Note: Descriptions are shown in the official language in which they were submitted.
LITHIIJM HALIDE RRINE PURIFICATION
The present inven~ion pertains to a method
for obtaining a high purity li~hium halide solution from
a resin/aluminate composite.
U.S. Patents 4,116,856; 4,116,858; and
4,159,311 disclose preparation of, and use for, ion-
-exchange resins having incorporated therein crystal~
line LiX 2Al(OH)3~ where X is halide, especially chlo-
ride.
When resin/aluminate composites are employed
to absorb Li+ ions from brines by having the Li+ taken
up into the aluminate crystals, there still remains in
the interstices of the composite some of the brine from
which the Li ions have been absorbed. When the brine
contains othe~ metal ions (i.e., other than alkali metal
ions) such as alkaline earth metal ions, then when an
a~ueous wash is employed to elute d~sired Li ions from
the crystals, the effluent also contains these other
metal ions which were in the interstices but which were
not in the crystal structure.
29, 770-F -1-
It has now been found that the use of a con-
centrated pure Na halide brine washing step before the
water washing step used for Li leaching, results in
washing out the me-tal ions (such as alkaline earth metal
ions) without leachlng out a significant amount of the
Li in the aluminate crystals. Then when water is
employed to leach out the Li~ in the aluminate crystals,
the only other metal ions present in any appreciable
amount in the effluent are those of the Na halide brine.
In accordance with the present invention,
concentrated, relatively pure, Na halide brine is used
as a pre-wash before the water leaching of Li~ values
from a composite comprising an ion-exchange resin
having incorporated therein a crystalline lithium halide
aluminate composition. Thus the Na halide which is
washed from the interstices of the composite by the
water leaching is mostly present in the initial por-
tions of the effluent and in decreasing amounts in the
subsequent portions of lithium halide solutionO Thus
there is obtained a lithium halide solution which con-
tains only Na halide as an impurity in significant
arnount.
The "resin/aluminate composites", used in the
present invention are preferably prepared by incorporat
ing crystalline LiX 2Al(OH)3 (where X is halide) into ion-
exchange resins, are as shown in the references listed
supra, including the heavily loaded resins described in
EP 0,029,253 and US 4,3~1,349.
29,770-F -2-
.~
;'; `~
~ 7~
As used hexein, the "halide" of the expres-
sion "Na halide" or 'ILiX" refers to Cl, Br, or I, with
Cl being preferred. For simplicity in this disclosure
the "halide" will be shown as the preferred chloride,
with the understanding that bromide or iodide is also
suitable.
The relatively pure Na halide brine, herein-
after referred to as NaCl brine for purposes of concise-
ness, may also be taken to mean an "alkali metal halide
brine" other than lithium halide brine, though a small
amount of Li+ in the NaCl brine is beneficial. It is
this relatively pure, but concentrate~, NaCl ~rine which
is used as the pxe-wash for the li~hium-loaded resin/-
aluminate composite before the water leaching step; the
water leaching step is done to remove much of the Li
values from the composite (but not all of the Li values),
thereby substantially "unloading" the composite. This
concentrated NaCl brine is preferably at or near satura-
tion, but any concentration above about 20 percent should
provide a reasonably efficient operation. ~t further
reduced concentrations the operation becomes less and
less efficient. From a practical standpoint to achieve
an economical and efficient operation, the concentration
is preferably from 24 percent to 26 percent.
The Li -containing aqueous solution or brine
solution from which Li+ ions are desirably removed are
those which are contaminated with metal ions other than
Li or alkali metal ions. These other metal ions are
usually alkaline earth metal ions, such as Mg , Ca
etc., but may also be vixtually any other non-alkali
metal cation.
29,770-F -3-
~ 7 ~h. ~
The Li -loaded resin/aluminate composites,
having in the interstices thereof any contaminating
metal cations other than alkali metal cations, would
ordinarily be derived from a process wherein a substan-
-tially "unloaded" composite is used to absorb or take
up Li ions from an impure brine, thereby becoming
"loaded" with Li ions. Obviously, though, any such
Li -loaded resin/aluminate having the "other" metal
cations in the interstices thereof may be used in the
present process. The purpose and intent of the pxesent
process is to remove the contaminating metal cations
from the interstices without removing a significant
amount of the Li~ ions; this is done by using a rela-
tively pure, preferably concentrated NaCl pre-wash
which leaves the Li+ ions in place, but which leaves
the interstices filled with the relatively pure NaCl
brine which i5 essentially devoid of the said contami-
nating metal cations.
