The present technology for processing copper sulfide ores involves pyro-metallurgical (smelting) and to some degree hydrometallurgical techniques. Although these techniques are efficient and economical, they cause air and water pollution. Environmental concern has led the Bureau of Mines to investigate an anhydrous chlorination technique to extract Cu, Fe, and elemental S from sulfide ore without formation of slags and offgases containing sulfur dioxide.

A chlorination process to extract lead and zinc from metal sulfide ores and concentrates was commercially developed in 1897 by Swinburne and Ashcroft. However, the chlorination process was not competitive with the simultaneously developed pyrometallurgical techniques.

Several recent references deal with the extraction of copper from sulfide ores by chlorination. Mervin and Donald White described treating pulverized copper ore with excess chlorine in a closed reactor at 300° to 700° C. The FeCl3 and S2Cl2 byproducts were removed by volatilization, and the residual product was leached with water to obtain a CuCl2 solution from which copper was electrowon. Spreckelmyer treated a mixture of pyrite, gangue materials (quartz, sand, or coke), and copper ore with chlorine to facilitate conversion of CuS to CuCl2. They performed the reaction in a closed tube using excess chlorine at temperatures between 300° and 400° C. After the FeCl3 and S2Cl2 were volatilized, the residue was leached with water and the copper was recovered from solution by cementation. Polinsky reacted sulfide ore with S2Cl2 dissolved in dichloroethane at 75° C. The resulting slurry consisted of insoluble iron-copper chlorides, gangue, and dissolved sulfur in dichloroethane. Subsequently the metal chlorides (solids) were filtered and heated at 300° C to volatilize FeCl3. The residue containing CuCl2 was leached with water, and the copper was recovered by cementation. The spent dichloroethane was distilled at 125° C to recover elemental sulfur. All these disclosures differ primarily in the chlorination procedure used in converting copper sulfide ores to metal chlorides. Additional steps were necessary to recover elemental sulfur. Adaptation of these processes to continuous operation appears to be very difficult.

The Bureau-developed process to extract copper from chalcopyrite involves three major steps as shown in figure 1.

Chalcopyrite is reacted with a controlled flow of chlorine at temperatures above the boiling point of sulfur (445° C) to form metal chlorides and elemental sulfur.

CuFeS2 + Cl2 → [CuCl, CuCl2, FeCl2, FeCl3] + S………………………………………(1)

The resulting sulfur is distilled from the reactor and recovered in elemental form while the molten chlorides drain from the bottom of the reactor.

The mixed metal chlorides in step 1 are pulverized and treated with O2. This results in a preferential oxidation of iron chloride while copper chlorides remain unaffected.

2FeCl3 + 3/2 O2 → Fe2O3 + 3Cl2……………………………………………(2)

2FeCl2 + 3/2 O2 → Fe2O3 + 2Cl2……………………………………………(3)

The free chlorine produced by this reaction is recovered and recycled to the chlorination step. The copper chlorides are separated from iron oxide by leaching and filtration.

The resulting copper chloride solution is electrolyzed to produce elemental copper and chlorine.

CuCl2 → Cu + Cl2…………………………………………………………………..(4)

2CuCl → 2Cu + Cl2………………………………………………………………..(5)

Again, the free chlorine produced in electrolysis can be recycled to chlorination in step 1. The main advantages of this chlorination technique are that usable elemental sulfur and iron oxide coproducts are produced while chlorine is recovered and recycled.

Preliminary work on the overall three-step process has been reported previously. This report concerns only the initial step; that is, the anhydrous chlorination of chalcopyrite. The second and third steps have also been reported separately.

In an effort to make an efficient chlorination step, a continuous operation was developed. This involves using a vertical shaft reactor equipped with a grid to support a bed of chalcopyrite pellets. Chlorine is continuously introduced into the bottom of the bed where the reaction takes place. The resulting copper-iron chlorides form a molten, complex mixture, which drains through the grid. Excess chlorine present in the lower reaction bed forms sulfur chloride (S2Cl2), which volatilizes along with a small quantity of excess FeCl3. However, these coproducts rise through the bed and become secondary chlorinating agents for the fresh pellets at the top of the bed. The flow of chlorine is adjusted to allow adequate reaction and to control the overall chlorination so that only elemental sulfur is vaporized.

