Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
11;~3~7
BACKGROUND OF THE INVENTION
This invention relates to the purification of zinc
sulphate solution and, in particular, relates to the removal
of chloride impurity from aqueous solutions of zinc sulphate
intended for use in the electrolytic recovery of zinc.
The presence of chloride in zinc sulphate solution,
which may be introduced into the solution during hydrometallur-
gical treatment of zinc sulphide concentrate, e.g., during
pressure leaching with sulphuric acid, has a deleterious effect
on electrodes during electrolysis. For example, leaching of a
typical Mississippi Valley-type concentrate containing 55~ Zn
and 0.08% Cl in a single pass of return acid from zinc elec-
trolysis may dissolve about 145 mg/L chloride. To avoid
deleterious effects of chloride during electrolysis, this
chloride should not exceed 100 mg/L. Use of return acid
initially containing 100 mg/L chloride would produce a leach
solution containing about 245 mg/L in the next pass and,
without chloride removal, the amount of chloride in solution
would increase as cycling of return acid continues. It is
therefore evident that a process is required in which all the
solution being used in electrolysis is treated to decrease the
chloride to not more than 100 mg/L or a portion of the elec-
trolyte is treated for substantially complete removal of
chloride and then mixed with untreated solution. The latter
is considered to be more economic.
It is noted in United States Patent 1,403,065 that
corrosion of electrodes is largely reduced if chloride is
removed from the electrolyte prior to the electrolysis of
zinc-bearing solution. A method of chloride removal is
provided wherein a portion of solution is withdrawn from the
~.
27
circuit, treated with a soluble silver salt to precipitate
silver chloride, and purified solution is returned to the
circuit. The silver chloride is reduced to metallic silver and
reconverted into a soluble salt for re-use. In more detail,
the portion of zinc-bearing solution to be treated is made
slightly acid with sulphuric acid and finely powdered silver
sulphate is added. The mixture is agitated for about one hour
and then rendered neutral or slightly basic to coagulate silver
chloride. The precipitate is allowed to settle and clear
solution is decanted and treated with zinc dust to remove traces
of silver. Sulphuric acid and zinc dust are added to bottom
solution and silver chloride filter cake to reduce silver
chloride to metallic silver. When reduction is complete, the
precipitate of metallic silver is agitated for some time in
slightly acid solution to ensure complete dissolving of excess
zinc dust. The silver precipitate is washed with water until
free of chloride and then dried and heated with pure sulphuric
acid to 250 - 300C to convert all the silver to sulphate. This
is ground to a fine powder for re-use. Each cycle takes nearly
3 days to complete and, in an ongoing operation, a large
inventory of costly silver is required. In 1936 AI~E Transac-
tions 121, pages 503-4, silver recoveries by this method are
reported to be 97.5 to 98 percent per cycle, i.e., silver losses
are about 2 percent per cycle.
SU~MARY OF THE INVENTION
We have found that chloride impurities can be substan-
tially removed from zinc sulphate solution by reacting the solu-
tion with elemental silver dispersed on a particulate inert
support material and with an oxidizing agent at a temperature
between about 60 and 100C, preferably between about 60 and 80C
-- 2
il~3Z27
to form silver chloride precipitate, separating purified zinc
sulphate solution from the support material containing silver
chloride by filtration and treating the silver chloride with a
reductant in an aqueous alkaline medium to conver the silver to
its elemental state for cyclic re-use in the process. More
particularly, our invention comprises treating zinc sulphate
solution with elemental silver in the presence of an oxidizing
agent in the form of ferric iron, e.g., ferric sulphate, to
precipitate silver chloride and form ferrous sulphate. In a
preferred modification of the invention, a small amount of
ferric iron that remains in the zinc sulphate solution as an
impurity after an iron precipitation step is used as the
oxidant and this ferric iron is maintained in adequate
abundance by stage-wise additions of hydrogen peroxide to
provide almost immediate re-oxidation of ferrous sulphate that
is formed.
It is the principal object of the present invention
to provide an efficient process for the purification of zinc
sulphate solut;on to a desired extent by the removal of chloride
impurity as a silver chloride precipitate with minimum loss of
silver.
