metals is an alternative to their pyrometallurgical or leaching treatment. It ...... Tuller, W. N. (Editor), The Sulphur Data Book, McGraw-Hill, 1954,. P· 5. 32. Sullivan ...
I
•
EXTRACTION OF COPPER FROM CHALCOPYRITE BY ANODIC DISSOLUTION E. N. Zeygolis* and S. R. B. Cooke** .ABSTRACT The extraction of copper from chalcopyrite (CuFeS ) by anodic dissolution in an acid electrolyte was studied. ·2The chalcopyrite anodes were massive and not in powder form, with a polished free surface area of 1 to 3 cm2. When an anodic current passed the chalcopyrite-electrolyte interface, the mineral decomposed uniformly, without forming intermediate sulfides. and the copper. iron, and sulfur were extracted in the atomic ratios 1:1:2, respectively. For the following net half-cell anodic dissolution reaction, Faraday's laws were found to apply to chalcopyrite at 100 percent anodic current efficiency and up to about 65.0 ma/cm2 anodic current density: CuFeS
2
+ ap+ ~ Cu 2+ + Fe 2+ + 2S 0 +be +
where a + b = 4 and p represents a hole. The extracted sulfur was orthorhombic with the exception of a very small fraction which was amorphous. The origin of ferric ion (Fe3+) detected in the electrolyte was not e'stablished. A comparison of the electrolytic and the auxiliary sulfuric acid-ferric sulfate leaching experiments showed that the dissolution rate by electrolysis was much greater than by acid leaching at room temperature and atmos.p heric pressure. The leachfng experiments alone confirmed other work which showed that chalcopyrite w~ very resistant to leaching at room temperature and pressure, and that it dissolVed re.adily at 180°C and at elevated pressures.
* E. N. Zevgolis,...,;.formerly Teachi_ng Associate, Department Metallurgical Engineering, University of Minnesota, is with Hazen Research Inc., G'olden, Colorado. ** S. R. B. Cooke, formerly Professor, Department of Metallurgical EngineeJ is now Professor, Department of Geology and Geophysics, University of Minnesota, Minneapolis, Minnesota ·
36
INTRODUCTION Electrolysis of mattes or natural sulfides for the recovery of their metals is an alternative to their pyrometallurgical or leaching treatment. It has been applied mostly to mattes, however, and not to natural sulfides because a molten matte may be cast into a mold to form a co~pact electrode with a rather well-defined conductivity, a relatively uncomplicated chemical behavior, and a simplified handling. On the other hand, natural sulfides raise the problems of electrode preparation, handling, heterogeneity of composition, and a wide fluctuation of electrical conductivity. To date, the only successful commercial applications of matte electrolysis are those employing nickel-matte anodes, as used by the International Nickel Company of Canada Ltd. (INCO)l- 4 and the Japanese Shimura Kako Ltd.s In 1966, Habashi6 prepared a survey on matte electrolysis and subsequently a few more articles have appeared in the literature 7 •8. Experimental electrolysis of mechanically compacted natural sulfides began a few years ago (1961) in Japan and work on lead sulfides, zinc sulfide, 1 0, and chalcopyrite 11 •12 concentrates has been reported. The present work was performed on specimens of shaped massive chalcopyrite. The main objective of the present work was to examine the possibility of extracting copper from chalcopyrite, by passing a current through this mineral a~ the anode of an electrolytic cell in an acid electrolyte. By knowin2 the electrolytic behavior of individual sulfides one might be successful in predicting and controlling the electrolysis of a sulfide concentrate or a matte. Calcopyrite dissolution, either by electrolyis or leaching, involves a solid/liquid system. The best approach ~o such a system consists of three steps. First, the general picture of dissolution of chalcopyrite by the so-called "wet" methods is obtained by considering the physical chemistry of dissolution in an aqueous electrolytic environment. Second, the electro-chemical. ,properties of the chalcopyrite-electrolyte interface are studied. Third, the electrolyis of the chalcopyrite is examined. PHYSICAL CHEMISTRY OF THE HYDROMETALLURGICAL DISSOLUTION OF CHALCOPYRITE A basic understanding of the chemical or electrochemical reactions during the dissolution of sulfides is obtained by using thermodynamics and chemical or electrochemical kinetics .