When water leaching is subsequently employed
to remove Li from the composite, the initial portions
of efflueIlt wash ou-t much of the NaCl and some of the Li ;
subsequent portions comprise substantially pure Li solu-
tions having NaCl as virtually the only contaminant. It
is preferable than not all the Li be removed from the
crys-talline LiX-2Al(OH)3 because removal of all the Li
can cause collapse or destruction of the crystalline
aluminate structure. Therefore it is best if the water
which is used for water leaching contains a small amount
of Li (say, at least about 80 ppm) as this prevents
total removal of Li from the aluminate structure.
29,770 F -4-
It is believed that the greater the concen-
tration of NaCl in the solution, th~ greater is the ten-
dency for the crystalline aluminate structure to adsorb,
and hold, the Li values. Upon water leaching, the con-
centration of NaCl in the interstices of the compositeis reduced and as the reduction is occurring, the Li
is becoming more and more leachable from the crystal.
One of the techniques by which Li values
are removed from aqueous solution is by precipitation
as lithium carbonate ~Li2Co3). If other metal cations,
such as alkaline earth metal cations, are present they
may also precipitate as carbonates along with the lith-
ium. Sodium carbonate does not precipitate with the
lithium carbonate, therefore NaCl in the Li+ solution
does not pose the same contamination problems as the
other metal cations. Furthermore, if NaCl is the only
con-taminant in the Li solution, the NaCl can be crystal-
lized by evaporation of most of the water, leaving a pure
concentrated LiCl solution; this would not be possible
with the alkaline earth metals.
The water used for water leachin~ the Li
~rom the aluminate structure in the composite should
be substantially devoid of other metal ca-tions (such
as alkaline eaxth metal cations) except that a small
amount of Li should be present, for purposes stated
supra. Also a small amoun-t of other alkali metal cat-
ions, such as Na may be tolerated. Deionized (soft-
ened) water may be used as well as distilled water or
other reasonably pure water, so long as there is no sub-
stantial amount of alkaline earth metal or other non-
-alkali metal cations present~
29, 770-F -5-
~ t~3
The following examples are given to illus-
trate certain embodiments, though the inv~ntion is not
limited to the particular embodiments illustrated.
In the following example, the water leaching
step is performed using deionized water containing about
68 ppm Li ions. The crude (impure) Li -containing feed
brine from which Li+ ions are to be extracted is a min-
eral brine fxom the Smackover deposit near Magnolia,
Arkansas which was substantially saturated with NaCl
and which contained, nominally about 280 ppm Li ions,
about 6~2 percent Na~, about 17.05 percent Cl , about
0.41 perc~nt K , about 0.31 percen-t Mg , about 3.3 per-
cent Ca , about 0.23 percent Sr~+, about 0.022 percent
B , about 20-30 ppm Mn , about 20-30 ppm Mo , and
about 9-10 ppm Cu ; its pH was about 5.8. The substan-
tially pure NaCl brine used as a pre~wash before water
leaching was saturated, contained abou-t 200 ppm Li ions,
and was about 6.3 p~. The resin/aluminatP composite was
a weak base anion-exchange resin, available commercially
fxom The Dow Chemical Company under the tradename DOWEX
MWA-1, into which had been incorporated crystalline
LiCl 2Al(OH)3. The vessel employed was a glass ion-
-exchange column of 120 cc volume which was jacketed
for temperature control by circulating fluid.
Exam~le 1
The ion-exchange column was filled with the
resin/aluminate composite which had been water leached
to substantially "unload" the composite with respect -to
29,770-F -6
--7--
its Li content. ~t a temperature of about 77C and a
flow ra~e of 10 cc/min., the crude feed brine was passed
through the composite bed until the total hardness (Mg
Ca , etc.~, as determined by s~andard versene method,
was the same entering and leaving the column. Then a
relatively pure saturated NaCl brine pre-wash was used
to displace the crude feed brine remaining in the inter-
stices of the composite, the pre wash was passed through
at 3.3 cc/min. until the amount of hardness in the efflu-
ent was less than 0.002 M. Following tha-t the wat~r
leaching step was done, using water flow at 3.3 cc/min.
