantitative
analysis of elements is described in Table 6.4. The presence of a Ca-S-O phase
(presumably CaSO4) tends to enclose partially or totally the titaniumoxygen
phase. The elemental mapping also confirmed the simultaneous occurrence of the
Ca-S-O phase with silicon and aluminium, which could also explain the reduction
of aluminium extraction at temperature above 100 ºC. Although calcium-
containing phases transform to their sulfate form easily, calcium dissolution is low
due to the fact that calcium sulfates have very low solubility in water, i.e. 2 g L -1
at 20 ºC [54].
10000 Untreated
7500 Slag III.SC
Counts
5000 W T
Y P T T
2500 M N N Y Ti P P
0
10 20 30 40 50 60 70
10000 M W 1 N H2SO4
7500 An
Counts
5000 An Y
An An
P B
2500
0
10 20 30 40 50 60 70
10000 An
3 N H2SO4
7500
Counts
G
5000 M W
B An T B
2500 An
0
10 20 30 40 50 60 70
Figure 6.7: XRD pattern of slag III.SC before and after HPAL with H2SO4 (sample at 150
ºC) (Y: yoshiokate (Al5.4Ca2.7O16Si2.7); N: nepheline (AlNa0.98O4Si); T: tricalcium
aluminate (Ca3Al2O6); Ti: calcium-silico-titanate (CaO5SiTi); An: anhydrite (CaSO4); W:
wollastonite (CaSiO3); G: gehlenite (Al2Ca2O7); P: perovskite (CaO3Ti); M: mayenite
(Ca12Al14O33); B: bassanite (CaSO4 0.5H2O)).
- 139 -
Figure 6.8: Back-scattered electron (BSE) image of the leached slag III.SC with H2SO4 at
150 ºC, and elemental distribution of Al, Ti, S, Ca, Si and O.
Table 6.4: SEM-EDX (average) semi-quantitative analysis results of elements in Figure 6.8.
Concentration,
Element
wt.%
Al 6±5
Ti 5±2
Na 0.5 ± 0.1
Ca 9±8
Si 7±5
S 7±5
O 34 ± 6
In Figure 6.9, the effect of the temperature and acid concentration on the
extraction of selected REEs (Sc, Y, La, Nd) after leaching with H 2SO4 is shown.
With low acid concentration, i.e. 1 N H2SO4, the extraction of scandium and
yttrium was similar between 60 – 100 ºC, i.e. 32 wt.% (ca. 4 mg L-1), but it was
reduced when the temperature was increased up to 180 ºC, presumably due to the
- 140 -
co-precipitation with jarosite (MFe3(SO4)2(OH)6, where M can be Na+, NH4+,

H3O+, Li+, K+), as scandium can replace iron in its crystalline lattice structure [51].
The extraction of lanthanum and neodymium was lower than 15 wt.%
over the whole temperature range. It must be noted that with 1 N H2SO4 the
extraction of REEs is very low due to the high pH of the solution (~ 2.5) [17].
Moreover, in acidic media, most of the REEs start dissolving only at pH values
below 2, while the dissolution of major elements starts occurring at pH values
below 5 [16]. The highest extraction yield of REEs was obtained with 3 N H2SO4
between 60 – 100 ºC. At temperatures higher than 100 ºC, a gradual reduction of
REEs recovery, with the exception of scandium, was observed. Several researches
have reported a reduction of the lanthanides dissolution due to the formation of a
sodium-lanthanide-double sulfate compound, i.e. NaLn(SO 4)2nH2O, which has
been observed in solid residues obtained after acid leaching of REEs-containing
compounds (e.g., phosphogypsum, monazite and eudialyte minerals, pre-
concentrated bastnasite mineral) with H2SO4 [55–58]. Although, sodium
dissolution described a substantial reduction at temperatures above 120 ºC (Figure
6.6), the sodium-lanthanide-double sulfate compound could not be detected by
SEM-EDX none by XRD in our sample, probably due to the low concentration of
the lanthanides and sodium (about 0.5 wt.%). Therefore, it is believed that
lanthanum and neodymium remain associated to the perovskite (CaTiO 3) phase,
which was not totally dissolved after leaching, and it is known that can host REE
in its matrix [25,59,60]. The yttrium (and other heavy rare-earths) double sulfate
is very soluble and, therefore, the decrease of yttrium extraction could be caused
by the adsorption on silicon- and/or aluminium-oxide minerals, such as mayenite
(Ca12Al14O33), yoshiokaite (Al5.4Ca2.7O16Si2.7), tricalcium aluminate (Ca3Al2O6)
and/or gehlenite (Al2Ca2O7), which were identified in the sample after leaching
(Figure 6.7). This phenomenon has been described earlier in the literature, and is
supported by the observed similarity in extraction behaviour with increasing
temperature between yttrium and aluminium, but also with silicon [61]. It is
believed that due to the similar behaviour with yttrium, lanthanum and neodymium
can also be adsorbed on the surface of the same mineral phases [62]. The reduction
of scandium leaching at temperatures above 120 °C could be caused by the
insufficient amount of acid produced by the chemical transformation of CaCO 3
into CaSO4. The formation of jarosite could also explain the low recovery of
scandium at high temperatures [51].
- 141 -
100
1 N H2SO4
Extraction, wt.%
80
60
40
20
0
40 60 80 100 120 140 160 180 200
Temperature ºC
100
Extraction, wt.%
80
60
40 3 N H2SO4
20
0
40 60 80 100 120 140 160 180 200
Temperature, ºC
Sc Y La Nd
Figure 6.9: Effect of the temperature and H2SO4 concentration on the extraction of selected
REEs (Sc, Y, La, Nd) from the slag III.SC (L/S: 10, t: 1 h).
Figure 6.11 describes the elemental distribution of the residue obtained

after HPAL with HCl, while the quantitative analysis based on the elemental
mapping by SEM-EDX is shown in Table 6.5. As it was observed before in the
residue obtained after leaching with H2SO4 (Figure 6.8), titanium was also found
simultaneously with silicon and aluminium in the residue obtained after leaching
with HCl. The low extraction of titanium could be caused by its chemical
association to the perovskite phase, which in this case was not totally decomposed
after HPAL with HCl (see Figure 6.12).
- 142 -
100 1 N HCl
Extraction, wt.%
80
60
40
20
0
40 60 80 100 120 140 160 180 200
100
Extraction, wt.%
80
3 N HCl
60
40
20
0
40 60 80 100 120 140 160 180 200
Temperature, ºC
Al Fe Si Ti Ca Na
Figure 6.10 depicts the effect of the temperature and HCl concentration
on the extraction of the major elements (Al, Fe, Ti and Si). High HCl concentration
led to a significant dissolution of aluminium and iron, which is further enhanced
by increasing the temperature. Silicon dissolution remained < 1 wt.% due to the
release of water as vapour at high temperature, which inhibits the generation of
silica gel (Re 6.6), similar to the case of HPAL with H2SO4.
M2 SiO4(s) + 4HCl(l) + (4𝑛 − 2)H2 O(l) Re 6.6

→ 2MCl2 ∙ 𝑛H2 O(l) + SiO2(filterable)
Figure 6.11 describes the elemental distribution of the residue obtained

after HPAL with HCl, while the quantitative analysis based on the elemental
mapping by SEM-EDX is shown in Table 6.5. As it was observed before in the
residue obtained after leaching with H2SO4 (Figure 6.8), titanium was also found
simultaneously with silicon and aluminium in the residue obtained after leaching
with HCl. The low extraction of titanium could be caused by its chemical
association to the perovskite phase, which in this case was not totally decomposed
after HPAL with HCl (see Figure 6.12).
- 143 -
100 1 N HCl
Extraction, wt.%
80
60
40
20
0
40 60 80 100 120 140 160 180 200
100
Extraction, wt.%
80
3 N HCl
60
40
20
0
40 60 80 100 120 140 160 180 200
Temperature, ºC
Al Fe Si Ti Ca Na
Figure 6.10: Effect of the temperature and HCl concentration on the dissolution of
Al, Fe, Ti and Si from the slag III.SC (L/S: 10, t: 1 h).
- 144 -
Figure 6.11: Back-scattered electron (BSE) image of the leached slag III.SC with HCl at
120 ºC, and elemental distribution of Al, Ti, Ca, Si and O.
Table 6.5: SEM-EDX (average) semi-quantitative analysis results of elements in Figure

6.11.
Concentration,
Element
wt.%
Al 5±4
Ti 4±6
Ca 2±7
Na 1 ± 0.2
Si 8±3
O 31 ± 8
- 145 -
10000 Untreated
7500 Slag III.SC
Counts 5000 W T
Y P T T
2500 M N N Y Ti P P
0
10 20 30 40 50 60 70
10000
7500 1 N HCl
Counts
T
5000
P T T
2500 P P
0
10 20 30 40 50 60 70
10000
3 N HCl
7500
Counts
5000
A P
2500 P
0
10 20 30 40 50 60 70
Position 2 (Copper)
Figure 6.12: XRD pattern of slag III.SC before and after HPAL with HCl (sample at 120
ºC) (Y: yoshiokaite (Al5.4Ca2.7O16Si2.7); N: nepheline (AlNa0.98O4Si); T: tricalcium
aluminate (Ca3Al2O6); Ti: calcium-silico-titanate (CaO5SiTi); W: wollastonite (CaSiO3); P:
perovskite (CaO3Ti); A: anatase (TiO2); M: mayenite (Ca12Al14O33)).
The effect of the temperature and the HCl concentration on the extraction
of scandium, yttrium, lanthanum and neodymium is shown in Figure 6.13. With
1 N HCl, the extraction yield of yttrium, lanthanum and neodymium significantly
increased when the temperature was increased. Neodymium and lanthanum
showed a comparable leaching behaviour, as they occur only in the +III valence
state and tend to have a similar mineralogical distribution in bauxite residue
[30,63]. However, scandium was scarcely leached out from the solid matrix with
low HCl concentration, due to its association to the tricalcium aluminate
(Ca3Al2O6) mineral phase. According to the XRD analysis shown in Figure 6.15,
Ca3Al2O6 was detected in similar scanned angles as that of the perovskite (CaO3Ti)
phase, not only in the unprocessed slag sample, but also in the solid residue
obtained after leaching with 1 N HCl. However, among the REEs, the scandium
concentration in the perovskite phase can be almost negligible [15]. Previous
reports have shown that it remains associated to aluminium compounds [20,23].
Ca3Al2O6 was not detected in the solid residue obtained after leaching with 3 N
HCl, presumably due to its total decomposition, which may explain the high REEs
recovery at temperatures > 100 °C. The decrease of scandium recovery above
120 ºC could be caused by its adsorption on silica, which started to precipitate at
temperatures above 100 ºC (Figure 6.10). This phenomenon has been reported in
the literature and could be confirmed by these results [64].
- 146 -
100
1 N HCl
Extraction, wt.%
80
60
40
20
0
40 60 80 100 120 140 160 180 200
Temperature ºC
100
Extraction, wt.%
80
60 3 N HCl
40
20
0
40 60 80 100 120 140 160 180 200
Temperature ºC
Sc Y La Nd
Figure 6.13: Effect of the temperature and HCl concentration on the extraction of selected
REEs (Sc, Y, La, Nd) from the slag III.SC (L/S: 10, t: 1 h).
REEs recovery by HPAL from other slags of bauxite residue smelting