..
Pictorialty, the thermodynamics of a sulfide-water system (MeS-H 0) can be 2 expressed in terms of Eh and pH diagrams. Garrels and Christ 13 constructed such an hh-pH diagram for the system Cu-Fe-S- H 0, but the necessary data for chalcopyrite 2 were estimate~ deduced from a pressure diagram of the system Cu-Fe-S 2-o 2 which 4 they also constructedl . However, they suggest that their pressure diagram be used only as a "working model" since their estimates are not reliable. Peters and MajimalS give the Eh-pH diagram for the same system, but since the origin ·of their
thermodyn~ic data was riot indicated, the free energy value for chalcopyrite, tlG 0 f(CuFeS ), was deduced from their paper in the 'f ollowing manner . The electro2 chemical reactionlS
CuFeS corresponds to Eh 0
=
0.29 volts. Eho
0
.tlG f (CuS)
or 0.29 = or 0 . 29
=
- CuS 2 -
+
+
Fe
2+
+ S0
orthorhombic
+
2e-
( 1)
Hence
= .tlGo
(2)
nP
0 2 0 0 .tlG f(Fe +) + tlG f(S ) (2) (23.06)
-
0
tlG f(CuFeS ) 2
(3)
-11.7- 20.3 + 0.0- .tlG 0 £(CuFeS ) 2 (2) ( 23 . 06)
0 and .tlG f(CuFeS 2) - 45.37 kal/mole. In a later publication, Peters 1S gave 0 . tlG £CCuFeS 2) = -4 2. 79 kcal/mole as being derived from the previously mentioned work of Garrels and Christ 1 j .
Sogina et a1. 17 suggested two forms of chalcopyrite as follows:
Chalcopyrite
I (CuS.FeS)
Atomic ~arge (pet) 2 ·cu + = 100 Cu+ =
100
0
tlG f (kcal/mol e) -35.0 -84.0
By using each of the above free energy values for chalcopyrite , ~he Eh-pH diagram of the system Cu-Fe-S-H2 0) has been constructed, as shown in Figu~ 1, with 0 lines 8, 8', 8" corresponding to 6G f(CuFeS2) = -35 . 0, -42.0 = C.-l!!) and -45.37 kcal/mole respectively. The detai Is of its construction are given elsewhere,19 • Since the diagram has been constructed by using thermodynamic data 1 7•18, the information it provides is limited; that is, although it shows the areas of stability, or instability, of the different components, it does not tell us anything about (1) the rate of a reaction, nor since Eh is a state function, should a reaction occur we know nothing about (2) the path of the reaction, i.e. , its partial steps - ·· Al t hough the boundary lines in Figure 1 have limited accuracy due to the assumptions made, one can easily see the possible oxidation or reduction products as functions of Eh and pH for the specified system. For example, in the dissolution of chalcopyrite without Fez03 precipitation, the conditions within the upper shaded area delimited by lines 22,24,25,31, and 34 must be satisfied.
38
0 .6
0.4
(f)
!J 0 > z
'' -0.2
''
''
' ' ®, ' ~ '4
0
-0.4
-0.6
r
Fe2+, H S ,Cu 2
0
0
' '-
c-2•, Fe
H.?S, Cu2 S
~ ___
'
'-
Q)r-----~'-.
