RI 8826

Bureau of Mines Report of Investigations/1983

Co/EX SX & IX Information\Versatic & Naphthenic

Neodecanoic Acid as a Co3*

Solvent Extractant

By L. R. Penner, Davis E. Traut,

and Jack L. Henry

UNITED STATES DEPARTMENT OF THE INTERIOR

Report of Investigations 8826

Neodecanoic Acid as a Co3*

Solvent Extractant

By L. R. Penner, Davis E. Traut,

and Jack L. Henry

UNITED STATES DEPARTMENT OF THE INTERIOR

James G. Watt, Secretary

BUREAU OF MINES
Robert C. Morton, Director
Library of Congress Cataloging in Publication Data:

Penner, L. R. (Larry R.)

Neodecanoic acid as a Co solvent extractant.

(Report of investigations / United States Dept. of the Interior, Bu

reau of Mines ; 8826)
Bibliography: p. 8-10.

Supt. of Docs, no.: I 28.23:8826.

1. Cobalt— Metallurgy. 2. Neodecanoic acid. •3. Extraction (Chem

istry). I. Traut, D. E. (Davis E.). II. Henry, Jack L. III. Title. IV.
Series: Report of investigations (United States. Bureau of Mines) ;
8826.

TN23.U43 622s [669'.733] 83-600274

 CONTENTS
Page

Abstract 1
Introduction 2
Experimental equipment and solutions 4
Results and discussion 5
Cobalt extraction 5
Reagent losses 6
Coextraction of ammonia 6
Conclusions 8
Bibliography 8

ILLUSTRATIONS

1. Reduction roast , ammonia leach method 3

2. Effect of equilibrium pH on cobalt extraction with 30vol pet NDA 5
3. Effect of pH on solubility of NDA in aqueous raffinate solutions 6

TABLE

1. Typical analysis for nickel-copper raffinate 5

 UNIT OF MEASURE ABBREVIATIONS USED IN THIS REPORT

°C degree Celsius mol mole

g gram pet percent

g/L gram per liter psig pound per square inch

gage
Ib pound
vol pet volume percent
min minute
 NEODECANOIC

By L.
ACID

R. Penner,
10
AS

Davis
A Co3*

E. Traut,
SOLVENT

and
EXTRACTANT

Jack L, Henry
*3

ABSTRACT

The Bureau of Mines conducted solvent extraction batch tests to evalu

ate the usefulness of neodecanoic acid (NDA) for extracting Co^+ from
ammoniacal leach liquors high in ammonium sulfate. This work is an out
growth of previous Bureau research in which technology was devised for
the recovery of cobalt, nickel, and byproduct copper from domestic lat-
erites. The previously devised technology uses an oxidative ammoniacal
sulfate leach to treat the laterite, and nickel and copper are then
recovered from the leach solution by solvent extraction and electro-
winning. This report deals with the subsequent solvent extraction of
cobalt from the nickel-copper raffinate solution.

The results showed that about 70 pet of the contained Co^+ can be
extracted batchwise at pH 8 using NDA (a 10-carbon-chain highly branched
carboxylic acid). However, NDA had a solubility of approximately 0.6 to
0.8 g/L in aqueous raffinate following extraction, and this high solu
bility makes it impractical for use with alkaline solutions that are low
in cobalt. In addition, the extracted cobalt species (predominately
cobaltic hexammine) is not amenable to the electrolytic recovery of
cobalt metal because of its stability.
This fundamental prob
extreme
lem, common to all extractants of the cobaltic ammines, is discussed,
and an alternative for recovering cobalt metal from loaded solvent by a
hydrogen-reduction method is also discussed.

1 Research chemist.
^Chemical engineer.
^Research supervisor.
Albany Research Center, Bureau of Mines, Albany, OR.
 INTRODUCTION

