Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The recovery of copper from synthetic leach solutions with iron powder Hughes, Garry Maurice 1970-12-31

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
[if-you-see-this-DO-NOT-CLICK]
UBC_1970_A7 H83_2.pdf [ 6.41MB ]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0081087.json
JSON-LD: 1.0081087+ld.json
RDF/XML (Pretty): 1.0081087.xml
RDF/JSON: 1.0081087+rdf.json
Turtle: 1.0081087+rdf-turtle.txt
N-Triples: 1.0081087+rdf-ntriples.txt
Original Record: 1.0081087 +original-record.json
Full Text
1.0081087.txt
Citation
1.0081087.ris

Full Text

THE RECOVERY OF COPPER FROM SYNTHETIC LEACH SOLUTIONS WITH IRON POWDER by GARRY MAURICE HUGHES B.Sc.(Eng), Queen's University, 1965 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in the Department of Mineral Engineering We accept this thesis as conforming to the required standard THE.UNIVERSITY OF BRITISH COLUMBIA In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and Study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Mineral Engineering The University of British Columbia Vancouver 8, Canada Date September T, 1970 i ABSTRACT Recent increases in the production of copper by acid-leaching techniques have depleted the supply of iron and de-tinned cans as sources of precipitant for copper from solution. Previous ..investigators of the iron cementation process have indicated that high purity iron powder, although more costly than detinned cans, has several properties which render it a potential precipitant of copper from leach solutions but few data have been published on the subject. The purpose of this thesis is to investigate the variables encountered in the cementation of copper with iron powder and to obtain data to substantiate indications that the iron consumption could be reduced and a higher grade cement copper could be produced which would compensate, at least in part, for the higher cost of the precipitant. A vertical column mounted on an inverted cone was used as the reaction vessel in the experiments. The pregnant solution was forced into the precipitator through the apex of the cone and the iron powder was added from the top of the column through a small diameter tube. The copper product was removed from the reaction vessel by the effluent solution and collected in a settling tank. Several series of experiments were conducted.to study the influence of variations in the' velocity of solution flow through the precipitator, the concen trations of free acid,.ferrous and ferric ion in the pregnant solution, and the temperature of the pregnant solution on the iron consumption and the grade of the cement copper product. A velocity of solution flow in the column of approxi mately ore centimeter per second was found to be suitable for iron powders retained on a .200 mesh screen. The capacity of the precipitator was reduced considerably with iron powder passing through a 2 00 mesh screen because iron of that size fraction was easily forced to overflow the precipitator before cementation.is complete. Neither temperature nor ferrous ion concentrations within the normal ranges of operation had any significant effect on the cementation process. Free acid con centrations up to twenty grams per liter were tested with no significant increase in iron consumption. The presence of ferric ion in solution during cementation enhanced the consumption of excess iron appreciably. Excess iron consumption was due to -the •reaction between metallic copper and ferric ions. The cupric ion produced in the reaction consumed additional iron by the normal cementation reaction. The reaction between metallic iron and ferric ions was not a major contributor to excess iron consumption. The consumption of ferric ion was not excessive at the beginning of experiments, when the surface area of metallic iron was at a maximum, but increased in approximate proportion to the increase in surface area, of metallic copper in the re action vessel. Although the experiments were conducted with batch additions of iron the results indicated that the cone-column precipitator could be operated continuously by adding precipitant at a rate equal to its consumption. With continuous operation the contact between metallic copper and ferric ion was Ill minimized which in turn reduced the iron consumption. The cement copper product of cementation from synthetic leach liquors with high purity iron powder was dried ir. alcohol and acetone and consistently contained more than ninety-five percent copper. xv: LIST OF CONTENTS Page ABSTRACT i LIST OF CONTENTS. . ;'...iv LIST OF TABLES ' vii LIST OF FIGURES and ILLUSTRATIONS . . . , ' ix ACKNOWLEDGEMENT S • .' . xi CHAPTER I INTRODUCTION 1". CHAPTER II "RECOVERY OF COPPER FROM SOLUTION THE ALTERNATIVES TO IRON CEMENTATION,,. 5 A Precipitation with Sulphur Dioxide. 5 B Precipitation with Hydrogen Sulphide 6 C Precipitation with Hydrogen 7 D Ammonia Leaching and the Recovery of Copper from Complex Ammonium Salts, , , .7 E Elect rowinning 8' F. Solvent Extraction 9. G Precipitation of Copper with Iron after Reduction to Cuprous Chloride 9 .-..CHAPTER III THEORETICAL CONSIDERATIONS OF THE CEMENTATION PROCESS. THE KINETICS OF CEMENTATION . . 11 CHAPTER IV REVIEW OF CEMENTATION METHODS 24 A History of CementationB Drum Precipitators 2 5 C Tank Precipitators  5 D Cone Precipitators 26 V Page > . E Launders . 29 F A Discussion on the Launder System.. 30 G The Product of Cementation Plants: Cement Copper. 35 • 1. Hydrolysis of Dissolved Cations 36 2. Solids in Suspension 38 3. Oxidation during Drying 3 9 4. Impurities in the Precipitant... • 40 CHAPTER V IRON POWDER: A SOURCE OF IRON FOR CEMENTATION 41 A Supply. ...... '. 41 B Advantages ' 44 C Iron Powder Precipitators 44 CHAPTER VI EXPERIMENTAL PROCEDURES 4 7 A Apparatus and Methods Employed to Control Variables..... 47 1. Analyses 52 2 . . Assays. . ' 5 3 -CHAPTER VII RESULTS 54 A Velocity of Solution through the Column 54 B Acid Concentration of the Pregnant Solution 5 9 C Temperature of the Pregnant Solution 61 D Ferrous Ion Concentration of the Pregnant Solution 6 3 E Ferric Ion Concentration of the Pregnant Solution 6 5 F Ideal Conditions for the Systems Employed 75 CHAPTER VIII EVALUATION OF THE PARAMETERS IN THE CONE-COLUMN PRECIPITATOR 76 1. Velocity of Solution Flow in the Column 7 9 vi Page 2. Length of Column 8 0 3. Rate of Feed to the Precipitator.... 