Equipment and Operating Procedures

Two types of chlorination apparatus each having a similar configuration were used. Initially a 10-cm-ID glass reactor was used to perform most of the experiments to determine the effects of operating variables. Next, eight reactors of 15 cm ID each were built from various commercial refractory materials and operated under production conditions.

10-cm Reactor

A schematic diagram of the 10-cm reactor is shown in figure 2. The upper large section, which contains the reaction bed, measures 10 cm ID by 35 cm long. A grid having 0.47-cm spacings and located in the bottom of the reactor supports the pellet bed. The section below the grid serves as a passage through which molten metal chlorides drain and enter the receiver below. A tube serving as a chlorine line and preheater is placed alongside the reactor. The upper section of the reactor consists of a pellet feed port and a side arm through which sulfur vapor is directed to a condenser and receiver. The reactor was made of Vycor, except for the chloride receiver and the sulfur condenser, both of which were made of Pyrex.

The chalcopyrite concentrate used for the experiments discussed in this report was obtained from the ASARCO smelter in Tacoma, Wash. Chemical analyses of the concentrate are shown in table 1. Screen analyses of this concentrate is shown in table 2.

In preparing pelletized feed, the chalcopyrite concentrate required an additional binder in order to form strong pellets. Consequently, starch, sugar, and sulfur were tested as pellet binders. Of these, sulfur was chosen for several reasons: (1) It forms an excellent binder, (2) it is compatible with the sulfide mineral and does not cause complications during chlorination, and (3) it is recovered along with the sulfur extracted from the sulfide mineral. Thus the pelletized feed used for the experiments contained sulfur additive , usually 5 pct.

These feed pellets , ranging from 0.6 to 1.9 cm diam, were prepared by adding sufficient water to the chalcopyrite-sulfur mixture on a rotating disk. The wet pellets were dried in an oven at approximately 60° C for 16 to 24 hr before using. However, during the final hour of drying, the oven temperature was increased from 60° to 120° C. The purpose was to melt the sulfur additive so that upon cooling, the binder cemented the mineral grains and formed hard, strong pellets. Chemical analyses of pelletized feed containing 5 pct sulfur additive is shown in table 1.

Prior to chlorination, the reactor was partially filled with pellets to approximately three-fourths of the selected bed height, which ranged from 10 to 30 cm. Nitrogen was used to purge the preheated bed to prevent pellet oxidation. After the operating temperature was attained, chlorine was introduced followed by filling the reactor with additional pellets to the desired height. During chlorination, the reaction bed slowly moved downward as the molten chlorides drained from the bed into the receiver below while sulfur distilled and passed into the condenser. Continuous chlorination was sustained by replenishing the bed with fresh pellets at 10- to 15-min intervals and by replacing the filled product-receivers. In most cases the reaction bed self-moved throughout the entire run. Occasionally obstructions occurred but were removed by light poking with a small metal rod.

A continuous chlorination experiment usually lasted 7 to 8 hr. Steady-state conditions were considered attained after ½ to 1 hr of operation when products were produced at a uniform rate. Occasionally after the completed experiment, the reactor was filled with nitrogen and held overnight at 300° C. On the following day, it was heated to the desired temperature and chlorine flow resumed. Continuous chlorination then proceeded as before but under a different set of operating conditions.

The metal chlorides were transferred from the receiver, allowed to cool to room temperature, ground, and sampled for chemical analyses. Handling and grinding of these metal chlorides were carried out under a nitrogen atmosphere to prevent moisture pickup. The molten sulfur that collected in a glass flask maintained at about 130° C was periodically poured into a Teflon-lined pan and cooled. The solidified sulfur was also pulverized and sampled for analyses.

15-cm Reactor

A total of eight reactors were built from the following list of commercially available refractory materials: Aztex brick, Puro-Tab castable, Puro-Cast castable, Mono-90 ramming mix, Hi Ram-80 ramming mix, and Green Cast-12 castable. The first five materials were manufactured by Kaiser Refractories; the last one, by A. P. Green. The size and shape of these eight reactors were similar but varied in height from 71 to 76 cm.

The Aztex brick reactor was constructed using Kaiser’s K-16, a special chlorine-resistant mortar. The refractory castables were mixed with specific amounts of water in a 100-lb capacity mixer and then poured into an agitated mold set on a platform vibrator. When working with the plastic ramming refractories, a pneumatic hammer was used in packing the materials into a mold. A schematic diagram of the latest reactor and its components is shown in figure 3.