Other slags than III.SC, presented in Table 6.2 and 6.3, were processed
by HPAL under the conditions that were optimised for slag III.SC (see section
6.3.2). To recapitulate, the highest extraction yield of REEs was obtained with an
acid concentration of 3 N, with both acids: HCl and H 2SO4. The optimal
temperature was defined in terms of the lowest concentration of aluminium, iron,
titanium and silicon in the solution, in order to enhance the selectivity for REEs
recovery. Thus, in the case of HPAL with HCl, the lowest aluminium/iron (Al/Fe)
and scandium/iron (Sc/Fe) leaching ratios (i.e. 1.1 and 0.9, respectively) were
obtained at 120 ºC with a high REEs extraction (Sc: 79 wt.%; Y, Nd and La > 97
wt.%) and very low concentration of titanium and silicon in the solution
(< 0.5 g L-1). In the case of HPAL with H2SO4, the lowest Al/Fe and Sc/Fe (i.e.
0.9 and 1.1) ratios were obtained at 150 ºC, at which the Sc/REEs ratio was the
highest one (i.e. 3) and, therefore, at high selectivity for scandium recovery
(REEs was considered as the sum of yttrium, lanthanum and neodymium
extraction).
The extraction of selected REEs (Sc, Y, La, Nd) from the remaining slags
by HPAL, with HCl and H2SO4, are summarized in Figure 6.14 and 6.16,
respectively. Similar to the behaviour of slag III.SC, HPAL with HCl showed a
high recovery of REEs. Meanwhile, with H2SO4, scandium was preferentially
- 147 -
dissolved over the other REEs, presumably due to the formation of a CaSO 4
product layer on the surface of the solid particles (see section 6.3.2, Figure 6.8).
The solid residue obtained after leaching with H2SO4 was highly crystalline due
to the formation of CaSO4 (Figure 6.17). The solid residue after leaching with HCl,
on the contrary, resulted in a quite amorphous structure, and unveiled the presence
of titanium-rich phases, which were formed even with a fast cooling rate. It must
be noted that Borra et al. was able to suppress the CaTiO3 formation by quenching
the slag produced during smelting of a solid residue with very low aluminium
content (< 1 wt.%), which was generated by alkali roasting followed by water
leaching. In this investigation, the use of a pre-treatment process prior to reductive-
smelting was not considered. According to Figure 6.15, titanium-rich phases (i.e.
perovskite, anatase, rutile) are less soluble in HCl-media, as their peaks were the
most evident in the corresponding diffractograms of XRD analysis. The highest
recovery of scandium (ca. 85 wt.%) was obtained from slag II subjected to a fast
cooling (slag II.FS in Table 6.2 and 6.3), which was the most amorphous one
among all the other slags (Figure 6.3) and, therefore, the easiest one to leach.
Therefore, the treatment of this slag by HPAL can be further optimized by
reducing the leaching temperature and/or the acid concentration. The figures also
show that, from slag I (formed by the mere use of CaO as flux during smelting of
bauxite residue), the extraction yield of scandium was limited up to 60 wt.%,
presumably due the formation of CaSO 4 during leaching, but also due to the
presence of grossite (aluminium rich) and laihuinite (iron rich) compounds (Figure
6.17), as scandium can be part of their corresponding lattice matrix. On the other
hand, scandium was mostly extracted from slags II and III, particularly the ones
subjected to slow cooling rates, which were characterised by the presence of
yoshiokaite and nepheline (both aluminium-silicate compounds), and leached with
H2SO4. This can be explained by the preferential decomposition of such
compounds in H2SO4 rather than with HCl [30]. In slag I, however, aluminium
was found together with calcium in the mineral phase gehlenite, but not associated
with silicon, which can also explain the low scandium recovery from this slag.
Therefore, it is believed that aluminium-calcium compounds may have a lower
solubility in acidic media compared to aluminium-silicate compounds, which may
limit the recovery of scandium and other REEs from the slags.
- 148 -
Fast cooling
100
80
Extraction, wt.%
60
40
20
0
I II III
Slag
Slow cooling
100
80
Extraction, wt.%
60
40
20
0
I II III
Slag
Sc Y La Nd
Figure 6.14: Extraction of selected REEs (Sc, Y, La and Nd) after HPAL with HCl of fast
and slow cooled slag from bauxite residue by reductive smelting (acid concentration: 3 N,
120 ºC, L/S: 10, t: 1 h).
Subjected to fast cooling Subjected to slow cooling
5000 10000
Slag I Slag I
7500
Counts
Counts
P
2500 A
5000
P P
P 2500 P
P
0 0
10 20 30 40 50 60 70 10 20 30 40 50 60 70
5000 10000
Slag II Slag II
7500
Counts
Counts
R
2500 5000
P P Lw A Gr
2500 Lw P Fe
P
0 0
10 20 30 40 50 60 70 10 20 30 40 50 60 70
5000 10000
Slag III Slag III
7500
Counts
Counts
2500 A 5000
P P A P
2500 P
0 0
10 20 30 40 50 60 70 10 20 30 40 50 60 70
Position 2q (Copper) Position 2q (Copper)
Figure 6.15: XRD pattern of different slags treated by HPAL with 3 N HCl at 120 ºC (L/S:
10, t: 1 h; A: anatase (TiO2); P: perovskite (CaTiO3); Fe: metallic iron; Lw: lawrencite
(FeCl2); R: rutile (TiO2); Gr: grossite (Al4CaO7)).
- 149 -
Fast cooling
100
80
Extraction, wt.%
60
40
20
0
I II III
Slag
Slow cooling
100
80
Extraction, wt.%
60
40
20
0
I II III
Slag
Sc Y La Nd
Figure 6.16: Extraction of major Al, Fe and selected REEs (Sc,Y, La and Nd) after
HPAL with H2SO4 of fast and slow cooled slags from bauxite residue by reductive
smelting (acid concentration: 3 N, 150 ºC, L/S: 10, t: 1 h).
- 150 -
Subjected to fast cooling Subjected to slow cooling
An L
5000 10000
An Slag I An
Gr 7500 Slag I
Counts
Gr An
An An B An
2500 5000 B
An B B
2500 An An
0 0
10 20 30 40 50 60 70 10 20 30 40 50 60 70
5000 L
10000 Gy
An
An Slag II An Slag II
7500 Gy
Counts
Counts
B
2500 Gr An Gr 5000 B
An An B
An Y
2500 An B
0 0
10 20 30 40 50 60 70 10 20 30 40 50 60 70
B B An An
5000 A 10000
Slag III An
G
Slag III
Gi 7500
Counts
Gr
2500 Q 5000 Y W
B An T B
Y 2500 An An An
0 0
10 20 30 40 50 60 70 10 20 30 40 50 60 70
Position 2q (Copper) Position 2q (Copper)
Figure 6.17: XRD pattern of different slags treated by HPAL with 3 N H 2SO4 at 150 ºC
(L/S: 10, t: 1 h; An: anhydrite (CaSO4); Gr: grossite (Al4CaO7); L: laihuinite
(Fe2+Fe3+·2(SiO4)2); Gi: gismondine (Ca2Al4Si4O16·9(H2O); Q: quartz (SiO2); B: bassanite
(CaSO4 0.5H2O); A: anatase (TiO2); T: tricalcium aluminate (Ca3Al2O6); G: gehlenite
(Al2Ca2O7); M: mayenite (Ca12Al14O33); Gy: gypsum (CaSO4·2H2O); Y: yoshiokaite
(Ca7.5Al15SiO32)).
Direct treatment of bauxite residue by HPAL versus HPAL of bauxite

residue slags and other acid leaching processes
The direct treatment of Greek bauxite residue by HPAL was performed
with HCl and H2SO4 in order, according to the diagram presented in Figure 6.18,
to evaluate the recovery of REEs in the presence of a high content of iron
(33 wt.%, Table 6.2). HPAL with H2SO4 (Figure 6.19) allows a relatively high
recovery of REEs up to 100 °C, without bringing too much iron into the solution
(< 10 wt.%), but with a substantial co-dissolution of silicon (ca. 60 wt.%). A
further increase of the temperature led to ca. 70 wt.% of scandium recovery,
presumably due to the decomposition of the iron-rich mineral phases, e.g.,
hematite, goethite, which are the main host minerals for scandium in this particular
bauxite residue [15]. Nevertheless, the recovery of other REEs was relatively low
(< 45 wt.%) due to the high dissolution of sodium (> 40 wt.%) that may lead to
the formation of double sulfates [55-58]. HPAL with HCl (Figure 6.20) allowed
the precipitation of silicon and titanium at temperatures above 100 ºC, beside high
recoveries for scandium (> 70 wt.%) and other REEs (La: 67 – 80 wt.%, Nd and
Y > 88 wt.%). The substantial co-dissolution of iron and aluminium, however, was
unavoidable at high temperatures and acidic conditions. Nevertheless, HPAL of
(untreated) bauxite residue should still be investigated in detail.
- 151 -
Bauxite Diluted mineral acid

residue (1 - 3 N H2SO4 and HCl)
High pressure
leaching
(60 - 180 ºC; L/S: 10; 1 h)
Solid
Filtration Residue
Liquid
Pregnant leach
solution
(REEs)
Figure 6.18: Flow sheet for high-pressure acid leaching of the bauxite residue sample.
100 Al Fe Si Ti Ca Na
Extraction, wt.%
80
60
40
20
0
40 60 80 100 120 140 160 180 200
100 Sc Y La Nd
Extraction, wt.%
80
60
40
20
0
40 60 80 100 120 140 160 180 200
Temperature ºC
Figure 6.19: High-pressure acid leaching of bauxite residue with 3 N H2SO4 at

different temperatures (L/S: 10, t: 1 h).
- 152 -
Al Fe Si Ti Ca Na
100
Extraction, wt.%
80
60
40
20
0
40 60 80 100 120 140 160 180 200
100
Extraction, wt.%
80
60
40 Sc Y
20 La Nd
0
40 60 80 100 120 140 160 180 200
Temperature ºC
Figure 6.20: High-pressure acid leaching of bauxite residue with 3 N HCl at

different temperatures (L/S: 10, t: 1 h).
It must be noted that the extraction yields obtained at temperatures

