See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/226752751

Solution-Chemistry Analysis of Ammonium Bicarbonate Consumption in

Rare-Earth-Element Precipitation

Article  in  Metallurgical and Materials Transactions B · October 2003

DOI: 10.1007/s11663-003-0031-z

CITATIONS READS

52 1,844

6 authors, including:

Zhenghe Xu Shengming Xu
University of Alberta Tsinghua University
473 PUBLICATIONS   21,718 CITATIONS    154 PUBLICATIONS   4,323 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Hydrocyclone Project View project

surface and interface science View project

All content following this page was uploaded by Z.A. Joe Zhou on 09 August 2018.

The user has requested enhancement of the downloaded file.

17_02-618B-B.qxd 9/11/03 11:48 AM Page 611

Solution-Chemistry Analysis of Ammonium Bicarbonate

Consumption in Rare-Earth-Element Precipitation
R. CHI, Y. HU, G. ZHU, S. XU, Z. ZHOU, and Z. XU

A solution-chemistry analysis is applied to estimating the consumption of ammonium bicarbonate in

the recovery of rare-earth (RE) elements from leachates of weathered clays. The theoretical analysis
shows that a two-step process is needed for recovering RE from the leachates of the weathered clays
by precipitation using ammonium bicarbonate. The first step is a precipitation at solution pH 5 to
remove impurities such as Fe and Al. The second step is to precipitate RE by adjusting the solution
pH above 8. The consumption of ammonium bicarbonate was found to depend on the concentration
of RE elements and impurities in the leachates. The total amount of ammonium bicarbonate con-
sumption for the entire process was determined experimentally, and the results showed an excellent
agreement with that calculated based on solution-chemistry analysis. The decomposition of H2CO3 was
identified as one of the main causes of ammonium bicarbonate overdose, accounting for up to 41 pct
in comparison to 20 pct consumption for the removal of impurities. The amount of ammonium bicar-
bonate required in terms of the NH4HCO3: RE2O3 (RE oxides) molar ratio was found to be 4:1 for
maximal RE recovery. An overall RE recovery around 90 pct was achieved with a product purity being
about 90 pct.

I. INTRODUCTION and Ca2
. In this communication, a solution-chemistry

RARE-EARTH (RE) elements are widely used as essen-
tial components in permanent magnets, catalysts, special steels,
superconductors, and fluorescent powders, to name a few[1]
applications. According to their atomic weight and properties,
the RE elements can be divided into light (La, Ce, Pr, and
Nd), medium (Sm, Eu, and Gd), and heavy (Td, Dy, Ho, Er,
Tm, Yb, Lu, and Y) RE elements.[1] In general, the light RE
elements are extracted from bastnasite and monazite minerals,
while the medium and heavy RE elements are disseminated
in weathered clay minerals as absorbed ions.[2–6] These RE
elements can be released into solutions by ion-exchange mech-
anisms using certain types of cations. The dissolved RE
elements can then be recovered by solvent extraction or
precipitation using either oxalic acid[7] or ammonium bicar-
bonate (NH4HCO3). Compared to precipitation, solvent extrac-
tion[8] requires more expensive equipment, and therefore, the
capital cost is higher. For industrial operations, RE precipitation
with ammonium bicarbonate is preferred, as ammonium
bicarbonate is less expensive than oxalic acid and does not
cause severe environmental consequences.
A commonly encountered problem in RE recovery using
ammonium bicarbonate precipitation is the excess consump-
tion of ammonium bicarbonate, with a molar ratio of
NH4HCO3 to RE2O3 greater than 2.5 times the stoichiometric
requirements. The exact reason for this excess requirement
of NH4HCO3 is most likely due to the precipitation of other
released non-RE metallic species such as Fe3
, Al3
, Mg2