DESCRIPTION OF PREFERRED EMBODIMENTS
The process of our invention will now be first
described with reference to a series of small scale tests for
treatment of zinc sulphate solutions containing 140 - 150 g/L
Zn, 3 - 5 g/L free acid as H2SO4 and 300 mg/L Cl with elemental
silver reagent dispersed on a particulate inert support material
in the presence of an oxidant in sufficient amount to provide
stoichiometric oxidation of the silver required to precipitate
-- 3 --
:~`
` `` ` 1~33~:Z7
the chloride as AgCl. Preparation of supported silver reagent
was effected by mixing the support material in water and adding
an appropriate quantity of silver nitrate solution. Sufficient
sodium chloride solution was then added to precipitate the
silver as silver chloride. The slurry was mixed for about
15 minutes and filtered on a pressure filter. The filter cake
was repulped with water and filtered again. The silver was
reduced to the elemental state by reaction of the aqueous
slurry under pressure at 95C with hydrogen in the presence of
sodium hydroxide. Chloride precipitation tests were carried
out in which stoichiometric quantities of silver comprised lO
or 20 percent of the dry weights of the beds containing silver
and support material. The fol]owing support matrix materials
were tested:
Diatomite ~fine powder)
Carbon (-12 + 40 mesh and powdered)
Fused alumina (15Q mesh)
Silicon carbide (180 grit powderl
Chloride precipitation tests were carried out between 60 and
80C with sufficient agitation of slurries to maintain good
suspension of the solids.
Although some differences due to matrix materials
were observed, choice of matrix material was not critical.
In the preparation and regeneration of diatomite beds, reduction
of precipitated silver chloride to elemental silver was more
rapid at 45 psig hydrogen pressure than the reduction of silver
chloride precipitated on fused alumina and silicon carbide beas.
For example, hydrogen regenerations of diatomite beds were
substantially complete in 60 minutes, while regenerations of
fused alumina beds were only about 75% complete in 60 minutes
_ 4 _
27
and about 97% complete in 120 minutes. At 200 psig pressure,
diatomite beds were completely regenerated in 30 minutes. At
45 psig hydrogen pressure, silicon carbide beds were similar to
fused alumina beds, but required 30 minutes at 200 psig
pressure for 97% regeneration.
Moisture retained in filter cakes obtained with fused
alumina and silicon carbide beds was 40 - 60% of the dry weight
of the beds, while, with diatomite beds, filter cake moisture
was about 100% of the dry weight. Zinc sulphate solution was
removed more readily from the silicon carbide and alumina beds.
Weight losses of diatomite beds due to dissolving of
some silica during regeneration in an alkaline medium were
observed. The other materials had an advantage over carbon
which had a tendency to absorb ions other than chloride ions.
In order to precipitate chloride ions as AgCl, an
oxidizing agent is required to oxidize the elemental silver to
the monovalent state. Several oxidizing agents were tested.
Potassium permanganate and ozone were effective oxidants.
However, when they were used, manganese, which is usually
present in the zinc sulphate solution, precipitated as
difficult-to-filter manganic oxides.
Ninety percent of the chloride was removed from
solution containing 300 mg/L Cl by precipitation with silver
at 70C in the presence of 25 g/L potassium peroxidisulphate
(K2S2O8). This method led to accumulation of K2SO4 in the
solution which could not be removed conveniently. With
operation at 70C and use of hydrogen peroxide in the amount of
seven times the stoichiometric requirement to oxidize the
elemental silver combining with chloride, 90 percent of the
chloride was removed from solution containing 300 mg/L Cl in
11;~ 27
20 minlltes. With less hydrogen peroxide, results were erratic.
For example, as little as 28 percent of the chloride was pre-
cipitated when 3 - 5 times the stoichiometric requirement of
hydrogen peroxide for the reaction were used according to
equation:
2Ag + H22 + 2HCl ~ 2 AgCl + 2H2O (1)
Titration of treated solutions with potassium permanganate
showed complete consumption of added hydrogen peroxide within
15 minutes. It was evident that the hydrogen peroxide was
decomposing rapidly in the presence of the elemental silver by
the reaction of equation 2:
- H22 > H20 + /,2 2 (2)
Successive additions of hydrogen peroxide did not result in
adequate precipitation of chloride.
Chloride precipitation using ferric iron as the
oxidant was successfully carried out by the following reaction:
Fe2(SO4)3 + 2Ag + 2HCl - ~ 2AgCl + 2FeSO4 + H2SO4 (3)
For the removal of 300 mg/L chloride, a stoichiometric
requirement of 470 mg/L ferric iron is indicated. With this
amount of ferric iron, 76 percent of the chloride was precipi-
tated in 30 minutes while 83 percent was precipitated in 90
minutes. With the use of twice this stoichiometric amount of
ferric iron, obtained by adding ferric sulphate, 96 percent
of the chloride was removed in 45 minutes.