The reactor measured 15 cm ID by 76 cm high with 7.5-cm wall thickness. Its upper and lower wall sections were heated by electric element while the midsection of the reaction zone was left unheated. The entire outer wall was insulated with 15 cm of vermiculite. Six thermocouples were evenly spaced and imbedded in the reactor wall to monitor the reactor temperatures during operation. A separate steel retort that measured 7.5 cm ID by 30 cm high was mounted above the reactor to serve as a preheater for the feed pellets. Exit gases from the reactor were directed into this preheater to prevent the feed pellets from oxidation. The metal chloride receiver was fabricated from mild steel and its inner surface was glazed with blue enamel. A rubber gasket formed a seal between the receiver and the reactor base. A small opening in the receiver provided a chlorine inlet to the reaction chamber.

Different approaches to solving the bed support problem were tried. The first six reactors were built with a cone-shaped, refractory hearth to support the pellet bed. Generally the hearth had a 5-cm-diam opening that was covered with either a small glass grid or with an inverted and slotted porcelain dish to prevent pellets from falling through. In another case a hearth with an uncovered 0.9-cm-diam opening was used. Although the hearth afforded simple construction and greater strength in supporting a heavy column of pelletized mineral, it was proven unsuitable in extended operation because of the restricted chloride outlet, which caused inconsistency in the free flow of chloride products. Consequently, a full-size grid similar to that of the 10-cm reactor was built and placed above the hearth as means to support the bed for the final two reactors.

The grid that supports the bed charge was constructed of high-density alumina brick sawed into slices measuring 1.25 cm wide by 5 cm high. Each slice was anchored to the reactor wall by a rammed refractory and supported by the original hearth with 5-cm opening. A 1-cm spacing was left between the slices to facilitate drainage of the metal chloride product.

The operating procedure for the 15-cm reactor was essentially similar to that described earlier for the 10-cm reactor except the preheated pellets (approximately 400° C) were fed to the reactor. Feed for this series of tests usually contained 5 pct sulfur as binder, but in some cases 7½ pct sulfur additive was used. In conducting tests with the 15-cm reactor, frequent poking was required to maintain bed movement.

Experimental Results and Discussion

10-cm Reactor

Tests in the 10-cm reactor were made to study the effects of temperatures, chlorine flow rate, bed height, and pellet size on the extraction of elemental S, Fe, and Cu from chalcopyrite. The effects of these variables were measured by the amounts of metallic impurities in the resulting sulfur product, the amounts of unreacted mineral in the chloride product, and the species of copper and iron chlorides obtained. The criterion for an efficient operation was essentially complete extraction and separation of elemental sulfur and of the chloride products.

The melting and boiling points of various metal chlorides that are extracted from chalcopyrite are shown in table 3. Operating conditions and analytical results are shown in table 4. Pertinent data are plotted in figures 4 to 7 and show the effects of operating conditions on elemental sulfur and metal chloride product.

The percentage of Bi, Zn, Fe , and Cu extracted from the feed material and condensed with the sulfur product are shown in figure 4. Increasing both the temperature from 550° to 750° C and the chlorine flow rate from 3 to 5 l/min generally increased the percentage of these metallic impurities extracted from the feed. Bismuth was increased from approximately 40 to 60 pct; zinc, from 0.3 to 9 pct; iron, from 0.03 to 0.3 pct; and copper, from 0.02 to slightly over 0.1 pct at 650° C but dropped slightly below 0.1 pct at 750° C.

The high level of bismuth extracted and transferred to the sulfur product is expected owing to the relatively low boiling point of BiCl3 (447° C). However, incomplete volatilization of BiCl3 and ZnCl2 at temperatures above their boiling points indicates the formation of metal chloride complexes of lower volatility. Copper and iron chlorides also form stable complexes, which, along with the ability of FeCl3 to act as a chlorinating agent, keep copper and iron impurities well below 1 pct in the sulfur product.

Thermal decomposition of CuCl2 may be the reason for the lower copper impurity in sulfur at 750° C than at 650° C. The CuCl3 alone decomposes at 993° C; however, it may well react with chalcopyrite at 750° C to form CuCl, which has a boiling point of 1,490° C.

