The extraction chromatography behavior of zirconium and hafnium
is studied using tri-n-octylamine (TNOA) and tri-caprylyl-
monomethyl ammonium chloride (Aliquat 336) as the stationary
phase supported on a styrene/divinylbenzene copolymer resin and
hydrochloric acid as the mobile phase. The effects of hydrochloric
acid concentration, extractant loading on the support, support
particle size, column dimensions, and hydrofluoric acid are
investigated. Microgram amounts of Zr and Hf are successfully
separated on a small size (
∅5 × 45 mm) column, on which Zr is
adsorbed and Hf is completely eluted with 8M HCl. The adsorbed
Zr is then eluted with 2M HCl. Aliquat 336 is found to be superior
to TNOA for the mutual separation of Zr and Hf. The stability of the
extraction chromatography resin is also examined.
Introduction
It is well known that the mutual separation of zirconium and
hafnium is very difficult because of the extreme similarity of these
two elements. Both solvent extraction and ion-exchange tech-
niques have been effectively applied to this mixture (1–5).
However, the solvent extraction technique, which requires multi-
stage counter-current mixer settlers using tributyl phosphate
(TBP) as extractant, operating at the industrial scale, is not con-
venient for analytical separations. Ion-exchange technology is a
very effective separation method for similar ions and has been
extensively and successfully applied in analytical separations. The
selectivity of the ion-exchange process depends on the properties
of the ion-exchanger being used and the composition of the
aqueous phase. In the case of two ions having the same charge
and very similar radii, the selectivity resulting from the properties
of the ion-exchanger (such as acidity or basicity and the degree of
crossing-linking) is not sufficient to ensure effective separation.
In such a case, an appropriate complexing agent has to be added
to the aqueous phase; the attained selectivity is then due to either
the differences in the stability constants or the different charges
or structures of the complexes formed. An alternative is extrac-
tion chromatography, which combines the advantages of the high
selectivity of solvent extraction and the high efficiency of ion-
exchange chromatography. It has been extensively applied to sep-
arations in analytical chemistry. Extraction chromatography
offers many opportunities for optimizing a separation by
adjusting stationary phase parameters (type and size of support
particles, type and concentration of the extracting agents used in
diluents, or type and degree of the loading of liquid or solid
extractants) and mobile phase parameters (flow rate, concentra-
tion of partitioning substances, type and concentration of com-
plexing or salting out agents).
Crawley (6) used TBP on kieselguhr for the separation of zir-
conium and hafnium in the HNO
3
–NH
4
NO
3
system, and Ueno
and Hoshi (7) reported the same agent absorbed on a celite sup-
port for this separation in hydrochloric acid. TBP is a very good
extractant for uranium, but the selectivity of TBP for zirconium
and haf-nium is not high. The separation factor of Zr–Hf using
the TBP–celite–HCl system was approximately 4 (7). Zr and Hf
were also separated in the system MIBK–[NH
4
SCN+(NH
4
)
2
SO
4
]
(8). Although Wang et al. (9) studied the extraction chromatog-
raphy behavior of Zr and Hf using sulfoxides as a stationary phase
and an NH
4
SCN–HCl solution as a mobile phase, the obtained
separation factor of 3.6 was not large, and in addition, the intro-
duction of NH
4
SCN salt into the aqueous mobile phase causes
complications.
As the selectivity of an extraction chromatography column is
determined by the nature of the extractant used as the sta-
tionary phase, it is important to select a suitable extractant for
effective separation. Moore (10) first reported the solvent
extraction and separation of Zr and Hf with long-chain
aliphatic amines (methyl-
n-octylamine) in chloride solutions.
The long-chain aliphatic amine compounds are also known as
liquid ion-exchangers and have been widely used in the solvent
extraction separation of various metal ions in nuclear fuel pro-
cessing and fission product recovery. Cerrai and Testa (11,12)
reported detailed investigations on the solvent extraction and
separation of Zr and Hf using tri-
n-octylamine (TNOA) and tri-
171
Abstract
Separation of Hafnium from Zirconium by Extraction
Chromatography with Liquid Anionic Exchangers
X.J. Yang
1,2
, C. Pin
1
, and A.G. Fane
2
1
Département de Géologie, UMR 6524-CNRS, Université Blaise Pascal, 5, rue Kessler, 63038, Clermont-Ferrand, France and
2
Department of
Chemical Engineering, The University of New South Wales, Sydney 2052, Australia
Reproduction (photocopying) of editorial content of this journal is prohibited without publisher’s permission.