The Bureau of Mines has previously con preextraction step to the electrowinning
ducted research on methods for recovering step. Direct solvent extraction of Co3+
nickel and cobalt from low-grade laterite would nonetheless be a considerable im
deposits in southern Oregon and northern provement because it would eliminate a
California (19, _26-_28).4 This research processing step and the cobalt metal
showed that the nickel and cobalt — and inventory costs associated with the pre
copper as well — can be recovered using extraction metal-reduction step.
the leach method illustrated in figure 1.
This method consists of the following Several classes of organic extractants
unit operations: have been tested for cobalt extraction.
Hydroxyoxime chelating extractants such
1. Selective reduction of nickel, as Henkel ?s LIX 63 (24) and hydroxyquino-
cobalt, and copper by roasting with car lines such as Sherex Chemical Co.'s Kelex
bon monoxide; 100 (22) have been used to extract Co2 +
values. Both chelating types have been
2. Controlled oxidation in a 2M ammon- used with carboxylic acids in synergistic
iacal sulfate leach liquor; mixtures optimized for nonalkaline feed
solutions containing Co2 + (^, 15) .
3. Liquid-solid separation for leach
liquor recovery; Bisphenol sulfide extractants have been
proposed , but only the extraction of
(_5_)
4. Magnesium and manganese removal by Co2+ values has been reported. Co2+ val
precipitation as phosphates; ues from both acidic solutions (pH 5) and
alkaline solutions (pH 11) have been
5. Leach liquor pH adjustment; extracted with di(2-ethylhexyl) phos
phoric acid (D2EHPA) (20-2O . Although
6. Solvent extraction of nickel and Co3 + has also been extracted with D2EHPA,
copper with Henkel Corp.'s LIX 64N5 extraction efficiency suffered severely
extractant followed by selective strip when sulfate values exceeded 0.4Af (21).
ping and electrowinning of nickel and
copper; and As a group, the most promising cobalt
extractants are the carboxylic acids.
7. Solvent extraction of cobalt with These acids have been the subject of
Henkel fs LIX 51 extractant followed by numerous studies, including studies on
stripping and electrowinning of cobalt. carboxylic mixtures (12) , naphthenic acid
(_7.)
, oleic acid (30) , and phosphino-
Before cobalt can be extracted from the carboxylic acid (17) . In these studies,
nickel-copper raffinate with LIX 51, Co3+ the acids were primarily used to extract
must be reduced to Co2+ (18). For this Co2+ from low-sulfate or nonsulfate aque
reduction, cobalt metal is used, but this ous feed solutions.
necessitates recycling of one-third of
the total cobalt metal produced. Extrac Extraction of Co2+ using neo acids —
tion of Co3+ would not necessarily lower highly branched carboxylic acids such as
overall current requirements in electro- Shell Chemical Co.'s Versatic 9, 10, and
winning because the reduction of Co3 + to 911 — has been studied extensively using
Co2+ would only be delayed from the both aqueous nitrate solutions (j^, 10,
25) and alkaline sulfate solutions O_,
4Underlined numbers in parentheses re 23, 29, jn-320. Extraction of Co3 + from
fer to items in the bibliography at the dilute ammoniacal nitrate solutions (16)
end of this report. and from 0.5A? sulfate solution using Ver
^Reference to specific products does satic acids (32) has been demonstrated.
not imply endorsement by the Bureau of Very little work, however, has been pub
Mines. lished on the extraction of Co3+ from
 Laterite

Crush

Coke

~
I
Calc;ine CO? i fc
5 P , PPi nre
L , ^u prc )ducer [^
i— r* — *• reduc :tion -^ uu '
FeS2 j
1
\ A; ;h 02
Cool Makeup NH4OH,
(NH4)2S 04, H20
i

Gr J * ]

' i
;
n •*. Leach
(NH -^^O-
| solution
Liquid -solids
fI
^
separ ation Recovery NH3

solution
"'
Produc I licjuur A

' Solution
H4H2P04 — ** | Impurity
,

—*" T s wash
Tlailmg •* — H20
;f NH4(Ma. Mn)P04
NH3 stripping *•»
i
{ NH3 Gan gue
Ni-Cu solvent
extra clion AL L

Roffinate r, xr
[
i Raffmate recycle
Spent
,
electrolyte
Co solvent
recycles extraction

Electr olytes I
Spent
electrolyte
recycle
r \f H
|
•^ \ Electrowinning |
'

Ni Cu Co

FIGURE 1. - Reduction roast, ammonia leach method.