R. CHI, Professor, is with Wuhan Institute of Chemical Technology,
Wuhan, China 430073. Z. ZHOU, Research Associates, and Z. XU, Pro-
fessor, NSERC/EPCOR/AERI Industrial Research Chair, are with the
Department of Chemical and Materials Engineering, University of Alberta,
Edmonton, AB, Canada T6G 2G6. Contact e-mail: zhenghe.xu@ualberta.ca
Y. HU, Professor, is with the School of Resources and Bioengineering,
Central South University, Changsha, Hunan, P.R. China 410083. G. ZHU
and S. XU, Associate Professors, are with the Institute of Nuclear Energy
Technology, Tsinghua University, Beijing, P.R. China 100084.
Manuscript submitted December 30, 2002.
approach is applied to a phenomenological analysis of a RE–
ammonium bicarbonate two-step precipitation system. The
analysis allows a quantitative description of various contri-
butions of ammonium bicarbonate consumption. With the
approach developed here, operating conditions for minimal
ammonium bicarbonate consumption with a maximized
purity of RE oxides are derived.
II. SOLUTION CHEMISTRY ANALYSIS
A. General Consideration
Ammonium bicarbonate (NH4HCO3) is a commonly used
nitrogenous fertilizer. It is produced by heating a mixture
of weak acid (H2CO3) and alkali (NH3H2O). The hydroly-
sis and ionization of ammonium bicarbonate when dissolved
in water are governed by the solution reactions given in
Table I. The thermodynamic data used in this communica-
tion are taken from References 8 through 12. These data are
obtained for the standard condition of 298.15 K and 1 atm
pressure, unless otherwise specified. The solution reactions
listed in Table I show that ammonium bicarbonate in water
undergoes a series of ionizations to release both H
and
OH. Among the equilibrium constants, K2h has the largest
value, indicating that the solution of ammonium bicarbon-
ate is always alkaline.
In a process of using ammonium bicarbonate as a pre-
cipitant to recover RE elements, it is extremely important
to know how much CO32 is present, as only CO32 reacts
with RE ions to form RE carbonate precipitate: (RE)2(CO3)3.
It is well known that carbonic acid is a weak acid. The
total concentration of soluble carbonic species ([CO32]T) is
given by

3 ]T  [CO3 ]aq
[HCO3 ]aq
[H2CO3]aq [1]
[CO2 2
where the subscripts T and aq represent the total and species
concentration in solution, respectively. The fraction of each

METALLURGICAL AND MATERIALS TRANSACTIONS B VOLUME 34B, OCTOBER 2003—611

17_02-618B-B.qxd 9/11/03 11:48 AM Page 612

Table I. Solution Reactions of Ammonium Bicarbonate[8–12]

Equilibrium
Solution Reaction Constant Reaction
NH4HCO3  NH
4
HCO3 1
NH
4  NH3
H
K1h  5.6  1010 2
HCO3
H2O  H2CO3
OH K2h  2.3  106 3
H2CO3  CO2↑
H2O 4
H2CO3  HCO 3
H


K1a  4.47  107 5
HCO 3  CO3
H
2

K2a  4.68  1011 6

species with respect to the total carbonate concentration can

then be expressed as

[CO2
3 ]aq K1aK2a
w0   [2] Fig. 1—Calculated species distribution of ammonium bicarbonate in water
[CO2
3 ]T [H
]2
K1a[H
]
K1aK2a as a function of solution pH.

[HCO3 ]aq K1a[H
]

w1   [3]
[CO2
3 ]T [H
]2
K1a[H
]
K1aK2a Table II. Precipitation Reaction of Leachate when
Ammonium Bicarbonate is Added[8–12]

[H2CO3]aq [H
]2 Precipitation Reactions pC

w2   [4]
[CO2
3 ]T [H
]2
K1a[H
]
K1aK2a 2RE3

3HCO3  pCRE23
 8.54  1.5 log A
RE2(CO3)3↓
3H

Ca2

HCO3  pCCa22
 10.2  log A
where 0
1
2  1. Using the thermodynamic data CaCO3↓
H

given in Table I, Eqs. [2] through [4] allow us to calculate RE3

3HCO3  pCRE13
 18.5
3pH
the distribution of individual species as shown in Figure 1. RE(OH)3↓
3CO2↑
It is readily seen from the calculated species distribution that Al3