< 100 °C are comparable to the ones reported by other researchers, who have also
investigated the direct treatment of bauxite residue by acid leaching. Thus, Borra
et al. have reported REE recoveries in the range of 70 – 80 wt.%, by treating the
bauxite residue via direct acid leaching with L/S-ratio of 50:1, in order to enhance
the extractability of REEs. Up to 50 wt.% of scandium can be recovered with less
than 5 wt.% of iron in the solution, while aluminium, silicon and titanium are
significantly dissolved [16]. In another study, the use of strong oxidative
conditions (e.g., addition of H2O2) to avoid silica polymerization has also been
reported [36]. According to the authors, by treating the bauxite residue at 90 ºC,
with an equal concentration of 2.5 M of H2O2 and H2SO4 directly, an extraction of
approximately 70 wt.% of scandium can be achieved with a substantial depletion
of silica dissolution, but about 35 wt.% of iron is also extracted in the process. The
method, however, allows the recovery of about 90 wt.% of titanium as [TiO-O]SO4
complex. The mixture of bauxite residue with concentrated mineral acids, at very
high solid-liquid ratio, followed by water leaching, has also been studied as a
process to avoid silica polymerization [30]. The application of this method has
demonstrated a recovery up to 40 wt.% of scandium (and up to 50 wt.% of other
REEs) with a significant co-dissolution of iron (about 25 wt.%), but the
concentration of REEs in the leachate can be enhanced by considering a multi-
stage leaching with the recirculation of the leach liquor [30].
- 153 -
As described in the previous section, during the treatment of slags from

bauxite residue smelting by HPAL, silicon remains undissolved in the solid
residue and, consequently, silica polymerization is avoided. Titanium also remains
in the solid residue, but its recovery after leaching would be beneficial.
Furthermore, HPAL of the slags allows to recover more than 90 wt.% of scandium
without bringing too much iron into the solution (< 3 g L-1). The recovery of other
REEs is high when using HCl. However, the large volume of effluents enriched
with HCl may represent the major concern in the process due to its high
corrosiveness. The use of glass-lined reactors and valves and pipes made with
high-performance chemically-resistance polymers can significantly increase the
capital cost. From this point of view, the recovery of scandium from the slags can
alternatively be done by using H2SO4, which is widely used in the industry and
allows a high extraction yield, although the recovery of other REEs is low. The
residues generated after HPAL are rich in silica and low in sodium content. They
can also be rich in CaSO4 when H2SO4 is considered in the process. The residue
can be further studied for their applicability in building materials or cementitious
binder.
6.4 Conclusions
Iron was successfully separated from Greek bauxite residue by using

different mixtures of coke, CaO and SiO2 at a temperature of 1500 °C. After
heating, the slag solidified slowly at room temperature resulted less amorphous
than the slag subjected to quenching (fast cooling). The slag III, subjected to slow
cooling, was used as a reference to selectively recover rare-earths by HPAL due
to its high content of silicon, aluminium and iron. HPAL with HCl allows an
extraction up to 90 wt.% of scandium, and about 95 wt.% of yttrium, lanthanum
and neodymium at temperatures above 100 ºC. On the other hand, by performing
HPAL with H2SO4 the extraction of scandium reached up to 95 wt.% at 150 ºC,
while the extraction yield of yttrium, lanthanum and neodymium was about
40 wt.%. The formation of CaSO4 during HPAL with H2SO4 hinders the
dissolution of TiO2, but also the dissolution of other REEs, with less effect on
scandium extraction. The extraction of lanthanum and neodymium is limited by
the formation of a double sulfate (NaLn(SO4)2nH2O) and/or by their adsorption
on the surface of silicon/aluminium-oxide minerals. Nevertheless, the formation
of sulfates result in an increase in the selectivity of scandium over the other REEs.
The optimal conditions determined from the reference slag were successfully
applied to other slags from bauxite residue smelting. HPAL with HCl at 120 ºC
allowed to obtain high extraction yields for yttrium, lanthanum and neodymium
(> 90 wt.%), but the slag II subjected to fast cooling (amorphous slag) resulted in
the highest extraction yield of scandium (about 85 wt.%). The treatment of slag I
and III, both subjected to slow cooling, with H 2SO4 at 150 ºC, resulted also in a
substantial scandium dissolution (about 90 wt.%), but with very low dissolution
of other REEs. About 70 wt.% of scandium can be recovered from the (raw)
bauxite residue by HPAL together with a high co-dissolution of other major
- 154 -
elements (aluminium, iron, titanium and silicon). Although the treatment of slag
from bauxite residue by HPAL has demonstrated better performance in terms of
REEs recovery than other reported technologies, the process still requires further
investigation to optimize the acid consumption, leaching time, L/S-ratio and
leaching temperature. An economic analysis should be developed in order to
assess the economic feasibility for bauxite residue and bauxite residue slag
processing via HPAL.
6.5 References

aluminium.org.
(2011) 33–45. doi:10.1016/j.hydromet.2011.02.006.
331. doi:10.1007/s40831-016-0060-x.
doi:10.1180/minmag.2016.080.052.
doi:10.1016/j.hydromet.2011.02.007.
doi:10.1016/j.jclepro.2015.02.089.
doi:10.1016/0003-2670(94)80250-5.
doi:10.1007/s40831-016-0068-2.
- 155 -
[9] A.S. Wagh, W.R. Pinnock, Occurrence of scandium and rare earth
elements in Jamaican bauxite waste., Econ. Geol. 82 (1987) 757–761.
doi:10.2113/gsecongeo.82.3.757.
doi:10.1016/0003-2670(95)00486-6.
[11] S. Reid, J. Tam, M. Yang, G. Azimi, Technospheric Mining of Rare Earth
Elements from Bauxite Residue (Red Mud): Process Optimization,
Kinetic Investigation, and Microwave Pretreatment, Sci. Rep. 7 (2017)
15252. doi:10.1038/s41598-017-15457-8.
[12] Z.R. Liu, K. Zeng, W. Zhao, Y. Li, Effect of temperature on iron leaching
from bauxite residue by sulfuric acid, Bull. Environ. Contam. Toxicol. 82
(2009) 55–58. doi:10.1007/s00128-008-9576-5.
[13] R.M. Rivera, B. Ulenaers, G. Ounoughene, K. Binnemans, Behaviour of
Silica during Metal Recovery from Bauxite Residue by Acidic Leaching,
pp. 547–556.
Tkaczyk, V. Vassiliadou, D. Panias, Modes of occurrences of scandium in
Greek bauxite and bauxite residue, Miner. Eng. 123 (2018) 35–48.
doi:10.1016/j.mineng.2018.04.025.
doi:10.1016/j.mineng.2015.01.005.
[17] R.M. Rivera, G. Ounoughene, C.R. Borra, K. Binnemans, T. Van Gerven,
Neutralisation of bauxite residue by carbon dioxide prior to acidic
leaching for metal recovery, Miner. Eng. 112 (2017) 92–102.
doi:10.1016/j.mineng.2017.07.011.
[18] G. Alkan, B. Xakalashe, B. Yagmurlu, F. Kaussen, B. Friedrich,
Conditioning of Red Mud for Subsequent Titanium and Scandium
Recovery – A Conceptual Design Study, World Metall. – ERZMETALL.
70 (2017) 5–12.
[19] K. Jayasankar, P.K. Ray, A.K. Chaubey, A. Padhi, B.K. Satapathy, P.S.
Mukherjee, Production of pig iron from red mud waste fines using
thermal plasma technology, Int. J. Miner. Metall. Mater. 19 (2012) 679–
684. doi:10.1007/s12613-012-0613-3.
- 156 -

doi:10.1007/s40831-015-0026-4.
[21] B. Yagmurlu, G. Alkan, B. Xakalashe, B. Friedrich, S. Stopic, Combined
SAF Smelting and Hydrometallurgical Treatment of Bauxite Residue for
Enhanced Valuable Metal Recovery, in: 35th Int. ICSOBA Conf.
[22] F. Kaußen, B. Friedrich, Reductive Smelting of Red Mud for Iron
Recovery, Chemie-Ingenieur-Technik. 87 (2015) 1535–1542.
doi:10.1002/cite.201500067.
Recovery of rare earths and major metals from bauxite residue (red mud)
by alkali roasting, smelting, and leaching, J. Sustain. Met. 3 (2017) 393–
404. doi:10.1007/s11837-016-2111-y.
2111-y.
[25] O.P. Shrivastava, N. Kumar, I.B. Sharma, Solid state synthesis and
structural refinement of polycrystalline La x Ca 1 – x TiO 3 ceramic
powder, Bull. Mater. Sci. 27 (2004) 121–126. doi:10.1007/BF02708493.
(2015) 2504–2521. doi:10.3390/polym7121528.
568–590. doi:10.3390/en7020568.
379. doi:10.1016/0022-3093(86)90503-X.
doi:10.1016/j.gca.2009.06.002.
doi:10.1016/j.mineng.2018.01.023.
- 157 -
[31] S. Kaya, Y.A. Topkaya, High pressure acid leaching of a refractory

lateritic nickel ore, Miner. Eng. 24 (2011) 1188–1197.
doi:10.1016/j.mineng.2011.05.004.
[32] S. Kaya, Y.A. Topkaya, Extraction Behavior of Scandium From a
Refractory Nickel Laterite Ore During the Pressure Acid Leaching
Process, in: Rare Earths Ind. Technol. Econ. Environ. Implic., 2015: pp.
177–188. doi:10.1016/B978-0-12-802328-0.00011-5.
[33] Y. Li, I. Perederiy, V.G. Papangelakis, Cleaning of waste smelter slags
and recovery of valuable metals by pressure oxidative leaching, J. Hazard.
doi:10.1016/j.seppur.2015.10.051.
[35] C.W. Bale, E. Bélisle, P. Chartrand, S.A. Decterov, G. Eriksson, A.E.
Gheribi, K. Hack, I.H. Jung, Y.B. Kang, J. Melançon, A.D. Pelton, S.
Petersen, C. Robelin, J. Sangster, P. Spencer, M.A. Van Ende, Reprint of:
FactSage thermochemical software and databases, 2010–2016, Calphad
Comput. Coupling Phase Diagrams Thermochem. 54 (2016) 35–53.
doi:10.1016/j.calphad.2016.07.004.
[36] G. Alkan, B. Yagmurlu, S. Cakmakoglu, T. Hertel, Ş. Kaya, L. Gronen, S.
Stopic, B. Friedrich, Novel Approach for Enhanced Scandium and
Silica Gel Formation, Sci. Rep. 8 (2018) 5676. doi:10.1038/s41598-018-
24077-9.
[37] S. Van Loy, K. Binnemans, T. Van Gerven, Recycling of rare earths from
lamp phosphor waste: enhanced dissolution of LaPO 4 :Ce 3+ ,Tb 3+ by
mechanical activation, J. Clean. Prod. 156 (2017) 226–234.
doi:10.1016/j.jclepro.2017.03.160.
[38] P. Baláz, Extractive Metallurgy of Activated Minerals, 10th ed.,
Amsterdam, 2000.
[39] G.S. Grest, M.H. Cohen, Liquid-glass transition: Dependence of the glass
transition on heating and cooling rates, Physical. 21 (1980) 4113–4117.
[40] P. Kokhanenko, K. Brown, M. Jermy, Silica aquasols of incipient
instability: Synthesis, growth kinetics and long term stability, Colloids
Surfaces A Physicochem. Eng. Asp. 493 (2016) 18–31.
doi:10.1016/j.colsurfa.2015.10.026.
doi:10.1016/j.hydromet.2014.08.001.
- 158 -

