Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
PC .1l/CAN
FIELD OF THE INVENTION
The present invention relates to the purification of
aqueous cobaltous solutions, and in par-ticular to the removal
of nickel from such solutions down to very low levels.
BACKGROUND OF THE INVENTION
In many hydrometallurgical processes for recovering
nickel and cobalt from their sulfidic or oxidic ores, interme-
diate agueous process streams are generated which contain
; dissolved nickel and cobalt. In order to produce high purity
cobalt from such solutions it is necessary ~o remove the nickel
therefrom down to very low levels. For example, coba~ pro-
ducts are desired, the purity of which is such that the cobalt
to nickel ratio is o the order of 1000 to 1. If such cobalt
is to be produced rom the purified solution, by hydrogen
reduction for example, the solution itself must exhibit a
purity of Co:Ni >1000:1 if the desired product purity is to
be achieved. It would be particularly advantageous to be able
to use an ion exchange procedure for purifying the cobalt-
containing solutions.
Many ion exchange resins availab}e commexcially are said
to exhibit some selectivity between nickel and cobaltr i.e.,
nickel will load more readily than cobalt lnto such resins.
The selectivity indicated by the published literature for
such resins is of a smaller degree than that which is achievable
between more dissimilar metal pairs, such as nickel and copper
or nickel and zinc. Nevertheless, some of the resins which
have been developed in recent years would seem capable of being
applied to the task of cobalt solution purification. Selection
of any resin is in practice based primarily on the Ni/Co
selectivity factor quoted for such resin, or calculable from
the quoted data. The selectivity factor, SNi~cO, is defined
as the ratio of distribution coefficients DNi and DCo, each
i6
of which in turn is the ratio of the concentration of metal
in the resin and in an aqueous ~solution in equilibrium there-
with. Thus the selectivity factor can be expressed as:
Ni/Co ~i
DCo
Ni]r x [C]a
wherein ~ ]r refers to the metal concentration in the resin,
and [ ]a refers to the metal concentration in the aqueous phase
.
; in equilibrium therewith.
In the case o~ some of recently developed resins, the
Ni/Co selectivity can be estimated from selectivity published
for each of these two metals with respect to a common reference,
typically calcium. When the claimed selectivity is examined
~- .
in this way, the most promising~of commercial resins would
seem to be a chelating resin having aminocarboxylic acid
functional groups r available from Rohm and Haas under the name:
: - .
Amberlite* IRC718. The selectivity factor for this resin is
indicated by the manufacturer's trade literature as being 54.4
(evaIuated from selectivity factors of 3100 and 57 quoted for
nickel and cobalt, respectively, with respect to calcium).
Indeed the use of such a resin for selective recovery of nickel
and cobalt from aqueous solutions is the subject of U. S.
patent 4,123,260 issued on 10/31/78 to Sefton and Kofluk.
It has now been found, however, that resins of the amino-
carboxylate type are not capable of removing nickel effectively
from concentrated cobaltous solution. Specifically, it was
determined that the selectivity exhiblted by a resin such as
Amberlite IRC718 is lower (by as much as an order of magnitudel
~ . , .
* Trademark
2-
in the presence o concentra-ted aqueous cobaltous solutions,
than the value :indicated by the published li-terature, which
value is probably related to resin performance in the presence
of dilute solutions such as effluent streams from which small
amounts of Ni or Co are to be removed.
OBJECT OF THE INVENTION
The present invention aims at providing a method whereby
substantially all of the nickel can be efficiently removed
from aqueous cobaltous sulfate solutions containing about 50
grams per liter (y/l) or more of cobalt in solution.
SU~MAR~ OF THE INVENTION
According to the invention a process is provided for
removing nickel from an aqueous cobaltous sulfate solution which
contains dissolved nickel and at least about 50 g/l of dissolved
cobalt~ comprising adjus~ing the pH of -the solution if
necessary to a value between about 2 and 6, and contacting
the solution at a temperature of about 20-60C with a chelating
ion-exchange resin having bis (2 - picolyl)amine functional
groups to sélectively load nickel onto the resin and produce
a purified solution from which high purity cobalt can be
recovered.
A commercial resin ~hich can be used for this purpose
is available from Dow Chemicals under the name: XF-4195. The
effectiveness of this resin in the process of the invention
is particularly surprising in view of the poor selectivity
indicated by the published data therefor. Thus data published
Eor nicke1 and cobalt absorption by this resin suggest a
selectivity factor of only 6.3, i.e., it would appear to be
far less selective than the Amberlite IRC718 resin~ Yet it
was found that, for reasons which are by no means clear, the
measured selectivity in the presence of concentrated aqueous
cobaltous solutions is not predictable from the published data
-3-
relating to behavior with dilute solutionst Indeed the meas-
ured selectivity of Dow's XF-4195 resin was found to be four
to five times greater than the value calculated for dilute
solutions, whereas the selectivity of the Amberlite IRC718 resin
was found to be about ten times smaller than its respective
dilute solution value.