alkaline solutions high in ammonium sulfate at pH 8 was used, no significant
sulf ate. change in saturation loading of Versatic
911 was found through 5 complete cycles
The similarityof the Co3+ feed solu of extraction and stripping (23) . This
tions used in the studies cited above to result indicated that solubility losses
the leach liquor product of the Bureau's were low at the higher (4/if) ammonium sul
reduction roast, ammonia leach process fate concentration. The aqueous solubil
(fig. 1) suggests that neo acids could be ity of 7- to 9-carbon-chain (C7 to Cg)
used as extractants for Co3+. The evalu aliphatic monocarboxylic acids was found
ation of NDA, a generic equivalent of to increase with increasing extractant
Versatic 10, as a potential Co3 + ex- concentration in the organic phase (_9_) .
tractant from ammoniacal leach liquors Similar data for NDA solubility in am
high in ammonium sulf ate is the subject moniacal sulfate solutions have not been
of the report. published,but are of critical import to
the economic application of NDA; these
One problem encountered with carboxylic data are therefore included in this
acids in general is relatively high aque report.
ous solubility. For the application
studied, this high solubilitywould pre An additional consideration in the com
sent two problems: (1) Loss of extract- mercial application of any solvent ex
ant to the aqueous phase would result in tractant is its selectivity. Carboxylic
high reagent makeup costs, and (2) con acids are nonselective unless pH is care
tamination of the aqueous phase would fully controlled (29). This study con
necessitate cleanup to avoid fouling of cerns treatment of the ammoniacal system
other unit operations. after removal of the divalent metals;
therefore, the most important impurities
Carboxylic acid solubility has been remaining in the system are nonmetallic
studied by several researchers. In one (mainly ammonia). Hence, there is a need
study, in which Versatic 911 was used to to study ammonia coextraction in the form
extract cobalt from ammonium sulf ate, of dissolved gas, an ammonium ion, or a
solubility losses of Versatic 911 were cobalt ic-ammine complex, as it affects
found to increase at higher pH levels and cobalt extraction using NDA.
decrease at higher ammonium sulfate con
centrations (1). In the same study, This report describes the -extraction of
cobalt extraction from 2M ammonium sul Co3+ using NDA from an ammoniacal sulfate
fate at pH 8 (values similar to those of solution devoid of nickel and copper.
the nickel-copper raffinate used in this The effects of pH on Co3+ extraction and
study) was accompanied by an economically NDA solubility losses to the aqueous
prohibitive solubility loss of Versatic phase are quantified. Also discussed is
911 (8 g organic phase per liter of aque ammonia coextraction in reference to its
ous phase contacted). However, in a effects on subsequent metal recovery.
different study, in which 4A/ ammonium
EXPERIMENTAL EQUIPMENT AND SOLUTIONS

Solvent extraction batch tests were only the purer (95 pet C10) prime grade
conducted using separatory funnels and was used.
automatic shakers. After agitation,
phase separation, and filtration, the Actual nickel-copper raffinate, the
aqueous phase was analyzed for cobalt by product stream of the process described
atomic absorption spectrophotometry . in the introduction, was used as the
aqueous feed with the pH adjusted to the
NDA in kerosene (Kermac 470B, Kerr- value of interest. A recent improvement
McGee Corp.) was used asthe organic to the process, contacting the raffinate
extractant. Although both prime and with an ion-exchange resin (IRC 718, Rohm
technical grade NDA are manufactured, & Haas Co.) prior to cobalt extraction,
leaves only a trace of
nickel and the TABLE 1. - Typical analysis for nickel-
zinc in
solution. Theresin-treated copper raffinate, g/L
nickel-copper raffinate, as received,
contained about 5 to 10 g/L free ammonia; Ion Analysis
its pH was about 9.0 to 9.5. (The free
ammonia content was determined by tit ra Ca. , 0.003
tion with H2S04 to the neutral point.) A Co. , .140
typical analysis of the ions present is Cu. , .0006
shown in table 1. Fe., .001
Mg., .004
The cobalt in the nickel-copper raf Mn., .0003
finate was about 85 pet cobaltic hex- Na., .052
ammine [Co(NH3)63+] t and the remainder NH3 5
was cobaltic pentammine [Co(NH-5)5S04
+
] .
Ni., .001
Identification was by high-pressure S04 197
liquid chromatography (HPLC) , a technique Zn. , .0007
that was developed for a sulfate system
and is similar to a technique described
for a carbonate system (13).

RESULTS AND DISCUSSION

The extent of cobalt extraction was extraction from nickel-copper raffinate

measured as a function of the pH of the is shown in figure vol pet2. For 30
equilibrated aqueous phase. Solubility NDA, the highest cobalt extraction
losses of the extractant and ammonia occurred at an equilibrium pH of about
coextraction could then be examined in 8.0. The nickel-copper raffinate, as
the pH range in which Co3 + is effectively produced, was about pH 9.0 to 9.5. After
extracted. a batchwise contact with NDA at A:0 = 2
to extract cobalt, the equilibrium pH of
COBALT EXTRACTION the solution was 8.0 to 8.3. Therefore,
maximum cobalt extraction was obtained
A series of single extraction contacts from the as-produced nickel-copper raf
was made at equilibrium pH values rang finate with little or no pH adjustment.
ing from about 6.8 to 8.9. Adjustment of
the nickel-copper raffinate pH to de
sired levels was accomplished by sparging 80
with anhydrous ammonia (to increase pH)
70
and by treating with 2.5M H2S04 (to lower
pH) . Neither pH-adjustment method appre z 60
ciably changed total volume nor sig
nificantly changed the sulfate content I 50
from its initial 2.0A/ concentration. The
40
equilibrium pH after extraction con E
<
tact was generally 0.5 to 1.0 pH units J
30
lower than the feed solution pH. This
was a result of ammonia coextraction 2 20 Aqueous feed:
and the transfer of hydrogen ions from o 0.14 g/L Co
0
NDA. 10 2M(NH4)2S04