3HCO3  pCAl13
 9.1
3pH
CO32 is the predominant species at a pH greater than 10. Al(OH)3
3CO2↑
Between pH 7 and 9, the main species is HCO3 , with a min- Fe3

3HCO3  pCFe13
 3.5
3pH
imal amount of CO 32 and H2CO3 (approximately 0.5 pct Fe(OH)3
3CO2↑
Mg2

2HCO3  pCMgl2
 17.3
2pH
each). At a pH below 7, H2CO3 is the predominant species.
Mg(OH)2
2CO2↑
These balances collectively control the amount of ammonium
bicarbonate needed to maintain a desired level of CO 32 in *Subscript 1 represents hydroxide, and subscript 2 represents carbonate.
The term A in this table is defined as ([H
]2
K1a[H
]
K1aK2a).
a given system, as will be discussed later.

B. Selective Precipitation The main impurities in the leachates of the weathered RE

A leachate of RE ores commonly contains 1 to 8 g/ L of
Re2O3, along with a small amount of impurities such as
Al3
, Fe3
, Ca2
, and Mg2
. When ammonium bicarbon-
ate is added into the leachate, the solution reactions given
in Table II are anticipated. It is evident that impurity ions
will precipitate with RE carbonate, depending on the amount
of carbonates available and relative reactivity. The incor-
porated impurity ions will affect the purity of the resultant
RE carbonate. In order to minimize the incorporation of
impurity ions in RE carbonate, it is highly desirable to
remove impurity ions prior to RE precipitation. This can
be accomplished by raising the pH of the solution with
ammonium bicarbonate to a desired value that is suitable
for impurity-ion precipitation with minimal coprecipitation
of RE elements. After removing impurity ions as metal
hydroxide precipitates, the RE precipitation as carbonates
can be realized by further increasing the solution pH to
provide sufficient carbonates. To define these two-step pre-
cipitation conditions, a solution-chemistry calculation is
conducted, as described below.
ores are Al3
, Fe3
, Ca2
, and Mg2
, with Al3
and Fe3

being the dominant ones. Al3
and Fe3
can be easily pre-
cipitated as Al(OH)3 and Fe(OH)3 by adjusting the pH.
Unfortunately, RE elements can also form RE(OH)3 above
a certain pH. It is, therefore, extremely important to choose
a suitable pH value to precipitate Al3
and Fe3
selectively,
while keeping RE ions in solution. To determine a suitable
pH range for selective precipitation of metal impurity ions,
a pC-pH diagram is constructed using the solubility-product
constants for competitive precipitation reactions given in
Table II.
From the pC-pH diagram shown in Figure 2, it can be seen
that the precipitation of RE ions is negligible at leachate pH
up to 5 adjusted with ammonium bicarbonate, while almost
all of Al3
and Fe3
are present as Al(OH)3 and Fe(OH)3 pre-
cipitates at pH 5. Theoretically, the residual Al3
and Fe3

concentrations are lower than 106 mol/L. It should be men-
tioned here that the pCRE2
3 line is on the left side of the
p CRE13
line, indicating that RE ions would preferentially
form RE carbonates, if there is a sufficient amount of