The purification process can be carried out in any con-
ventional manner including use of a fixed bed or a moving bed
of resin, and the solution can be processed in either of batch
or continuous modes of opera~ion. While the process may be
carried out at room temperature F we prefer for kinetic reasons
to maintain the resin bed at about 50-60C. The solution pH
is preferably adjusted to between 4.5 - 5.5 Eor best results.
Under such conditions we have found it possible to achieve
purification so as to produce solutions wherein the cobalt to
nickel ratio exceeds 1000.
The invention will now be specifically described by way
of examples.
EXAMPLE 1
Four series of tests were carried out usin~ Dowls XF~4195
resin to purify solutions having cobalt contents ranging from
49 to 127 g/lO In each series of tests the feed solution
was a synthetic sulfate solution having a fixed cobalt content,
containing about 50 g/l of sodium sulfate and varying amounts
of nickel between 0.19 and 1.5 g/l. Each tes~ was a batch
test wherein the solution pH was adjusted to 2.0 at 50C,
mixed with resin in a phase ratio (solution/resin) of 10 and
maintained in agitated contact with the resin at 50C for a
24 hour duration. At the end of that time the distribution of
nickel and cobalt between the phases was determined by assaying
the phases. Table 1 below shows the results determined in
the first series of tests, wherein the feed solution contained
49.3 g/l of cobal~.
T~BI.E 1
Analysis of Equl. ibrated Phases (g/l)
P queous Resin Ni/Co
Ni Co Nl Co
0.095 ~6.2 1.79 31.3 27O8
0.170 ~6.2 3.19 30.6 28.4
0.27846044 4.37 28.6 25.5
0O37046.44 5.29 28.6 23O2
0O54546O66 7.75 26.4 25O1
0~67546.67 9.69 26.3 2504
The three other series of tests were carried out with
initial solutions containing 74~ 97.8 and 127 g/l of cobalt
respectivelyO Table 2 below provides by way of su~nary r the
average value of SNi/Co determined from the six tests of each
series O
;~ TABLE 2
~eed (g/l) Average S~i/CO
49 D 3 28O9
7~ 2~.7
~: 97.8 2503
. . 127 33.2
,:~
The above resul.ts clearly show that good selectivity
can be obtained throughout the cobalt concentration range
tested. The average selectivity factors determined are to be
contrasted with the value o~ 6.3 suggested by available data
for the resin.
- EXAMPLE 2
. By way of comparison two further series of b:atch tests
were carried out in which it was attempted to purify con-
centrated cobaltous solutions with different resins. In each
series five tests were conducted with initial solutions
'
~119~1~
wherein the cobal-t content was 96.6 g/l in all cases, and the
nickel content was, respecti~ely, 0.25, 0.5, 0.75, 1.0 and
1.25 g/l. Each test involved a contacting time of 24 hours
as in the case of Example 1, the pH in this case was 2.5 and
the temperature 60Co The first series of these comparative
tests, the results of which are shown in Table 3 below, were
carried out using the Amberlite IRC-718 resin. The second
series of tests were conducted with another resin of the same
type commercially available from Bayer A.G. under the name
Lewatit* TP207, and the results obtained are given in Table 4.
TA~LE 3
. Analysis of __
Equilibrated Phases (g/l) .
_
: Aqueous Resin Ni/Co
_ .
Ni Co Ni Co
0.27 91~2 0.83 54.2 5.22
0.43 gl.4 1036 52.2 5.60
0.63 91.4 1.88 51.9 ~.28
0.79 . 9~.0 2.21 45.9 5.61
: 1.16 92.6 3.~5 40.0 6.96
..
6: . TABLE 4
Analysis of
~ Equilibrated Phases (g/l)
;~: Aqueous Resin Ni/Co
_ _
Ni CoNi Co
__ _
0.20 89.8 0.98 67.9 6.
0.36 89.7 1.71 69.4 6.1~
0.52 90.1 2.45 65.2 6.49
0.68 90.6 3.39 60.3 7.51
0.98 89.7 4.58 68.5 6.21
It will be seen from the results of Table 3 and 4 that
neither of the resins: Amberlite IRC-718 and Lewatit TP207
____ ____________
* Trademark
exhibits a selec-tivity in the presence of concentrated cobalt-
ous solutions comparable to that which the bis (2 - picolyl)amine
resin was found to exhibit.