80 4. Amount of Precipitant in the Reaction Vessel.... 81 CHAPTER IX SUMMARY AND CONCLUSIONS 82 1. Suggestions for Further Experimentation., 84 LITERATURE CITED -86 APPENDIX; 9 0 vii LIST OF TABLES Table Page ' 1 Summary of Operating Data from Copper Leaching Operations 31 la Fe:Cu Calculations for Several Leaching Operations 31a 2 Analyses of Leach Liquors 36 3 Chemical Analyses of Iron Powders 4 2 4 Screen Analyses for Iron Powders A, B, and C and Data to Accompany Figures 15, 16 and 17. 44' 5 Data from Tests with Variations in the Velocity of the Solution through the Column with Iron Powder "A" (-10 0+2 00 mesh) 91 6 Data from Tests with Variations in the Velocity of the Solution through the Column with Iron Powder "A" (-200 mesh) 92 7 Data from Tests with Variations in the Acid Concentration of the Pregnant Solution with Iron Powder "A" (-100+200 mesh) 93 8 Data from Tests with Variations in the Acid Concentration of the Pregnant Solution with Iron Powder "A" (-200 mesh) 94 9 Data from Tests with Variations in the Temperature of the Pregnant Solution with Iron Powder "A" (-100+200 mesh) 95 10 Data from Tests with Variations in the Temperature of the Pregnant Solution with Iron Powder "A" (-200 mesh) 96 • 11 Data from Tests with Variations in the Ferrous Ion Concentration of the Pregnant Solution with Iron Powder "A" (-100+200 mesh) 97 12 Data from Tests with Variations in the Ferrous Ion Concentration of the Pregnant Solution with Iron Powder "A" (-200 mesh) , 98 13 Data from Tests with Variations in the Ferric Ion Concentration of the Pregnant Solution with Iron Powder "A" (-100+200 mesh) 99 Vll 1 Table 14 Data from Tests with Ion Concentration of with Iron Powder "A" Page Variations in the Ferric the Pregnant Solution (-200 mesh) 100 / vix . LIST OF FIGURES AND ILLUSTRATIONS • Figure Page 1 Apparatus Used by Nadkarni et al and Nadkarni and Wadsworth for the Cementation of Copper and Tron 13 2 Copper Concentrations against Time for Various Sized Fractions of Cast Iron Pellets 16 3 Influence of Metallic Copper in Accelerating Iron Consumption in the Presence of Ferric Iron 20 4 Sketch of Kennecott's Cone-type Precipitator for Detinned Cans 2 7 5 Variation in the Content of Leach Liquors during the Passage through the Launder System at Rio Tinto, Spain 3 7 6 Sketch of the Precipitator Used in the Experiments 4 9 7 Velocity of Solution versus percent Copper in Overflow and Fe:Cu Ratio with -10 0+2 0 0 mesh Iron Powder "A" 56 8 Velocity of Solution versus Percent Copper in Overflow and Fe:Cu Ratio with -2 00 mesh Iron Powder 5 6 9 Acid Concentration of Pregnant Solution versus Percent Copper in Overflow and Fe:Cu Ratio with -100+200 mesh Iron Powder "A" , 61 10 ; Acid Concentration of Pregnant Solution versus Percent Copper in Overflow and Fe:Cu Ratio with -2 0 0 mesh Iron Powder "A" 61 11 Temperature of Pregnant Solution versus Percent Copper in Overflow and Fe:Cu Ratio with -100 +2 00 mesh Iron Powder "A" 6 3 12 Temperature of Pregnant Solution versus Percent Copper in Overflow and Fe:Cu Ratio with -200 mesh Iron Powder "A" 6 3 13 Ferrous Ion Concentration of Pregnant Solution versus Percent Copper in Overflow and Fe:Cu Ratio with -100 + 200 mesh Iron Powder "A" 65 >.x Figure Page 14 Ferrous Ion Concentration of Pregnant Solution versus Percent Copper ir the Overflow and Fe:Cu Ratio with -200 me-oh Iron Powder "A" 65 15 Ferric Ion Concentration oi Pregnant Solution versus Percent Copper in Overflow and Fe:Cu . Ratio with -100+200 mesh Iron Powder "A" 67 16 Ferric Ion Concentration of Pregnant Solution ; versus Percent Copper in Overflow and Fe:Cu Ratio with -200 mesh Iron Powder "A" 67 17 Cu++ and Fe+++ Ion Concentrations versus Time with Iron Powder "A" -69 18 Cu++ and Fe+ + + Ion Concentration, versus Time with Iron Powder "B" 6 9 19 Cu++ and Fe+++ Ion Concentration versus Time with Iron Powder "C"  9 Photograph 1 1 Minute after Start of Experiment 71 2 5 Minutes after Start of Experiment 71 3 11 Minutes after Start of Experiment 71 4 15 Minutes after Start of Experiment 71 5 18 Minutes after Start of Experiment 72 6 2 0 Minutes after Start of Experiment 72 ~7 25 Minutes after Start of Experiment 72 8 Settling Tank 7xi ACKNOWLEDGEMENTS Ti.e author is grateful to Dr. CL. Emery for his advice and arrangements of financial assistance. He is indebted also to.those employees of Biometals Corporation Ltd., particularly Messrs. A.A. Mayling and N. Larsen, who carried out many of the tedious analyses, and to Biometals Corporation Ltd. for its financial support and loan of equipment.. 1 CHAPTER I INTRODUCTION Increases in the costs of labour and decreases in the reserves of high grade copper sulphide deposits have revived the arts of leaching copper from oxidized ores with acid and the recovery of copper from solution by cementation on scrap iron. Prior to the acceptance of flotation, the acid-leaching technique was the most efficient method of recovering copper from acid soluble ores. Gravitational methods were employed to recover copper sulphides but tailings containing more than one percent copper were quite common. Interest in leaching was almost eliminated by the development of efficient flotation methods which were more applicable to the sulphide ores; however several companies continued to use their leaching facilities to recover copper from the "waste" oxides particularly in the south western United States. Leachable copper deposits were rarely found in Canada where relatively insoluble chalcopyrite is the dominant copper mineral, Most of the copper mineral in the south-western United States is mined' by open pit methods. Because of the decreasing grade of material mined coupled with increasing labour costs and pit depth it became necessary to increase both the tonnage mined and the scale of mining equipment used. Stripping ratios rose and in many cases the waste represented a much larger tonnage than the ore, This material which was considered