Journal of Chromatographic Science, Vol. 37, May 1999
Downloaded from https://academic.oup.com/chromsci/article-abstract/37/5/171/302600
by guest
on 27 February 2018
Journal of Chromatographic Science, Vol. 37, May 1999
172
caprylyl-monomethyl ammonium chloride (Aliquat 336) in a
hydrochloric medium, and later separated 9.7 mg of Zr from
0.3 mg of Hf using a TNOA–8M HCl–5% concentrated HNO
3
system on a cellulose powder column (13). However, high con-
centrations of HCl pose a problem for the cellulose powder sup-
port (14). Das and Chattopadhyay (15) reported briefly on the
solvent extraction and chromatographic separation of nio-
bium, zirconium, and hafnium with Aliquat 336. These earlier
studies suggest that extraction chromatography using TNOA or
Aliquat 336 is potentially promising for the mutual separation
of Zr and Hf; however, systematic studies of the parameters
affecting the separation are lacking. The aim of the present
study is to systematically investigate the separation of Zr–Hf by
extraction chromatography using TNOA and Aliquat 336. The
studied parameters include the loading of the extractant, sup-
port particle size, column dimensions, composition of eluents,
and stability of the resins in order to explore the potential appli-
cation of the method for the separation of Zr–Hf from geolog-
ical materials.
Experimental
Materials and reagents
TNOA, trioctylmethyl ammonium chloride (marketed as
Aliquat 336, molecular weight 404.17), and cyclohexane were
obtained from Aldrich Chemical (USA). The ready-to-use
Levextrel resins contain TNOA or tricaprylylmethyl ammonium
chloride (a mixture of C
8
–C
10
chains with C
8
predominating; also
marketed as Aliquat 336, but referred to as Aliquat 336M here-
after in order to distinguish it from the previous Aliquat 336).
Levextrel resin is a copolymer of divinylbezene (DVB) with
styrene-containing organic extractant, where the organic extrac-
tant is added before the polymerisation reaction starts (17,18).
The support for the preparation of solvent-impregnated resins
(SIR) was Amberchrom CG-300cd, a highly crosslinked
styrene/divinylbenzene copolymer, which was obtained from
Supelco (USA). The spherical beads of Amberchrom resin offer
exceptional mechanical and chemical stability. The physical
properties of Levextrel resins and Amberchrom CG-300cd resin
are listed in Table I.
The weight ratios of TNOA over the support in the Levextrel
resins are 1:9, 1:2, and 1:1, and that of Aliquat 336M is 1:2.
Zirconium and hafnium standard solutions (Aldrich) were 990
and 1002 µg/mL, respectively. All other chemicals
were of analytical grade, and all the chemicals
were used as received. Water with 18M
Ω
•
cm
resistivity that was purified using a Milli-Q
(Millipore, Milford, MA) system was used
throughout.
Apparatus
Polypropylene columns (inner diameter
∅3.5,
3.8, 5, and 9 mm) with a 30 µm polyethylene frit at
the bottom were used for separation work. An
inductively coupled plasma atomic emission spectrometry (ICP-
AES) instrument (Jobin-Yvon 70 II, Longjumeau, France) was
used for the measurement of Zr and Hf concentrations in the
solutions collected fractionally from the columns.
Preparation of resins
The extraction chromatography resin was obtained using two
approaches: the solvent impregnating method (direct adsorption
of the extractant into macroporous polymeric supports), pro-
ducing the SIR resin, and the suspension polymerisation method
(polymerisation of styrene and divinyl copolymers in the presence
of the extractant, developed by Kroebel and Meyer [17,18]), pro-
ducing the Levextrel resin. The SIR resin was prepared by impreg-
nating 10 g of Amberchrom CG-300cd resin in a 50–100-mL
cyclohexane solution containing 5 g TNOA or Aliquat 336, stir-
ring the mixture slightly, and evaporating it at room temperature
(20°C) to dryness. This usually takes approximately 5 hours. The
Levextrel resin was prepared at the Beijing Institute of Chemical
Engineering and Metallurgy, Beijing, China. The Levextrel resin
was sieved, and the 75–120 mesh size (125–200 µm) was used in
this work.
Determination of the distribution and separation coefficients
of Zr(IV) and Hf(IV) in an HCl medium using the
batch-equilibrium method
Resin (0.20 g) was placed into a 10-mL cylindrical vessel, and
5.0-mL portions of aqueous solutions containing 50 µg/mL Zr
and Hf in various HCl concentration were added. The vessel was
closed and shaken vigorously for 30 min. After filtering, the con-
centrations of Zr and Hf in the aqueous phases were measured by
ICP-AES. The distribution coefficient (
D) of Zr and Hf can be cal-
culated using the calculation
D = [( C
0
–
C
e
)
C
e
](
V/ m)
Eq 1
where
C
0
and
C
e
are the initial and equilibrium concentrations of
Zr and Hf in the aqueous solutions, respectively;
V is the volume
(mL) of aqueous solution; and
m is the amount (g) of resin. The
separation factor (
β) is the ratio of distribution coefficients.
