Dr.-Ing. Faustino G. Prado, Prof-Ing. Faustino L. Prado, P.E., Ch.Eng.

PRADO & ASSOCIATES, INC., Florida 33604, USA

Ir. Martin Vandebriel

REZINAL NV, Zolder, B-3550 Belgium

________________________________________________________________________

The use of ammonia-ammonium carbonate (AAC) aqueous liquors was conceived in 1861, when Ernest Solvay rediscovered and perfected an old process now known as the Ammonia Soda Process. Its application for the recovery of zinc can be traced back to 1880, in Germany, in the works of C. Schnabel. Since then, it has been extended to other metals such as copper, nickel, cobalt, and manganese. Prado & Associates' involvement with the AAC system covers half a century. Starting in 1952, in Nicaro, Cuba, it has included research, development, design, start-up and operation for nickel, cobalt, zinc and copper projects.

Following Dr. Prado's 1985 publication on commercial-scale recovery of high-purity zinc oxide, Prado & Associates was approached by REZINAL NV of Zolder, Belgium. Their purpose was to commercialize the recovery of zinc as a "high-purity zinc oxide" from various zinc-loaded metallurgical dusts. From 1987 to 1990, a comprehensive process was defined and tested on laboratory, bench and semi-pilot scales, including preliminary design and feasibility/environmental studies. Due to political changes occurring in Europe at that time, the project was temporarily deferred.

During the last decade Prado & Associates has been involved in other tests, studies and evaluations related to zinc recovery from electric furnace dusts (with and without pyrometallurgical concentration) and secondary copper smelters.

Recently, REZINAL management has reactivated the project, taking into account the current residue supply situation, market requirements, environmental regulations and the conversion of all residues into an acceptable, marketable form. This paper presents an updated review of the results and potential applications and incorporates our recent experience.

The use of ammonia-ammonium carbonate (AAC) aqueous solutions was conceived in 1861 when Ernest Solvay (1865) rediscovered and perfected an old process now known as the Ammonia Soda Process. Initial interest in using AAC for metals extraction was documented in notes from an 1871 AIME meeting and dealt with the recovery of native copper. About the same time, C. Schnabel (1880) applied the technology to zinc recovery. The first installation following his ideas was built at Lautenthal, in the Harz mountains to process a mixture of zinc oxide, lead oxide, some cupric oxide, and pellets of enriched lead, produced in steaming liquated zinc-silver-lead crust of the Parkes process. After the Lautenthal plant had been in operation for some time, a similar one was built at Hoboken-les-Anvers.

In 1913, S. E. Bretherton published his experiences removing zinc from roasted concentrates at Afterthought mine, Shasta County, California. In 1916, recovery of copper from Calumet and Hecla tailings (2000 t/d) started at Lake Linden, Michigan. At the same time, Kennecott Copper began operations to leach copper from copper carbonate ores in Tamarak, Alaska. The plant reached a capacity of 1000 t/d in 1922. Both plants have produced in excess of 400 million pounds of copper from tailings alone, in addition to half as much again from secondary materials after 1942 when it was tried out for the Ward Production Board scrap resulting from the manufacture of small-arms, ammunition shells, and other material.

The first commercial plant to use AAC solutions for the recovery of nickel and cobalt from lateritic iron ores was built at Nicaro, Cuba in 1944 and is still in operation. Since then, other plants using AAC for nickel and cobalt have been built in Sered (Czechoslovakia), Yabulu (Australia), Nonoc (Philippines), Niquelandia (Brazil) and Punta Gorda (Cuba). International Nickel Co. uses the AAC process for the production of high purity nickel oxide.

In the quest for an economical and environmentally acceptable reclamation of metallurgical dusts and other zinc-containing residues, our goals have been the recovery high-purity zinc oxide, the complete recycling of all reagents, and conversion of residues into marketable or acceptable disposal forms.

Our learning process started fifty years ago in the Research Department of the Nickel Recovery Plant at Nicaro, Cuba, under the teaching authority of Professor Caron, from Delft University, who had developed the Ammonia-Ammonium Carbonate (AAC) process for Nickel recovery. The Nicaro Plant established in 1944, bears his name (Ni-Caro). Later on (1969-1971) we had the opportunity to be in charge of the supervision and evaluation of the pilot work for the Nickel/Cobalt project at Yabulu (Townsville, Queensland), followed by the development of the complete process design and detail engineering.

