Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
~223~D~2
Field of the Invention
This invention is directed to the recovery of copper from
aqueous ammonia Cal solutions by solvent extraction. A mixed extract ant
composition his been found to give improved results.
S l3ackground and Prior Art
Solvent extraction has been the usual method employed for the
recovery of copper from acidic and alkaline leach solutions, and numerous
extraction processes have been developed which usually are followed by
electrGwinning to produce high quality cathode copper.
lo In the processing of alkaline solutions containing copper the
extraction process has involved contacting relatively low grade solutions
of up to 20 g Curl with certain oxide, substituted quinoline or allcyl~
phosphoric acid extract ants, or by a mixture of extract ants, e.g. an
organophosphoric acid plus an ~hydroxy ogle
Electronic equipment today is mostly built from components
soldered onto electronic printed circuit boards. This industry has had a
very rapid expansion with the advancement of computer technology and its
demands. Circuit boards consist of a deposit of copper which is covered
by the lay-out circuit pattern. The copper not covered by the lay-out it
excess, and must be removed by treatment with a solution This leaching
or etching process can be accomplished using ammonia Cal solutions, con-
twining free ammonia, and one or more ammonium salts, erg. NH4Cl, as well
as possible oxidizing agents (H202, Nikko). By recycling, the copper
content in the etching solution can be raised to a level of about 150-170
g Curl. The copper must be recovered and the leach solution be suitable
for recycle to leaching.
For the processing of copper etch solutions, one presently used
technique of solvent extraction employs as extracting reagent, a butted-
kitten (trade name LOX 54 of Henkel Inc., see: United States Patent No.
4,083,758). In this patented process, the copper is reduced from a level
of about 150 g Curl in the leach etch liquor, down to about lo g Curl
after solvent extraction. That is, 2/3 of the copper 1B recycled to the
leach circuit in the alkaline NH~,Cl medium. The H2SU4 strip liquor goes
on to electrowinning. There is an obvious drawback to such a process, in
that the copper level is only reduced to 100 g/L in the extraction This
I
mean that the leaching and solvent extraction circuits are going to be
larger, and more expensive, than if the leach liquor could be reduced to
a much lower copper level, eon that the recycled etch solution is more
Pavourable for the subsequent leach of the circuit boards upon recycle to
the leaching stage.
This invention results from efforts to improve the extraction
of copper while allowing ready stripping and recycle of the extract ant.
Summary of the Invention
This invention provides a process of recovering copper from an
aqueous ammonia Cal solution thereof by solvent extraction, comprising:
(a) contacting the copper-loaded ammonia Cal solution with an organic
liquid phase comprising:
i) a tertiary carboxylic acid of the formula:
IH3
Cluck KIWI
R2
where Al - alkyd of 1 to 3 carbon atoms
R2 _ alkyd of 3 Jo 5 carbon atoms,
and, if) a alkali salicylaldoxlme of the formula:
HC=N-OH
Jo 011
.
where R - alkyd of 8 to 10 carbon atoms,
until the copper in the ammonia Cal solution is substantially extracted
into the organic phase,
(b) separating the two liquid phases, and
(c) recovering copper from the organic phase.
Preferably, the carboxylic acid is present in greater amounts than the
oxide, in an organic liquid delineate. The loaded organic phase desirably
is scrubbed to recover ammonia and then the copper stripped with aqueous
acid solution, the strip solution being subsequently fed to electron-
: nine for copper recovery.
!
~2~3~2
The invention includes a mixed extract ant composition comprise
in:
i) a tertiary carboxylic acid of the formula:
Clue
Rl---C--COO~I
lo
where Al = alkyd of 1 to 3 carbon atoms
R2 = alkyd of 3 to 5 carbon atoms,
and, ii) a alkali salicylaldoxime of the formula:
HC=N-OH
~011
where R - alkyd of 8 to 10 carbon atoms.
Description of Drawing
The single drawing it a flow sheet (schematic) illustrating a
preferred process for copper recovery from copper etch or scrap solution
with recycle features.