The effluent from the column was taken as
cuts for analysi.s of density, molar hardness, and ppm
L,i . Table I demonstrates the elution pattern for one
cycle of loading/brine wash/leaching. It will be read-
ily understood by persons skilled in ~hese arts that
the "unloaded" composite may ke employed in numerous
such cycles.
29,770-F 7~
7~.~
TABLE I
Effluent
Cuts
Vol. Density Hardness Li
No. cc (q/cc) (Molar~ (ppm) R_marks
1 50 1.137 - - "Feed" brine in
2 50 1.186 - -
3 50 1.182 - -
4 48 1.190 1.144
10 5 - _ _ _
6 ~
7 - 1.207 1.156 - begin brine wash
8 50 - 1.076
9 25 1.162 0.504
- 1.144
11 25 1.14h 0.032 270
12 25 - 0.012 210
13 25 - 0.002 190
14 25 - - 170
15 25 - - 150 begin water leach
16 25 - - 145
17 25 - - llS
18 25 - - 100
19 25 - - 1324
20 25 - - 1375
21 25 - - 675
22 25 - - 490
23 25 - - 385
24 25 - - 320
25 25 - - 290
The peak cuts (19-24) contained very little hard-
ness, if any.
Example 2 - Compaxative Example
This example demonstrates the effect of omlt-
ting the NaCl pre-wash of the present invention. The pro-
cedure of Example 1 above is substantially followed except
29,770-F -8-
that the exchange column bed was 170 cc volume, the resin
composite contained a higher loading of the LiC1 2Al(OH13
crystalline material such as prepared substantially in accordance
with the process disclos~d in Canadian Patent No. 1j144,699, the
temperature of the column was 90C, the flow ra-te was 2.5 percent
of bed volume per minute, and the water wash contained about 60
ppm Ii . After passing enough of the Smackover brine through
the resin/aluminate composite to "load" the composite with Li
values, the water leaching step was conducted without a NaCl
solution pre-wash. As can be seen from the data in Table II
below, the "peak" cuts for Li+ concentration contained a substantial
amount of hardness values.
TABLE II
EfE n _Cuts
Molar
No. Vol.cc Hardness ppm Li Remarks
1 50 1.29 350 water leaching
2 10 1.29 350
3 10 1.30 350
4 10 1.31 350
1.29 350
6 5 1.29 40Q
7 5 1~34 450
8 5 1.32 700
9 5 1.14 2700
0.75 6250
11 5 0.42 7600
12 5 0.24 6900
13 5 0O14 6000
14 5 0.07 50Q0
0.06 4350
16 5 0.05 3800
17 5 0O03 3300
18 5 0.02 300
_ 9 _
., ,~
--10--
Exam~le 3
This example is run substantially as in Exam~
ple 1. The Smackover brine was pumped down-flow through
a 120 cc bed of lithium chloroaluminate in DOWEX- ~A~l
resin until the effluent contained the same concentration
of Li+ as the feed; this was tested by standard atomic
absorption methods. A pre-wash of pure, saturated NaCl
brine was passed through the bed at a flow rate of 2.75
percent of bed volume per minute and the effluent col-
lected in cuts. After about 0.83 bed volumes of thepre-wash, the bed was flushed with deionized water
containing 280 ppIn Li for 1.66 bed volumes and the
effluent collected in cuts. Then the flow of Smackover
brine was resumed at a flow rate of 8.2 percent bed
volumes per minute to recharge or "reload" the resin
,composite with lithium. Table III shows the elution
data.
29~770 F -10-
`1~.` ~
TABLE II
Effluent Cuts Molar +
No. Vol.cc Hardness Ppm Li Remarks
1 10 1.128302 Sat'd. NaCl Yre-wash
2 10 3~7
3 10 375
4 lQ 377
1.148377
6 10 1.020376
10 7 10 ~.670424
8 10 0.~32408
9 10 0.272~06
~.188532
11 10 0.144511 Begin water~wash
(280 ppm Li )
12 10 0.108527
13 10 0.07Z513
14 10 0.036496
0.048479
2016 10 0.02~3315
17 10 0.0721650
18 10 0.0524975
lg 10 0.0284275
0.0203575
2521 10 - 2825
~2 10 0.0122425
23 10 0.0022075
24 10 0.0061800
- 1525
3026 10 - 1375
27 10 - 1250
28 10 - 1175
29 10 - 1075
0.006992
3531 10 0.004971 Begin filtered
Smackover brine
32 10 0.0081000
33 10 0.008928
34 10 0.008793
~035 10 0.184383
36 10 0.368~38
37 10 0.496 45
38 10 0.640 34
39 10 0.736 28
4540 10 0.~16 24
29,770-F -ll-
-12-
Of the above cuts, 18 21 were taken as the
product cut, having about 3912 ppm Li~ and low hardness.