In 1973, Dr. M. A. Hewedi, at our suggestion and supervision, produced and published an extensive search on The NH₃ -CO₂ -H₂O Systems at Atmospheric Pressure in Nonferrous Extractive Metallurgy [1]. From mid 1974 to Dec. 1975, we assisted a small facility at Terminal Island, Los Angeles, California, in removing and processing a large stock pile of brass scrap, recovering, via AAC, copper metal and zinc oxide.

During 1979-80, in Trenton, Michigan, we assisted Huron Valley Steel (HVS) as consultants for the production of zinc oxide from ashes and drosses of the automobile recycling industry, using AAC, with emphasis on the precipitation of the Basic Zinc Carbonate. Using most of the equipment already available from an unsuccessful process alternate, a plant was assembled with an annual capacity of 14 million kilograms of zinc oxide. The HVS operation confirmed the capability of the process to produce a zinc oxide of high purity. It also confirmed the calculated energy requirements as well as the ability to operate continuously at commercial scale, with high zinc concentrations in the liquor. In addition, it demonstrated the capability to produce a wide range of physical properties, such as controlled particle distribution, large and adjustable active surface (BET), and a free flowing product. Dr. F. G. Prado's paper presented at the First International Symposium on Metals Recycling, in 1985, reflected the Michigan experience. [2]

In 1988, Sumitomo Metals published a paper on Pilot and Commercial Plant Operations at Tsukijima Works, following very closely the Prado scheme. As raw material they used relatively pure zinc dusts (91 to 99 % Zn) from vapor galvanizing operations. This work resulted in the eventual construction of a zinc oxide plant currently in commercial operation. [3]

In 1989-1990 the Prado Group conducted tests for REZINAL NV, Zolder, Belgium, for the recovery of zinc as high purity oxide from a wide range of industrial residues, mostly metallurgical dusts. After bench scale tests to confirm performances and quality, a small demonstration unit was assembled. The unit was run continuously for 6 hours every day, producing several hundred kilos of zinc oxide from different dust sources. Some of the results were presented in P&A paper to the Second International Symposium on Metals Recycling, (1990) [4]. After the process development program, a feasibility study and engineering package were prepared. In 1991 the construction of the new zinc oxide plant was postponed indefinitely for market reasons.

During 1991-1993 the Prado Group developed technology for the separation and recovery of zinc, lead, cadmium, copper and tin, from secondary copper smelter dusts, using Ammonia-Ammonium Carbonate (AAC) and Ammonium Acetate (AmAc) systems [USA - confidential client]. The results were summarized in P&A the paper presented at the Third International Symposium on Metals Recycling, Point Claire, 1995. [5]

Following their findings, the Prado Group has continued its activities in this field, expanding the scope into the partial cleaning of EAF dusts, recovering the zinc, lead and cadmium values, and making the residues amenable for internal recycling. Some of the results were presented at the International Symposium on Extraction and Processing for the Treatment and Minimization of Wastes, TMS, San Francisco, 1994. [6]

Some metals (M) are capable of reacting with ammonia (NH)₃ to form ammines [M(NH₃)]ⁿ. The AAC Process is in theory very simple and selective. In AAC liquors, only those metals capable of forming ammino compounds will be dissolved. The solubility will be a function of the stability constant of their Ammonia Complexes. [7]

M + nNH₃ = M(NH₃)ⁿ

Kn = [M(NH₃)ⁿ]/[M ][NH₃]ⁿ

According to their Kn values, the elements that may achieve different levels of solution, depending of their present structures and the composition of the leaching liquors, are

Cu⁺ , Ca⁺⁺, Mg⁺⁺, Cu⁺⁺, Zn⁺⁺, Cd⁺⁺, Co⁺⁺, Ni⁺⁺, Fe⁺⁺, Mn⁺⁺, Co⁺⁺⁺

Fe⁺⁺ is removed from solution by oxidation. When NH₃ is removed from the leaching liquor any of the above elements will be rendered insoluble as soon as the threshold of its Kn value is reached. The goal is the recovery of zinc, since Zn is more electropositive than Cu, Cd, Co, Ni, Fe, Mn, Co. Zinc metal will reduce and force their precipitation in the metallic state (cementation). Due to their low ammine stability, Ca and Mg solubility decrease in the early stages of ammonia removal. Any ammonia and ammonium carbonate left with the residue after leaching may by removed and recovered by drying it at 120/140°C.

In the presence of the Ammonia and Ammonium Carbonate liquor

ZnO + 2NH₄.OH = [Zn(NH₃)₂]++ + 2H₂O

[Zn(NH₃)₂].2OH + [(NH₄)]₂.CO₃ = [Zn(NH₃)₄]₂.CO₃ + 2H₂O

and when the ammonia is removed...