Detailed Description and Preferred Embodiments
The feed solution can be any aqueous ammonia Cal copper-contain-
in solution. Frequently, suitable feed solutions result from etching or
plating operations and have a copper loading of about 130-170 g Curl and
a pal of about 8-9.5. Other metals will not normally be present in sign-
flcant amounts in copper etch solutions. If other metals are present and
co-extract with the copper, they can be separated at a later stage. The
prows it able to extract a high percentage of the copper and leave less
than about 2 g Queue in the aqueous phase raffinate.
The carboxylic acid (i) of the extract ant is an alpha-disubsti-
tuned alkanoic acid having a total of from 7 to if carbon atoms, prefer-
ably 8-10 carbon atomizer a mixture thereof. A preferred carboxylic acid
extract ant it that known as Voyeuristic lo or Voyeuristic 911 [trademark] from
Shell Canada. Other tertiary carboxylic acids and mixtures of this type
: could be used.
I
The selected salicylaldo~ime (H) extract ant is the alkali
derivative, the alkyd group being straight or branched and having from
~-10 carbon atoms. The 5-nonyl derivative is preferred. The derivative
known as cargo P5100 or P5300 [tra-lemarkl from Imperial Chemical
5 Industries it of this type and has been found very suitable.
A variety of water-immisclble organic solvents may be used as
delineates for the tertiary carboxylic acid and aldoxime extracting no-
agents. As known in the art, such delineates should be substantially
wat~r-immiscible, should dissolve the extraction reagents and metal come
plexus thereof, and should not alter the extraction function. Suet- dill-
ens can be aliphatic or aromatic liquid hydrocarbons, halogenated hydra-
carbons, ethers and other liquid petroleum derivatives. Examples include
Tulane, zillion, fuel oil, carbon tetrachloride, 2-ethylhexanol and
especially kerosene, the latter being preferred.
The carboxylic acid (i) is present in the organic phase
(including delineate) normally in greater amounts than the aldoxime and
usually within the concentration range of about 10 to about 50~ by wt.
The aldoxime (ii) is present usllally within the rflnge of about 25% down
to about 5% by wit- (tile lesser amounts normally would be used with
amounts ox carboxylic acid in the upper portion of its range). Suffix
client of (1) and (ii) should be used to effect the desired extraction.
The aqueous to organic (A/0) phase ratio during the extraction
normally will be within the range of about l/l to 1/10. The actufll phase
ratio can be varied depending Oil the feed concentration, extraction
apparatus used, temperature, etc., as known to the art.
It has been found that ammonia co-extracts with the copper to a
significant extent and, for economy, it is preferred to recover and no-
cycle the extracted ammonia. An aqueous copper chloride scrubbing soul-
lion containing preferably about Lowe g Curl at a pi of about 0.5-1.5 has
been found most suitable. Some copper is removed with the ammonia, but
both the recovered ammonia and removed copper can readily be recycled.
The loaded solvent phase is stripped, most suitably with a
mineral acid solution, preferably sulfuric acid. Hydrochloric acid or
phosphoric acid could also be used. The strip solution suitably contains
from about 120 to about 200 g H2S04/L, preferably about 150-180 g/L.
I assay
-- 5 --
Some So will remain With the solvent phase after stripping
(when ~12S04 it used for stripping) and to allow for continual recycle, a
scrllb of the solvent phase to remove sulfate has been found desirable.
Aqtleous Nl~4Cl has been found quite effective as a sulfate scrub solution
(where the etch ablution is chloride based). From 2 to 10~ Nil Of soul-
Sheehan could be used, with about 5% by wt. preferred.
The solvent phase is recycled to contact the ammonia Cal feed
solution again, while the loaded strip solution is fed to a copper
electrowinning stage. The spent electrolyte from electrowinning is suit-
able for recycling to the acid stripping stage.
The following test procedures, examples and flow sheet are illustrative.
Example Procedure and Feed Analyses
Feed solutions for most of the bench tests were an actual etch
solution obtained from a plant in West Germany, analysis 146 g Clout.,
163.6 g Of AL with a pi 8.5. The solvents were diluted in Shell 140
[trademark] aliphatic kerosene. Separator funnels were used in bench
shake-out tests, and the continuous circuit was run on a 10-stage mixer
settler unit (Bell Engineering Co.). The continuous testing in the mixer
settler circuit employed a synthetic copper etch solution because of the
quantity required. Unless otherwise noted, all tests were at ambient
temperature.