D
Zr
β = ——––
Eq 2
D
Hf
The purification factor (PF) corresponding to the initial solu-
tions by the batch-equilibrium method can be calculated using
Table I. Physical Properties of Amberchrom and Levextrel Resins
Surface
Mean
Approximate
Skeletal
Chemical
area
Porosity
pore size
Mesh size
pore volume
density
Resin
structure
(m
2
/g)
(%, v)
(Å)
(µm)
(mL/g)
(g/mL)
CG-300cd
polyaromatic
700
55–75
300
80–160
1.66
1.08
Levextrel
polyaromatic
80–345
65
180–800
125–200
NA*
NA*
* NA, data not available.
Downloaded from https://academic.oup.com/chromsci/article-abstract/37/5/171/302600
by guest
on 27 February 2018
Journal of Chromatographic Science, Vol. 37, May 1999
173
[(
C
0
–
C
e
)Zr/(
C
0
–
C
e
)Hf]
PF
Zr
= ———————————
Eq 3
(
C
0
, Zr/
C
0
, Hf)
(
C
e
, Hf/
C
e
, Zr)
PF
Hf
= ———————
Eq 4
(
C
0
, Hf/
C
0
, Zr)
Determination of exchange capacity
A small-size column (
∅3.5 × 50 mm) and 0.1 g of resin were
used for the determination of the exchange capacity of Zr by the
breakthrough curve (i.e., the Zr concentration change of the
effluent with effluent volume) that was obtained by loading the
feed solution until the concentration of Zr in the effluent reached
the same concentration of zirconium in the feed solution. Then,
the feeding was interrupted, and the column bed was washed with
an eluent that displaces the interstitial volume without stripping
the Zr retained in the stationary phase. When no Zr was found in
this eluate, another eluent (2M HCl) That was able to strip the Zr
retained on the column was passed through the column until no
Zr was detected in this stripping eluent. The equilibrium
exchange capacity (EEC) for Zr was obtained by measuring the
concentration of Zr in the collected volume of this stripping
eluate:
CV
EEC = ———
Eq 5
m
where
C is the concentration of Zr in the eluate and V is the
volume of eluate.
Column separation
The resin, suspended in water, was transferred into a chro-
matographic column, and the resin bed was pressed slightly
with a glass rod to eliminate air bubbles. A 30-µm polyethy-
lene frit was placed on top of the column bed to prevent the
disturbance of the resin during the elution phase. The
column was preconditioned with 8–10 mL of the same con-
centration of hydrochloric acid as the first eluant. A solution
containing a Zr–Hf mixture to be separated was then loaded
into the column. The elution separation was carried out with
eluants composed of various concentrations of hydrochloric
acid as mobile phases using gravity flow. The concentrations
of Zr and Hf in the fractions of the effluents were determined
using ICP-AES, and the elution curves were drawn by plot-
ting the concentration against the effluent volumes. The
purification factor (PF) of the column method is calculated
using
(
C
Hf
/
C
Zr
)(eluate 1)
PF
Hf
= ———————
Eq 6
(
C
Hf
/
C
Zr
)feed
(
C
Zr
/
C
Hf
)(eluate 2)
PF
Zr
= ———————
Eq 7
(
C
Zr
/
C
Hf
)feed
where eluate 1 is the Hf eluate obtained by passing 8M HCl
through the column, and eluate 2 is the Zr eluate obtained by
passing 2M HCl through the column.
Results and Discussion
Effect of HCl concentration on distribution coefficient and
separation factor
The distribution coefficient as a function of HCl concentration
ranging from 6 to 10M obtained by the batch method is shown in
Figure 1. The plots of log
D versus log [HCl] give linear relation-
ships. It has been assumed that the extraction mechanism of
Zr(IV) and Hf(IV) by tertiary amine (TNOA) and quaternary
ammonium salt compound may be represented as
TNOA:
R
3
N
resin
+ HCl
aq
=
R
3
NH
+
× Cl
–
resin
Eq 8
Figure 1. Dependence of log D on log [HCl]. 1:2 Levextrel TNOA resin (A),
1:2 SIR Aliquat 336 resin (B), 1:2 Levextrel Aliquat 336M resin (C). Aqueous
phase, 5 mL of 50 µg/mL Zr and Hf in various HCl concentrations; resin phase,
0.20 g.