5[Zn(NH3)₄]₂.CO₃ ----> [ZnCO₃]₂.[Zn(OH)₂]₃ + 3CO₂ + 20 NH₃

the Basic Zinc Carbonate is washed and dried and the carbon dioxide and ammonia, absorbed in water, are returned as Fresh Leach Liquor to the leach circuit.

The zinc-bearing residues already available in the market, (some of them constituting a serious disposal problem) offer a wide spectrum of chemical composition and physical properties. They can be grouped in three principal categories:

1. Zinc ashes & drosses

2. Furnace dusts ("EAF", converters, etc.)

3. Scrap

Usually, zinc ashes and drosses are treated in processes such as ball-milling, for the recovery of the metal fraction, leaving a fine residue rich in impurities. This residue is characterized by a relatively high Zn content (50 to 70%), partially in the metallic state. Their particle size is approximately 50% below 74µ.

Zinc-bearing Metallurgical Furnace dusts are extremely fine. Raw iron and steel dusts are, presently, outside of our scope since the presence of zinc ferrites reduce the zinc extraction and recovery by the AAC liquor. The Waelz-type crude zinc oxide falls within our scope of review. Flue dusts derived from secondary smelters, brass mills, remelters, etc., are usually low in zinc (30 to 40%). When copper alloyed with zinc is present, its recovery requires a more elaborated processing circuit.

Zinc scrap is often remelted for zinc recovery, generating residues for the other two categories. Alloyed zinc (brass ) is also suited for the AAC technology if the facility to process it is properly designed.

The following elements, Zn, Sn, Pb, Cu, Fe, Al, Si, Ca, Mg, Na, K, S, Cl, F, C and O account for virtually 100% of the mass of the dusts from nonferrous operations. In the ferrous residues Ni, Co, Mn, Cr and P may also be found.

A characterization study on dusts originating from different nonferrous furnace operations show that more than 98% of the dust occurs principally as <0.5μm particles which are loosely agglomerated into larger masses. They consist principally of ZnO together with minor amounts of ZnCl₂, PbCl₂, PbSO₄, SnO₂, and Zn₂SnO₄. The remaining coarser fraction includes metals and alloys, oxides and spinels, Ca and Ca-Fe silicates, Pb and Zn sulfates, Pb, Zn and alkali chlorides, and carbonaceous material. The presence, in converter dusts, of numerous bubble-like particles of an organic phase containing both S and C is interesting.

The Waelz-type -pyrometallurgically enriched- ferrous dusts follow closely the above particle size distribution. Rich in free ZnO, they also contain iron-forming zinc ferrites or spinels. The amount of ZnO relative to spinel increases very markedly with decreasing particle size. There is also compositional difference in the Zn/Fe ratio between the spinels in the magnetic and nonmagnetic fractions. The spinels remain insoluble in the AAC liquors but they are amenable to partial recovery from the bulk of the residue by magnetic separation.

REZINAL in Zolder, Belgium, which started from ashes in 1975 and old zinc from roofing in 1982, produces Zinc Metal for the market (Galvanizer metal - remelted metal), Zinc metal for distillation (off grade metal) and Fine Ashes. Since the added value comes from the metal not from the fines, from the beginning Rezinal has looked for added value on fine ballmilled zinc ashes. Beginning in 1986 zinc prices for fines started going down. Rezinal, facing a non favorable situation on the outlet market for fines started to search for a new process for treating fines.

The Process Needs were established as follows:

• must be able to treat Rezinal fines
• must be able to treat low value zinc residues on the market to build up plant capacity
• end product cannot be technical oxide
• chlorides cannot be discarded as salt water but has to be disposed of as a reusable product
• use less chemicals as possible by recycling them
• make no secondary zinc products
• bring out the balance of metallic components as byproducts that can be sold, if possible, with added value.

In a first survey, Rezinox explored the information from the Metallurgical Recycling Convention, Fort Lauderdale, 1985, Florida. The Ammonia-Ammonium Carbonate was chosen for being an existing and proven technology for Ni as well as for Zn and Cu, and Prado & Associates has proven to be capable of transferring the technology for Zn applications.

Rezinal is interested in the processing of dust such as:

1. Mixture of Coreco ashes (in house)

2. Fine galvanizes ashes (in house)

3. Brass smelters (imported)

4. Waelz oxide (imported)

In the original process concept, based on accumulated experience and the possibility of discarding the residual liquors loaded with chlorides and soluble sulfates, a dust mixture (1, 2, 3) was leached with fresh leach liquor, without a previous chlorides removal. The metallic zinc available in the mix assisted in keeping the unwanted heavy metals in the residue.