Extraction
Screening of Selected Extract ants
Selected extract ants, previously tried for the extraction of
copper from ammonia Cal solutions, were contacted with the feed liquor for
5 minutes in separator funnels. The results of these screening tests
are given in Table l and indicate that these extract ants, by themselves,
were capable of reducing the feed copper concentration from 146 g/L only
to about 100 g/L. The Acorga P5100 and Voyeuristic 91l gave the best per-
pheromones of the extract ants evaluated in tilts system. The 25% Acorga
P5100 appeared to provide for the lowest raffinate while the 40~ Voyeuristic
911 gave the highest loading of copper. To attempt to improve on such
results, further tests with a mixture of the two extract ants were carried
; 35 out (in the hope of finding a synergistic extraction system).
:~L2~232~
TILE 1
Screening of Selected Extract ants - Feed 146 g Curl
. .. . .. .. .
Extract ant in Extraction
kerosene Dlluent A/O Ratio Rail Organic Cut Extracted*
_ _ _ g Curl g Curl
20% DUPE + 5% TBPl 1/3 96.5 9.66 33.9
__ I_ _
25~ Versatlc 911 1/3 76.5 21.8 47.6
40% Voyeuristic 911 1/3 53.1 28.9 63.6
40% Voyeuristic 911 1/75 9.17 16.7 93.7
., . ......... , . .. . .
10% Acotga P5100 1/4 111.0 6.74 24.0
25% Acorga P5100 1/3 105.0 10.0 28.1
25% Acorga P5100 1/7.5 3.57 1~.3 97.6
25% LOX 632 1/3 197.0 10.5 26.7
25% LOX 63 1/7.5 71.6 8.75 51.0
25% LOX 63 1/10 66~3 7.32 54.6
.
* % Extraction based 0l1 raffinate assay
1 DUPE = Di-2-ethylhexylphosphoric acid
TOP 3 Tributylphosphate
2 LOX 63 - trade name for an alpha-hydroxy oxide of Henkel Inc.
~2~:3~:~2
Example 1 - Mixed E~tractant System
A series of tests were conducted in which the extract ant system
comprised various proportions of the Voyeuristic 911 and the Acorga P5100.
The mixed solvent phase was contacted with the ammonlacal copper etch
solution at room temperature for 3 minutes at 0/A ratio of 10/1. The
results are given in Table 2, indicating that a mixture of 40% Voyeuristic
911 10% Acorga P5100 in Shell 140 kerosene can produce a raffinate of
1.5 g Curl. These results indicate a synergistic extract ant had been
found. Subsequent tests used this solvent composition.
TABLE 2
Extraction Using Mixed Solvent
% in Mixed Extract ant
Voyeuristic 911 Acorga P5100 Raffinate % Extraction
g Queue, _ _ __ _
1.49 98.9
4.75 96.5
2.58 98.1
_ _ _ _ _
lo Example 2
Examination of Loading Properties of
Solvents and Solvent Mixtures for NH3
Tests were performed on the components of the solvent mixture,
and compared with the mixture, to determine whether NH3 was loading
preferentially on one of thy components. The results, shown in Table 3,
indicate that the carboxylic acid, Voyeuristic 911, is extracting more NH3
than is the Acorga P5100. Therefore) it might be considered to decrease
the Voyeuristic 911 component. The synthetic feed solution contained
134.7 g Curl and 154 g NH3/L and 270 g NH4Cl/L. The A/O ratio was 4/1,
purposely high to determine whether NH3 was preferentially extracted.
The data indicated a relationship between Cut and NH3 being extracted
:~2232~2
TABLE 3
Extractioll of NH3
Extract ant % in % Extracted
kerosene Delineate Cut NH3
40% Voyeuristic 911 7.72 6.49
25% Acorga P5100 3.64 3.25
40% V911 10% Acorga 7.72 5.34
Exile 3
Additional screening tests Jo determine the relative extraction
of Cut and NH3 were performed on several solvents. The data in Table 4 '
indicates again that the Acorga P5100 has the greatest discrimination of
Queue of those solvents evaluated. The synthetic feed solution
contained 120 g Curl, 120 g ~13tL and the tests were performed at A/O
ratio of 1. The Acorga P5300 differed from the P5100 in that increased
amounts of nonylphenol were present in the former.