612—VOLUME 34B, OCTOBER 2003 METALLURGICAL AND MATERIALS TRANSACTIONS B

17_02-618B-B.qxd 9/11/03 11:48 AM Page 613
Fig. 2—pC-pH diagram of RE3
, Al3
, Fe3
, and Ca2
.
carbonates available in the solution. As shown in Figure 1,
however, the amount of soluble carbonates is minimal at
pH  8, suggesting that RE ions would remain in solution at
a pH below 8. It is clear that selective precipitation of impu-
rities as metal hydroxides from leachate of the weathered
RE ores can be achieved by adjusting the leachate pH to 5
with ammonium bicarbonate. The ammonium bicarbonate is
used as a pH modifier, as it would simplify the subsequent
RE precipitation as RE carbonates.
After removing the impurities as solid metal hydroxides
by solid-liquid separation, the precipitation of RE2(CO3)3
can proceed. To precipitate RE ions as solid carbonates, a
sufficient amount of carbonates is required. From the solu-
tion speciation diagram shown in Figure 1, it is evident
that only when the solution pH is greater than 8 can suffi-
cient amounts of bicarbonate be introduced into the solu-
tion. As a result, RE precipitation needs to be performed at
a solution pH greater than 8, adjusted, preferably, by ammo-
nium bicarbonates, which function as a pH modifier as well
as a carbonate source.
C. Ammonium Bicarbonate Consumption
1. Impurity removal
For convenience, we assume that all impurities given in
Table III precipitate when the pH is adjusted to 5 using
ammonium bicarbonate. In this case, the amount of ammo-
nium bicarbonate (HTp1) used for removing impurities in a
liter of leachate is
HTp1  Hp1
Hkp1 [5]
where Hp1 and Hkp1 represent, respectively, the stoichiometric
requirement for the precipitation of impurities and the amount
of ammonium bicarbonate needed for raising and maintaining
the pH at the desired value to ensure complete impurity
removal. According to the reactions given in Table II, Hp1
can be expressed by
n precipitation.
H p1  3[Al3
]
3[Fe3
]
...  a Zi Mi [6]
i1
where Zi and Mi represent the ionic charge and the molar
concentration of ion i, respectively. During the precipitation
METALLURGICAL AND MATERIALS TRANSACTIONS B
of impurity hydroxides, carbon dioxide (CO2) evolves and
leaves the solution as gases, as shown in Table II. To cal-
culate Hkp1 only a basic hydrolysis reaction (Reaction [3]
in Table I) is considered here. From this consideration,
[OH  ]2 Kw
Hkp1  [HCO3 ]aq 
w1K1a
K3w 1 [H
]21
K1a[H
]1
K1aK2a 2
 [7]
K21a[H
]31
where [H
]1 represents the equilibrium H
concentration
required for removing impurities in the first step. The con-
sumption of ammonium bicarbonate in removing impurities
is, therefore, given by
n
H TP1  Hp1
Hkp1  a ZiMi
i1
K3w 1 [H
]21
K1a[H
]1
K1aK2a 2

[8]
K21a[H
]31
2. RE-carbonate precipitation
After removing impurities, the solution contains RE ele-
ments and a small amount of calcium. Rare earths and calcium
will precipitate as carbonates, provided that a sufficient
amount of carbonates are available. The total consumption
of ammonium bicarbonate (HTp2) can be divided into three
parts: (1) precipitation reaction; (2) carbonic acid decompo-
sition; and (3) raising the pH value to ensure a complete RE
precipitation.
1. Precipitation of RE ions. Based on the stoichiometric
requirement of the precipitation reactions, the consumption
of ammonium bicarbonate (Hp2) used for precipitating
RE ions is given by
3 1 n
H p2  [RE3
]
[Ca2
]
...  a ZjMj [9]
2 2 j1
2. Carbonic acid decomposition. From the precipitation
reactions presented in Table II, it is evident that the pre-
cipitation of carbonates releases H
ions. To keep a con-
stant solution pH, H
ions released have to be consumed.
These H
ions combine with HCO3 to become CO2
(g) and H2O. The amount of bicarbonate (Hd2) consumed
as such is given by
3 1 n
H d2  [RE ]
[Ca ]
...  a ZjMj
3
2