After removing the ammonia to precipitate Basic Zinc Carbonate, BZC was washed with slightly ammoniated fresh water to remove chlorides and sulfates. The washed BZC product achieved Cl- levels below 100 ppm, in spite of levels of 20 gpl in the liquor. Cu, Cd, Pb, Fe in the final ZnO were each kept below 10 ppm. The BZC, as well as the ZnO after calcination, were of free flowing character and a particle size distribution (100% smaller than -140µ and 50% between 40 and 4 µ). (Table 2)

The current environmental situation does not allow the discharge of the residual liquors. It also requires, for some markets, a BZC with a finer particle size distribution. The composition of the residues needs a more favorable market horizon.

The new approach includes a pre-washing step before leaching. The dusts, alone or as a mixture, depending on its chemical composition and the required quality for the residue, are pre-washed with the liquors from the stripping after complete NH3 separation. The liquor is adjusted with sodium carbonate to a pH slightly above 7.0 in order to convert all the sulfates and chlorides of zinc and lead into insoluble carbonates. Fresh or recovered water is used for in-filter wash. The residual wash liquor, practically free of heavy metals but loaded with sodium and potassium chlorides as well as soluble sulfates is sent to the water-recovery area.

During the warm washing, operation the metallic zinc present in the dust will be oxidized with the evolution of hydrogen. The capacity for cementation of Cu, Cd, etc., is lost. Nevertheless, such capacity is partially reduced during the leaching if the metallic zinc enters the liquor after the other metals are already in solution. The in house dust (Samples 1 and 2, Table1), very low in chlorides and high in metallics, properly introduced into the process, overcomes the problem and decreases the purification cost. Once the bulk of chloride ions is removed by the prewash, after leaching, the Cl- concentration in the product liquor falls to 2 g/l or less. Under such conditions the washing requirements for the precipitated BZC are substantially minimized in order to achieve the required purity level.

The washed dusts are leached, in a two-stage countercurrent fashion, with the Fresh Leach Liquor (FLL) This liquor is produced from the recovered NH₃ and CO₂ from the BZC precipitation and the calcination steps, as well as the recycled recovered water. The FLL is the adjusted to the desired level of NH₃ and CO₂. The bulk of the ZnO, largely in the submicron fraction is dissolved in the first 15 minutes. Additional retention time is provided to achieve an effective extraction. Also most of the cadmium, some ferrous iron, the copper in oxide form and other minor amounts of metals such as lead and tin are dissolved. The solids are removed by filtration, washed with FLL and sent to drying where the vapors containing NH₃ and CO₂ are returned to the FLL recovery area. Table 1 presents the composition of residues from different sources as well as their weights. The percent of zinc recovered is also shown.

Metallic zinc dust or metallic zinc-rich dust are added to the Pregnant Liquor (PL), to cement (insolubilize) all the other heavy metals. If the Cd and/or Cu are at reasonable levels, they may be recovered here. The Pregnant Liquor (PL) is preheated to start the removal of NH₃ and initiate the BZC precipitation. With the first BZC also precipitates some of the impurities such as Ca and Mg. It is probable that the Fe and Pb that still appears in the PL are not in solution but in extremely fine particles that are trapped with the initial precipitate. These solids are removed by filtration and returned to the leaching stage.

The complete removal of NH₃ also completes the BZC precipitation. This removal is performed by rising the temperature of the Zinc bearing liquor through the introduction of steam that at the same time acts as a carrier or stripper of the mixture of vapors generated. Since the early development work for the design of the Nicaro Plant, it was found that the precipitation of the Basic Nickel Carbonate can produce a wide spectrum of particle size, starting with almost colloidal material when the concentration of Ni is small and a very fast speed of precipitation.

An increase in concentration and retention time will assist in obtaining solids amenable to fast settling, better filtration rates and less water retained by the cake. [8] The nickel plants using AAC operate with low concentrations and use bubble cap columns for the operation. Their BNC slurry produced under such parameters, with particles below 74 microns, requires an elaborated thickening before filtration.

Lately, INCO developed, following what has been used by the manganese recovery industry, a precipitation scheme based on a series of kettles instead of column plates, increasing the retention time and achieving a larger particle size, fast settling and very easy to filter and wash. [9] The Basic Zinc Carbonate (BZC), with a chemical structure identical to Basic Nickel Carbonate (BNC), follows the same pattern of behavior during precipitation. In our design we have used the kettles scheme for the BZC precipitation up to a point, when the Zn concentration is too low, and the final stripping is conducted in a side-to-side plate column.