TABLE 4
Comparison of Solvents for Co-Extraction of Cut and NH3
Solvent Composition Loaded Solvent (g/L) Queen
% in Kerosene Cut N1~3 Ratio
40% Voyeuristic 10% Acorga P5100 38.3 19.0 2.01
40% erratic + 10% Acorga P5300 39.0 18.0 ~.17
40% Voyeuristic 911 36.6 16.0 2.29
25% Acorga P5100 17.8 4.0 4.45
25% Aeon a P5300 _ 11.4 5.0 2.28
g ' _ i
Example 4
Kinetics and Extraction of Cut and NH3
A series of tests were conducted over a contact time period of
5 seconds to 300 seconds to determine whether a significant difference in
extraction of either Cut or NH3 occurred with time. The results showed
a constant extraction of about 30% Cut and 15% ~13 Over this time period
investigated. The 0/A phase ratio was 1 in all tests. The extraction
was rapid an constant for both Cut and NH3.
I
Example 5 - Scrubbing of NO
Screening of Scrubbing Agents for NH3 Removal
The continuous tests in a mixer settler circuit (see En. 8
below) indicated that considerable ammonia (up to 40%) was being extract-
Ed with the copper. If not scrubbed anti recovered before copper s~rip~pLng the al~nonia would be lost to the acid strip electrolyte. Therefore,
it would be preferable to selectively scrub and recover the ammonia with-
out stripping the copper at the same time. Most of the tests were not
sampled and analyzed as either the phase separation time was very long or
the scrub solution was colored indicating that copper also was removed.
The results shown in Table 5 indicate that of the solutions tested, a 10
g Curl copper chloride solution at pal about 0.5 resulted in the removal
of 6.59 g NH3/L after one stage at A/O phase ratio of l/1. The contact
time was for 3 minutes at room temperature and the loaded solvent con-
twined 32 g Curl.
Scrubbing with Quickly
A loaded solvent containing 26.7 g Curl was scrubbed 5 success
size times, A/O ratio 1/1, with a Quick solution at pi 0.70 containing
8.53 g Curl. The data in Table 6 indicates that ammonia, together with
some copper, are removed by scrubbing. One to two stages should be surf-
ficlent. In a process flyweight, this scrub solution or a bleed, would
eventually be returned to the extraction circuit.
TABLE 6
Successive Scrubbing for NH3 Removal
8.53 g Curl as Quick at pi 0.7
Contact Scrub Aqueous (g/L)
Cut NH3
l 3.78 5.90
2 12.4 1.68
3 13.9 0.78
4 15.0 0.31
15.3 ~.09
.
~L2232~2
-- 10 --
TABLE 5
Scrubbing of NH3
Scrub Solution Phase Separation Scrub Aqueous Rough (g/L)
_ Time (seconds) Raffinate Cut N113
Ill OWE M> 600 clear, colorless
0,5 6 blue, hazy
1.0 6 blue, hazy
H20 > 600 clear, colorless
NH40H > 600 clear, colorless
NH4Cl 0.25 M> 300 clear, colorless
0.50 > 300 clear, colorless
It > 300 clear, fight violet
Quick OWE g/L > 300 clear, colorless
2.5 g/L > 300 clear, colorless
5 glue 300 clear, colorless
10 g/L> 300 clear, colorless
pi 0.53* 6 clear, colorless 7.07 6.59
1.0* 5 blue, hazy 33.8 6.52
* Aqueous scrub feed 9.58 g Curl
~2~2~;~
Exhume 6 - SOL
tripping of 40% Voyeuristic + 10% Acorga in Kerosene
This loaded solvent, containing 29.5 g Curl, was contacted at
varying phase ratios with 15Q g 712S()4/1. strip solution for 3 minutes.
The equilibrium data obtained are glen in Table 7.