as 2 waste because the values could not be economically recovered by flotation methods usually contained a considerable quantity of copper. At Kennecott Copper Corp.'s 3ingham Canyon operation, for example, material containing 0.24 percent copper is being piled on waste dumps at a rate of approximately 200,000 tons per day. Kennecott is one of the companies which has continued to recover some of the copper from its low grade material by acid leaching and iron cementation. Recently, the leaching and iron cementation facilities were expanded to a capacity of 200 tons of cement copper per day (7). Kennecott has introduc ed several changes in its leaching techniques which include the use of autotrophic bacteria to dissolve copper from sulphides, improved heap design and circulation systems, and cone-type copper precipitators. The cone-type precipitators, which have replaced the launder-type cementation plant, employ the same iron precipitant, namely detinned cans. Since Kennecott re quires between four and five hundred tons of detinned cans per day and several other major producers require more than one hundred tons per day it is quite possible that the demand for this form of scrap iron will surpass the supply in the near future. The purpose of this thesis is to investigate the use of iron powder as an alternate source of precipitant for the recovery of copper from dilute leach solutions. Many forms of scrap iron with small surface areas could be potential preci-pitants if ground to a powder,but iron powders can also be produced, chemically from iron ores. Iron powder is more costly than detinned cans;' however, several investigators have indi-* Ground cast iron costs 2.5-3.0<r7lb f.o.b. Chicago/11 npti'nnpH rfltin ro^f ?.n-?.»/1h f.o.b. T.os Angelas. C al i f.  3 cated that the use of iron powders could reduce the iron re quirements for cementation sufficiently to compensate for the extra cost. If iron powder is used as the precipitant it may be possible to produce a higher grade cement copper which could be used in powder metallurgy thus eliminating the costs of smelting ana refining. A cement copper powder containing more than ninety-eight percent copper has a dollar value of approxi mately twice that of the contained copper in a concentrate or lower grade cement copper. In Chapter II of this thesis the author briefly discusses several alternatives to iron cementation which have been used or proposed as methods of recovering copper from solution. Chapter III is a literature survey of the theoretical considerations of the iron cementation process. Although the process has been used for several hundred years the most signi ficant contributions to the theory of the subject have been made in the last two or three years. Most of the cementation procedures which have been employed to recover copper from solution with iron, including the cone-type precipitators recently developed by Kennecott, are discussed in Chapter IV. The available operating data and theory have been employed to analyze the launder system which is the oldest and most popular method of recovering copper from acid leach solutions with iron. Chapter V 'is an introduction to the topic of iron powders and includes a brief description and analysis of the precipitators which have been proposed or utilized to recover copper from dilute leach solutions with iron powder'. A description of the experimental procedures and the apparatus employed is presented in Chapter VI. The results of the experiments are presented and dis cussed in Chapters VII and VIII. The final chapter summarizes the results of the ex perimental work and contains suggestions for further research necessary should' the methods proposed in this thesis be applied to a commercial operation. 5 CHAPTER II ' RECOVERY OF COPPER FROM SOLUTION THE ALTERNATIVES TO IRON CEMENTATION. Several alternatives to the iron cementation process have been used or proposed as methods of recovering copper from solution. The more important methods are discussed in this chapter primarily to illustrate the role of. iron cementation. The methods of removing copper from solution which are discussed in this chapter are as follows: (A) Precipitation.with Sulphur Dioxide (B) Precipitation with Hydrogen Sulphide (C) Precipitation with Hydrogen (D) Ammonia Leaching and the Recovery of Copper from Complex Ammonium Salts (E) Electrowinning (F) Solvent Extraction (G) Precipitation with Iron after Reduction to Cuprous Chloride. (A) Precipitation with Sulphur Dioxide Sulphur dioxide at elevated temperature (320-330°F) and pressure (200 psi) will reduce copper sulphate to metallic copper according to the following reaction: CuS04 + S02 + 2H20 •* Cu(») + 2H2S04 (1) but,, the process is applicable only to strong solutions and would be impractical for dilute leach liquors^^)* Sulphur dioxide is also used to precipitate copper as cuprous chloride from solutions containing cupric chloride and cupric sulphate^Q) according to equations (2) and. (3). 2CuCl2 + S02 + 2H20 -> 2CuCl + H2S04 + 2HC1 (2) 2CuS04+2NaCl+S02 -*• 2H20+2CuCl+2H2S04+Na2S04 (3) At the Mantos Blancos operation in Chile(20) Sas containing fourteen percent sulphur dioxide moves counter-currently to the solution in steel spray towers reducing the copper content of the solution from thirty-five grams per litre to thirteen grams per litre. The smelter converts the chloride to metallic copper with lime according to the equation: 4CuCl+2CaC03 + C -»• HCu(s)+2CaCl2 + 3C02 (4) The CaCl2 is tapped from the reaction vessel as slag and con tains about one percent copper. Since the concentration of the cupric ion in leach liquors from heap leaching operations is within the range of 0.3 to 10.0 grams per liter sulphur dioxide precipitation is not applicable unless the concentration of copper in solution is increased substantially. (B) Precipitation with Hydrogen Sulphide Hydrogen sulphide has been used to precipitate coppe from dilute solutions as copper sulphide; however, further " ^u(s) anc^ ^e(s) denotes metallic copper and iron respectively. processing is necessary to produce a high purity copper powder (10). Cementation of copper on iron produces