Can. Metall. Q. 25 (1986) 201–209.
[43] L. Shi, S. Ruan, J. Li, A.R. Gerson, Desilication of low alumina to caustic
liquor seeded with sodalite or cancrinite, Hydrometallurgy. 170 (2017) 5–
15. doi:10.1016/j.hydromet.2016.06.023.
[44] Y. Zhang, Y. Hua, X. Gao, C. Xu, J. Li, Y. Li, Q. Zhang, L. Xiong, Z. Su,
M. Wang, J. Ru, Recovery of zinc from a low-grade zinc oxide ore with
high silicon by sulfuric acid curing and water leaching, Hydrometallurgy.
166 (2016) 16–21. doi:10.1016/j.hydromet.2016.08.010.
[45] W. Wang, C.Y. Cheng, Separation and purification of scandium by
solvent extraction and related technologies: A review, J. Chem. Technol.
Biotechnol. 86 (2011) 1237–1246. doi:10.1002/jctb.2655.
[46] T. Falayi, F. Ntuli, F. Ndubisi, Kinetic and thermodynamic parameters of
silica leaching from Camden power station fly ash, in: Adv. Environ.
Agric. Sci., 2015: pp. 241–248.
[47] B.I. Whittington, R.G. McDonald, J.A. Johnson, D.M. Muir, Pressure
acid leaching of arid-region nickel laterite ore: Part II: Effect of ore type,
Hydrometallurgy. 70 (2003) 31–46. doi:10.1016/S0304-386X(03)00043-
4.
[48] M.W. Colby, A. Osaka, J.D. Mackenzie, Effects of Temperature on
Formation of Silica Gel, J. Non. Cryst. Solids. 82 (1986) 37–41.
doi:10.1016/0022-3093(86)90108-0.
[49] A. Seidel, Y. Zimmels, Mechanism and kinetics of aluminum and iron
leaching from coal fly ash by sulfuric acid, Chem. Eng. Sci. 53 (1998)
3835–3852. doi:10.1016/S0009-2509(98)00201-2.
[50] S. Sangita, N. Nayak, C.R. Panda, Extraction of aluminium as aluminium
sulphate from thermal power plant fly ashes, Trans. Nonferrous Met. Soc.
China (English Ed. 27 (2017) 2082–2089. doi:10.1016/S1003-
6326(17)60231-0.
[52] C.R. Borra, J. Mermans, B. Blanpain, Y. Pontikes, K. Binnemans, T. Van
Gerven, Selective recovery of rare earths from bauxite residue by
151–159. doi:10.1016/j.mineng.2016.03.002.
[53] G. Alkan, B. Yagmurlu, B. Xakalashe, S. Stopic, B. Friedrich,
Enhancement of Sc and Ti extraction rates from Fe-depleted slag by
hydrogen peroxide and sulfuric acid leaching, in: Y. Pontikes (Ed.), 2nd
Int. Bauxite Residue Valoris. Best Pract. Conf., Athens, 2018: pp. 255–
261.
- 159 -
[54] A. Myerson, Handbook of Industrial Crystallization, second, Boston,

2002. https://www.google.com/#cns=1.
[55] M. Kul, Y. Topkaya, I. Karakaya, Rare earth double sulfates from pre-
concentrated bastnasite, Hydrometallurgy. 93 (2008) 129–135.
doi:10.1016/j.hydromet.2007.11.008.
[56] P. Meshram, B.D. Pandey, T.R. Mankhand, Process optimization and
kinetics for leaching of rare earth metals from the spent Ni-metal hydride
batteries, Waste Manag. 51 (2016) 196–203.
doi:10.1016/j.wasman.2015.12.018.
[57] P. Davris, S. Stopic, E. Balomenos, D. Panias, I. Paspaliaris, B. Friedrich,
Leaching of rare earth elements from eudialyte concentrate by
doi:10.1016/j.mineng.2016.12.011.
[58] R.D. Abreu, C.A. Morais, Purification of rare earth elements from
monazite sulphuric acid leach liquor and the production of high-purity
ceric oxide, Miner. Eng. 23 (2010) 536–540.
doi:10.1016/j.mineng.2010.03.010.
[59] F. Zheng, F. Chen, Y. Guo, T. Jiang, A.Y. Travyanov, G. Qiu, Kinetics of
Hydrochloric Acid Leaching of Titanium from Titanium-Bearing Electric
Furnace Slag, JOM. 68 (2016) 1476–1484. doi:10.1007/s11837-015-
1808-7.
doi:10.20944/preprints201801.0288.v1.
[61] M. Kosmulski, Adsorption of trivalent cations on silica, J. Colloid
Interface Sci. 195 (1997) 395–403. doi:10.1006/jcis.1998.6034.
[62] W. Piasecki, D.A. Sverjensky, Speciation of adsorbed yttrium and rare
earth elements on oxide surfaces, Geochim. Cosmochim. Acta. 72 (2008)
3964–3979. doi:10.1016/j.gca.2008.05.049.
Chem. Geol. 55 (1986) 313–322.
[64] J. Ma, Z. Wang, Y. Shi, Q. Li, Synthesis and characterization of lysine-
modified SBA-15 and its selective adsorption of scandium from a
solution of rare earth elements, RSC Adv. 4 (2014) 41597–41604.
doi:10.1039/C4RA07571D.
- 160 -
Chapter 7 Comparative analysis for REEs extraction
7. COMPARATIVE ANALYSIS FOR RARE

EARTH EXTRACTION FROM BAUXITE
RESIDUE
Direct acid leaching, as it has been discussed in the previous chapters, is

characterized by the high acid consumption and the high co-dissolution of large
amounts of iron, titanium, aluminium and silicon, which decrease the efficiency
of the further separation process due to their co-adsorption on the ion-exchange
resins, for instance, during solvent extraction or ion exchange. The eventual
elution of these impurities requires large amounts of acids; this increases the
process cost and it also requires large amount of bases for neutralisation further on
in the process [1]. A comparative analysis between the processes developed in this
PhD Thesis can give significant hints on the feasibility for further studies through
an integrated extraction-separation process at pilot-plant scale.
7.1 Metal extraction and selectivity
Table 7.1 and 7.2 summarise the extraction yields and concentrations of
major (Al, Fe, Ti) and minor (Sc, Y, La, Nd) elements obtained with the different
processes studied in this PhD Thesis. The values given in both tables are based on
experimental results obtained under optimized conditions. The flow sheets and
mass balances of the conventional direct acid leaching method with H 2SO4 and
HCl are shown in Figure 7.1.
- 161 -
Table 7.1: Overview of different leaching processes (as absolute values) studied in this PhD Thesis considering H 2SO4 as a lixiviant (basis: 1000 kg
of bauxite residue/slag).
Extraction yield Concentrations Concentration REEs

Process (wt.%) (g L-1) (mg L-1)
Fe Al Ti Sc Y Nd La Fe Al Ti Sc Y Nd La
DLx 3 26 28 39 58 21 23 1 3 1 5 4 3 2
CO2 N-Lx 2 21 29 36 56 23 20 1 2 1 4 4 2 2
DD 1 Lx 4 30 23 24 15 12 13 3 6 2 6 2 3 2
DD M Lx 5 22 21 20 13 10 12 10 12 4 14 6 5 8
HPAL BR 40 50 5 68 34 11 12 13 5 0 7 2 2 1
S-HPAL BRS 82 70 0 91 16 9 9 3 13 0 11 2 1 1
DLx: Direct leaching (chapter 3); CO2 N-Lx: CO2-neutralisation followed by acid leaching (chapter 3); DD 1 Lx: dry
digestion with single-stage leaching (chapter 5); DD M Lx: dry digestion with multiple-stage leaching (chapter 5);
HPAL BR: high-pressure acid leaching of bauxite residue (chapter 6); S-HPAL BRS: smelting followed by high-pressure
acid leaching of bauxite residue slags (BRS) at 150 ºC (chapter 6).
- 162 -
Table 7.2: Overview of different leaching processes (as absolute values) studied in this PhD Thesis considering HCl as a lixiviant (basis: 1000 kg of bauxite
residue/slag).
Extraction yield Concentrations Concentration REEs

Process (wt.%) (g L-1) (mg L-1)
Fe Al Ti Sc Y Nd La Fe Al Ti Sc Y Nd La
DLx 3 30 30 45 70 50 38 2 5 2 5 4 3 2
DD 1 Lx 4 23 2 30 47 33 30 3 4 0 7 7 6 7
DD M Lx 4 24 3 32 50 34 32 7 13 1 19 19 17 18
HPAL BR 74 40 7 88 99 99 77 24 4 0 9 7 6 11
S-HPAL BRS 84 91 0 79 99 98 98 3 17 0 12 11 11 17
DLx: Direct leaching (chapter 3); DD 1 Lx: dry digestion with single–stage leaching (chapter 5); DD M Lx: dry digestion with
multiple-stage leaching (chapter 5); HPAL BR: high-pressure acid leaching of bauxite residue (chapter 6); S-HPAL BRS:
smelting followed by high-pressure acid leaching of bauxite residue slags (BRS) at 120 ºC (chapter 6).
- 163 -
Figure 7.1: Flow sheet and mass balance of the conventional direct acid leaching (DLx)
process with H2SO4 (left) and HCl (right) as leaching reagents.
The REEs recovery from the non- and neutralised bauxite residue was
not significantly different. Yttrium, in particular, showed the highest extraction
yields because it was not affected by the CO2-neutralisation and, therefore, ca.
60 wt.% of yttrium was recovered from the neutralised material. The flow sheet
and mass balance of the neutralisation-leaching method is shown in Figure 7.2.
The neutralisation process investigated in this PhD Thesis allowed to sequester
about 24 g of CO2 per kg of bauxite residue with high pressure (PCO2 = 30 bar) and
temperature (150 ºC). This value is relatively low compared to the sequestration
yield reported in the literature with prolonged period of time (i.e. about 40 – 50 g
of CO2 per kg of bauxite residue with contacting period > 3 h [2,3]). Nevertheless,
the leachate obtained after neutralisation is rich in sodium and can be recirculated
back in the Bayer process, while the residue obtained after acid leaching can be
used for building materials. The low extraction yields obtained from the
neutralised bauxite residue were caused by an insufficient amount of acid devoted
to leaching, as consequence of secondary chemical reactions (i.e. transformation
of calcite into bassanite and dissolution of silicate compounds), but also due to the
- 164 -
reduction of the diffusion leaching rate as CO2-neutralisation leads to an increase