EX~MPLE 3
A sulfate solution containin~ 50 g/l Co and 14.3 g/l
Ni was treated with the XF-4195 resin in a column wherein 1.3 1
of resin were contained as a bed 4.2 cm in diamete.r and
91 cm deep. The solution, which had a pH of 5.0 was made to
flow upwards through the column at the rate of 1 cubic meter
per hour per square meter of column cross section ~m3~m2/h)
and the bed was maintained at 50C. The e.~fluent exiting ~rom
the column was analyzed for cobalt and nickel and Table 5
below shows the assays of fractions of effluent collected as
the test progressedO The fractions are expressed in multiples
of the resin bed volume (BV).
.
TABLE 5
.' Effluent fractioI Effluent Assay g/l Co/Ni ratio
~` collected (BV) Co Ni~in fraction
0 to G.5 0.001 0.001 ~_ .
:~ 0.5 to 0.7 o.ool OoOOl __
0.7 to 0.8 OoO01 0.001 __
.~ 0.8 to 0.9 5O~6 OoO01 5860
0.9 to 1.0 47.3 0.001 47300
.1.0 to 1.1 58.0 0.002 29000
1.1 to 1.2 60.5 0.005 12100
1.2 to 1.3 62.0 0.013 4769
1.3 to 1.4 60~0 0.052 1154
1.4 to 1.5 60.0 0.133 451
1.5 to 1.75 58.0 0.600 97
1.-i5 to 2.0 55.5 2.75 21
2.0 to 2.25 52.5 6.25 8.4
2.25 to 2.5 51.0 9.50 5.4
2.5 to 2.75 50.0 11.7 4.1
2.75 to 3.0 50.5 12.5
3.0 to 3.5 50.0 13.5 3.6
3.5 to 4.0 ~ 50.0 14.~5 3.5
:- -7-
:`
' .
f~l~
Because of the displacement of hydrogen ions in the resin by
metal ions, the loading of the resin with either cobalt or nickel
causes a release of hydrogen ions into the aqueous solution. As
a result the measured pH of the effluent varied from about 1.5
in the early ractions where substantial loadin~ was taking
place to 5.0 after 4 BV had been collected. It is clear from
Table 5 that even with the very high nickel content present in
the initial solution (in which the Co/Ni ratio was about 3.5)
up to 4 BV of effluent were collected before "breakthrough" was
evidenced by an effluent composition corresponding to that of
the aqueous feed.
EX~MPLE 4
Three column tests were conducted with a sulfate solu-
tion assaying about 100 g/l Co, 1.16 g/l Ni and having a pH o~ 6.3
measured at 2,2C. In each case the solution was passed through
a 0.8 m long resin bed at the rate of 3 m3/m2/h, and the bed
, - ~
temperature was 50CO One of the resin beds contained the
XF-4195 resin xeferred to in Examples 1 and 3. The second bed
contained the Amberlite IRC-718 resin referred to in Example 2,
while the thixd bed contained another recently developed Dow
Chemicals resin available commercially under the designation
XF-4196 which contains N-(2 - hydroxyethyl)~2-picolylamine
functional groups. The nickel assay of various ef~luent
fractions collected in the tests is shown in Table 6. The per-
formance of the XF-4196 resin was found to be comparable to
that of the Amberlite IRC-718 resin, whereas the results obtained
using the XF-4195 resin were vastly superior to both of the
other resins.
. . .
~ ~ .
~ -8-
; ~.
TABLE 6
Effluent Fraction Ni (g/l) in fraction for test with:
Collected (BV) XF-4195 XF-4196 Amberlite
resin resin IRC-718 resin
_ ,
0 to 0.5 0.0~3 0.058 0.045
0.5 to 1.0 0O096 0.253 0~204
1~0 to 1.5 ~.123 0.424 0~373
1.5 to 2.0 0.160 0.523 0.532
2.~ to 2.5 0.210 0.626 0.667
2.5 to 3.0 0.290 0.747 0.773
3.0 to 4.~ 0.420 Q.878 0.~82
4.0 to 5.0 0.600 1.000 0.995
_ _ . ._
.
The results are illustrated yraphically in the accom-
panying drawing which shows a plot of nickel removal as a
function of the amount of effluent collected, the nickel
removal being expressed in terms of the ratio o the nickel
content of the effluent fraction to the nickel content of the
feed solution. The drawing clearly shows the superiority of
the XF-4195 resin~
The invention has been described with reference to
preferred embodiments thereof. It will be understood that
various modifications may be made to the details of these
embodiments without departing from the scope o the invention
which is defined by the appended claims~