Fe .• H2 S, Cu
~---'-----..1..'----"------'-1
0
'
-··
3
2
4
5
pH
FIGURE 1. Eh-pH DIAGRAM OF THE SYST£t.i CU. Fe-S-H 20
Temperature
= 25°C
Pressure
= 1 atmospher e
6
39
This diagram is similar to that of Peters and MajimalS except that there is no field for bornite (CusFeS 4 ), since the available data for this mineral 17 give lines coinciding with that of chalcopyrite as shown by the following equations: 2 (S) Line 9: 5Cu 2S + 2Fe + + 3H2S = 2Cu FeS + 6H+ + 2e 5 4 Line 10:
2+ cu Fes + 4Fe·+ 6H S 5 4 2
Line 11:
CuSFeS 4 + H2S
= SCuS
with AG 0 f(Cu FeS ) 5 4 Line
8:
Cu 2S + 2Fe
2+
= 5CuFeS 2
+ Fe
2+
= -76.2
+ 2H
+ 12H+ + 4e-
+
+ 4e
-
(6)
(7)
kcal/molel7, and
+ 3H S = 2CuFeS 2 2
= 6H+
+ 2e-
(8)
0
with AG f(CuFeS 2) = -35.0 kcal/mole. Figure 1 is valid for a deaerated sys tem. Furthermore, pyrite (FeS 2) and digenite (Cu S ) have not been taken into account. 9 5 As has been indicated, thermodynamics alone is insufficient to help understand the dissolution of chalcopyrite. In fact, to find reaction rates and propose mechanisms one needs the help of kinetics. Kinetics will determine (1) how the different factors affect the hyd~ometall~rgical extraction of a . given metal from a given SUlfide and (2) how much each Of these factors affects the extraction. In other words, kinetics attempts to provide the answers to such questions as: I
(A) How does an acid solution accelerate the leaching, · i.e:: what is the role of the hydrogen ion? Also, what is the relationship between acidity and dissolution rate? The Eh ~ diagram shows that an acidic (and oxidative) environment is needed for dissolution of chalcopyrite, but it does not give any information about acceleration of dissolution with increasing acidity 2 0. (B) Do electrochemical reactions contribute significant ly to sulfide dissqlution, and if so, what are these reactions and to what extentd o they affect the dissolution rate? Can these electroehemical reaction(s) in a given sulfide be accelerated and to what extent by imposing an external current so that dissolution becomes a matter of practical value? (C) What is the effect of the electronic structure of sulfides on their surface reactivity, i.e., does it make any difference-for the anodic dissolution of chalcopyrite--if the mineral is an n-type or p-type semiconductor?
40
(D)' In what fonn does the extracted sulfur appear and how may it affect the rate of dissolution?
(E) Does chalcop~ite decompose in steps as in the case of chalcocite (Cu S) ,2l,22 or is there only a single step? 2 The foregoing are some of the questions the metallurgist faces. This paper represents an attempt to ans~er such questions in the dissolution of chalcopyrite . The Relationship between Leaching and Electrolyis of Sulfides The current thinking regarding oxidation mechanisms responsible for the leaching of most sulfides in aqueous solutions accepts the e l ectrochemical approach 1 5, 2 3. This is summarized by the following net oxidation reaction: MeS
~
Me 2+ + S0 + 2e -
(9)
Assuming that this prevails in sulfide dissolution, then the conditions for this reaction to proceed are: (1) enough energy to break the bonds among the atoms of the sulfide and (2) a means to carry the electrons away from the mineral. I n leaching chalcopyrite, for example, the oxidizing reagent would provide the energy required for the breaking of the lattice bonds and would also carry the electrons of the reaction away from the mineral . Thus, in the reaction: CuFeS
2
~
Cu 2+ + Fe 2+ + 2S 0 + 4e -
(10)
oxidants such as Fe 3 + or 0 would provide the energy and would "pick up" the electrons released, accordtng to the equations: llG
or
2H+ + 1/2 0
2
1
< 0
+ 2e- ~ H 0 2
(11) (12)