A
B
C
Downloaded from https://academic.oup.com/chromsci/article-abstract/37/5/171/302600
by guest
on 27 February 2018
Journal of Chromatographic Science, Vol. 37, May 1999
174
2R
3
NH
+
× Cl
–
resin
+ ZrCl
6
2–
aq
=
(R
3
NH
+
)
2
× ZrCl
6
2–
resin
+ 2Cl
–
aq
Eq 9
Aliquat 336:
2R
3
R'NCl
–
resin
+ ZrCl
6
2–
aq
=
(R
3
R'N)
+
2
× ZrCl
6
2–
resin
+ 2Cl
–
aq
Eq 10
where “resin” refers to the resin phase, “aq” is the aqueous phase,
R
3
N is TNOA, R
3
R'N
+
Cl
–
is Aliquat 336, R is the octyl group, and
R’ is the methyl group. The distribution coefficients at a given
HCl concentration are in the order TNOA > Aliquat 336M >
Aliquat 336, showing that the extraction power is in this order. It
has been demonstrated that the extraction power of Aliquat 336 is
generally lower than that of TNOA and thus of tri-
i-octylamine
(TIOA) (16). The separation factor as a function of HCl concen-
tration is given in Figure 2. The maximum separation is obtain-
able at an HCl concentration of 8M, and the separation factor is in
the order Aliquat 336M (
β = 27) > Aliquat 336 (β = 23) > TNOA
(
β = 12), which is much larger than the literature values obtained
by TBP (
β = 4) (7) and sulfoxides (β = 3.6) (9). The Zr and Hf con-
centrations in the 8M HCl aqueous solution after the batch equi-
librium using Levextrel Aliquat 336M were 7.51 and 42.3 µg/mL,
respectively (initial concentration, 50 µg/mL). Therefore, the
purification factors obtained by the batch method in 8M HCl were
5.52 for Zr and 5.63 for Hf. It is expected that the separation will
be greatly improved by the column method, because numerous
equilibria occur in the column during the elution. The column
separation will be discussed later.
Breakthrough curve and exchange capacity
The breakthrough curves for Zr with Aliquat 336M and Aliquat
336 resin are shown in Figure 3. Using Equation 5, the exchange
capacities of Levextrel Aliquat 336M resin and Aliquat 336 SIR
Figure 2. Separation coefficients as a function of HCl concentration: l, TNOA;
n
n
, Aliquat 336; u, Aliquat 336M.
Figure 3. Breakthrough curves of zirconium on the Levextrel Aliquat 336M
resin and SIR Aliquat 336 resin columns (0.1 g of resin). l, Levextrel Aliquat
336M resin;
l
l
, SIR Aliquat 336 resin.
Figure 4. Separation profiles of Hf and Zr on a
∅5- × 45-mm column (loading,
50 µg Zr and 50 µg Hf): 1:1 Levextrel Aliquat 336M (flow rate, 0.24 mL/min)
(A); 1:2 SIR Aliquat 336 (flow rate, 0.24 mL/min) (B); 1:1 Levextrel TNOA
(flowrate, 0.2 mL/min) (C).
A
B
C
Downloaded from https://academic.oup.com/chromsci/article-abstract/37/5/171/302600
by guest
on 27 February 2018
Journal of Chromatographic Science, Vol. 37, May 1999
175
resin are 24.3 and 30.8 mg Zr/g of resin, respectively.
Column separation
Figure 4 shows the column separation of Hf from Zr using
Aliquat 336, Aliquat 336M, and TNOA as stationary phases and 8M
HCl as the mobile phase to elute Hf. The separations of Hf from
Zr with Aliquat 336 and Aliquat 336M are excellent using 10 mL
of 8M HCl. However, the elution curve with Aliquat 336M seemed
to show more tailing than with Aliquat 336, and the elution peak
of Hf was very broad (only approximately 50% of the amount of Hf
loaded was eluted with 15 mL of 8M HCl from the TNOA resin
column). This can be explained by the higher distribution coeffi-
cient of the TNOA resin than that of the Aliquat 336 resin (Figure
1). Hence, a larger volume of 8M HCl will be needed to separate
Hf from Zr on the TNOA column. The recovery and purification
factors in different initial Zr/Hf ratios of the feed solutions
obtained by the Aliquat 336M columns are summarized in Table
II (10 mL of 8M HCl eluate was collected for recovering Hf, and 10
mL of 2M HCl eluate was collected for Zr). The purification fac-
tors are much larger than those obtained by the batch-equilib-
rium method.
It should be noted that the 5–8 mL of effluent of the condi-
tioning phase (10 mL of 8M HCl) became turbid, and white spots
were observed in the columns during the course of elu-
tion/separation. The turbidity was assumed to be caused by the
drainage of some of the extractant from the column. Many more
spots appeared in the SIR resin columns than in the Levextrel
resin columns. As a result, the flow rate decreased gradually. This
could be attributed to the instability of the resin (i.e., the losses of
the organic substance from the support). The stability of the SIR
and Levextrel resins will be discussed in more detail later.