a higher grade product with a lower initial capital expenditure. (C) Precipitation with Hydrogen Sherritt-Gordon Mines Ltd. employs hydrogen reduction to recover copper from solution after ammonia pressure leaching (12). Nickel, cobalt and copper are present in the leach liquors. The copper is removed at elevated temperature and pressure to produce a high purity copper powder. Hydrogen is presently being used to precipitate copper from acidic-leach solutions at an operation in the south western United States; however, no literature has been published by the company. The process is successful in that a high grade copper powder is being produced but the conditions required for precipitation are so corrosive that expensive stainless steel reaction vessels and valves must be used to minimize maintenance problems. (D) Ammonia Leaching and the P%ecovery of Copper from Complex Ammonium Salts. Copper dissolved in ammoniacal solution forms complex ammonium salts, which hydrolize readily unless an excess of ammonia is present. The process is applicable to copper carbonate ore (malachite) according to the equation: Cu2C03(OH)2 + 4(NH4)2C03 * 2Cu(NH3)l+C03+3C02 + 5H2 (5) Copper in acid leach solutions is complexed with ammonia according to.the equation: 8 CuSO +4NH3 Cu(NH3)4S04 • (6) Through simple heating to drive off the excess ammonia, the complexes can be converted to the copper salv according to the equations: CuCNH -)4C03 -*• CuO + 4NH3+C02 (7) Cu(NK3.)4S04 + H20 •* CU0 + CNH4)2S04+2NH3 (8) If carbonate has been used for the negative radical the final copper product will be cupric oxide since.cupric carbonate will decompose in solution. The ammonia and carbon dioxide are re covered in scrubbing towers. This process is applicable to oxidized deposits con taining excessive percentages of acid-consuming minerals. The product requires additional treatment by smelting and refining. The capital cost of process equipment is high in comparison to that required for cross cementation and the operating costs are high due to. losses of ammonia. (E) Electrowinning The electrowinning method of recovering copper from leach solutions of relatively high copper concentration has been well known and practiced for many years. Among the advantages of the method is the recovery of a product that is equal in purity to the best refined cathode copper. The process also regenerates one equivalent of sulphuric acid for each equivalent of copper produced. . The method is not applicable to dilute leach liquors from a low grade leaching operation as the 9 current efficiency is very low in the presence of high concen trations of impurity ions, and the initial capital costs are high, Electrowinning may be economically feasible if used in con j unct .'.on with a solvent extraction process to eliminate the interfering ions and to increase the concentration of copper in solution. (F) Solvent Extraction The choice of a suitable organic reagent for solvent extraction would permit the isolation and concentration of cupric ions providing an excellent pregnant solution for an electrowinning plant." The first reagents chosen were not adaptable to the wide variation in pll values encountered in a leaching operation; however, this problem appears to be solved^j.. Ranchers' Exploration and Development Corp. have recently started the installation of a solvent extraction plant at its' Bluebird copper mine in Miami, Arizona-; however, operating data will not be available for at least one year. The high capital costs of a solvent extraction-electrowinning plant would require a leachable deposit of very large tonnage to warrant the installation costs. -(G) Precipitation of copper with Iron after Reduction to Cuprous Chloride. At Chuquicamata, Chile Q^J, the copper occurring in solution as copper sulphate and cupric chloride in the waste electrolyte is reduced to the insoluble cuprous chloride by an * E S M.J. October 196 7. P. 118. 10 excess of cement copper in the presence of chloride ions ac-cordirg to the following reactions: Cu( s) + CuS04 + FeCl2 -> 2CuCl+FeS04 (9) Cu, ,+CuCl0 2CuCl (10) (s; 2 The cuprous chloride is dissolved in a strong ferrous chloride brine from which•copper can be precipitated according to the following equation: 2CuCl+FeCl„(brine)+Fe, . + 2Cu,»+2FeCl_ (11) 2 (s) (s; 2 Several other methods of recovering copper from leach solutions have been attempted but without commercial success. Kennecott Copper Corp., which is considered one of the most progressive of the large copper producers, has recently committed itself to several years of iron cementation with detinned cans and it is very unlikely that any of the other major producers will accept radical changes in their production methods. It appears, that for the near future, copper producers will seek a different source of iron for their cementation plants rather than a different method of recovering copper from solution. 11 CHAPTER III THEORETICAL CONSIDERATIONS OF THE CEMENTATION PROCESS . THE KINETICS OF CEMENTATION The cementation process was applied successfully for several hundred years with little understanding of the principles involved. By the early nineteen hundreds it was well recognized that^^5 1. careful mass balances must be kept on influent and effluent streams, 2. free acid should be limited to prevent excess iron consumption, 3. large surface areas of iron are necessary for rapid cementation, 4. velocity of stream flow is important, 5. aeration of liquors by tumbling through air is objectionable because the increased oxidation increases iron consumption, 6. reactions involved in the process occur faster within limits, at higher temperature, 7. ferric ion is a major contributor to iron consumption, In 1944 excessive iron-consumption at a cementation plant in the United States initiated an investigation by Wartman and Robertson^^) wh° presented the following reactions to explain the process of cementation : CuS04+Fe(s) - Cu(s)+FeS04 (12) Fe2(S04)3+Fe(s) + 3FeS04 (13) H2S04 + Fe(s) + FeSO^ + r^ (14) According to reaction (12) 0.88 pounds of iron would produce one pound of copper; however, In actual practice iron consump tion was several times the theoretical figure. The excess iron consumption was attributed to reactions (13) and (14). Wartman and Robertson found that reactions (12) and (13) proceed at approximately the same velocity while reaction (14) was much slower and could be diminished by controlling the contact time. It was also pointed out that increased agitation is beneficial in removing;hydrogen bubbles and barren layers of solution at the iron surface as well as removing contaminants resulting from the hydrolysis of iron salts. Until the recent work of Nadkarni et al,ors and v I b) Nadkarni and Wadsworth^ 2^) very few investigations into the cementation of copper on iron were conducted. Several invest!