Incidentally, Equation 12 may explain how a "more acidic" solution accelerates the leaching (Item A above). The two requirements for dissolution stated above, do not place any constraints either upon the source of energy or upon the means of the electron transfer; th~J instead of having the oxidant (solvent) in solution, the requisite energy could be provided through an external source, i.e. , a potential difference, and the removal of the electrons via an external circuit , al ways provided that the mineral is sufficiently conductive. In other wor ds, the principles remain the same; the difference between leaching and electrolysis consists of the manner in which the required energy is provided and the mode
41
of the electr6rt transfer. By' making a sulfide the anode of an electrolytic cell, the electrowinning process would be represented as follows: Anode:
(9)
Me 2+
Cathode:
Net Reaction:
+
2e-
MeS
-+
-+
Me 0
Me 0
+
(13)
s0
(14)
As stated in the introduction, the electrowinning approach has been applied in a number of cases but without correlation with leaching. In summary, and with respect to the basic principles governing the dissolution of sulfides, there exists in most cases a direct correlation between leaching and electrolysis . Part of the objective of this work was to study how far these principles can be applied to the dissolution of chalcopyrite, i.e., to ascertain if it is possible to dissolve the sulfide by applying a potential difference across the chalcopyrite-electrolyte (Cp/El) interface and removing the electrons via the mineral . The Chalcopyrite
· vs~ietal
Cathode Galvanic Cell
Assume the following galvanic cell, which is also shown schematically in Figure 2: 5 1' 2 3 4 1 Me I Me I Me II Chalcopyrite Electrolyte Me III Cp El Cu Cu Cu Hg Figure 2(b) shows the phase sequence and Figure 2(c) the galvanic cell voltage which is the potential difference between the two terminal phases 1 and. 1' , i.e., £
= ~~
= ~Mei(l)
- ¢Mei(l 1 )
(15)
Figure 2(c) also shows the discontinuities of electrical potentials ~-. between two consecutive phases i, j; the signs and magnitudes of these discontiftrtities are arbitrary, because the galvanj c potentj al di ffe:rence of two different phases has never been meas~red 24 and only the difference:
-··
£
=
6~
=
cell voltage
(15)
is a measurable quantity. However, although the potential~, i.e., the galvani ~ or inner potential , is constant within a phase, its change 6~ from one phase to another is not as sharp as shown in Figure 2(c), due to (1) the space charge region 25which is the portion of electrical double layer within the solid side
2
_Me I
Men
Me I
Mem
CHALCOPYRITE (C P}
@ ......._-CONTAINER
1ELECTROLYTE
2
Mel
Mel!
3
4
Cp
E.l
( Ei}
5
I'
I
Meml
Mel
------~----------~------------
r
I
I
---tl- - - ~ - (/)
!J
'--
I I I I
_J -
I I
-
-
-
..J - - I
I
I _J -
-
- ...~ - .•
I
~=A4>
I
:!
="-
'
0
~ z
-&-
X FIGURE 2.
PHASE AND POTENTIAL DIAGRAM OF THE GALVANIC CELL (SCHEMATIC)
43
of an interface and (2) the compact and diffuse portion of it within the liquid side of an interface when a liquid phase exists, but at present it would not be of any importance to present them in Figure 2. A metal Me III or another auxiliary electrode is necessary in order to close the electric circuit.
. . PROPERTIES OF THE CHALCOPYRITE-ELECTROLYTE INTERFACE Among the five interfaces of Figure 2(b), the most important ones are the chalcopyrite-electrolyte and the · e1ectrolyte-copper where, respectively, dissolution and deposition occur; our interest is at the chalcopyrite-electrolyte interface and we shall concentrate on it. As was pointed out in the preceding article, chalcopyrite is an n-type semi conductor, the entire chalcopyrite-·e lectrolyte system is similar to oth er n-type semiconductor-electrolyte systems, and the anodic electrochemical behavior of chalcopyrite resembles the behavior of other n-type semiconductors. As expected then, a limiting anodic current il1 m appears in chalcopyrite at the exhaustion point of th'e free positive carriers (the holes). This is shown in Figure 3, where the i vs ECp curve can be expressed by the equation:
c
+
i RT i3F ln -:-1
RT ln(l F
0
where
c
a
-
i
.lim)
(16)
1
= a constant
the transfer coefficient (Ot::
(/)
z
UJ 0 ~
z
w
0: 0:
:;)
u u
0 0
z
ct
0 .1 4 6 8 !0 ANODIC POTENTIAL (Ecpl vs S.C.E. IN VOLTS
FIGURE 3.