Effect of extractant loading on the support
The effect of extractant loading on the column separation was
examined using 1:9, 1:2, and 1:1 Levextrel TNOA resin, and the
results are shown in Figures 4 and 5. The separation of Hf from Zr
can be effectively performed with 1:2 TNOA (Figure 5B), whereas
the extractant loading of 1:9 is not enough to retain Zr on the
column (Figure 5A). Using a 1:1 ratio, only approximately 50% Hf
loading was eluted with 15 mL of 8M HCl (Figure 4C), reflecting
the high affinity of the extractant for Hf and Zr. These results
show that 1:2 extractant loading has the optimum column effi-
ciency. Grosse-Ruyken and Bosholm (19) reported that the min-
imum heights of the equivalent theoretical plate (HETP) values at
different di-(2-ethylhexyl) phosphoric acid (HDEHP) loadings on
the 0.09-mm diameter silica gel were obtained at the HDEHP/
Figure 5. Effect of the extractant loading on the separation of Hf and Zr: 200
µg Zr and Hf loaded on the
∅9- × 136-mm column containing 2.9 g of 1:9
Levextrel TNOA resin (flow rate, 0.47 mL/min) (A); 50 µg Zr and Hf loaded on
the
∅5- × 45-mm column containing 0.40 g of 1:2 Levextrel TNOA resin (flow
rate, 0.38 mL/min) (B); 50 µg Zr and Hf loaded on the
∅5- × 45-mm column
containing 0.59 g of 1:1 Levextrel TNOA resin (flow rate, 0.2; see Figure 4) (C).
Table II. Purification Factor Obtained by the Column Method*
Found (µg)
Recovery (%)
Mixture (µg)
8M HCl (10 mL)
2M HCl (10 mL)
8M HCl
2M HCl
Purification factor
Hf
Zr
Hf
Zr
Hf
Zr
Hf
Zr
Hf
Zr
Column 1
50
50
48.4
0.41
0.56
49.9
96.7
99.9
118
89.2
100
5
102
0.13
0.58
4.77
102
95.4
40.4
164
50
2.5
51.2
0.07
0.38
2.34
102
94.0
36.6
123
Column 2
10
100
9.50
1.18
0.25
94.5
95.0
94.5
80.5
38.3
10
50
9.74
0.49
0.12
48.7
97.4
97.0
99.4
81.2
1
†
40
†
0.93
0.27
0.05
40.8
93.0
102
138
20.4
* Column dimension, 80 mm
× 3.8-mm i.d. ; resin type, Levextrel Aliquat 336M; amount, 0.4 g; conditioning, 8 mL 8M HCl. The purification factors by the batch-equilibrium method were
5.52 for Zr and 5.63 for Hf.
†
The natural Zr/Hf ratio in geological materials is about 40:1.
A
B
Downloaded from https://academic.oup.com/chromsci/article-abstract/37/5/171/302600
by guest
on 27 February 2018
Journal of Chromatographic Science, Vol. 37, May 1999
176
support weight ratio of 1:2. Ueno and Hoshi (7) investigated the
effect of the TBP-celite weight ratios of 1:1, 1:2, and 1:4 on the
separation of Hf from Zr and also found that the 1:2 ratio was the
optimum extractant loading for Hf–Zr separation.
Effect of column dimension and support particle size
Bed length is one of the most important parameters in extrac-
tion chromatography; elution volumes, separation time, and
the number of theoretical plates increase linearly with bed
length, and column resolution increases with its square root.
Extraction chromatographic experiments are usually run under
gravity flow. When very similar ions are involved, very long
lengths of column are usually necessary to obtain a satisfactory
resolution. An increase in bed length leads to a proportional
increase in pressure needed to force the mobile phase through
the column. Most columns used in chromatography for analyt-
ical purposes have bed lengths ranging between 50 and 150 mm,
mainly determined by the resolution needed and the time that
can be tolerated for analysis. The support particle size is of
importance to the column performance in both gas and liquid
chromatography (LC), because it significantly affects the sur-
face area and the HETP. However, the particle size of the support
also has a significant influence on the flow resistance. Figure 6
Figure 6. Effect of column bed length and support particle size on flow rate.
The column diameter was 6 mm. l, 125–200 µm 1:2 Levextrel TNOA resin; n,
80–160 µm 1:2 SIR TNOA resin.
Figure 7. Effect of column bed diameter on flow rate. The column bed height
was 45 mm with SIR Aliquat 336 resin.