of the average particle size (see Chapter 4). Consequently, it is believed that the
extraction yields from the neutralised bauxite residue can be enhanced by leaching
the bauxite residue with HCl at high temperatures (> 25 °C), as the formation of
bassanite and the eventual polymerization of silica can be avoided, but also the
diffusion leaching rate can be improved.
Figure 7.2: Flow sheet and mass balance of the neutralisation-leaching (CO2 N-Lx) process
with H2SO4 as leaching reagent.
The flow sheet and mass balance of the dry digestion method, considering
one single-stage of leaching, is shown in Figure 7.3. The method avoids the co-
dissolution of silicon due to the low amount of water used during the digestion
stage. It must be noticed that in the first stage, about 0.8 kg of water per kg of
bauxite residue are used to agglomerate the particles, i.e. formation of a paste,
while the conventional leaching process demands approximately 6 – 7 times that
amount of water to form a slurry. Dry digestion with multi-stage leaching
considered five stages, in which fresh bauxite residue was contacted repeatedly
with the leach solution of the previous cycle. Although less amount of metal was
extracted from the bauxite residue by one single-stage dry digestion, a multi-stage
process allows to obtain a more concentrated leachate in terms of REEs with a
substantial reduction in water consumption. In 1 cycle, the water consumption is
about 5.8 kg per kg of bauxite residue, but in the following cycles extra water was
added just to maintain the same liquid-to-solid ratio (i.e. L/S: 5). Hence, the overall
water consumption of the multi-stage process is about 3 kg per kg of bauxite
residue, which is a reduction of 60% with respect to the conventional acid leaching
- 165 -
process. The extraction yields obtained by dry digestion with H2SO4 and HCl were
also different. The average extraction of REEs with H2SO4 was ca. 14 wt.%, while
the extraction yields achieved with HCl were between 30 – 50 wt.%.
Consequently, dry digestion with HCl allowed to obtain a more enriched leachate
in REEs than with H2SO4, particularly in combination with multi-stage leaching.
Nevertheless, the main drawback of this process is the high co-dissolution of iron
and aluminium, i.e. 7 and 13 g L-1, respectively, which is about four times higher
the concentration obtained during conventional acid leaching.
Figure 7.3: Flow sheet and mass balance of the dry digestion (one single-stage, DD 1 Lx)
process with H2SO4 (left) and HCl (right) as leaching reagents.
The experimental results obtained with the HPAL method show that the
recoveries were low at 1 N and below, whereas no further increase in recovery was
observed at concentrations beyond 3 N, with both mineral acids (see chapter 6).
Because no data points are available between 1 and 3 N, it is expected that the acid
- 166 -
consumption can be further optimized between these acid concentrations. Hence,

additional experiments must be conducted. For our purpose the acid consumption
was calculated based on the leaching results obtained at 3 N. The flow sheet and
mass balance of the HPAL method considering the direct treatment of bauxite
residue, with H2SO4 and HCl, is shown in Figure 7.4. The direct treatment of
bauxite residue by HPAL leads to an extraction of scandium and other REEs of
more than 75 wt.%, particularly with HCl, but iron and aluminium are also
strongly dissolved. The iron concentration was about 24 g L-1, while the average
aluminium concentration was about 6 g L-1. The concentration of titanium (and
silicon) in the leach solution was < 0.5 g L-1. HPAL with H2SO4 allows a better
selectivity for scandium over the other REEs, as scandium concentration was
about 4 times higher than the concentration of yttrium, lanthanum and
neodymium. The average concentration of REEs in the leach solution obtained
with HCl was about 9 mg L-1, which is three times higher than the average
concentration of REEs in the leach solution obtained with conventional direct acid
leaching, but not as high as the concentration achieved by dry digestion with multi-
stage leaching.
Figure 7.4: Flow sheet and mass balance of the high-pressure acid leaching (HPAL BR)
process for bauxite residue processing with H2SO4 (left) and HCl (right) as leaching
reagents.
- 167 -
The separation of iron from the REEs by reductive smelting, followed by

HPAL of the corresponding slag, allows to obtain the highest extraction yields of
REEs, particularly with HCl (80 wt.% of scandium and more than 95 wt.% of
yttrium, lanthanum and neodymium). HPAL with H2SO4 leads to a high extraction
of scandium (ca. 90 wt.%) but not of the other REEs (extraction < 20 wt.%),
presumably due to the formation of a double sulfate (NaLn(SO 4)2×nH2O) and/or
due to the adsorption on the surface of silicon/aluminium-oxides compounds (see
chapter 6). The flow sheet and mass balance of the HPAL method considering the
treatment of bauxite residue slag, with H2SO4 and HCl, is shown in Figure 7.5.
HPAL of the bauxite residue slag with HCl also demonstrated a substantial
enrichment of REEs in solution with a range in concentration between 11 –
17 mg L-1, with lanthanum concentration being the highest. It must be noticed that
the HPAL corresponds to a single-step process and, consequently, the circulation
of the leach solution is not considered, and it can be unpractical due to the high
temperatures and pressures required. The HPAL method, similar to the dry
digestion method, avoids silica polymerization but also leads to a substantial
depletion of titanium dissolution.
Figure 7.5: Flow sheet and mass balance of the high-pressure acid leaching (HPAL BRS)
process for bauxite residue slag processing with H2SO4 (left) and HCl (right) as leaching
reagents.
- 168 -
7.2 Comparative preliminary economic analysis
A preliminary comparative economic analysis of the studied processes

was developed to assess the economic feasibility of further tests at large scale, for
instance, pilot tests. The results are based on 50,000 tonne of bauxite residue
processed per year. It has been assumed that the slag generated during the
reductive smelting process represents approximately 80 wt.% of the feed material,
as about 90 wt.% of iron can be recovered as metallic iron (i.e. pig iron) during
the smelting process. Meanwhile, according to the literature, about 99 wt.% of Sc
and 40 wt.% of Ti can be recovered from the leach solution as Sc 2O3 and TiO2 by
considering a recovery process that includes solvent extraction (with
D2EHPA/TBP/Shellsol), H2SO4-based scrubbing and calcination [4]. The
recovery of iron is considered only with the high-pressure acid leaching of the slag
produced from bauxite residue smelting, as iron was separated before acid
leaching as metallic iron. Meanwhile, the recovery of titanium by HPAL process
is not considered as it precipitates at high temperatures, thus remaining in the solid
residue after acid leaching [5]. Aluminium is not recovered because it remains
dissolved into the leach solution. The production rates of metallic-Fe, Sc2O3 and
TiO2 with the H2SO4- and HCl-based leaching processes studied in this PhD
Thesis are shown in Table 7.3 and 7.4, respectively.
Table 7.3: Production rate (tonne year-1) of metallic-Fe, TiO2 and Sc2O3 with H2SO4-based
leaching processes studied in this PhD Thesis.
Product DLx CO2 N-Lx DD 1 Lx DD M Lx S-HPAL

Fe - - - - 14,850
TiO2 335 335 266 1,214 0
Sc2O3 7 7 4 18 14
DLx: Direct leaching; CO2 N-Lx: CO2-neutralisation followed by acid leaching;
DD 1 Lx: dry digestion with single-stage leaching; DD M Lx: dry digestion with
multiple-stage leaching; S-HPAL: smelting followed by high-pressure acid
leaching of bauxite residue slags.
- 169 -
Table 7.4: Production rate (tonne year-1) of metallic-Fe, TiO2 and Sc2O3 with HCl-based
Product DLx DD 1 Lx DD M Lx S-HPAL

Fe - - - 14,850
TiO2 347 23 173 0
Sc2O3 8 6 29 12
DLx: Direct leaching; CO2 N-Lx: CO2-neutralisation followed by acid leaching; DD 1
Lx: dry digestion with single-stage leaching; DD M Lx: dry digestion with multiple-
stage leaching; S-HPAL: smelting followed by high-pressure acid leaching of bauxite
residue slags.
The economic analyses considering H2SO4 and HCl as leaching reagents

are shown in Table 7.6 and 7.7, respectively. It has been reported that the capital
cost of a conventional direct acid leaching process is about 16,250 USD per tonne
of Sc produced, while the operating cost can be approximately 250,000 USD per
tonne of Sc produced. For the recovery process, the capital and operating cost has
been reported to be approximately 18,750 and 3,000,000 USD per tonnes of Sc
produced, respectively [6]. The capital cost entails an estimated amount to be
considered as investment, and includes the purchase and installation of new plant
equipment. The operating cost, on the other hand, represent the estimated cost of
utilities (e.g., electricity, water, steam, fuel), direct labour, maintenance, supplies
and laboratory. The cost of royalties, patents and licenses have not been considered
in this preliminary economic evaluation [7]. The operation cost of bauxite residue
has been estimated to be between 4 – 12 USD/tonne. Hence, the consumption of
bauxite residue is represented as a benefit, and therefore as a positive value, as the
material is a liability to the company [8,9]. Other assumptions for the calculations
are (i) a product purity of 99.99% for Sc2O3 and 99% for TiO2, (ii) a range of
historical prices for H2SO4, NaOH, Sc2O3 and TiO2 (see Table 7.5), and (iii) the
capital and operating cost of the extraction processes of neutralisation-leaching
and dry digestion have been calculated with an increase of 30% over the costs of
a conventional leaching process, while for the smelting-HPAL process an increase
of 50% was considered. The following processes were evaluated: direct leaching
of bauxite residue (DLx), CO2-neutralisation followed by acid leaching (CO2 N-
Lx), dry digestions with a single-stage leaching (DD 1 Lx), dry digestion with
multiple-stage leaching (DD M Lx), and smelting followed by high-pressure acid
leaching of the produced slag (S-HPAL). Among the REEs, the recovery of Sc
entails the highest benefits [10].
- 170 -
Table 7.5: Source of prices for chemicals and products used in this preliminary calculation.
Compound Max (USD) Min (USD) Year Source

H2SO4 300 per tonne 90 per tonne 1988 - 2018 [13]
Fe 475 per tonne 185 per tonne 2010 – 2018 [14]
Sc2O3 6000 per kg 1500 per kg 1991 - 2014 [6]
3500 per 2000 per
TiO2 2016 - 2018 [15]
tonne tonne
- 171 -
Table 7.6: Comparative economic analysis (in USD year-1) of different processes studied in this PhD Thesis considering H2SO4 as a lixiviant.
Value DLx CO2 N-Lx DD 1 Lx DD M Lx S-HPAL

Costs
Extraction
Capital cost -39,325 -46,010 -30,674 -127,806 -110,019
Acid -6,000,000 -6,000,000 -6,000,000 -6,000,000 -6,000,000
Operating cost -605,000 -707,850 -471,900 -1,966,250 -423,150
Recovery
Capital cost -44,921 -40,429 -26,953 -112,303 -83,784
Operating cost -6,738,188 -6,064,369 -4,042,913 -16,845,469 -12,567,555
Benefits
Bauxite residue 500,000 500,000 500,000 500,000 500,000
TiO2 1,173,986 1,173,986 931,092 4,250,639 0
Sc2O3 44,097,727 39,687,954 26,458,636 110,244,318 82,247,728
Fe 7,053,750
Margin with highest prices 32,344,279 28,503,282 17,317,290 89,943,129 70,616,970

Margin with lowest prices 2,967,847 2,434,180 1,274,273 9,638,188 8,824,674
- 172 -
Table 7.7: Comparative economic analysis (in USD year-1) of different processes studied in this PhD Thesis considering HCl as a lixiviant.
Value DLx DD 1 Lx DD M Lx S-HPAL

Costs
Extraction
Capital cost -44,241 -38,342 -204,490 -95,511
Acid -11,820,000 -11,820,000 -11,820,000 -11,820,000
Operating cost -680,625 -589,875 -3,146,000 -367,350
Recovery
Capital cost -50,536 -33,691 -179,685 -72,735
Operating cost -7,580,461 -5,053,641 -26,952,750 -10,910,295
Benefits
Bauxite residue 500,000 500,000 500,000 50,000
TiO2 1,214,468 80,965 607,234 0
Sc2O3 49,609,943 33,073,295 176,390,908 71,401,873
Fe - - - 7,053,750
Margin with highest prices 31,148,548 16,118,711 135,195,217 55,239,732