ANODIC POTENTIAL VS CURRENT DENSITY FOR THE CHALCOPYRITE-ELECTROLYTE SYSTEM
12
45 ...
We notice t~at, in genera}, the space charge region has been very much neglected as a part of the electrical double layer in the flotation and hydrometallurgy of minerais; and (3) that the rate-controlling step when an anodic current passes the interface ~s either the cha·rge transfer across the space charge region (when i < il1m) or the hole supply at the interface (when i = ilim). An extensive discussion along these lines is given in a separate paper27
ELECTROLYSIS STUDIES According to :Equation 14 the electrowinning of elements from sulfides seems to be very simple, at least in principle, resulting in metal depcsition at the cathode and elemental sulfur at the anode. Furthermore, the practical reasons supporting res earch on chalcopyrite electrolysis are: (I) Conventional smelting of sul f ides faces a steadily growing pollution problem, due to the sulfur dioxide produced. ( 2) As will be shown later ~ leaching of chalcopyrite at low temperture and pressure by sulfuric·acid~ferric sulfate solutions is yery slow. At high temperature and pressure, it is greatly accelerated and some variation of this procedure coul d successfully be applied.
(3)
Electrolysis bYPasses several steps neces s ary in smelting.
(4)
Chalcopyrite has almost metallic conductivity.
(5) When the extracted sulfur is in elemental form, its handling is much easier and cheaper than that of sulfur dioxide and sulfuric acid.
EXPERIMENTAL The chalcopyrite electrode preparation, the experimental assembly~ the procedure and the analytical methods employed are described els~~here! 9 ,L 7 .
RESULTS AND DISCUSSION
-··
All the preliminary experiments, which were qualitative in nature, were positive. In other words, when an anodic current was passed across the chalcopyrite-electrolyte interface,i.e., positive charge-carriers from the solid to the electrolyte, the chalcopyrite dissolved. Most of the copper was deposited on the copper cathode and the iron remained in solution, in agreement with their standard electrode potentials. The extracted sulfur ·remained as a coati.ng at the anode.
6
~
In, the subse~uen~~quantitative experiments, the solutions were analyzed for Cu+, Cu2+, Fe+, and Fe 3+, and the quantity of elemental sulfur rema1n1ng on the anode was determined. These data were essential to determine (1) the dissolution products (2) the stoichiometry of the net half-cell anodic reaction and (3) the dissolution rate. (1) The Dissolution Products: No cu+ was ever detected in the electrolyte using the 2-2'-biquinoline method 29 . Copper was alwabs present as Cu2 + and was ~etermined by the diethyi .dithiocarbamate method3 . Ferric (Fe3+) and ferrous (Fe 2+) ions were both present and were determined by the 0- phenanthrol i ne method after a minor modification 19 . According to its X-ray diffr acti on pattern, the elemental sulfur was orthorhombic but it also contained a minor amount of amorphous sulfur which was detected by its insolubility in carbon disulfide . The primary dissolution products were: Cu2 +, Fe~+, and S0 ; it is not clear whether Fe3~ which appeared in solution in all experiments, is a primary or secondary oxidation product, that is, if