TABLE 7
ova Ratio Equip. pi Organic Strip % Cut Stripped
A. ___.. ___._____ _ I Cull g Curl
5/1 7.75 - 0.97
3/1 3.8 - 1.06
1/1 1.38 1.43 - 95.1
1/3 1.25 1.58 12.0 94.6
1/5 1.2 0.19 7.45 99.3
Effect of H2S04 Concentration on Stripping
Subsequent stripping tests were performed at varying concentra-
lions of H2S04 to determine whether any selectivity over ammonia or
chloride occurred. The data in Table 8 are for a loaded solvent contain-
in 30 g Curl contacted in a single stage at an 0/A ratio of 3/1. The
results indicate that as the strip cold concentration is increased from
135 to 180 g H2S04/L, the copper concentration is increased from 27 to 60
g/L in the strip, while the NH3 decreases slightly from 28.8 to 23.5 g/L.
Chloride it stripped with increasing acidity. Also at the higher strip
acidity, there is less interface emulsion.
TABLE o
Strip Feed O/AEquil. Strip Solution (g/L) Remarks
g H2S0~/L Ratio pi Cut NH3 Of
,
135 3/1 3.42 27.328.8 0.16 Interface emulsion
150 3/1 3.20 39.727.9 - Some emulsion.
Aqueous phase cloudy
180 3/1 2.85 59.723.5 2.51 Small emulsion. Some
white crystals.
. . -----
~L2~3~
- 12 -
Example 7 - Scrubbing of Solvent for S04 Removal
Normally, in the design of the circuit, following stripping of
copper with H2S04, the stripped solvent would be returned to the extract
lion ~rcuit. Because of the possible presence of S04 anion Oil the
~otvent due to stripping, such So would usually require removal before
contact with the an~onlum chloride feed solution containing copper
(otherwise a buildup of So would occur in the etch solution). Scrub
tests using water (A/0 ratio 1/1) and 5% N114Cl (A/0 ratio 3/1) at pi 7.5
were performed on the ~2S04-strlpped solvent, using an A/O ratio ox 1.
The results in Table 9 of five successive contacts indicate the S04 is
removed by scrubbing, and that aqueous clue is a most effective scrub
solution.
TAO 9
Water Scrubbing for S04 Removal
-
Contact Scrub Solution (g SLY)
Water 5% clue at put 7.5
1 0.23 OWE
2 0.21 0.5
3 0.17 0.23
4 0.20 0.077
0.20 0.011
Example 8 - Continuous Testing
Mixer Settler Data
__ _
Based on early bench data (before scrubbing and stripping were
optimized) a continuous circuit was evaluated on a Bell Engineerirlg mixer
settler unit. The object was to test one possible flow sheet whereby 5
stages of extraction (0/A ratio 4/1), 3 stages of stripping (0/A ratio
1.3/1) with 150 g H2S04/L and 2 stages of water scrubbing (to remove
S04 ) would be evaluated. Based on the data of those tests, with some
further work it would be possible to optimize the operating parameters.
The solvent mixture consisted of OWE Voyeuristic 911 10% Acorga P5100
diluted in Shell 140 kerosene. The initial 3 hour run was at room
temperature, with no organic recycling (Table 10), while the subsequent
~223;~
test was for an 8 hour duration, with the organic recycled (Table lo).
Also, only a single scrub stage was used in the latter run. The Cynthia-
tic aqueous feed contained 115 g Curl., 135 g N~13/L and 151.6 g Of AL.
The second, longer run, operated extremely well, with no obvious chemical
problem nor phase disengagement problems.
De~radatlon Tests on Recycled Solvent
After recycling the solvent mixture 29 times through extract
lion, stripping and scrubbing, the solvent was compared to fresh solvent
to determine whether degradation had occurred. The equilibrium isotherms
indicated very little change had occurred, within experimental error, as
a result of recycling.
Electrowinning of Copper Strip Solution
A quantity of strip liquor from the continuous run was subjected
to electrowinning. The feed solution corltained 35 g Curl, 2.21 g NH3/L,
and 2.27 g Of AL. Before electrowinnlng, the strip solution was passed
through a bed of activated carbon to remove any organic present (optional
step), These preliminary electrowinning jests at about 20 amp/ftZ pro-
duped smooth, compact deposits and reduced the concentration to 15 g
Curl. No additive was necessary.