[10]
2 2 j1
where the subscript d2 denotes the ammonium bicarbonate
consumption due to carbonic acid decomposition. Unfortu-
nately, this portion of ammonium bicarbonate consumption
cannot be avoided, although it is not related to RE-carbonate
3. Raising the pH. In order to ensure a complete precipi-
tation of RE ions as carbonates, sufficient ammonium
bicarbonate is required to raise the pH to a desired
value. For RE-carbonate precipitation, the amount of
VOLUME 34B, OCTOBER 2003—613
17_02-618B-B.qxd 9/11/03 11:48 AM Page 614

ammonium bicarbonate needed for raising the pH (Hkp2)
is given by
1Ksp(Nd) 1
H kp2    3.14  10 6
w0[CRE] w0
where Ksp(Nd) is the solubility-product constant of Nd2(CO3)3
and [CRE] is the RE concentration at complete RE-carbonate
precipitation. In Eq. [11], Ksp(Nd) was chosen to represent the
average solubility-product value for entire RE elements as
commonly practiced.[1,7] This approximation is justified, as
Nd represents one of the three richest RE elements (ca. 25 pct)
and has the highest solubility product of RE carbonates (5.36
 1026) compared with the other two richest RE elements
(La2O3 and Y2O3), as indicated in Table IV. This approxi-
mation results in an upper limit in estimating ammonium
bicarbonate consumption.
The total amount of ammonium bicarbonate (HTp2) for
precipitating RE ions is, then, given by the summation of
the three components, Hp2, Hd2, and Hkp2, i.e.,
n
H Tp2  a Zj Mj
III. EXPERIMENTAL PROCEDURES
A. Materials
[11] The leachates of RE-containing clay minerals from four
different mines, Ningde (ND), Wuping (WP), Jianning (JN),
and Qingliu (QL), all in China, were examined. The chem-
ical analysis showed that the mother lachates contain
substantial amounts of aluminum, iron, magnesium, and
calcium, with a relatively constant acidic pH (Table III).
The RE component is distributed among a wide range of
RE minerals (Table IV). The determined distribution of
RE elements by X-ray fluorescence spectroscopy is shown
in Table IV. It is evident that La, Nd, and Y account for
about 80 pct of the total RE elements. The RE concentra-
tion in leachates was determined by a gravimetric method,
using oxalic acid as a precipitant, and was checked by a
volumetric method, using EDTA as a titrant.[5] All the
experiments were conducted at room temperature (25 °C)
and ambient pressure.
B. Precipitation Procedure

1 [H
]22
K1a[H
]2
K1aK2a 2

j1
A 1-L leachate was transferred to a 1.5 L Plexiglas con-
1/3
Ksp(Nd)
a b [12]
tainer. A desired amount of ammonium bicarbonate was added

in the solid form into the leachate under stirring. For remov-
K1aK2a [CRE]2/3
ing the impurities, the slurry pH was adjusted by ammonium
where [H
]2 is the H
concentration at complete RE-car- bicarbonate to 5 to precipitate Al3
, Fe3
, and Mg2
ions.
bonate precipitation. To facilitate precipitate-liquid separation, 2 mL 0.1 pct poly-
The total ammonium bicarbonate consumed for the whole acrylamide flocculant solution was added to the suspension.
process (HT) can be evaluated by the summation of HTp1 and The flocculated precipitates were recovered by filtration
HTp2. The ammonium bicarbonate efficiency (ABCE) can and washed with a dilute ammonium bicarbonate solution
then be calculated by
ZREMRE
ABCE (pct)   100 [13]
HTp1
HTp2 Table III. Concentration (mmol/dm3) of Major Elements in
It should be noted that, to simplify the derivation and the and pH of the Leachates Taken from Four Different Clay
subsequent calculation, the species concentration instead of Mines
activity was used in this study. Considering that the solu- Mine RE Al Fe Mg Ca pH
bility constants of RE carbonates (Ksp) are extremely small,
in the order of 1026, the use of concentration is justified, ND 39 4.4 4.1 0.26 2.1 3.6
WP 41 4.8 4.3 0.42 2.4 3.4
as the residual RE and carbonate concentration is extremely
JN 38 4.2 3.7 0.30 3.1 3.5
low. As a result, the effect of ionic strength on the calculated QL 37 4.5 3.9 0.32 2.8 3.5
ABCE is negligible.