0
Extraction tests consisted of 5-min 6.5 7.0 7.5 8.0 8.5 9.0
contacts at an aqueous-to-organic ratio EQUILIBRIUM pH
(A:0) of 2 and were conducted at ambient
temperature (20° to 25° C). The in FIGURE 2. - Effect of equilibrium pH on cobalt
fluence of equilibrium pH on cobalt extraction with 30 vol pet NDA.
 REAGENT LOSSES 3.0

of the fundamental requirements for

One ,2.5 KEY
oPure NDA
a successful solvent extraction process o 30 vol pet NDA
is that the extractant must have a low 2.0
solubility in the aqueous phase. In
addition, the extractant phase must rap 1.5
idly and completely separate from the
aqueous phase when mixing is stopped. 1.0
Solubility and entrainment losses can
exact a heavy by toll
increasing reagent .5
makeup and exit-stream cleanup costs. If
organic impurities in raffinate solutions 0
are not completely contamination
removed, 7.0 7.5 8.0 8.5 9.0 9.5 10.0
of other unit operations of the process EQUILIBRIUM pH
and/or the environment may result. En
trainment losses of the organic phase are FIGURE 3. - Effect of PH on solubility of NDA
a physical result of the mixing process. in aqueous raffinate solutions.
Currently, entrainment losses are con
trolled by a variety of tools such as to $6 per pound of cobalt extracted at pH
coalescers and skimming tanks. Of 8.0 to 8.25. Solubility losses to strip
greater concern when NDA is used is re acid, although much lower, (0.078 g/L at
agent loss due to the solubility of NDA pH 1.3), would add to this total. The
in cobalt raffinate solution. Therefore, cost of removing entrained organic con
batch tests were run to determine the tent from the raffinate would also add to
solubility of NDA in cobalt raffinate the total cost of phase separation and
solutions. reagent makeup. At these levels of re
agent loss, NDA cannot be considered an
Samples of both pure NDA and 30 vol pet economically feasible cobalt extractant.
NDA in kerosene were contacted with 2Af
ammonium sulfate nickel-copper raffinate Figure 3 shows a lower aqueous solubil
solutions at room temperature. The aque ity for vol pet NDA than for pure NDA,
30
ous solutions were adjusted to a variety which is consistent with other findings
of pH values between 7 and 10, the range (9). This suggests that the unacceptably
found to be efficient for cobalt extrac high solubility of NDA could be overcome
tion. After contact, the aqueous phases through the development of synergic ex-
were carefully filtered to remove en tractants, other diluents, or other chem
trained organic phases. The remaining ical modifiers that might suppress the
organics in the aqueous solution were solubility of NDA in aqueous solution.
extracted with CC14 and analyzed by in However, the suppression of NDA solubil
frared spectroscopy . ity was not studied.
Figure 3 shows the solubility of NBA6 COEXTRACTION OF AMMONIA
as a function of pH and extractant con
centration. In the range of highest In addition to the high solubility of
cobalt extraction, pH 8.00 to 8.25, the NDA, which from an economic standpoint
solubility of NDA using a 30 vol pet alone prohibits its use with nickel-
solution about 0.56
was to 0.81 g/L. copper raffinate solution, there is the
Assuming 70-pct cobalt
a extraction from problem of ammonia coextraction. Coex-
a solution containing 0.14 g/L Co and a traction of ammonia, which can result in
$0.76/lb cost for NBA, the value of re fouled process equipment, is a problem
agent loss to solubility alone is from $4 encountered with the use of all Co3"1"
extractants, including NDA, in ammoniacal
"Kerosene solubility was almost negli- systems .
gible — 0.002 g/L.
 REAGENT LOSSES 3.0