Margin with lowest prices 1,694,605 -446,959 10,915,793 6,105,827
- 173 -
The profit margins of these processes are shown in Figure 7.6. With the
highest market price, the margins are not going below zero, but they can be very
low when the prices of metals are reduced. Dry digestion with single-stage resulted
in the lowest profit margin due to the low recoveries of TiO 2 and Sc2O3, but also
due to the high acid consumption. The margin, however, is significantly improved
by considering the dry digestion process with multi-stage leaching due to the high
consumption of bauxite residue, which allows a high production of TiO2 and Sc2O3
(see Table 7.3 and 7.4). Multi-stage leaching not only allows to suppress the silica
gel formation, but also allows to decrease the large volumes of effluents generated
during direct leaching, which also leads to an improvement of the profit margin. It
must be noted that dry digestion of bauxite residue entails a significant co-
dissolution of aluminium and iron that may require a large number of processing
steps and consume large amount of chemicals in the subsequent downstream
process, e.g., solvent extraction and/or ion exchange, which may increase the
overall operational cost [11]. However, the separation of aluminium and iron
before leaching can help to surpass the possibly elevated cost of the process by
enhancing the profit margin of the dry digestion process with multi-stage
circulation of the leach solution. Titanium can be recovered simultaneously with
the REEs.
The treatment of bauxite residue by HPAL is restricted by the high co-
dissolution of major metals, i.e. iron, aluminium and titanium, which can
significantly affect the efficiency of the subsequent separation process (e.g.,
solvent extraction and/or ion exchange). However, the separation of iron from the
REEs by reductive-smelting allowed a better selectivity for REEs recovery
compared to the extraction yields achieved by leaching the bauxite residue
directly. This also allows to obtain a positive profit margin, particularly when
H2SO4 is considered as a lixiviant for the leaching process. However, the profit
margin is severely reduced when the metal prices are very low.
The economic analysis of the neutralisation-leaching process resulted in
a lower profit margin compared the profit margin obtained by treating the bauxite
residue directly by acid leaching. This is because of an increase of the operating
cost, as the process requires the installation of a neutralisation stage. The lower
scandium recovery and higher water consumption, compared to the conventional
leaching method, also limit the enhancement of the profit margin. However, the
recirculation of the leachate obtained from the neutralisation stage and the leach
solution can improve the economics of the process. During the re-circulation of
the leach solution, however, silica gel formation can also be expected due to an
increase in silica concentration. This gel formation can be suppressed by treating
the neutralized bauxite residue with the dry digestion process. Although high-
temperature leaching of the neutralised bauxite residue can also suppress silica gel
formation, the process can be unpractical due to the need of using high-pressure
vessels for both stages, i.e. neutralisation with high CO2-pressure and temperature,
but also for acid leaching.
- 174 -
1×108 Margin high prices H2SO4

Margin low prices
8×107
USD/year
6×107
4×107
2×107
0
DLx
CO2 N-Lx
DD 1 Lx
DD M Lx
HPAL BRS
1×108 HCl
Margin high prices
8×107 Margin low prices
USD/year
6×107
4×107
2×107
0
DLx
DD 1 Lx
DD M Lx
HPAL BRS
Figure 7.6: Profit margin considering H2SO4 and HCl as lixiviants reagents in the different
Although HPAL and the dry digestion method allow to suppress the silica
gel formation and titanium leaching, the high co-dissolution of aluminium and/or
iron still may represent some problems for the downstream processing of the leach
liquor. Nevertheless, the recovery of alumina and iron before leaching can lead to
a substantial improvement of the profit margins, as it was reported by Borra et al.
[10]. However, the process demands large capital investment as it requires two
main processing steps at high temperatures, which leads to a significant energy
consumption, i.e. 3 – 5 GJ per tonne of bauxite residue. On the contrary, although
the leach solution obtained from the dry digestion process can also contain a high
concentration of major metals, the process works at ambient temperature, which
represents its most remarkable advantage. Furthermore, the leach solution can be
treated by solvent extraction with D2EHPA or TBP (as it was considered in this
preliminary economic evaluation), and/or ion-exchange chromatography with
Dowex 50W-X8 resin, as an alternative to Shellsol, which have demonstrated high
selectivity of scandium over base metals like iron, aluminium and/or titanium
[4,12].
- 175 -
The residues generated after HPAL or dry digestion are rich in silica and
low in sodium content. They can also be rich in CaSO4 when H2SO4 is considered
in the process. The residue, therefore, can be further studied for their applicability
in building materials or cementitious binder.
7.3 Summary
By considering an integrated extraction-separation process, among the

leaching methods studied in this PhD Thesis, the use of H2SO4 as a lixiviant for
leaching resulted in higher profit margins than the HCl-based leaching due to the
low acid consumption. The treatment of bauxite residue by dry digestion with
multi-stage circulation of the leach solution appears to be economically the most
interesting leaching alternative. The process allows to reduce significantly the
volume of effluents due to the low amount of water required for the process. Gel
formation does not occur and titanium can be recovered simultaneously with the
REEs. High-pressure acid leaching of slag from bauxite residue smelting also
avoids the polymerization of silica gel, but titanium remains in the solid residue
after leaching. The process still requires further studies to decrease the acid and
water consumption.
It must be noted that the economic analysis developed in this PhD Thesis
represents just an outline of the expected profit margins by considering
experimental conditions at laboratory scale. A more comprehensive cost analysis
than the one developed during this research must be performed for an integrated
extraction-separation process under optimized conditions.
7.4 References

recovery from various resources: A review, Hydrometallurgy. 108 (2011)
100–108. doi:10.1016/j.hydromet.2011.03.001.
176 (2010) 1044–1050. doi:10.1016/j.jhazmat.2009.11.146.
[3] D. Bonenfant, L. Kharoune, R. Hausler, P. Niquette, CO2 Sequestration
by aqueous red mud carbonation at ambient pressure and temperature,
Ind. Eng. Chem. Res. 47 (2008) 7610–7616.
[4] W. Wang, Y. Pranolo, C.Y. Cheng, Recovery of scandium from synthetic
red mud leach solutions by solvent extraction with D2EHPA, Sep. Purif.
Technol. 108 (2013) 96–102. doi:10.1016/j.seppur.2013.02.001.
- 176 -

doi:10.1016/j.seppur.2015.10.051.
[6] R.P. Narayanan, L.-C. Ma, N.K. Kazantzis, M.H. Emmert, Cost Analysis
as a Tool for the Development of Sc Recovery Processes from Bauxite
Residue (Red Mud), ACS Sustain. Chem. Eng. 6 (2018).
[7] R.D. Pascoe, Capital and operating costs of minerals engineering plants -
a review of simple estimation techniques, Miner. Eng. 5 (1992) 883–893.
doi:10.1016/0892-6875(92)90255-8.
016-0060-x.
331. doi:10.1007/s40831-016-0060-x.
2111-y.
[11] G.D. Fulford, G. Lever, T. Sato, Recovery of rare earth elements from
Bayer process red mud, US5030424 A, 1991.
doi:10.1021/ie011047b.
[13] US. Bureau of Labor Statistics, Producer Price Index by Commodity for
Chemicals and Allied Products, retrieved from FRED, Federal Reserve
Bank of St. Louis; https://fred.stlouisfed.org/series/WPU0613020T1
(access on 23.11.2018).
[14] https://www.steelmint.com/pigiron-prices-global
[15] Europe TiO2 two-year run of price hikes ends as supply improves,
https://www.icis.com/explore/resources/news/2018/08/09/10249110/euro
pe-tio2-two-year-run-of-price-hikes-ends-as-supply-improves/ (access on
23.11.2018)
- 177 -
Chapter 8 Conclusions and future perspectives
8. CONCLUSIONS AND FUTURE

PERSPECTIVES
8.1 General conclusions and findings
Bauxite residue contains several critical metals and, due to the large
volume of material produced each year, is associated with a substantial
management cost, whereas spills have led to major environmental incidents,
including the Ajka disaster in Hungary. Making the bauxite residue useful for other
applications in combination with recovery of rare-earth elements (REEs) can
contribute to solving the environmental and storage issues associated with its
management. Therefore, different leaching processes were studied within the
framework of this PhD Thesis to extract REEs from bauxite residue. The main
conclusions of this study are summarized in this section.
Mytilineos S.A. - Aluminium of Greece (Agios Nikolaos, Greece). Iron(III) oxide
was the major oxide in the bauxite residue followed by alumina, calcium oxide,
silica, titanium oxide and sodium oxide. The main minerals were hematite,
goethite, gibbsite, diaspore, boehmite, calcite, portlandite, perovskite, rutile,
anatase, cancrinite and chamosite. Among the REEs, the concentrations of
scandium, yttrium, lanthanum and neodymium were considered due to their high
concentrations.
The recovery of REEs from bauxite residue was studied by three acid
leaching methods: 1) neutralisation-leaching; 2) dry digestion-leaching; 3) high-
pressure acid leaching (HPAL) preceded by reductive smelting. These methods
were compared with the direct conventional acid leaching method.
The effect of CO2 gas as a neutralisation reagent of the bauxite residue
slurry was studied to reduce the amount of acid needed during leaching. During
neutralisation at room temperature, only 2 – 4 wt.% of sodium reacts with the
carbonic acid formed during the dissolution of CO2 in water, while less than
5 wt.% of calcium was dissolved. Consequently, the pH after neutralisation was
reduced from 10.7 to 9.3 only. High-pressure neutralisation, at high temperature,
allowed to decrease the pH further down to ca. 8.7 with a substantial dissolution
of sodium (about 27 wt.%), but it led to the precipitation of calcite, as an increase
up to 8 wt.% was observed. Neutralisation of bauxite residue with CO 2 gas,
however, did not enhance the extraction of REEs from bauxite residue during acid
leaching due to their association with major metals, but also as consequence of the
lack of acid caused by the transformation of CaCO 3 into CaSO4 and the
decomposition of silicate compounds. Further analysis by electron-probe micro
analysis (EPMA) demonstrated that scandium remains associated to iron,
- 179 -
particularly to hematite, and aluminium mineral phases in the untreated and