Table IV. The Partitioning (Percent) of RE Elements from Four Different Mines

Mine La2O3 CeO2 Pr6O11 Nd2O3 Sm2O3 Eu2O3 Gd2O3 Tb2O3

28
Ksp (10 ) 82.5 1.08 33.7 536 26.0 29.7 2400 —
RE2(CO3)3
Molar mass 325.8 312.2 1021.5 336.5 348.7 350.0 362.5 365.9
ND 35.68 0.41 6.93 24.03 4.16 0.90 3.92 0.53
WP 25.29 0.57 6.18 21.72 4.76 0.85 5.80 0.72
JN 24.37 3.72 7.12 27.17 5.52 1.15 4.94 0.67
QL 30.14 2.14 5.94 20.24 3.58 0.54 3.91 0.58
Mine Dy2O3 Ho2O3 Er2O3 Tm2O3 Yb2O3 Lu2O3 Y2O3 Y2O3
Ksp (1028) 8400 — 44.1 — 3380 — 9.35 CeO2
RE2(CO3)3
Molar mass 373.0 377.9 382.5 385.9 394.1 397.1 225.8
ND 2.56 0.31 1.52 0.30 1.24 0.30 17.60 0.39
WP 4.02 0.43 1.98 0.30 1.17 0.30 26.24 0.69
JN 3.71 0.30 1.56 0.30 0.57 0.30 18.99 0.45
QL 2.92 0.43 1.81 0.30 0.77 0.30 25.63 0.59

614—VOLUME 34B, OCTOBER 2003 METALLURGICAL AND MATERIALS TRANSACTIONS B

17_02-618B-B.qxd 9/11/03 11:48 AM Page 615

(5 pct by weight). The RE elements were then precipitated
with ammonium bicarbonate by raising the pH of the puri-
fied leachate to higher than 8. The RE precipitate was
separated using the same procedures described in removing
impurities. To ensure that the process had reached equilib-
rium, all the precipitation experiments lasted for 2 hours,
unless otherwise stated. A 2-hour precipitation period was
proven to be sufficient, as shown by the results presented
later in Figure 5. The obtained RE precipitates were roasted
at 900 °C for 2 hours to convert RE carbonates to RE oxides.
The solid weight was determined, which allows RE recov-
ery and RE loss (wt pct) to be calculated, by comparing the
weight of the RE oxides with the values in the feed leachate
(Table III).
IV. RESULTS AND DISCUSSION
Typical RE recovery and residual impurity concentration as
a function of ammonium bicarbonate addition (NH4HCO3:
RE2O3) are shown in Figure 3. At an ammonium bicarbon-
ate–to–RE2O3 molar ratio below 1, increasing the NH4HCO3
dosage significantly decreased the amount of impurities left
in the leachate, while it increased the RE recovery only
marginally. With an ammonium bicarbonate –to–RE2O3 molar
ratio of 1, for example, more than 80 pct of the impurities
were removed from the leachate, while only about 16 pct
of RE ions precipitated. At a molar ratio greater than 1, the
impurity content in the leachate decreased only marginally,
while the RE recovery increased significantly by further
increasing the ammonium bicarbonate–to –RE2O3 molar
ratio. Achieving a RE recovery of up to 96 pct required an
ammonium bicarbonate–to–RE2O3 molar ratio greater than
4:1. This ratio is similar to the calculated theoretical value
of 3.7:1, as indicated in Table V. It is interesting to note
that the results obtained with the four different ores collapsed
on a single curve, suggesting a universal phenomenon in
RE recovery.
The characteristic curves in Figure 3 appear to suggest a
two-step precipitation process, i.e., impurity metal hydroxide
precipitation at an ammonium bicarbonate–to–RE2O3 molar
ratio of 1, followed, after impurity precipitates removal, by
an RE-carbonate precipitation which completes at an ammo-
nium bicarbonate–to–RE2O3 molar ratio of 4. To further con-
firm this suggestion, the leachate pH was measured during the
ammonium bicarbonate addition, and the results are shown in
Figure 4. The pH of the leachate increased with increasing
NH4CO3:RE2O3 molar ratio. At the molar ratio of 1, the leachate
pH was found to be around 5. Based on the species-distribution
diagrams in Figures 1 and 2, all the Al3
, Fe3
, and Mg2
ions
were precipitated as hydroxides at this pH. When the ratio of
NH4HCO3:RE2O3 increased to higher than 3, the pH of the
leachate increased to 8. At this pH, a sufficient amount of car-
bonate is present in the leachate to coprecipitate calcium and
RE, as suggested by the speciation diagrams shown in Figures 1
and 2. Therefore, the pH of the solution can be used as a marker
to determine the condition of RE precipitation.
From the previous observation, it is evident that the
added ammonium bicarbonate not only reacts with RE ions
to form RE-carbonate precipitates, but it is also consumed
to adjust the pH of the leachate and to precipitate impurity