of the fundamental requirements for

One ,2.5 KEY
oPure NDA
a successful solvent extraction process o 30 vol pet NDA
is that the extractant must have a low 2.0
solubility in the aqueous phase. In
addition, the extractant phase must rap 1.5
idly and completely separate from the
aqueous phase when mixing is stopped. 1.0
Solubility and entrainment losses can
exact a heavy by toll
increasing reagent .5
makeup and exit-stream cleanup costs. If
organic impurities in raffinate solutions 0
are not completely contamination
removed, 7.0 7.5 8.0 8.5 9.0 9.5 10.0
of other unit operations of the process EQUILIBRIUM pH
and/or the environment may result. En
trainment losses of the organic phase are FIGURE 3. - Effect of PH on solubility of NDA
a physical result of the mixing process. in aqueous raffinate solutions.
Currently, entrainment losses are con
trolled by a variety of tools such as to $6 per pound of cobalt extracted at pH
coalescers and skimming tanks. Of 8.0 to 8.25. Solubility losses to strip
greater concern when NDA is used is re acid, although much lower, (0.078 g/L at
agent loss due to the solubility of NDA pH 1.3), would add to this total. The
in cobalt raffinate solution. Therefore, cost of removing entrained organic con
batch tests were run to determine the tent from the raffinate would also add to
solubility of NDA in cobalt raffinate the total cost of phase separation and
solutions. reagent makeup. At these levels of re
agent loss, NDA cannot be considered an
Samples of both pure NDA and 30 vol pet economically feasible cobalt extractant.
NDA in kerosene were contacted with 2Af
ammonium sulfate nickel-copper raffinate Figure 3 shows a lower aqueous solubil
solutions at room temperature. The aque ity for vol pet NDA than for pure NDA,
30
ous solutions were adjusted to a variety which is consistent with other findings
of pH values between 7 and 10, the range (9). This suggests that the unacceptably
found to be efficient for cobalt extrac high solubility of NDA could be overcome
tion. After contact, the aqueous phases through the development of synergic ex-
were carefully filtered to remove en tractants, other diluents, or other chem
trained organic phases. The remaining ical modifiers that might suppress the
organics in the aqueous solution were solubility of NDA in aqueous solution.
extracted with CC14 and analyzed by in However, the suppression of NDA solubil
frared spectroscopy . ity was not studied.
Figure 3 shows the solubility of NBA6 COEXTRACTION OF AMMONIA
as a function of pH and extractant con
centration. In the range of highest In addition to the high solubility of
cobalt extraction, pH 8.00 to 8.25, the NDA, which from an economic standpoint
solubility of NDA using a 30 vol pet alone prohibits its use with nickel-
solution about 0.56
was to 0.81 g/L. copper raffinate solution, there is the
Assuming 70-pct cobalt
a extraction from problem of ammonia coextraction. Coex-
a solution containing 0.14 g/L Co and a traction of ammonia, which can result in
$0.76/lb cost for NBA, the value of re fouled process equipment, is a problem
agent loss to solubility alone is from $4 encountered with the use of all Co3"1"
extractants, including NDA, in ammoniacal
"Kerosene solubility was almost negli- systems .
gible — 0.002 g/L.
 Ammonia coextraction in an ammoniacal- discussion was included on how this would
cobaltic system may occur in several affect cobalt recovery.
different ways. Ammonia may dissolve in
the organic phase as ammonia gas, it The presence of ammonia in the ex
may form ammonium neodecanoate, or it tracted species portends problems in
may be extracted as a cobaltic ammine conventional st ripping-elect rowinning
complex. techniques for cobalt metal recovery. No
satisfactory methods for separating the
Tests showed that most of the free cobalt from the ammonia were found be
ammonia or ammonium neodecanoate can be cause of the stability of the hexammine
removed from NDA through a combination of complex. If the cobaltic ammine species
water scrubbing, air sparging, and heat is extracted without breaking up the com
ing of the organic. However, these steps plex, and subsequent treatments fail to
would add to expense of the total extrac break up the complex, the ammine group
tion process. But even these if
costs reports to the electrolyte when stripped.
were acceptable, the removal of ammonia Ultimately, the electrolytic process will
as coextracted ammines proves far more reduce the cobalt to metal and release
formidable. Cobaltic ammines are so the ammine group to be immediately neu
stable that ammine coextraction is a tralized by the H2S04 contained in the
fundamental problem which must be solved electrolyte. The resultant acid consump
for any Co3+ extractant in an ammoniacal tion would be economically prohibitive.
system.
with this ammonium
In addition, in the
As mentioned previously, the cobalt electrolyte, the electrowinning cell
species in the nickel-copper raffinate would soon clog up with the crystallized
and the extracted cobalt species in NDA double sulfate-salt of ammonia and cobalt
were identified by as primarily
HPLC [(NH4)2S04'CoS04'6H2Oj . A bleed stream
cobaltic hexammine. This conclusion was could not be used to control this salt
supported by an H2S04 material balance buildup since 6 mol of ammonia would be
and by a chemical analysis of the cobalt released for each mole of cobalt pro
valence state which showed predominantly duced. Therefore, a cobalt recovery