neutralised bauxite residue. Neodymium and lanthanum may also remain
associated to hematite, as they tend to occur simultaneously in similar mineral
phases as that of cerium, which was also found associated to iron and aluminium.
Nevertheless, their extraction yields could be enhanced by increasing the acid
concentration. The extraction of yttrium was not affected by the neutralised with
CO2, but it is believed that yttrium can remain associated to cancrinite and/or
chamosite, which were not completely dissolved by H 2SO4.
During direct acid leaching of the non-neutralised and neutralised bauxite
residue, between 50 – 65 wt.% of silicon was dissolved. With the dry digestion
process, only about 5 wt.% of silicon was dissolved because the hydrolysis of silica
was avoided. The maximum extraction yield of REEs was ca. 45 and 20 wt.% with
HCl and H2SO4, respectively. The lowest aluminium/iron leaching ratio (i.e. 2:1)
was obtained under the acid consumptions of 788 g HCl and 412 g H 2SO4 per kg
of bauxite residue. Under these acid consumptions, the dry digestion method with
re-circulation of the acid leaching solution allowed to increase significantly the
concentration of the REEs in solution. In a five-stage leaching process with HCl,
the REEs concentration increased from 7 to ca. 19 mg L-1, more than twice the
concentration observed by a single-stage method. Meanwhile, in a five-stage
leaching process with H2SO4, the scandium concentration increased from 5 to
14 mg L-1, while the concentration of yttrium, lanthanum and neodymium
increased from an average value of 3 to ca. 8 mg L-1. Unfortunately, the
concentration of major metals was also enriched. Nevertheless, dry digestion with
multi-stage of the leaching solution allows a high consumption of bauxite residue
with a substantial reduction in water consumption, i.e. a reduction ca. 60 wt.%
with respect to the conventional acid leaching method. The preliminary economic
analysis showed that the simultaneous recovery of REEs and titanium oxide with
a H2SO4-based dry digestion method is crucial to ensure a positive profit margin.
The separation of iron from the REEs by reductive smelting, followed by
HPAL of the corresponding slag, allowed to obtain the highest extraction yields
of REEs. The concentration of REEs in the slags was increased by a factor of about
1.4 compared to the concentration of REEs in the bauxite residue sample. A slag
subjected to slow cooling with high content of silicon, aluminium and iron was
used as a reference to determine the optimal leaching conditions. The highest
extraction yield of REEs was obtained with an acid concentration of 3 N, while
with 1 N the extractions were very low. Beyond a 3 N concentration, no further
increase in the extraction yield was observed. Therefore, the acid consumption can
be further optimized between the acid concentrations of 1 and 3 N. With HCl-
based HPAL, the lowest aluminium/iron, Al/Fe, and scandium/iron, Sc/Fe,
leaching ratios (i.e. 1.1 and 0.9, respectively) were obtained at 120 ºC with a high
REEs extraction (Sc: 79 wt.%; Y, Nd and La > 97 wt.%) and very low
concentration of titanium and silicon in the solution (< 0.5 g L-1). In the case of
H2SO4-based HPAL, the lowest Al/Fe and Sc/Fe (i.e. 0.9 and 1.1) ratios were
obtained at 150 ºC, at which the selectivity of scandium over other REEs was the
highest (i.e. 3). The treatment of the slag with H2SO4 allowed to extract ca.
- 180 -
90 wt.% of scandium, while most of the other REEs remain undissolved in the
solid fraction after leaching. The mechanism behind the low extraction yield
achieved by yttrium, neodymium and lanthanum was not studied thoroughly
during this PhD Thesis, but it is believed that by carrying out the HPAL process
with H2SO4 as a leaching reagent, these elements can be adsorbed on the surface
of silicon/aluminium-oxides compounds and/or be associated to a sodium-
lanthanide-double sulfate compound (NaLn(SO4)2×nH2O), which has been widely
reported in the literature, but was not detected in our samples by SEM-EDX during
the course of this research. Due to the low concentration of iron in the slag
(< 5 wt.%), its concentration in the leach solution was much lower (< 2 g L -1)
compared to the concentration in the leach solution of the bauxite residue
(24 g L-1 with HCl, and 13 g L-1 with H2SO4). The extraction of aluminium from
the slag was also very high (14 – 17 g L-1) with both mineral acids. Due to the
enrichment of aluminium in the slags, its concentration in the leach solution was
3 – 4 times higher than the concentration obtained after the direct processing of
the bauxite residue. The concentration of silicon and titanium, however, remained
< 0.5 g L-1. The preliminary comparative economic analysis shows that HPAL
with H2SO4 is the most interesting method for scandium recovery from the bauxite
residue, as iron is also recovered before leaching by reductive smelting. However,
further studies are required to decrease the acid consumption.
The preliminary economic analysis shows that the dry digestion process
with multi-stage leaching is economically the most interesting leaching alternative
due to the high consumption of bauxite residue and the low consumption of water.
Although, the profit margin is four times lower to the profit margins reported in
the literature (when half of the metal price is assumed), the process does not
consider the pre-treatment of bauxite residue to recover base metals, but it can be
done in a subsequent separation step, which avoid the installation of high-
temperature processes. However, further investigation is required, as the obtained
reduction in water consumption may vary as a function of the chemical and
physical properties of the different bauxite residue samples.
In summary, the main limitations of the direct treatment of bauxite
residue by acid leaching are the consumption of large amounts of acid during
neutralisation of the alkaline bauxite residue, the high co-dissolution of major
metals, particularly iron and aluminium, and the decomposition of silicate
compounds that leads to the polymerisation of amorphous silica. The HPAL and
the dry digestion processes allow to suppress the silica gel formation and titanium
dissolution, but the high co-dissolution of aluminium and iron may still represent
a drawback for the downstream processing of the leach liquor.
- 181 -
8.2 Outlook
In this PhD Thesis, three different leaching processes were studied for
the extraction of REEs and other metals from bauxite residue. Preliminary cost
calculations were performed to assess the economic feasibility. However, these
processes must be studied further, for instance, in pilot-scale setups to allow a
detailed techno-economic evaluation. Furthermore, different leach solutions were
generated from these processes. The recovery of REEs from these leach solutions
must be further studied by different separation technologies, e.g., solvent
extraction, ion exchange, precipitation, neutralisation, hydrolysis etc., as part of
an extraction-separation process. In this way, the cost comparison between
different processing alternatives can be evaluated by combining the best attributes,
i.e. lowest cost, highest selectivity, highest recovery, of the leaching processes
studied in this work.
The reductive smelting of bauxite residue allows the separation of the
majority of the iron content from the REEs, which is beneficial for limiting the
amount of iron that can be leached. This process can then be followed by dry
digestion or high-pressure acid leaching, resulting in a solution that contains REEs
and other bulk elements (minor amount of iron, aluminium, calcium). The use of
a D2EHPA and TBP as extractant in combination with Shellsol as a solvent can
be considered to effectively separate scandium from the leach liquor by solvent
extraction [1]. The use of inorganic titanium(IV) phosphates ion exchangers
requires special attention, as it has been demonstrated to be especially selective to
recover scandium from bauxite residue leachates rich in iron, aluminium and
calcium [2]. Although ionic liquids (ILs) have shown a great potential for
application in hydrometallurgy, their high viscosity represents the main drawback
in process design. However, the use of solid supports for ionic liquid (or supported
ionic liquids, SILP) has demonstrated promising results in terms of REEs
separation from the base elements in a single chromatography step, comprising
H3PO4 and HNO3 as eluting agents. Fast adsorption kinetics indicated a potential
applicability of the SILP to a large-scale process [3].
Further work required on the neutralisation-leaching method is the
enhancement of the scandium and other REEs recovery. Neutralisation of bauxite
residue with CO2 can be followed by acid leaching with HCl at high temperature,
so that silica polymerization can be avoided while HCl can be collected after
evaporation. The dry digestion method after neutralisation of bauxite residue
should also be studied. The use of microorganism for neutralisation of bauxite
residue can be studied as an alternative technology before acid leaching. Bauxite
residue neutralisation with acetic acid or HCl should be further investigated, as
they can dissolve a high amount of calcium, which can be further used to produce
CaCO3 with addition of CO2.
High-pressure acid leaching of slags from bauxite residue smelting
showed a high recovery of REEs with a very low co-dissolution of iron, titanium
and silicon, but aluminium still may represent an issue for further processing of
- 182 -
the leachate. Therefore, further research must be carried out not only to optimize
the experimental conditions for treating a specific slag, but also to fully understand
the occurrence mode of REEs in the slags generated from bauxite residue smelting.
The use of dry digestion on the treatment of the slags should also be studied.
The presence of naturally occurring radioactive material (NORM) in
bauxites residues, in particular 238U and its decay products, 232Th and its decay
products, and 40K, require special attention. NORMs are present in low
concentration, but previous research reported in the literature have already
demonstrated that the separation of major elements, such as iron, can lead to their
enrichment in the post-processed solid residue. Therefore, their concentrations
need to be followed in the solid and liquid streaming arising in the developed
process.
8.3 References
[1] W. Wang, Y. Pranolo, C.Y. Cheng, Recovery of scandium from

synthetic red mud leach solutions by solvent extraction with D2EHPA,
Sep. Purif. Technol. 108 (2013) 96–102.
doi:10.1016/j.seppur.2013.02.001.
[2] W. Zhang, R. Koivula, E. Wiikinkoski, J. Xu, S. Hietala, J. Lehto, R.
Harjula, Efficient and Selective Recovery of Trace Scandium by
Inorganic Titanium Phosphate Ion-Exchangers from Leachates of Waste
Bauxite Residue, ACS Sustain. Chem. Eng. 5 (2017) 3103–3114.
[3] D. Avdibegović, M. Regadío, K. Binnemans, Recovery of scandium(III)
from diluted aqueous solutions by a supported ionic liquid phase (SILP),
RSC Adv. 7 (2017) 49664–49674. doi:10.1039/C7RA07957E.
- 183 -
APPENDIX
I. Supplementary material of Chapter 3
Table A-I.1: Full mineralogical composition analysis (wt.%) of the bauxite residue
samples.
HP HPT
Phase Chemical formula BR NBR
NBR NBR
Dicalcium-silicate (C2S) Ca2SiO4 11 8 2 3
Hydrogrossular Ca3Al2 (SiO4)1.53(OH)5.88 7 7 9 6
Grossular Ca3Al2(SiO4)3 5 8 10 8
Tricalcium-silicate (C3S) Ca3SiO5 4 4 3 4
Bayerite -Al(OH)3 4 3 3 3
Na6(Al6Si6O24)·
Sodalite 3 2 1 1
xNaOH·(8−2x)H2O
Calcium phyllo-dodeca-
Al12CaO27Si4 2 2 3 2
alumotetrasilicate
Rutile TiO2 2 1 3 3
Na6Ca2Al6Si6O24(CO3)2·
Cancrinite 1 1 4 7
2H2O
Total 95 96 95 96
Unidentified/amorphous 5 4 5 4
BR = non-neutralised bauxite residue, NBR = neutralised bauxite residue at ambient conditions (qCO2:
0.25 L/min, T: 25 ºC, L/S: 5), HP NBR = high-pressure neutralised bauxite residue (PCO2: 30 bar, T:
25 ºC, L/S: 5), HPT NBR = high-pressure and high-temperature neutralised bauxite residue (PCO2:
30 bar ,
T: 150 ºC, L/S: 5).….
- 185 -
Table A-I.2: Normalised mineralogical composition (wt%) of the bauxite residue samples
(cf. Table A-I.1).
HP HPT
Phase Chemical formula BR NBR
NBR NBR
Dicalcium-silicate
Ca2SiO4 12 8 3 3
(C2S)
Hydrogrossular Ca3Al2 (SiO4)1.53(OH)5.88 7 7 9 6
Grossular Ca3Al2(SiO4)3 5 8 11 9
Tricalcium-silicate
Ca3SiO5 5 4 3 4
(C3S)
Bayerite -Al(OH)3 4 3 3 3
Na6(Al6Si6O24)·
Sodalite 3 2 1 1
xNaOH·(8−2x)H2O
Calcium phyllo-
dodeca- Al12CaO27Si4 2 2 3 2
alumotetrasilicate
Rutile TiO2 2 2 3 3
Na6Ca2Al6Si6O24(CO3)2·
Cancrinite 1 1 4 7
2H2O
Total 100 100 100 100