Fig. 3—Effect of ammonium bicarbonate amount (NH4HCO3:RE2O3, mole Fig. 4—Effect of the ammonium bicarbonate amount (NH4HCO3:RE2O3
ratio) on RE recovery (pct) and impurity removal (wt pct of the feed). molar ratio) on the pH of leachates.

Table V. Ammonium Bicarbonate Consumption for Leachates from Four RE Mines (mmol/l)

Rd* ABCE Calculated NH4HCO3: Experiment NH4HCO3:

Mine HTp1 Hd2 HpRE HT (pct) (pct) RE2O3 (Mole Ratio) RE2O3 (Mole Ratio)
ND 26.0 59.9 58.5 144.4 41.48 40.51 3.70:1.00 4:1
WP 28.1 63.9 61.4 153.4 41.66 40.03 3.74:1.00 4:1
JN 24.3 60.0 57.0 142.3 42.16 40.76 3.74:1.00 4:1
QL 25.8 58.2 55.4 139.4 41.74 39.74 3.77:1.00 4:1
*Rd (pct)  Hd2/HT, and HT  Htp1
Htp2

METALLURGICAL AND MATERIALS TRANSACTIONS B VOLUME 34B, OCTOBER 2003—615

17_02-618B-B.qxd 9/11/03 11:48 AM Page 616
Fig. 5—Effect of stirring time (min) on impurity removal (pct) and RE
loss (pct).
ions. To evaluate the contributions of ammonium bicar-
bonate consumption in RE precipitation, Eqs. [8], [10],
[11], and [13] can be used to assess HTp1, Hd2, HpRE, HT,
Rd, and ABCE. Using the constants given in Tables I and
II, the calculated results are shown in Table V. It can be
seen from this table that the main consumption of the
ammonium bicarbonate is the decomposition of H2CO3,
accounting for up to 41 pct. As a result, the ABCE is only
about 40 pct. It was noted that the consumption of the
ammonium bicarbonate determined experimentally is
slightly higher than the theoretical values. It appears that
additional bicarbonate was needed to replenish bicarbon-
ate loss due to decomposition of H2CO3 and pH varia-
tions during precipitation. The use of concentration instead
of activity in the solution-chemistry calculation may also
contribute to the observed underestimate. Although it only
accounts for about 20 pct of the bicarbonate consumption
for impurity precipitation, this amount would vary from
ore to ore, as it is highly dependent on the level of impurity.
As mentioned previously, impurities such as Al3
, Fe3
,
and Mg2
are anticipated to form hydroxide precipitates at
a pH around 5, adjusted with a 1:1 ammonium bicarbon-
ate:RE2O3 molar ratio. Impurity removal and RE loss (weight
percent) in the first stage of precipitation at pH 5 was, there-
fore, studied. The results in Figure 5 show that the impurity
removal reached about 70 pct after a 1-hour reaction. Fur-
ther increasing the reaction time did not increase the impurity
removal. Accompanied by removing the impurity was a
decrease in RE loss from 7.5 to 5 pct, suggesting that a longer
reaction time was beneficial to RE recovery. It should be
noted that due to a high concentration of RE elements in
the leachates, the amount of RE precipitates accounts for
40 pct of the total precipitates. These precipitates remain of
economical value and should, therefore, be collected for fur-
ther treatment to recover the enclosed RE elements.
After removing the major impurities, RE ions can be
recovered as relatively pure RE carbonates by adding ammo-
nium bicarbonate to a desired pH level. Figure 6 shows RE
recovery and the purity of the precipitates formed in the