process based on electrolytic reduction
is not technically feasible without an
Identification of the extracted species ammonium sulf ate-cobalt separation in the
is important to any effort to electro- strip acid-electrolyte. The necessity
lytically recover cobalt metal from an for such a separation could nullify any
ammoniacal sulfate feed solution. The advantage in using a solvent extraction
problems associated with extracting the process to extract Co3+ over the cur
whole cobalt complex may extend to other rently proposed process (reduction of
liquid or solid extractants unless novel Co3+ to Co2+ followed by extraction with
methods for breaking up the complex prior LIX 51).
to stripping are discovered. Although
others have reported the extraction of An alternative to electrolyticrecovery
Co3+, there has been a lack of discussion is hydrogen reduction method in which a
a
on identification of the extracted spe cobalt— loaded extractant is subjected to
cies and subsequent cobalt metal re temperatures from 150° to 300° C and
covery. For example, extraction of hydrogen pressures from 200 to 2,000
Co(NH3)63+ using certain cation-exchange psig. Neo acids such as NDA are ideal
resins has been reported (11); but candidates for this method since they
although the report claimed that 6 mol of yield acceptable cobalt metal powder
ammonium sulfate would be recovered for without undergoing degradation. Tests
each mole of cobalt recovered, no indicate that hydrogen reduction of
 Ammonia coextraction in an ammoniacal- discussion was included on how this would
cobaltic system may occur in several affect cobalt recovery.
different ways. Ammonia may dissolve in
the organic phase as ammonia gas, it The presence of ammonia in the ex
may form ammonium neodecanoate, or it tracted species portends problems in
may be extracted as a cobaltic ammine conventional st ripping-elect rowinning
complex. techniques for cobalt metal recovery. No
satisfactory methods for separating the
Tests showed that most of the free cobalt from the ammonia were found be
ammonia or ammonium neodecanoate can be cause of the stability of the hexammine
removed from NDA through a combination of complex. If the cobaltic ammine species
water scrubbing, air sparging, and heat is extracted without breaking up the com
ing of the organic. However, these steps plex, and subsequent treatments fail to
would add to expense of the total extrac break up the complex, the ammine group
tion process. But even these if
costs reports to the electrolyte when stripped.
were acceptable, the removal of ammonia Ultimately, the electrolytic process will
as coextracted ammines proves far more reduce the cobalt to metal and release
formidable. Cobaltic ammines are so the ammine group to be immediately neu
stable that ammine coextraction is a tralized by the H2S04 contained in the
fundamental problem which must be solved electrolyte. The resultant acid consump
for any Co3+ extractant in an ammoniacal tion would be economically prohibitive.
system.
with this ammonium
In addition, in the
As mentioned previously, the cobalt electrolyte, the electrowinning cell
species in the nickel-copper raffinate would soon clog up with the crystallized
and the extracted cobalt species in NDA double sulfate-salt of ammonia and cobalt
were identified by as primarily
HPLC [(NH4)2S04'CoS04'6H2Oj . A bleed stream
cobaltic hexammine. This conclusion was could not be used to control this salt
supported by an H2S04 material balance buildup since 6 mol of ammonia would be
and by a chemical analysis of the cobalt released for each mole of cobalt pro
valence state which showed predominantly duced. Therefore, a cobalt recovery
process based on electrolytic reduction
is not technically feasible without an
Identification of the extracted species ammonium sulf ate-cobalt separation in the
is important to any effort to electro- strip acid-electrolyte. The necessity
lytically recover cobalt metal from an for such a separation could nullify any
ammoniacal sulfate feed solution. The advantage in using a solvent extraction
problems associated with extracting the process to extract Co3+ over the cur
whole cobalt complex may extend to other rently proposed process (reduction of
liquid or solid extractants unless novel Co3+ to Co2+ followed by extraction with
methods for breaking up the complex prior LIX 51).
to stripping are discovered. Although
others have reported the extraction of An alternative to electrolyticrecovery
Co3+, there has been a lack of discussion is hydrogen reduction method in which a
a
on identification of the extracted spe cobalt— loaded extractant is subjected to
cies and subsequent cobalt metal re temperatures from 150° to 300° C and
covery. For example, extraction of hydrogen pressures from 200 to 2,000
Co(NH3)63+ using certain cation-exchange psig. Neo acids such as NDA are ideal
resins has been reported (11); but candidates for this method since they
although the report claimed that 6 mol of yield acceptable cobalt metal powder
ammonium sulfate would be recovered for without undergoing degradation. Tests
each mole of cobalt recovered, no indicate that hydrogen reduction of
cobaltic-ammine loaded NDA is technically still prohibit its use. Nonetheless, the
feasible, but the relative cost compared hydrogen-reduction method has been dis
to electrolytic recovery has not been cussed in several published reports (_2-^,
studied since the solubility of NDA would 14, 33).