Figure 5 shows that increasing the bed height from 10 to 30 cm lowered the percentage of Bi, Zn, Fe , and Cu extracted with sulfur product. Bismuth was lowered from approximately 40 to 30 pct; zinc, from 3.0 to 0.25 pct; iron, from 0.35 to 0.05 pct; and copper, from 0.3 to 0.03 pct. The improved quality of the sulfur product is most likely due to increased contact time between metal chlorides and chalcopyrite pellets in the deeper bed.

The presence of metal chlorides in the sulfur product does not create an operating problem unless their quantity is sufficient to plug the reactor outlet or coat the condenser wall and impair heat transfer. These metal chlorides are insoluble in molten sulfur. They were easily removed by filtration through coarse paper at 120° to 130° C. As shown in table 4, analyses indicate that chloride content was lowered by a factor of approximately 100 after the raw sulfur was filtered. Thus the purity of the filtered sulfur was increased to a value in excess of 99 pct.

A small exhaust stream, usually amounting to about 10 pct of the volume of chlorine input, was given off during chlorination. Analyses showed that these exhaust gases consisted of 80 to 90 vol-pct SO2 with the balance being nitrogen and moisture. Entrained air, residual moisture, and slight oxidation of the pelletized feed cause 5 to 10 pct loss of sulfur as SO2. However, careful drying of the pelletized feed in an industrial operation would minimize SO2 formation. In addition, the small quantity of SO2 generated could easily be scrubbed.

The extent of chlorination of chalcopyrite is determined by the analyses of sulfur left in the metal chloride product. One pct S in the metal chloride fraction corresponds to a 96-pct conversion of chalcopyrite to copper and iron chlorides. As shown in table 4, over 96 pct conversion is obtained as less than 1 pct S is found in most cases.

Small pellets are a major factor contributing to higher sulfur in the metal chloride product. This is because larger numbers of small pellets fall through the grid than when larger pellets are used. In all of the tests conducted with large pellets (0.95 to 1.25 cm diam), the resulting metal chloride contains less than 1 pct S. However, when small pellets (0.64 to 0.95 cm diam) were used (three tests), the metal chloride products contained 1.55 to 1.63 pct S.

Figure 6 shows that increasing the chlorine flow from 3 to 5 l/min causes an increase of the sulfur level in the metal chloride product from an average 0.5 pct to 0.8 pct over the temperature range from 550° to 750° C. Increased chlorine flow rate produced higher reaction rates and correspondingly yielded a larger quantity of molten metal chlorides , which easily flushed an appreciable amount of unreacted chalcopyrite into the chloride receiver. As the unreacted mineral mixed with the molten metal chloride in the receiver, a slow chlorination took place forming volatile sulfur chloride and lower valent metal chlorides. A faint but distinctive odor of sulfur chloride was always detected, even after the chloride product was transferred for cooling.

The weight ratio of chlorine to copper and iron indicates the relative amounts of cupric and ferric versus cuprous and ferrous chloride compounds. Inspection of figure 7 shows that the weight ratio of chlorine to copper and iron decreases with increasing temperature and increasing chlorine flow rate. High temperatures are believed to cause decomposition of higher valent copper and iron chlorides to lower valent chlorides. At 700° C and a chlorine flow of 3 l/min, the metal chloride product corresponds to CuCl2 + FeCl2. In all cases, the chlorine ratio lies between the “ic” and “ous” chlorides of copper and iron.

Samples of chalcopyrite pellets, sulfur product, and metal chloride product were fire-assayed for gold and silver. The results indicate that these precious

metals remained with the metal chloride product.

Only a trace quantity of lead was detected in the sulfur product. Lead chloride has a low vapor pressure at the operating temperatures used and therefore remains with the molten metal chlorides. The low lead content of chalcopyrite pellets (0.17 pct) makes lead analysis of the reaction products unnecessary.

15-cm Reactor

The primary purpose of tests conducted in the 15-cm reactor was to evaluate various refractories as chlorinator material. Methods of evaluation include length of time for the reaction products to penetrate the reactor wall and visual examination to determine the extent of deterioration. The secondary purpose was to compare the reaction products obtained from the 15-cm reactor with those from the 10-cm reactor.