The remaining cuts were used as follows, most of them
being wsed in a concentrating cycle, which data is shown
in Table IV below.
Cuts 1~6 were recycled back to the feed tank.
Cuts 7, 8 and 9 were combined and flowed through a resin/-
aluminate bed to "load" the Li into the resin. Cuts
10~13 were combined and passed through the resin bed.
Cuts 14-15 were combined and passed through the resin
bed. Then 33 cc of fresh make-up brine was passed
through to replace portions of cuts 7-15 which had
been ].ost or taken out through samplings. Cuts 16
and 17 were combined, saturated with NaCl and passed
-through the bed. Cuts 18~21 were saved as product
samples. Cuts 22-23 were combined and then passed
through the bed. A make-up quantity (34 cc) of 280
ppm Li+ deionized water was passed through the bed.
The data are in Table IV.
29,770-F -12-
~13-
TABLE IV
Effluent Cuts Molar
No. Vol.cc Eardness ppm Ll Remarks
. . . _
1 10 1.120 350 Previous cuts
through bed
2 10 ~ 392
3 10 1.164 401
4 10 - 408
5 10 ~.132 413
10 6 10 - 429
7 10 0.888 447
8 10 0.700 471
9 10 0.524 514
0.380 533
15 11 10 0.296 565
12 10 0.232 56~
13 10 0.164 579
14 10 0.128 571
0.080 575
20 16 10 0.06~ 564
~.7 10 0.052 573
18 10 0.076 731
19 10 0.100 3950
0.052 5440 Cuts 20-23
25 21 10 0.028 4675 taken as prod-
22 10 0.024 3925 uct cuts, com~
23 10 0.016 3400 bined, had +
24 10 0.012 2900 4360 ppm Li
0.106 2575
30 26 10 0.00~ 2325
27 10 0.008 2250
28 10 0.008 ~075
29 10 0.008 20~0
0.2~8 814
35 31 10 0.~72 346
32 10 0.612 178
29,770~F ~13-
-14-
Example 4
For this test, to show further concentratlng
of Li values by performing a mu3tiple number of absorp-
tions and desorptions of the Li+ values, selected cuts
are taken from the cuts shown in Table IV above.
This preferred method of operating, then,
involves utilizing portions of the effluent in succeed-
ing runs. This could be accomplished by storing the
effluent in a coil or a succession of tanks to approxi-
mate the gradlents. Thus the only "new" feed neededfor each elution cycle would be a make~up amount of
fresh deioniæed water (containing a small amount of Li ),
equivalent in volume to the product cut, and enough NaCl
to resaturate a portion of the gradien-t to be called
"reflux". By successive operations, using Li-contain-
ing cuts from prior runs, the concentration of the prod-
uct is increased a~ove that possible in a single cycle.
Referring to the cuts of Table IV above, the
first 6 cuts, being relatively high in hardness and low
in Li, are recycled back to the Smackover brine feed.
Cuts 7, 8 and 9 are combined and are used as the first
solution pumped into a column to elute an e~uivalent
volume of Smackover brine. Cuts 10, 11 and 12 are com-
b:ined and pumped through after the 7,8,9 mixture. This
is then followed with a solution made by mixing 13, 14,
15 and 16; then 17, lB and 19 are resaturated with NaC1
and pumped into the column. Cuts 20, 21, 22 and 23 were
saved as product. Cuts 24-30 were combined to follow
the "refluxll mixture of 17,18,19. Next a volume of 280
ppm Li deionized water (eguivalent to volume of prod-
uct cuts) follows the 24-30 cut mixture. The effluent
29~77G-F -14-
~15-
'7~
for all this was collected in cuts as before to use in
a subsequent concentrating cycle. This procedure was
followed until by the fifth cycle, the product cut had
increased from 3912 ppm Li (first cycle) to 5531 ppm
Li (fifth cycle), with little or no increase in total
hardness.
29,770-F -15-