The BZC is filtered and washed with light ammoniacal water to remove residual chlorides in the wetting product liquor. The filtrate is sent to final stripping and the washings go to the FLL preparation.

The Drying, if performed below 140°C, removes any residual ammonia, which is then recycled to FLL. The calcination, starting at 400°C, removes all the CO₂, but also reduces the surface area of the product. Both the CO₂ and NH₃ are collected and returned to the FLL column. The BZC and the ZnO after drying and calcination retain the structure and particle size distribution ( 50% between 140µ and 40µ and nothing below 4µ).

The FLL Area follows what has been standard practice for more than a century: a set of absorption columns following the latest characteristics of current commercial operations.

The bulk of the water is recovered by traditional methods (reverse osmosis, multiple stage evaporation, etc.) and the concentrated brine is crystallized for the mixed chlorides recovery. The stripping column discharge is settled, the sludge returned to the leach circuit and the liquor concentrated by the selected method (reverse osmosis, evaporation, etc.) into a brine from which salts are crystallized (item 5c in Table 1). The water returns to FLL and other points of use.

Since the bulk of the Zinc Oxide in the market is produced by a zinc vapor-burning technology, the specifications indicate a particle size in the submicron range. The so called Bayer alternate, produced by chemical precipitation, is offered with a particle size distribution all minus 30µ and 15% below 1µ. Rezinox ZnO ground with a Pallman Stitmühle reproduced such distribution.

Microscopically the BZC appears as an agglomerate with a coral-like structure, made up by primary crystals of submicron size, and very friable. A dispersion test in water, by ultrasonics, produced a granulation 100% under 200 mesh (74µ), 82% under 375 mesh and 12% smaller than 1µ. We have recently run promising tests that indicate, based on the known behavior of our solutions under different processing parameters, that the Bayer particle size distribution can be obtained within the proposed process scheme and the same selected equipment.

The Ammonia-Ammonium Carbonate (AAC) has been proven to be a selective way for the recovery of several metals, with commercial history in zinc, copper and nickel. The increasing production of metallurgical residues, their disposal and environmental regulations, as well as conservationism and economics, has led the search for processes to solve or improve such questions. And here we have one that has been used for more than a century and proven once and again to have the following attributes:

• selective in extraction
• reagents easily and completely recovered
• reagent capable of high metal loading
• high purity solutions obtainable
• equipment corrosion minimal
• materials of construction conventional
• compatible with strict environmental controls (zero effluent)
• competitive energy requirements

[1] Hewedi, M. A., & Engle, L. F., "The NH3-CO2-H2O System at Atmospheric Pressure in Nonferrous Extractive Metallurgy", Intl. Symp. On Hydrometallurgy, AIME, 1973, pp 806-858

[2] F. G. Prado, J. P. Dempsey, B. W. Wiegers, "High Purity Zinc Oxide Production (AAC) from Residues in Automobile Scrap Recycling", International Symposium on Recycle and Secondary Recovery of Metals, Fort Lauderdale, Florida, 1985, pp. 183-193.

[3] T. Hashimoto, T. Owaga, T. Kasai, "Development of a High Purity Zinc Carbonate Production Technology", Sumitomo Metal: Vol. 40 No. 2, 1988, pp. 45-54

[4] F. G. Prado, "High Purity Zinc Oxide from a Wide Range of Industrial Residues", International Symposium on Recycling of Metals and Engineering Materials, TMS, 1990, pp. 529-537

[5] Dr. F. G. Prado and F. L. Prado, "Secondary Copper Smelter Dusts: an Environmentally Friendly Reclamation", Third International Symposium on Recycling of Metals and Engineering Materials, Point Clear, Alabama, 1995

[6] F. G. Prado & F. L. Prado, "EAF Dusts: A Viable Complete Minimization, Extraction and Processing for the Treatment and Minimization of Wastes", TMS 1993, pp. 543-553.

[7] J. Bjerrum, "Metal Ammine Formation in Aqueous Solution", P. Haas & Son, Copenhagen, 1957

[8] R. C. Hills, Lectures, Unpublished.

[9] Illis, A., Nowlan, G. C., and Koehler, H. J., "Production of Nickel Oxide from Ammoniacal Process Streams", Canadian Inst. Of Min. & Met. Transactions. Vol. LXXIII, 1970, pp. 44-5

Back to Prado & Associates' Main Home Page