Conceptual Flow sheet
In order to provide for decreased capital and operating costs,
and to achieve a more compact and efficient circuit for the treatment of
copper etch solutions, the flow sheet shown in the attached drawing was
developed.
In one typical case, spent etch solution, after being through
the leaching process, and containing 140-170 g Curl, can be extracted by
a mixed extrsctant. Our work has shown that a mixture of two extract ants
result in synergistic extraction (lye. better than either of the two
extract ants alone), whereby a low level of 1-2 g Curl or less can be
achieved during extraction. The mixture found preferred for the typical
etch solution that has been investigated is comprised of 40% tertiary
carboxylic acid of about 10 carbon atoms, plus 10% of 5-nonyl salicylald-
oxide (a chelating agent), dissolved in a kerosene Dylan. The proper-
lions of the two extract ants coo be varied to compensate for the copper
content in the leach solution, but it appears favorable to use the
9.~23~
-- 14 --
I ox
it ., , a
a _.
i.- O .
g I a
o Jo
it I I
by I
I V I l l
Jo . I I it
I_
it i,. o , o o
o I- ox TV
a Pi
it Do O O
a us O O it I I l O
I
O O I I I I I
an ,-i TV O 'O I I
Z I a Cal Us IT to ED Jo
l I ) I Ed lit Us
Id (T) I okay Jo
Ed rod Cal
a ;~, !~; ox
Jo O I O 0 try
~11 Pi; C_) O t`-) O Roy O us O
I I- O ' O
I- ill O, o
U Jo
Us O O O or i
bSi aye æ z; I go
o o iYi o o i
i-i it
Us I'm it Us it
a) -i l mu
.-i O . i o I
arc
O I-- l I El--
it in
I? Jo Jo it
I 3 D C
o it ox
ri E-l Jo I
Otto I g ,_~ Jo to
Roy i i- r it i
I pa v I
on
ire I ¢ n ¢
I U X O Jo Ox) O
it n _
I
i
- 15 -
carboxylic acid as the major constituent compared to the chelating
reagent. The raffinate, containing about 1 g Curl is recycled to the
initial etching or leaching.
Depending upon the impunities that might be in the etch 801u-
lion, undo which might co-extract to a limited extent, a scrubbing stage
may be added to provide for maximum copper purity. The solvent is then
contacted with the copper sulfite - ll2S0~ spent electrolyte from
electrowinning to provide a strip solution of 40-60 g Curl, the value
depending upon the requirements of the electrowinning circuit. Following
stripping any entrained solvent that might be deleterious to electron
winning is removed, e.g. in a coalesce stage or by passing through a bed
of activated carbon, prior to electrowinning. The stripped solvent is
scrubbed for the removal of sulfite prior to recycling Jo the extraction
stage and the contact of fresh copper etch solution.
In the schematic flow sheet, tile etch feed solution is thus-
treated as being prepared by dissolving electronic copper scrap in aqueous
ammonia and filtering. Such a typical weed may contain 150 g Curl and
160 g C1 AL at pi 8.4. This solution is fed countercurrent to the mixed
extract ant through the extraction (5 mixer settler stages illustrated),
with the raffinate at less than 1 g Curl being recycled to the etch or
leach operation. The loaded solvent is scrubbed (2 stages illustrated)
with a 10 g Curl solution acidified to pi 0.5 to recover ammonia which it
recycled This NO scrub solution can be obtained by diluting a bleed
from the loaded strip liquor and adjusting the pi to 0.5 approximately.
The ammonia-free solvent phase is stripped (3 stages illustrated) with
spent electrolyte from electrowinning typically containing between ~0-40
g Queasily and 150-180 g H2S04/L~ The strip liquor is treated to remove
any entrained organic phase, e.g. with activated carbon, and when fed to
a copper electrowinning stage. As shown, a bleed can be withdrawn from
the electrowinning feed to provide the copper-containlng (10 g Curl)
scrub solution. After stripping, the depleted solvent phase is scrubbed
with anonym chloride solution to remove any sulfate, and the solvent
extract ant recycled as shown with dotted line.
' 35 I,