Figure 8. The elution curves of Hf and Zr with a large-dimension column (
∅9
× 136 mm): 3.0 g of 1:2 Levextrel TNOA resin (flow rate, 0.59 mL/min) with an
elution of 50 mL 8.5M HCl and then 40 mL 4M HNO
3
(A); the same column
as in A with an elution of 40 mL 8M HCl and then 30 mL 2M HCl (B); 4.8 g of
1:1 Levextrel TNOA resin (flow rate, 0.75 mL/min) (C); and 3.2 g of SIR TNOA
resin (bed lenth, 120 mm; flow rate, 0.22 mL/min) (D). Zr and Hf loadings were
200 µg. Particle sizes were 125–200 µm (A–C) and 80–160 µm (D). The inset
shows the Hf elution peak between 1- and 20-mL fractions, drawn to a larger
scale.
A
B
C
D
Downloaded from https://academic.oup.com/chromsci/article-abstract/37/5/171/302600
by guest
on 27 February 2018
Journal of Chromatographic Science, Vol. 37, May 1999
177
shows the effect of column bed length and support particle size
on the flow rate. The particle size of the Levextrel resin ranged
from 125 to 200-µm diameter, and that of the support for the
SIR TNOA resin ranged from 80 to 160-µm diameter; the
smaller resin (SIR TNOA) gave a lower flow rate as expected.
Figure 7 shows the effect of bed diameter on the flow rate. The
flow rate decreased drastically when the bed diameter decreased
from 9 to 3.8 mm.
Figure 8 shows the separation chromatogram for the larger-
dimension column (
∅9 × 150 mm) using different diameter
resins (125–200-µm diameter for 1:1 and 1:2 Levextrel TNOA
resins and 80–160 µm diameter for 1:2 SIR TNOA resin). For the
diameter of 125–200 µm resins (Figure 8, flow rate = 0.6–0.75
mL/min), in the first effluent fraction of 15 mL of 8M HCl, a por-
tion of 1–3% of the Hf amount loaded on the column was
detected, whereas in the following fraction of 10 mL, no Hf was
detected, indicating that Hf leakage occurred at the beginning of
elution. This elution of Hf amounting to 1–3% of the loaded
amount was presumably caused by channeling in the column due
to the larger size of resin and column diameter and the low dis-
tribution coefficient of Hf. The bulk of the Hf was eluted together
with the Zr (i.e., no separation occurred) when the mobile phase
was changed to 2M HCl. In the case of smaller size resin (80–160
µm diameter) (Figure 8D, flow rate = 0.22 mL/min, much smaller
than the previous cases), no obvious peak for Hf in the first frac-
tion of 15 mL was observed, indicating that there was no leakage
of Hf. Following this, the Hf was eluted separately from Zr with
8M HCl, resulting in excellent separation. It can be concluded
that for efficient operation, smaller resins, smaller diameter
columns, and slower mobile phase flow rates are preferred. In
general, the channeling phenomena led to the loss of resolution
and may occur when large-diameter columns and large support
particle sizes are used, because in such a case it is difficult to pack
the columns uniformly and homogeneously. Therefore, it is good
practice to reduce column diameters as much as allowed by
experimental limitations. A bed diameter of approximately 5–50
particle diameter (i.e., column internal diameters from 2.5 to 7.0
mm) for 50–150-µm particles is a compromise used frequently
(16). Recently, a more detailed investigation on the repro-
ducibility of column performance in LC and the role of the
packing density was reported by Stanley et al. (20).
Effect of hydrofluoric acid on separation
Hydrofluoric (HF) acid is commonly used as a complexing agent
to stabilize Zr and Hf solutions. In addition, when dealing with
geological materials and minerals, HF acid is the most convenient
way to create the decomposition. The main purpose of this study
was to explore the possibility of the present extraction chromato-
graphic system to separate of Zr and Hf from geological materials
and minerals. Therefore, the effect of HF acid on the elution
behavior of Hf and Zr was investigated and is shown in Figure 9. It
can be seen that no separation was achieved when a small amount
of HF acid was present in the solution; the elution peaks of Zr and
Hf completely overlapped (the amounts eluted in the 8M HCl were
99.8% for Hf and 97.3% for Zr). This is because Hf and Zr form
much more stable complexes with the fluoride ion than the chlo-
ride ion, and the fluoride complexes are not extractable by the
amine compounds. This system was applied to “real world” sam-
ples (e.g., geological materials) and found that the successful sep-
aration could be obtained in the case of complete removal of fluo-
ride ions from the solution. On the other hand, the separation
became very complex because of the presence of the major ele-
ments Al, Fe, and Ti, which are difficult to separate from Zr and Hf.
The details will be published elsewhere (21).