-, gators have studied the cementation of copper on metals other than iron and their experiments and conclusions are summarized in the paper by Nadkarni and Wadsworth. The. recent experiments by Nadkarni et al were con ducted in the apparatus shown in Figure .1. The reaction vessel was a three-necked distillation flask of 2,000 ml capacity. Experimental data were obtained by measuring the amount of copper and iron ions in solution at successive time intervals. Nitrogen was bubbled through the solution before and during the nitrogen Fig. 1 - Apparatus used by Nadkarni et al. and by • J Nadkarni and Wadsworth for the concentration of. copper on iron • 14 course of a run to prevent oxidation of the precipitated copper by atmospheric oxygen. The stirring speed was expressed in re volutions per minute and represented an arbitrary measure of the agitation. Several different- types of iron were used for the tests; however most of the results published applied to sheets to SAE 1020 mild steel (area = 4.65 sq. in.). The authors pointed out that the mechanism by which' the overall reaction at a solid-liquid interface can occur is generally considered to go by the following steps: (1) diffusion of reactants to the surface; (2) adsorption of reactants on the surface; (3) chemical reaction at the surface; (4) desorption of products from the surface; (5) diffusion of products away from the surface and that any of the above steps may be rate controlling. Dif fusion of reactants to the surface is controlled by the stirring speed and may be eliminated at sufficiently high stirring speeds. It was also suggested that diffusion through a boundary layer could be the rate determining step. + + + + A plot of ( Cu ) / ( Cu )•• versus time for different stirring speeds showed that the rate of the cementa tion reaction was first order, proportional to the surface area of the iron, and increased with speed of stirring until a maximum rate was observed. The data followed the integrated form of the following first-order rate equation: d + + [Cu++] = -CCu ]KQA (15) 15 where t = time in minutes r.Cu++] = concentration of copper ions in grams per litre A = area in square inches 1-2 K = rate constant = min- in o The authors found that at low speeds the precipitated copper was spongy and adherent. At medium speeds the copper peeled off in bright strips and at high speeds finely divided copper was produced and continually removed from the surface. At high speeds, however, the solution turbulence was not strong enough to keep the surface of the iron sheet completely free from the copper deposit; instead, the precipitated copper, clung to the surface as a few scattered "islands" that peeled off in patches. The "islands" were, attributed to inactive impurities in the iron and were eliminated in tests with high purity iron. The experimental activation energy was obtained by a least-square fit on an Arrhenius plot [ln(rate) vs 1/T°K], The average experimental activation energy, combining results from several types of iron, was 5.8 +_ 1.6 kilocalories per mole which was interpreted as representing diffusional control through a limiting boundary film. The data obtained by Nadkarni et al from tests with different mesh sizes of cast iron are plotted in Figure 2. Temperature, stirring speed, pH (1.8), initial con centration of copper (0.570 gpl), and weight of iron added were kept constant. For particles smaller than 60 mesh, all the iron was consumed. The Fe:Cu ratio was 1.8:1. The precipitat ed copper was in the form of soft irregular spheres with a black hole in the centre., presumed due to the graphite in the 16 cast iron. For particles larger than 60 mesh the precipitated copper was in the form of hard nodules with iron in the centre. The precipitated copper formed a protective layer and essential ly blocked further precipitation. 06 05 OA fc v 03 e c • ~ 02 O Ol 0 -0 60 120 180 TIME (minutes) Fig. 2 - Copper concentration vs time for various-sized fractions of cast iron pellets. (Nadkarni et al) Nadkarni and Wadsworth used the apparatus shown in Figure 1 with thin rectangular samples (2.0" x 0.61" x 0.10 3") of Armco Magnetic iron of 9 9,9+ percent purity as the precipitant. One end of each sample was coated with an inert material to provide a small region for mechanical support and to give a constant effective surface area of 2.0 inches squared. A series of experiments were carried out at various stirring speeds with an initial cupric ion concentration of one gram per litre, a temperature of 45°centigrade, a pH of approxi mately 2.. 5 , and under a pure nitrogen atmosphere at an ambient pressure of 0. 