A comparative evaluation of six refractory materials is shown in table 5. The duration of tests is the total hours of chlorination before heating element failure occurred. Wall penetration caused corrosion and failure of the electric heating elements. Although external heating may have hastened the chloride penetration, the extent of deterioration of the interior wall was the criterion for evaluating the relative resistance of various refractory materials to attack by molten salt and chlorine environment. Visual examination indicates that all castable materials suffered corrosion, whereas the plastic refractory and especially the ramming mix (Hi Ram-80) showed good resistance to chlorine and to the reaction products.

As mentioned earlier, difficulties were encountered when a hearth was used as bed support for the first six reactors. The restricted chloride outlet was frequently plugged, causing inconsistency in the chloride product recovery. Consequently data from these tests were not used for comparison.

Fairly smooth operations were achieved after a full size bed support grid was used in reactors 7 and 8. Reactor 7 was operated for a total of 55 hr including one extended 15-hr test and five additional tests of 8 hr each.

Five tests were conducted with reactor 8 for a total of 40 hr. Failure of the support grid prevented further testing.

Note.—Chlorinators 1 to 6, hearth used to support bed of pelletized charge.
Chlorinators 1 to 5, electric element covered entire outer wall surface.
Chlorinators 6 to 8, electric element covered top and bottom of outer wall surface only.

Although the firebrick slices used as the grid were still in good condition after several tests, they were not secured firmly to the reactor wall. The extra poking required to maintain bed movement may have contributed to dislodging the brick-slices creating a large opening to permit excessive transfer of unreacted chalcopyrite pellets to the metal chloride container.

Test results obtained from operating reactors 7 and 8 constructed of high alumina ramming mix are shown in table 6. These results are similar in most cases to those obtained previously in the 10-cm chlorinator as reported in table 4. However, some differences are noted that warrant further discussion.

The metal chloride product in the 15-cm reactor contained a higher percentage of chlorine than that obtained from the 10-cm reactor. This may have resulted because chlorine passed through the metal chloride receiver. The molten product, cooled under a chlorine atmosphere, was partially converted to higher valent copper and iron chlorides. However, the unreacted mineral that fell into the chloride receiver was not completely chlorinated as a relatively high sulfur level in the metal chlorides was noted in several cases.

Greater percentages of sulfur were noted in the metal chlorides from the final two tests in the 15-cm reactor than any of the tests previously made. Although the use of small size pellets may have contributed to these increases, the main cause was most likely failure of the support grid, which permitted excessive quantities of unreacted chalcopyrite pellets to fall into the chloride receiver.

As previously mentioned, the bench-scale reactors were made of refractory materials and were externally heated to maintain reaction temperature. In potential commercial operations, refractory-lined reactors enclosed within a steel shell would be used. The reaction between chlorine and chalcopyrite is highly exothermic and the reactor shell would be water cooled. Cooling the shell would also maintain lower wall temperatures thus preventing attack of the refractory liner and the steel shell by molten salts and chlorine atmospheres.

In the bench-scale tests, a refractory designated as Hi Ram-80 ramming mix exhibited the best resistance against molten metal chlorides and elemental sulfur in a chlorine environment. In addition, the refractory possessed good mechanical strength; a sledge hammer and a splitting wedge were required to break the reactor open for visual inspection. Unfortunately, the project was terminated before this promising refractory could receive further testing in larger scale, process-development units.

Conclusions

Continuous chlorination of chalcopyrite concentrate to extract copper-iron chlorides and high-purity elemental sulfur was demonstrated in small-scale, process-development units.

Test results show that over 96 pet of the chalcopyrite was reacted by anhydrous chlorination in a shaft-type reactor. Good separation of metal chloride and sulfur products was achieved.

The valence state of copper and iron in the metal chloride product depended on the reaction temperature and chlorine flow rate used.

Sulfur product purity depended on reaction temperature, chlorine flow rate, and depth of pellet bed. Higher reaction temperatures in combination with increased chlorine flows generally increased the percentage of metal chloride volatilized with the elemental sulfur. After filtration, elemental sulfur of greater than 99 pet purity was obtained.

Different refractory materials were evaluated in tests made under actual chlorinating conditions. A high-alumina ramming mix (Hi Ram-80) exhibited the best resistance to attack by molten chlorides and elemental sulfur in a chlorine environment.

Further testing of the refractory material is recommended in larger scale units. In addition, various techniques of bed support should be investigated.