In downstream neutralization of zinc sulphate solution,
the presence of 100 - 200 mg/L ferric iron is acceptable. It
assists in removing trace impurities such as As, Ge and Se.
Larger amounts of ferric iron tend to impede the fil~ration
which follows the neutralization. A test was made in which
use of 200 mg/L ferric iron, i.e. 0.43 or about one-half the
1133227
stoichiometric requirement for removal of 300 ml/L chloride,
was supplemented by slowly sparging the slurry with an excess
of gaseous oxygen. After a retention time of 90 minutes, only
56 percent of the chloride was precipitated. Tests in which
the 200 mg/L ferric iron were supplemented by stagewise
additions of hydrogen peroxide resulted in the removal of as
much as 85 percen~t of the chlorine within 30 minutes, a third
of the time required as compared to the use of oxygen as an
oxidant. Additional tests revealed that oxidation of ferrous
iron with hydrogen peroxide, according to Equation 4, is
almost quantitative and instantaneous under conditions
prevailing during chloride removal:
2 FeSO4 + H2S4 + H22 > Fe2(So4)3 + 2H2
We then discovered that staged additions of stoichiometric
amounts of hydrogen peroxide during chloride precipitation by
the reaction of Equation 3 resulted in regeneration of ferric
iron in situ by the reaction of Equation 4 before decomposition
of hydrogen peroxide by the reaction of Equation 2 occurred.
This is valuable in the treatment of zinc sulphate solutions
2Q obtained by pressure leaching of zinc sulphide concentrates
wherein solutions containing 100 - 200 mg/L ferric iron are
obtained. By successive regeneration of this ferric iron by
additions of hydrogen peroxide to oxidize ferrous iron formed,
additions of ferric iron that may lead to downstream filtra-
tion problems are not required~
Also in small scale tests, formaldehyde and hydrogen
regenerations of silver chloride containing beds of precipitant
were carried out in aqueous suspensions containing sodium
hydroxide. With formaldehyde treatments at 70C, conversions
to elemental silver were virtually 100 percent within 15 minutes
`` 11;~;~27
according to the following equation:
2 AgCl + HCHO + 2NaOH > 2 Ag + 2NaCl + HCOOH + H2O (5)
Although formaldehyde is an effective reductant of silver
chloride, it is costly and there are environmental problems
associated with the disposal of waste solutions containing
formaldehyde and the formic acid produced by the reaction.
Regeneration with hydrogen at 100C and 45 - 200 psig pressure
in an alkaline solution was found to be reasonably rapid and
produced no noxious by~products. This reaction proceeds
according to the following equation:
2 AgCl + H2 + 2NaOH > 2 Ag + 2NaCl + 2H2O (6,
Larger scale tests were then carried out in which
beds comprising the foregoing matrix materials and 10% or 20%
silver were used cyclically wherein washed filter cakes from
chloride precipitation were repulped to 1.5 litres with water,
sodium hydroxide was added and the slurry was treated in an
autoclave with hydrogen at 95C for re-use. Hydrogen pressures
of 45, 100 and 200 psig were applied for periods of 0.25 to
20 hours in different tests. Treated slurries were filtered
and the beds containing regenerated silver precipitant were
washed with water. In the precipitation of chloride, 25 litre
batches of zinc sulphate solution containing 300 mg/L chloride,
200 mg/L ferric iron and 3 to 5 g/L free acid as H2SO4 were
stirred with beds containing elemental silver for periods of
15, 30 or 60 minutes at temperatures between 60 and 80C during
which time stage-wise additions of stoichiometric quantities of
hydrogen peroxide for the oxidation of FeSO4 formed by Equation
3 were made. The temperature could be raised to the boiling point
of the solution without need of a pressure vessel. At tempera-
tures below 60C, longer reaction times were found to be
11;~3~'7
required. The slurries were then filtered and the filter cakes
were washed and weighed.