86 at in. At high stirring speeds first order kinetics were followed; however, at lower speed, a deviation from .first order occurred which was attributed to the build up of copper on the surface of the iron, increasing the diffusion path length. The specific rate of the reaction, which is the geometric slope of the first order plot multiplied by -2.3/A, where A is the area of the sample, increased with an increase in stirring speed until it reached a constant figure for high stirring speeds. Nadkarni and Wadsworth suggested that the dependence of reaction rate on stirring speed could result from either a surface controlled chemical reaction or diffusion through a limiting boundary film. The authors presented an argument, based primarily upon diffusion and temperature co efficient data, supporting diffusion through a limiting boundary film. A series of tests with solutions having different initial copper concentrations showed that first order kinetics were followed but the specific rate decreased slightly as the initial concentration increased. The specific rate was found to increase slightly with a decrease in pH. The authors presented the hypothesis that the variation has to do with a change in the structure of the diffusion medium, which is strongly hydrogen bonded and may be + influenced by increased concentrations of H^O ions. They pre sented, an alternate explanation that the increased hydrogen ion concentration results in more effective surface cleaning of the 18 iron surfaces under steady state conditions. The results of Nadkarni and Wadsworth supported the findings of Nadkarni et al in that the kinetics were fir^t order and followed the equation: d(Cu++) — = -ACCu )KQ (16) At lower stirring speeds, where bulk solution diffusion is involved, the cementation rate followed the equation: d(Cu++) ++ - = -ACCu )K n1/2 (17) dt o where n is the stirring soeed in r.p.m. The constant K in-eludes the diffusion coefficient, the kinematic viscosity of / the solution and units to convert r.p.m. to angular velocity. The value of KQ at 'i5°C, determined at maximum stirring rates, -3 -1 _? is approximately 3.4- x 10 mm. in . The authors proved, from theoretical considerations of electrochemistry, that the back reaction due to the presence of ferrous iron in solution would not be expected to influence the cementation kinetics. They conducted a series of experi ments which substantiated their argument. Experiments with and without ferric ion in solution showed that the actual rate of cementation based upon copper removal from solution was almost the same in both cases but the presence of ferric ions drastically increased the excess iron used. The ferric ion reaction rate was found to be appre ciably faster than the cementation reaction. The hypothesis that metallic copper enhances the con-sumption of metallic iron by reacting directly with ferric ion in solution according to the reaction: Cu(s) + 2Fe + + + Cu + + + 2Fe + + (18) was checked experimentally by measuring the rate of reaction between ferric ion and metallic iron with and without metallic copper in the reaction vessel. Their results are presented in Figure 3, where the total change in iron A(Fe++), which is the total iron concentration minus the initial, is plotted for any time t. Four separate experiments were carried out and in no instance was cupric ion added initially. All pertinent data are shown in Figure 3. The lower curves of Figure 3 clearly show that metallic copper increases the rate of reaction of the ferric ion in close proportion to the available surface area. The upper curves are plots of the concentration of cupric ion in solution versus time. At the maxima for each curve, the rate of change of cupric ion concentration is zero, representing conditions where the rate of cupric ion production is just equal to the rate of re moval of cupric ion due to the cementation reaction. The slope of these curves at any time t was expressed by the authors in the equation: • d(Cu++) +++ ++ — = APi> (Fe )K ' - A_ (Cu )K (19) dt Cu(s) o Fe(s) o where A0 is the surface area of the copper and K 1 is the Cu o specific rate for the metallic copper-ferric ion reaction. The equation represents competing reactions dependent upon the sur-r ^ CP + o 0-4 0-2 0 0-4 C7> O 0-2 0 @^-Q—g-==;CB)^_0_ -o- 45°C,pHi = 2-10 0 Cu area 25-4 cm2 50-8 cm2 76-2 cm2 (Fe+++); = I-l g/1 , Fe area = 13-1 cm' = 076 aim. 80 TIME (minutes) 120 160 Fig. 3 - Influence of metallic copper in accelerating iron consumption in the presence of ferric ion. 21 face areas of the metallic copper and iron respectively. The shapes of the curves are also dependent upon the ferric ion-metallic iron reaction, which helps diminish the ferric ion concentration and consequently the rate of reaction with metallic copper. At the maxima Ap (Fe++)K ' = Ar (Cu++)K (20) • Lu o re o and for times greater than those corresponding to the maxima the ferric ion concentrations diminish to zero and the cupric ion concentrations diminish due to the cementation reaction. Nadkarni and Wadsworth conducted a series of tests under atmospheres of nitrogen, pure oxygen, air, and pure hydrogen. With a nitrogen atmosphere -and different hydrogen ion concentrations the increases in excess iron consumption corresponded to conditions of low cupric ion concentrations resulting from extended reaction times. Excessive iron con sumption occurred at. cupric ion concentrations lower than 0.15 grams per litre and increased with a decrease in the initial pH. Tests in an oxygen atmosphere showed that the ferrous iron in solution due to the cementation reaction reached a maximum and then decreased with time. The excess iron con sumption due to oxygen was explained by the following reaction which is dependent upon the amount of copper in the solution Cu(s)+2H+ + l/2 02 •> Cu + + + H20 (21) The cupric ion formed would . consume .additional iron by the cementation reaction. Reaction (21). is enhanced by the oxida tion of ferrous iron to ferric according to the equation: 2Fe + + + 2H+ + l/2 02 •» . 2Fe + + + + H20- (22 ) whereupon the ferric ion reacts with metallic copper as explain ed previously. Oxygen also consumes iron by the direct reaction Fe(s) + 2H+ + l/2 02 -> Fe + + + H20 (2^) The data revealed that approximately twenty-five percent of the excess iron consumed in pure oxygen is due to reaction (23) and seventy-five percent is due to reactions (21) and (22). The findings of Nadkarni et al and Nadkarni'and Wadsworth have been used extensively in the analyses of the results of the experiments presented in this thesis by the present author and are summarized below for further reference. Nadkarni et al (1) The rate of the cementation reaction is first order and is proportional to the surface area, of the iron pre cipitant according to equation d + 4. + + —[Cu+ + ] = -[Cu ]K A at w (2) The rate of the cementation reaction increases with agitation up to a limit after which it is independent of the degree of agitation. (3) Particles of cast iron which are retained on a 60 mesh screen do not react completely during cementation but retain a small nodule of iron at the center. 