Stability and lifetime of resin
As mentioned above, instability phenomena were observed with
the resin used. The stability characterisation of Levextrel and SIR
resins were further examined by measuring the capacity varia-
tions of resin during operation. Figure 10 shows the capacity of
Levextrel Aliquat 336M resin and SIR Aliquat 336 resin as a func-
tion of elution cycle number. In the capacity measurement, no
preconditioning was carried out for the first run. It can be seen
that the capacity for Zr drops 29% for the Levextrel Aliquat 336M
resin and 50% for the SIR Aliquat 336 resin between runs 1 and
Figure 9. Effect of hydrofluoric acid on the elution behavior of Hf and Zr.
Conditions: column,
∅5 × 45 mm; 1:2 Levextrel TNOA resin; loading, 1 mL
of 8M HCl–0.05M HF containing 50 µg Zr and Hf; flow rate, 0.36 mL/min.
Figure 10. Exchange capacity for Zr as a function of operation run number.
The column was
∅3.5 × 30 mm filled with 0.1 g resin. Each run included 8
mL of 10M HCl for conditioning (except the first run), 40 mL of feed solution
in 10M HCl of 100 µg/mL Zr, 10 mL of 2M HCl for eluting Zr from the column,
and 10 mL of water for washing the column. l, SIR Aliquat 336 resin (the
flowrate decreased from 0.43 mL/min at the beginning to 0.12 mL/min at the
end);
l
l
, Levextrel Aliquat 336M resin (the flowrate decreased from 0.18
mL/min at the beginning to 0.05 mL/min at the end).
Downloaded from https://academic.oup.com/chromsci/article-abstract/37/5/171/302600
by guest
on 27 February 2018
Journal of Chromatographic Science, Vol. 37, May 1999
178
2, and thereafter there was little change in capacity. One of rea-
sons for this difference between runs 1 and 2 could be the pre-
conditioning prior to run 2. As has been mentioned previously
that at the preconditioning stage, some emulsion formation in
the effluent was observed, indicating loss of extractant. The flow
rate of the Levextrel Aliquat 336M resin column decreased from
0.43 to 0.12 mL/min, and that of the SIR Aliquat 336 resin
column decreased from 0.18 to 0.05 mL/min. The decreases in
flow rate and capacity reflect the loss of extractant from the sup-
port. The stability of a chromatographic column includes phys-
ical stability and chemical stability. Losses of extractant from
the column may derive either from dissolution into the eluants
or drainage by the eluant of undissolved portions of extractant
scarcely retained on the supporting material. They may result in
the undesirable presence of extractant in the eluate and also in
the variation of the characteristics of the column. When a
second, more retained element is involved (in the case of this
work, Zr) and loaded at a high ratio, leakages of extractant could
also include losses of the extracted compound and give rise to
contamination of the eluate and decreases in the purification
factor of the first, less retained element or incomplete recovery
of the second element in the subsequent steps. In general, a
maximum loading of no more than 20% of the column capacity
is recommended in extraction chromatography. In the experi-
ments shown in Table II, the resin amount in the column was
0.4 g, hence the Zr loading was between 0.003–1%. In addition,
10 mL of conditioning phase (8M HCl) aged the column.
Therefore, the instability of the resin has no significant effect on
the purification factors when the metals loading is very low (no
more than 1% of the maximum capacity) and aging action is
taken.
From Figure 10, it can also be seen that the Levextrel resin is
more stable than the SIR resin. This may be attributed to the dif-
ferent ways the extractant was introduced and held on the sup-
port. For the Levextrel resin, the extractant was added before the
polymerisation started, and as a result, a large amount of the
extractant was adsorbed and fixed in the network or matrix of the
porous copolymer. Figure 6A may be further evidence of this,
because no adsorption of Zr was observed on the 1:9 Levextrel
TNOA resin. For the SIR resin, which was prepared “in house”, a
large amount of the extractant was adsorbed onto the surface of
the resin by physical adsorption, and some of the extractant
entered the cavities of the porous support by capillary forces.
Consequently, the Levextrel resin should be more stable than the
in-house SIR resin. From Levextrel Aliquat 336M resin (Figure
4A) and SIR Aliquat 336 resin (Figure 4B) and Figure 8, it can be
seen that the tailing of Hf and Zr on the Levextrel resin is more
serious than on the SIR resin. This verifies the above analysis.
Conclusion
The experimental results showed that TNOA and Aliquat 336
are good extractants for the zirconium–hafnium separation in an
extraction chromatography system. Aliquat 336 is more selective
than TNOA. The separation factors obtained ranged between 36
and 164 for 1–100 µg of Hf and 2.5–100 µg of Zr. Channeling in
the column is more serious with larger column dimensions and
support particle sizes. An optimum extractant loading ratio of 1:2
(the weight ratio of extractant over the support) was identified for
separation purposes. No separation can be achieved when the feed
aqueous solution contains hydrofluoric acid. The stability of the
Levextrel resin is better than the solvent impregnated resin pre-
pared in house using the conventional method.