extraction process. It is evident that the precipitation is rapid.
A RE recovery of 90 pct with an RE purity of 86 pct is
accomplished. It is clear that a two-stage process is beneficial
for recovering high-purity RE elements using ammonium
616—VOLUME 34B, OCTOBER 2003
Fig. 6—Effect of stirring time (min) on RE recovery (pct) and RE purity (pct).
bicarbonate as a precipitant. This study also shows that the
required ammonium bicarbonate consumption does not
depend on the source of the ores.
V. CONCLUSIONS
1. Solution-chemistry analysis shows that two steps are
needed to recover RE elements from a leaching liquid
of weathered ores by precipitation using ammonium
bicarbonate: removal of impurities followed by precipi-
tation of RE elements. In the first step, the pH of the
solution should be adjusted to 5 with ammonium bicar-
bonate, and, in the second step, the pH of the solution
should be adjusted to above 8.
2. Ammonium bicarbonate consumption depends on the RE
and impurities concentration. The total ammonium bicar-
bonate consumption calculated for the entire process
agrees quite well with the experimental values.
3. The main consumption of the ammonium bicarbonate is
the decomposition of H2CO3, up to 41 pct total con-
sumption. While the ABCE from this study is about 40 pct,
up to 20 pct of the ammonium bicarbonate is consumed
for removing impurities.
4. The recovery of RE elements obtained from this study
is around 88 to 91 pct, with the purity of the product
being 90 pct. For maximizing the RE recovery, a
sufficient amount of ammonium bicarbonate is required,
or the NH4HCO3:RE2O3 mole ratio has to be kept at 4:1.
ACKNOWLEDGMENTS
We acknowledge the financial support in the form of
research grants from the National Outstanding Youth Sci-
ence Foundation of China (RC) and the Natural Sciences
and Engineering Research Council of Canada (ZX).
REFERENCES
1. R. Chi and D. Wang: Beneficiation and Extraction of Rare Earth
Ore, Science Press, Beijing, 1996, pp. 130-50.
2. H.W. Nebitt: Nature, 1979, vol. 27 (5), pp. 206-11.
3. L.R. Duddy: Chem. Geology, 1988, vol. 30 (4), pp. 363-73.
METALLURGICAL AND MATERIALS TRANSACTIONS B
17_02-618B-B.qxd 9/11/03 11:48 AM Page 617
4. F. Laufer, S. Yariv, and M. Steiberg: Clay Minerals, 1984, vol. 19,
pp. 134-49.
5. R. Chi, G. Zhu, Z. Zhou, and Z. Xu: Metall. Mater. Trans. B, 2000,
vol. 31B, pp. 191-96.
6. R. Chi: in Rare Earths Extraction: Preparation and Applications,
R.G. Bautisa and M.M. Wong, eds., TMS, Reno, NV, 1989, pp. 269-78.
7. R. Chi and Z. Xu: Metall. Mater. Trans. B, 1999, vol. 30B, pp. 189-95.
8. R. Chi, J. Xu, P. He, and Y. Zhu: Trans. Nonferrous Metals Society
of China, 1995, vol. 5 (4), pp. 36-41.
METALLURGICAL AND MATERIALS TRANSACTIONS B
9. Louis Meites: Handbook of Analytical Chemistry, 1st ed.,
McGraw-Hill Book Company, Inc., McGraw-Hill, New York,
NY,1963.
10. David R. Lide: Handbook of Chemistry and Physics, 80th ed., CRC
Press LLC, Boca Raton 2000, pp. 8112-14.
11. CRC Handbook of Chemistry and Physics, 66th ed., CRC Press Inc.,
Boca Raton, FL, 1985.
12. A.E. Martell and R.M. Smith: Critical Stability Constants, Plenum
Press, New York, NY, 1976.
VOLUME 34B, OCTOBER 2003—617