23 Nadkarni and Wadsworth (1) The specific rate of reaction increases slightly with a decrease in pH. (2) Ferrous ions do not interfere with the kinetics of c erne ntit ion. (3) Ferric ions increase ..the iron consumption by re-dissolving metallic copper. (4) Excessive iron consumption by the metallic iron-hydrogen ion reaction occurs below a cupric ion concen tration, of 0.15 grams per liter. (5) An oxygen atmosphere enchances iron'consumption. (6) A hydrogen atmosphere reduces iron consumption but the reduction is not significant over practical time intervals. 24 CHAPTER IV REVIEW OF CEMENTATION METHODS A. HISTORY OF CEMENTATION The history of copper cementation with iron dates back to about 1500 A.D.^i)* Ri° Tinto in Spain produced copper commercially by precipitation on iron during the sixteenth century, and "As early as 1G70, a concession is recorded a's having been granted for the recovery of copper from the mine waters at Rio Tinto, Spain, presumably by precipitation with iron. In 1752 also, the method is recorded as being used, this time to pre cipitate the copper from the liquors obtained from the leaching of the leaner ores after natural cementation.",go) Rio Tinto used the launder system to recover the copper from solution. The solutions were fed by gravity through a system of troughs filled with any available form of scrap iron and the precipitated copper which fell to the bottom of the trough, was removed with hand shovels. The launder method of cementation was employed on a small scale in the Butte, Montana area as early as 1888 and, in 1910, production of copper by this method amounted to 2 ,2 79 tons(25)« Greenwalt (T. 3) refers to sponge iron as a precipitant for copper in England as early as 1837 but scrap iron .and detinned cans have been the principal precipitants to the present. More recently, the cementation process has been, con ducted in a wide variety of equipment which was all designed 25 to provide suitable contact between the iron and the copper-bearing solutions. Systems employing drums, tanks and inverted cones have been used; however, the launder system still re mains the most•popular. All of the methods mentioned are described in this chapter. Only the launder system is discussed in detail as very little data have been published about the other methods. B. DRUM PRECIPITATORS At the Mantos Blancos operation^) scrap iron is used in inclined cementation drums of 7,900 gallons capacity to precipitate residual copper after precipitation with sulphur dioxide. The Miami Copper Company^uses wood stave drums, nine feet in diameter and twenty-eight feet long, with rubber linings, to precipitate copper from their entire leaching opera tion. A one hundred horsepower motor is required to drive the drums at a speed of eight rpm. The copper, in the discharge from the drums, is separated from the iron by magnetic separa tion and flotation. The grinding action of the drums produces fine particles, some colloidal in size, which are difficult to recover. The drums are difficult to feed and clean, and require large driving mechanisms. No operating data have been published. C. TANK PRECIPITATORS At Bisbee , ArizonaQy^, circular tanks twenty-four feet in diameter by ten feet high, placed in two rows of six, in series, are used to precipitate copper from the leach liquors with scrap iron as the precipitant. The tanks have false bottoms to support the scrap iron. An acid-proof stirring arm underneath the false bottom operates from a shaft in the centre of the tank to agitate the solution when necessary and to remove the precipitated copper. The discharge on the bottom of the tank is partly conical to assist in this operation. The liquor enters the tank alongside the agitator shaft and is delivered underneath the false bottom, flowing up through the iron where it is discharged over a peripheral launder to the next tank in the series. • Mo data are available from the opera tion of the tanks; however, the principles employed are similar to those used by the Kennecott cone-type precipitator which was developed a few years ago and is described in this paper. D. CONE PRECIPITATORS The most significant mechanical improvement in the precipitation of copper from dilute leach solutions was the development of the cone-type precipitator by the Kennecott Copper Corporation ^ 6,29)* Figure 4 is a sketch of the Kennecott precipitator developed for cementation with detinned cans as the precipitant. The precipitator consists of a cir cular tank into which is mounted an inverted cone. The outer tank contains a forty-five degree, sloped, false-bottom floor extending from one side of the tank to a bottomjSide discharge at the opposite side. The rim of the inner cone is extended to the tank wall by a heavy gauge, stainless steel screen. The cone supports a pressure manifold.that consists of six vertical legs, ig. 4 - Sketch of Kennecott's cone-type precipitator for detinned cans. 28 with each leg containing a series of nozzles directed inward from the tangent to the cone and upward from the angle of the legs of the manifold. The nozzles are arranged in such a manner so: to create a vortex when the copper bearing solutions are pumped through the manifold into the cone. The cone and the area of the tank above the stainless steel screens are filled with shredded detinned iron scrap such as is commonly used in the precipitation of copper. The precipitation cone is a self-cleaning, continuously operating unit. The copper precipitates settle down through the steel screen and accumulate on the sloped false bottom of the tank. The copper can be dis charged intermittently with the use of a pneumatically operated valve or bled continuously through a small diameter pipe into a thickener or holding basin. The cone precipitator requires very little manual labour as it is amenable to mechanization. After preliminary field tests, Kennecott personnel predicted a cement copper which would assay 90-95 percent copper and a reduction of iron consumption from 2.33 pounds per pound of copper with the