Acknowledgments
This work was carried out at the CNRS, Université Blaise Pascal
under the support of the CNRS - K.C. Wong Fellowship (France)
to X.J. Yang. The authors would like to thank Dennis Jenke,
whose critical comments greatly improved the manuscript.
References
1. J.L. Hague and L.A. Machlan. Separation of hafnium from zirconium
by anion exchange. J. Res. Natl. Bur. Std.
A65: 75–77 (1961).
2. F.W.E. Strelow and C.J.C. Bothma. Anion exchange and a selectivity
scale for elements in sulfuric acid media with a strongly basic resin.
Anal. Chem.
39(6): 595–99 (1967).
3. J. Korkisch. Modern Methods for the Separation of Rare Metal Ions.
Pergamon Press, Oxford, U.K., 1969, p. 415.
4. B. Nandi, N.R. Das, and S.N. Bhattacharyya. Solvent extraction of
zirconium and hafnium. Solvent Extr. Ion Exch.
1: 141 (1983).
5. J. Hara. Solvent Extr. Ion Exch. 10: 239(1988).
6. R.H.A. Crawley. Separation of zirconium and hafnium by reversed-
phase chromatography. Nature
197: 377–78 (1963).
7. K. Ueno and M. Hoshi. The separation of zirconium and hafnium by
tributyl phosphate–celite reversed-phase partition chromatography.
Bull. Chem. Soc. Jpn.
39: 2183–87 (1966).
8. J.S. Fritz and R.T. Frazee. Reversed-phase chromatographic separa-
tion of zirconium and hafnium. Anal. Chem.
36: 1324–26 (1964).
9. Q.Wang, K. Taunoda, and H. Akaiwa. Extraction crhomatographic
behaviours of zirconium and hafnium using sulfoxides as stationary
phases. Anal. Sci.
11: 909–913 (1995).
10. F.L. Moore. Long-chain amines: versatile acid extractants. Anal.
Chem.
29: 1661 (1957).
11. E. Cerrai and C. Testa. Extraction and separation of zirconium and
hafnium by means of liquid anionic exchangers in a hydrochloric
acid medium (II): Extraction with tri-n-octylamine. Energia Nucleare
6(12): 768–80 (1959).
12. F. Bonfanti, E. Cerrai, and G. Ghersini. Extraction and separation of
zirconium and hafnium by means of liquid anionic exchangers in a
hydrochloric acid medium (III): Extraction with tricaprylyl-
monomethyl ammonium chloride. Energia Nucleare
14(10): 578–85
(1967).
13. E. Cerrai and C. Testa. The use of tri- n-octylamine–cellulose in chem-
ical separations. J. Chromatogr.
6: 443–51 (1961).
14. G.S. Katykhin. Inert supports in column extraction chromatography.
In Extraction Chromatography, Journal of Chromatography Library,
Vol. 2, T. Braun and G. Ghersini, Eds. Elsevier, Amsterdam, the
Netherlands, 1975, Chap. 5.
15. N.R. Das and P. Chattopadhyay. Solvent and reversed-phase extrac-
tion chromatographic separation of niobium, zirconium and
hafnium with Aliquat 336. Bull. Chem. Soc. Jpn.
61: 4423–26
(1988).
16. Extraction Chromatography, Journal of Chromatography Library,
Vol. 2, T. Braun and G. Ghersini, Eds. Elsevier, Amsterdam, the
Downloaded from https://academic.oup.com/chromsci/article-abstract/37/5/171/302600
by guest
on 27 February 2018
Journal of Chromatographic Science, Vol. 37, May 1999
179
Netherlands, 1975.
17. R. Kroebel and A. Meyer. Ger. Offen., 2162951, 1973.
18. H.W. Kauczor and A. Meyer. The structure and properties of
Levextrel resin. Hydrometallurgy
3(3): 63–73(1978).
19. Grosse-Ruyken and J. Bosholm. Partition chromatography of rare
earths by bis-(2-ethylhexyl) phosphoric acid. I. Effect of the working
conditions on the separative efficiency of the column. J. Prakt. Chem.
25(1–2): 79–87 (1964).
20. B.J. Stanley, C.R. Foster, and G.Guiochon. On the reproducibility of
column performance in liquid chromatography and the role of the
packing density. J. Chromatogr. A
761: 41–51 (1997).
21. X.J. Yang and C. Pin. Separation of hafnium and zirconium from Ti-
and Fe-rich geological materials by extraction chromatography.
Anal. Chem.
71(9) (1999).
Manuscript accepted April 5, 1999.
Downloaded from https://academic.oup.com/chromsci/article-abstract/37/5/171/302600
by guest
on 27 February 2018
Dostları ilə paylaş: |