CONCLUSIONS

Solvent extraction of Co3 + from ammoni- sulfate leach solutions unless further
acal sulfate leach liquor was studied studies produce adequate solubility
using the extractant NDA. Although NDA suppressors.
extracted about 70 pet of the cobalt con
tained in 2M sulfate liquor, two major Extraction and stripping of the co-
problems were encountered: excessive baltic ammine complex by NDA essentially
extractant solubility losses to the aque did not alter the complex. Since the
ous phase and the lack of practical meth complexed cobaltic ammine is extremely
ods for separating cobalt from its coex- stable, the ammine could not be separated
tracted ammine complex. from the cobalt until electrolytic reduc
tion took place. Moreover, the freed
The loss caused by the high
reagent ammine would ultimately foul the electro
solubility of in the cobalt raffinate
NDA lyte by precipitation caused by ammonium
was prohibitive at the pH where cobalt sulfate buildup. Stripping and electro-
extraction efficiency was highest. Be winning of cobalt metal from Co3+-loaded
cause this loss was large (0.56 to 0.81 g NDA is thus impractical. Researchers
NDA per liter of raffinate) compared using other extractants for Co3+ need to
with the dilute cobalt levels in the feed be aware of these problems because they
solution tested, NDA is not considered to are not limited to NDA.
be a practical extractant for ammoniacal
BIBLIOGRAPHY
1. Ashbrook, A. W. The Extraction of 5. Burrows, R. C. (assigned to Ashland
Metals From Ammonium Sulphate Solution Oil Canada Ltd., Canada). Organic Sol
Using a Carboxylic Acid — I. Cobalt. J. vent Extraction of Cobalt From Aqueous
Inorg. and Nucl. Chem. , v. 34, No. 5, Solutions Containing Nickel and Cobalt.
1972, pp. 1721-1737. Can. Pat. 1,001,422, Dec. 14, 1976.

2.Burkin, A. R. Physical Chemistry 6. Crabtree, H. E., and N. M. Rice. A

of Metal Production by Reduction of Spectrophotometric Study of the Organic
Loaded Organic Solvents by Hydrogen. Phase Complexes Formed in the Extraction
Paper in Physical Chemistry of Process of Cobalt II
With Carboxylic Acids. Sol
Metallurgy (Richardson Conf., July, vent Extr. Proc. Int. Solvent Extr.
1973). Inst. Min. and Metall. , London, Conf., v. 1, 1974, pp. 791-822.
1974, pp. 43-47.
7. Flett, D. S., and A. W. Fletcher.
3. Production of Some Metal The Separation of Nickel From Cobalt by
Powders by Hydrometallurgical Processes. Solvent Extraction With a Carboxylic
Powder Metall., v. 12, No. 23, 1969, Acid. Trans. AIME, v. 247, No. 4, 1970,
pp. 243-250. pp. 294-299.