BR = non-neutralised bauxite residue, NBR = neutralised bauxite residue at ambient conditions (qCO2:
0.25 L/min, T: 25 ºC, L/S: 5), HP NBR = high-pressure neutralised bauxite residue (PCO2: 30 bar, T:
25 ºC, L/S: 5), HPT NBR = high-pressure and high-temperature neutralised bauxite residue (PCO2: 30
bar , T: 150 ºC,
L/S: 5).….
- 186 -
II. Supplementary material of Chapter 5
25 25
20 Nd 20
Nd extraction, wt.%
La extraction, wt.%
La
15 15
10 10
5 5
0 0
0 10 20 30
Sc extraction, wt%
Figure A-II. 1: Correlation between extraction of lanthanum (La) and neodymium (Nd) with
scandium (Sc) in H2SO4-based dry-digested bauxite residue sample with (L/S: 5, tdigestion:
experiments performed in triplicate.
- 187 -
15
(a)
Al Ti Si
Concentration, g L-1
10
0
0 1 2 3 4 5 6
Stage
15
(b)
Al Ti Si
10
0
0 1 2 3 4 5 6
Stage
Figure A-II. 2: Concentration of Al, Ti and Si (expressed in g L-1) in the leachate after water
leaching of each stage. Dry digestion was performed with (a) 788 g HCl/kg BR and (b) 412
g H2SO4/kg BR (L/S: 5, tdigestion: 24 h, tleaching: 24 h, 200 rpm)
- 188 -
III. Supplementary material of Chapter 6
Table A-III.1: Rare earth elements (in mg kg-1) in the bauxite residue and in the slag samples generated after smelting experiments.
Bauxite Slag I Slag II Slag III

Element
residue Fast cooling Slow cooling Fast cooling Slow cooling Fast cooling Slow cooling
Sc 97.8 ± 1.7 125.5 ± 2.0 122.5 ± 4.9 117.4 ± 1.4 128.3 ± 1.7 128.1 ± 1.5 124.0 ± 1.6
Y 65.7 ± 1.6 117.4 ± 1.1 118.2 ± 5.2 92.4 ± 0.7 105.2 ± 0.8 107.1 ± 1.6 105.6 ± 1.3
La 143.9 ± 3.0 211.9 ± 5.2 206.9 ± 18.0 159.2 ± 1.3 167.9 ± 2.7 174.8 ± 1.1 168.7 ± 1.5
Ce 352.7 ± 3.7 451.2 ± 10.5 447.6 ± 29.7 355.2 ± 1.5 370.0 ± 3.9 390.1 ± 3.6 379.2 ± 2.1
Pr 18.1 ± 0.6 35.0 ± 0.8 35.5 ± 1.6 29.0 ± 0.5 31.0 ± 1.5 32.4 ± 0.6 31.2 ± 0.3
Nd 64.8 ± 2.6 126.1 ± 2.1 126.9 ± 5.2 103.7 ± 1.0 111.2 ± 3.9 116.3 ± 2.2 111.8 ± 0.7
Sm 12.7 ± 0.5 23.2 ± 0.4 23.4 ± 0.8 19.1 ± 0.0 20.4 ± 0.2 21.7 ± 0.1 21.0 ± 0.1
Eu 2.7 ± 0.1 5.0 ± 0.4 4.8 ± 0.2 3.9 ± 0.0 4.2 ± 0.0 4.5 ± 0.1 4.4 ± 0.0
Gd 16.4 ± 0.7 27.1 ± 0.3 26.9 ± 1.2 21.8 ± 0.1 23.8 ± 0.4 25.2 ± 0.1 24.6 ± 0.2
Tb 19.1 ± 0.8 8.4 ± 0.1 6.5 ± 3.0 2.6 ± 0.1 2.7 ± 0.4 2.6 ± 0.1 3.1 ± 1.0
Dy 11.8 ± 0.6 19.4 ± 0.0 19.3 ± 0.8 15.8 ± 0.1 17.6 ± 0.3 18.5 ± 0.3 18.2 ± 0.2
Ho 2.5 ± 0.1 4.0 ± 0.0 4.1 ± 0.2 3.3 ± 0.0 3.7 ± 0.1 3.8 ± 0.0 3.8 ± 0.0
Er 8.0 ± 0.3 12.4 ± 0.0 12.7 ± 0.5 10.2 ± 0.1 11.7 ± 0.3 12.0 ± 0.2 11.9 ± 0.1
Tm 1.2 ± 0.0 1.8 ± 0.0 1.8 ± 0.1 1.5 ± 0.0 1.7 ± 0.0 1.7 ± 0.0 1.7 ± 0.0
Yb 8.2 ± 0.3 12.0 ± 0.2 12.6 ± 0.5 9.8 ± 0.2 11.9 ± 0.2 11.8 ± 0.1 11.8 ± 0.1
Lu 1.6 ± 0.1 1.9 ± 0.0 2.0 ± 0.1 1.5 ± 0.0 1.8 ± 0.0 1.8 ± 0.0 1.8 ± 0.0
- 189 -
30 H2SO4
Concentration, g L-1 25
Al Fe Si Ti Ca Na
20
15
10
0
BR
FC
SC
SC
I. F
I.S
.F
I I.
II .
.
III
III
S-
S-
S-
S-
S-
S-
Sample
30 HCl
Al Fe Si Ti Ca Na
25
20
15
10
0
BR
FC
SC
C
I. S
.S
I.F
.F
II .
I I.
I II
III
S-
S-
S-
S-
S-
S-
Sample
Figure A-III. 1: Concentration of major elements (Al, Fe, Ti, Si, Ca and Na) after HPAL
with H2SO4 (at 150 ºC) and HCl (at 120 ºC) of bauxite residue and bauxite residue slags
(acid concentration: 3 N, L/S: 10, t: 1 h).
- 190 -
30 H2SO4
Concentration, mg L-1
25
Sc Y La Nd
20
15
10
0
BR
FC
SC
SC
I. F
I.S
.F
I I.
II .
.
III
III
S-
S-
S-
S-
S-
S-
Sample
30 HCl
Concentration, mg L-1
25 Sc Y La Nd
20
15
10
0
BR
SC
C
I. F
I.S
.F
.S
II.
II .
III
I II
S-
S-
S-
S-
S-
S-
Sample
Figure A-III. 2: Concentration of selected REEs (Sc, Y, La and Nd) after HPAL with H2SO4
(at 150 ºC) and HCl (at 120 ºC) of bauxite residue and bauxite residue slags (acid
concentration: 3 N, L/S: 10, t: 1 h).
- 191 -

PHD Thesis - Final

Uploaded by

Copyright:

Available Formats

You might also like

PHD Thesis - Final

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PHD Thesis - Final

Uploaded by

Copyright:

Available Formats

INNOVATIVE TECHNOLOGIES FOR

RARE EARTH ELEMENT RECOVERY

Rodolfo Andres Marin Rivera

Bauxite residue is a by-product that originates during the Bayer process

Articles in international scientific conferences

Bauxite residue (also called red mud) is a by-product produced in the

1.1 Turning the bauxite residue into an opportunity

1.2 Limitations for rare-earth elements recovery from bauxite residue

1.3 Scope of the PhD Thesis

Neutralisation Acid leaching PLS I.2

PLS: Pregnant leach solution; HPAL: High pressure acid leaching

occurrence of naturally occurring radioactive materials (NORM) in bauxite

process residues: a critical review, J. Clean. Prod. 99 (2015) 17–38.

2.1 From bauxite to bauxite residue

Bauxite residue is mainly composed of compounds that have a low

2.2 Occurrence of rare-earth elements in bauxite residue

Figure 2.2: EPMA-WDS quantitative elemental mapping of a reacted LREE particle of

2.3 Rare-earth elements recovery from bauxite residue

Biometallurgical methods have also been investigated for the recovery of

2.4 Towards an efficient recovery of REEs from bauxite residue

Neutralisation of bauxite residue before acid leaching

Separation of major metals from bauxite residue

Figure 2.4: Conceptual process flow sheet of roasting-magnetic separation-leaching for

Solid-state reduction of bauxite residue by microwave hydrogen plasma

Smelting of bauxite residue (normally performed at temperatures

Silicon dissolution during acid leaching

During acid leaching of bauxite residue, silicate compounds can dissolve

M2 SiO4(s) + 4HCl(l) + H2 O(l) → 2MCl2(aq) + Eq. 2.1

H4 SiO4 (aq) + H2 O → H3 SiO4− + H3 O+ Eq. 2.2

Si𝑛 O𝑚 OH + H4 SiO4 → Si𝑛+1 O𝑚+2 OH + 2H2 O Eq. 2.3

In order to avoid the dissolution of silica and, consequently, the eventual

Figure 2.7: Conceptual process flow sheet of sulfation-roasting-leaching. Adapted from

High-pressure acid leaching (HPAL) can also limit the silicon

2.5 Existence of naturally occurring radionuclide activity in bauxite

238U 232Th 40K

The EU Radiation Protection Guideline 112 has recommended the range

The recovery of REEs by direct acid leaching of bauxite residue is limited

[1] World Aluminium and the European Aluminium Association, Bauxite

2018: pp. 132–133. doi:https://doi.org/10.3133/70194932.

Ochsenkühn, P. Georgiou, N. Defteraios, F. Tsopelas, M. Ochsenkühn-

[55] Y. Zhang, Y. Hua, X. Gao, C. Xu, J. Li, Y. Li, Q. Zhang, L. Xiong, Z.

[66] M.K. Hamdy, F.S. Williams, Bacterial amelioration of bauxite residue

[79] G. Alkan, B. Yagmurlu, S. Cakmakoglu, T. Hertel, Ş. Kaya, L. Gronen,

Hamburg, Ger. 2 – 5 October, 2017, 2017: pp. 603–614.

[112] W. Duyvesteyn, System and method for recovery of scandium values

[125] V. Cuccia, A.H. De Oliveira, Z. Rocha, Radionuclides in Bayer Process

issues arising from the introduction of (NORM) residues in building

ABSTRACT - The present work considers the neutralisation of bauxite residue

Based on the published paper

The objective of this paper is to evaluate the recovery of valuable metals

3.2 Material and methods

The bauxite residue studied in this paper was kindly provided by

3.3 Results and discussion

Characterisation of the bauxite residue

Mineralogical analysis allowed the identification of several mineral

Table 3.3: Normalised mineralogical composition of the bauxite residue sample.

Phase Chemical formula wt.%

Dicalcium-silicate (C2S) Ca2SiO4 12