In one test, a bed containing 22.75 g elemental
silver in 91 g fine fused alumina matrix material was prepared
as previously stated and added to 25 litres of impure zinc
sulphate solution containing 200 mg/L ferric iron and 300 mg/L
chloride. The slurry was agitated at 70C while 36 ml hydrogen
peroxide solution containing 100 g/L H2O2 were added in stages
over the first 10 minutes of processing. After 30 minutes the
slurry was filtered. The filtrate assayed 15 mg/L chloride,
indicating that 95 percent of the chloride in the
impure solution had been removed. The bed containing the
silver chloride was washed on the filter and then transferred
to an autoclave where the volume was adjusted to 1.5 litre by
addition of water, 12.65 g sodium hydroxide was added and the
autoclave was pressurized with hydrogen gas to 45 psig. The
temperature was raised to 95C and strong agitation was
provided to ensure adequate gas incorporation. Stirrïng was
continued for 2 hours. Then the pressure was released and the
slurry was filtered and washed. Chloride remaining in the bed
was determined to be 0.15 g, i.e., 98% of the chloride retained
by the bed from the previous treatment of the zinc sulphate
solution was removed in the hydrogen regeneration. The bed was
then ready for another cycle of chloride precipitation and
silver regeneration. With the same amount of silver comprising
10% of another fused alumina bed, there were 4 cycles with
45 psig hydrogen pressure during reganeration followed by
2 cycles at 100 psig and one at 200 psig hydrogen pressure.
Regeneration was more rapid at elevated pressures, being
95 percent complete in 15 minutes at 200 psig. Operation
"` 11332~7
with a silicon carbide bed containing 10 percent silver was
similar, with regeneration at 200 psig providing 97 percent
regeneration of the silver in 30 minutes.
In a test wherein 22.75 g elemental silver comprised
10 percent of a 227.5 g bed having a diatomite matrix material,
the bed was mixed with 25 litres impure zinc sulphate solution
containing 200 mg/L ferric iron and 300 mg/L chloride. In a
cyclic operation having 6 chloride removal and silver regenera-
tion steps, chloride removals ranged between 85 percent and
91 percent for the first 4 steps which had 30 minute retention
times, and were 82 percent in the 5th and 6th steps when the
retention time was decreased to 15 minutes. Regeneration with
hydrogen at 45 psig of silver chloride in the bed after each
chloride removal step was 97 percent complete within 45 minutes
when compared with formaldehyde regenerations of portions of
the beds in three of the regeneration steps. FormaIdehyde has
been previously shown to regenerate all the silver chloride in
the bed. In this sequence of chloride removals and silver
regenerations, losses in weight of the diatomite totalling
18 percent of that initially present was believed to be due to
dissolving of part of the diatomite silica in the sodium
hydroxide in the regeneration steps. Therefore it is
advantageous to use more refractory materials such as fused
silica, sand, fused alumina and silicon carbide powder as
matrices for the elemental silver precipitant. These materials
also retain less water than diatomite or carbon and are more
easily washed after precipitation and regeneration.
The foregoing tests indicate that about 90~ of the
chloride in impure zinc sulphate solution containing about
300 mg/L chloride may be removed in about 15 minutes by
-- 10 --
11~3~27
treatment between about 60 and about 100C with vigorous
mixing of a slurry comprising the solution, elemental silver
dispersed on an inert particulate support material and an
oxidizing agent in the form of ferric iron. Less than stoichio-
metric quantities of ferric iron provide effective oxidation
when ferrous iron formed in the reaction is oxidized to ferric
iron by stage-wise additions of hydrogen peroxide during the
mixing period. Silver chloride precipitated in the bed of
support material may be converted to the elemental silver by
hydrogen reduction in an aqueous alkaline medium with vigorous
stirring at elevated temperature and pressure, 95~ regeneration
of elemental silver being obtained in 15 - 30 minutes at about
100C and 40 - 200 psig pressure. The particulate inert
support material may be diatomite, fused silica, sand, fused
alumina or silicon carbide powder. Beds comprising silver
supported on these matrices may be used cyclically. Silver
losses to hydrogen reduction filtrate (purified zinc sulphate
solution) and wash waters have been estimated to be about 0.25
percent per cycle.
Since chloride removal and bed regeneration reactions
are rapid, continuous treatment of zinc sulphate solution
produced for electrolytic recovery of zinc may be carried out
without maintaining a large inventory of silver-containing
reagent.
It will be understood, of course, that modifications
can be made in the embodiments of the invention described
herein without departing from the scope and purview of the
invention as defined by the appended claims.
, . ,