it comes dir ectly from the mineral or from the oxidation of Fe 2+ after it has passed in the electr olyte. (2) The Stoichiometry of the Net Half-Cell Anodic Reaction : To find the stoi chiometry of the net anodic reaction, the amounts of copper, iron, and sulfur extracted were needed. Since part of the copper stayed in solution and part of it deposited on the cathode, a simple way to determine the total extracted copper was t o el i minate the cathode in the electrolytic cell by using a salt bridge and an auxiliary electrode in a second container, so that all copper and iron remained in solution, in the reaction cell~ The results so obtained are given in Table l. The atomic ratio for Cu2 + :Fe + was cl ose to 1:1. The amounts of copper and iron extracted are. given in F.igure 4. In another experiment, the elemental orthorhombic sulfur was extracted in carbon disulfide, the residual amorphous sulfur was crystallized b~ heating to 105° C (the transformation point of amorphous to monoclinic sulfur 1 ) and re-extracted in carbon disulfide. Evaporation of the carbon disulfi de left a residue of clean sulfur which was weighed. The atomic ratio Cu:S was found to be approximately 1:2. Hence, the atomic ratios are Cu:Fe:S = i:l : 2. In additi on, microscopic examination of the anodes after electrolysis did not show any other copper sulfides, and other l eaching studies32 ha~shown a homogeneous dissolution for chalcopyrite. Hence, the decomposition of chalcopyrite by electrolysis is homogeneous and not in steps as in the case of chalcocite (.Cu s)?•2.1•22•32. 2 Since holes participate in anodic chalcopyrite dissolution27 , the mineral decomposes as follows: CuFeS 2 + ap +
~
Cu 2"' + Fe 2+ + 2S 0 + be
(18)
where a + b = 4 and p+ represents a hole. Ferric ion (Fe 3+) does not appear in Equation 18 because, as will be shown later, the anodic current efficiency is 100 percent, assuming only Cu 2 + and Fe2 + produced . If chalcopyrite were of the form Cu 2S·Fe 2S3 17 , that is, if the iron were in the ferric state, then the copper appearing in solution should be in the cuprous state. This was not the case. It is therefore suggested that the Fe 3+ found in the electrolyte
47 TABLE 1.
..
ATOMIC RATIO Cu2+:Fetot DURING ANOQIC DISSOLUTION OF CHALCOPYRITE TempeTature = 30°C . Initial anodic cuTrent intensity = 42 milliamperes Anodic surface area~ 6 . 46 square centimeters
13(mg) Time Cu2 +
(Hours)
Fe tot
Cu 2+ : Fe tot Atomic Ratio
0.25 0.50
2. 72 4.98
1.99
1. 20
3 . 45
1. 27
0.75
6.92
6.36
0.96
1.00
9.75
7.81
1.10
1. 25
11.38
9.81
1. 02
1.50
13. 72
12.06
1. 00
1. 75
14.37 16.69
0.96
2.00
15.71 19.60
2.25
21.50
18.06
1. 02
2.50
22.84
20.76
0.96
2.75
2S . 77
22 . 08
1.02
3.00
28.1S
23.41
1. OS
3.25
30.40
26.06
1.02
3.50
35.45
28.S9
1.09
3.75
34 . 30
30.29
0.99
4.00
36.30
27.78
1.14
6.00
5S.08
46 . 03
LOS
8. 00
75.S O
62 .49
1.06
10.00
99.54
80.22
1.09
12.00
124 . 51
96.22
1.14
14.00
142.55
111.00
1.13
22.33
217 . 80
169.02
221.30
176.60
1.13 1.10
219.70
183.56
1. OS
28.00
225.08
188 . 52
1. OS
29.75
235 . 92
191.95
1. 08
31.58
231.32
211. 52
0.96
36.33
233.87
189.10
1. 08
24.00 26 . 00
1. 03
~
~.
48
-
40 E= 12,volts T=30°C
0'
E
Cu
c w 30 ~ u
2•
l
~ 0
0
~
10
00
2 t (h)
4
3
E = 12 volts T= 30°C ~200
E
0
w
~
u
0 ~
0
:E