4. Burkin, A. R. , and J. E. Burgess. 8. Flett, D. S., and D. W. West. Ex

Production of High Purity Metal Powders traction of Metal Ions by LIX 63/Carbox-
by Direct Pressure Reduction of Loaded ylic Acid Mixtures. Solvent Extr. Proc.
Extractants. Paper in Proceedings of the Int. Solvent Ext. Conf., v. 1, 1971,
First Annual Meeting of Canadian Hydro- pp. 214-223.
metallurgists (Ottawa, Oct. 28-29, 1971),
pp. 51-62.
 Gindin, L. M.
9. Separation of Met 18. Nilsen, , R. E. Siemens, and
D. N.
als Extraction
by Exchange With Aliphat S. C. Rhoads. Solvent Extraction of Co
ic Acids in Alkali. Russ. J. Inorg. balt From Laterite-Ammoniacal Leach
Chem. Eng. Transl. , v. 5, No. 8, 1960, Liquors. BuMines RI 8419, 1980, 23 pp.
pp. 906-910.
19. Rhoads, S. C., D. N. Nilsen, and
10. Haffenden, W. J. Law-
, and G. J. R. E. Siemens. Solvent Extraction of
son. Effect of Ammonium Salts on the Nickel, Cobalt, and Copper From Laterite-
Solvent Extraction of Metal Ions With Ammoniacal Leach Liquors. Solvent Extr.
Liquid Cation Exchangers. Paper in Ad Proc. Int. Solvent Extr. Conf., v. 2,
vanced Extractive Metallurgy (Symp. , Lon 1977, pp. 446-452.
don, 1967). Inst. Min. Metall. , London,
1968, pp. 678-685. 20. Ritcey, G. M. , and A. W. Ashbrook.
Separation of Nickel and Cobalt in Ammo
11. Hatch, W. R. Removal of Cobalt niacal Solutions by Liquid-Liquid Extrac
From Ammoniacal Process Leach Solutions tion. Trans. Inst. Min. and Metall. ,
by Conversion-Ion Exchange Process. v. C 78, No. 751, 1969, pp. 57-63.
Paper in Proceedings of the Eleventh Com
monwealth Mining and Metallurgy Congress 21.Ritcey, G. M. , A. W. Ashbrook, and
(Hong Kong, May 1978). Inst. Min. and B. H. Lucas. Development of a Solvent
Metall., London, 1979, pp. 655-677. Extraction Process for the Separation of
Cobalt From Nickel. CIM Bull., v. 68,
12. Hawes, W. B. (assigned to Shell No. 753, Jan. 1975, pp. 111-123.
Research Ltd., London, England). A Pro
cess for the Extraction of a Metal Value 22. Ritcey, G. M. , and B. H. Lucas.
or Values. Brit. Pat. 959,813, Jan. 3, Extraction and Separation of Cop
1964. per, Nickel, Zinc, and Cobalt From Am
moniacal Solution Using Kelex 100.
13. Hurst, F. J. Separation of Cobalt CIM Bull., v. 68, No. 754, Feb. 1975,
From Nickel in Ammonia- Ammonium Carbonate pp. 105-113.
Solutions Using Pressurized Ion Exchange.
Hydrometall. , v. 1, 1976, pp. 319-338. 23. Process for the Recovery
of Cobalt From Ammoniacal Solutions Using
14. Kunda, and R. Hitesman.
W. , The a Carboxylic Acid. Solvent Extr. Proc.
Reduction of Cobalt From Its Aqueous Int. Solvent Extr. Conf., v. 1, 1971,
Ammine Ammonium Sulphate System Using pp. 463-475.
Hydrogen Under Pressure. Hydrometall. ,
v. 4, 1979, pp. 347-375. 24. . Separation of Copper from
Nickel and Cobalt by Liquid-Liquid
15. Lakshmanan, V. I., and G. Law- J. Extraction From Ammoniacal Solution.
son. Extraction of Cobalt by Kelex 100 CIM Bull., v. 65, No. 721, 1972,
and Kelex 100/Versatic 911 Mixture. J. pp. 46-50.
Inorg. and Nucl. Chem., v. 35, 1973,
pp. 4285-4294. 25. Shibata, J. , S. Nishimura, and
S. Mukai. Effect of Ammonium Salts on
16. Lawson, G. L. , and B. J. Pridden. the Solvent Extraction of Some Transition
Extraction of Copper, Nickel, and Cobalt Metals With Liquid Cation Exchangers. S.
With Versatic Acid From Ammoniacal Solu Nippon Kinz. Gakk. , v. 38, pt. 2, 1975,
tions. Solvent Extr. Proc. Int. Solvent pp. 55-73.
Extr. Conf., v. 1, 1974, pp. 711-723.
26. Siemens, R. E., and J. D. Corrick.
17. Michlmayr, M. J. (assigned to Process for Recovery of Nickel, Cobalt,
Shell Oil Co., New York, NY). Liquid/ and Copper From Domestic Laterites. Min.
Liquid Extraction of Cobalt Values. U.S. Congr. J.,
v. 63, 1977, pp. 29-34.
Pat. 3,728,366, Apr. 17, 1973.
10

27. Siemens, R. E., and P. C. Good, Nickel. U.S. Pat. 3,380,801, Apr. 30,
(assigned to U.S. Dep. Interior). Pro 1968.
cess for Recovery of Non-Ferrous Metals
From Oxide Ores and Concentrates. U.S. 31. Van der Zeeuw, A. J. Metal Ex
Pat. 3,929,468, Dec. 30, 1975. traction With Carboxylic Acids — I. Com
position of Complexes With Ni, Co II, and
28. Siemens, R. E., P. C. Good, and Fe III. Hydrometall. , v. 4, 1979,
W. A. Stickney. Recovery of Nickel and pp. 21-37.
Cobalt From Low-Grade Domestic Laterites.
BuMines RI 8027, 1975, 14 pp. 32. . Metal Extraction With Car
boxylic Acid — II. Effect of Temperature
29. Spitzer, E. L. , J. Radder, and on the Extraction of Cu, Ni, Co II, Zn,
H. M. Muys. Extraction Tests on Aqueous Fe III, and Co III. Hydrometall., v. 4,
Metal Sulphate Solutions Using Ver- 1979, pp. 39-50.
satic 9. Trans. Inst. Min. and Metall. ,
v. C 75, No. 718, 1966, pp. 265-266. 33. Van der Zeeuw, A. J. (assigned to
Shell Oil Co., New York, NY). Copper,
30. Williams, R. E., and J. H. Bayless Nickel, Cobalt and Iron Separation Pro
(assigned to Esso Production Res. Co., cess. U.S. Pat. 3,701,650, Oct. 31,
Houston, TX) . Separation of Cobalt From 1972.

^U.S. GOVERNMENT PRINTING OFFICE: 1983-705-020/89 INT.-BU.OF MINES, PGH..P A. 27196