launder system to 1.5 8 pounds per pound of copper with the cone Q g)• A recent publication^y^ reports an iron consumption of 2.5 pounds of iron per pound of precipitate with anticipation of a reduction to 2.0 pounds. The average grade of the preci s'; pitate is 85 percent copper. A.E. Back of Kennecott Copper Corporation developed a cone precipitator to recover copper from dilute leach liquor with particulate (sponge) iron^j. The precipitator is similar in design to the inverted cone described above, however, the * Estimated theoretical iron Consumption / •1.4:1.0 29 sides of the cone are at an angle of 6 5 to 8 5 degrees in the bottom section (the reaction zone) and the angle is decreased in the upper section to increase the cross-sectional area and decrease the flow. The flow is adjusted to 0.5 feet per minute at the overflow to ensure that precipitated copper Is not forced to overflow the cone. A system of "burpers" (baffles) near the top of the cone directs hydrogen, formed in the acid-iron reaction, to the centre of the cone where it activates valves to control the input of iron to the cone and the discharge of copper from the cone. A twenty foot diameter cone will handle 1,000 gpm of solution. The iron consumption is lower with this cone using particulate Iron because very little acid is con sumed in the operation; however, no operating data have been published. E. LAUNDERS The launder system used by Rio Tinto almost three hundred years ago is still the most popular method of preci pitating copper from dilute leach liquors. A wide variety of launder shapes and sizes have been employed. Many of the plants in use today were constructed-several decades ago and the original criteria for design have long been obscured by numerous expansions and alterations. Regardless of the design of the plant the method of operation is quite standard. Monninger£23) described the operation by dividing it into five steps as follows: (1) Charging of the metallic iron to the precipitation launders by hand or mechanical devices. 30 (2) Introduction of the pregnant solution to the metallic iron where the actual precipitation takes place and is allowed to proceed until near-total consumption of the cupric ion. (3) Removal of the precipitated copper from the launders by washing or a mechanical loader. (H) Washing and screening of the precipitated' copper through a trommel screen for removal of soluble salts, foreign solids, and any unconsum-ed iron. (5) Drying the copper precipitate. Improvements to the original launder system have been recent, and limited to the introduction of mechanical loading devices for the iron and copper and the injection of solutions through pipes at the bottom of the launders (activated launders). Mechanical devices have minimized the labour requirements; however, the launder system continues to consume two to four times the theoretical iron requirements. F. A DISCUSSION ON THE LAUNDER SYSTEM Table 1 is a summary of the available operating data from several of the large leaching operations which have used, or are now using, the launder system to recover copper from their leach liquors. In the cases cited the barren solution is either void of sulphuric acid and ferric ion or contains very little which explains the corresponding high iron consumptions:' The relatively low iron consumption at Anaconda's Weed Heights operation is due to the high copper concentration in comparison See Table la, Page 31a. TA3LE 1 - SUMMARY OF OPERATING DATA FROM COPPER LEACHING OPERATIONS. OPERATION REF. Cu+ + H2S0i4 pH + + Fe Fe % Copper in • Cgpl) (gpi) (gpl) ' (gpl) Pre g. Barren P re g. Barren Preg. Barren Preg. Barren Preg. Barren Product Fe :Cu Ray ' • (4) 1.0 0.08 2.0 0 . 3 2.5 3. 5 1.1 3.7 1.2 0.12 na 1.8 Chino • (4). 1.5-2. 3 0.1-o. 3 0.6 0.2 2.5 3.5 0.9 3.5 0.6 0.2 na na Silver 3. (4) 1.09 0 .006 0 . 6 0.08 2. 3 3.6 0 . 01 2.. 0 8' 0 . 57 tr. na 1.5 Inspirat.'n 1933 (2) 4.6 na na na na ' na na na 5-11 na . na 1.86 196 3 (15) 3.2 0.2 2.0 0 . 0 na na 6 .4 9.9 0.2 0.0 ' na 1.4 Rio Tinto 1943(1) ( 32) 2.1. 0.04 10 . 7 9 . 9 na na 15. 5 19. 8 1.4 0.1 62-80 1.56 . Anaconda ' (Butte) 3.2-1957(ii) (26) 0 . 3 na na na 3.4 na (tote .1 Fe+ + S Fe * 0.5) 70 1.75 1966(iii) 0 . 16 na na na 2.1 na 0. 21 na 0.1 na na na (Weed Hts) 1954 (29) 24 . 4 0.09 5.9 o. 6 na na 3*5 38. 8 8.9 0.0 na • 1.25 1963(iv) (4) 22 na 5 na na na 3 na 7 na 8 3 1.22 Ohio Cu 19 2 5 (v.) (3) 2.0 0.6 na na na na 0. 15 na 0 . 3 na 60-90 1.1 Esperanza (vi) (4) 1.0 tr. .0.2 0.1 2.8 3. 75 0.01 1.1 0.1 0 . 01 na 1.3 5 (i) Bi 0.002 (ii) 7.5 gpl solids (iii)' 1.7 gpl solids (iv) solution diluted to 14 gpl Cu (v) CaSO^ 1.4 gpl, MgS04 2.0 gpl (vi) 0.12 gpl insol. TABLE la - Fe:Cu CALCULATIONS FOR SEVERAL LEACHING OPERATIONS. Operation Cu Extracted (gpl) H2S04 Con sumed (gpl) Fe + + + Consumed . (gpD Reported Fe :Cu Calculated Fe :Cu @ 10 0% Fe Calculated Iron Factor (% metallic Fe in precipitant) Calculated Fe :Cu @ 85% Iron Factor Ray 0.92 1.7 1.08 1.8:1.00 2.52:1.00 -.- 2.97 Silver Bell 1.09 0.52 0.57 1.5:1.00 1.41:1.00 0. 94 1.61 Rio Tinto 2.05 0 . 8 1.3 1.56 :1.00 1.41 :1.0-0 (reported 1. 44 :1. 00) 0.91 1.66 Inspiration 1963 3.0 2.0 .0.2 1.4:1.00 1.29:1.00 0.925 1.52 Weed Heights 1954 24. 3 5.3 8.9 1.25 :1.00 1.19 :1.00 0.95 1.4 0 Comments The calculated figures are based on the figures presented in Table 1 and equations (12), (13) and (14). Since the author has neither a thorough knowledge of the methods used to obtain the figures in Table 1 which were reported by the respective operations(possibly yearly averages) nor the complete composition of the . solutions, the above calculations are unrealistic, For example, at the Ray operation 1.6 of the total 1.7 gpl of H2S0^ consumed in the (Cont'd on page 31b) 31b launder system must be attributed to side reactions involving (Al+++, Fe+++, etc.) in order to obtain the iron consumption quoted from the reference. The iron consumption should be quoted in terms of precipitant required per unit weight of metallic copper produced, but, it can be quoted in terms of precipitant per unit weight of cement copper, metallic iron per unit weight of metallic copper or metallic iron per unit weight of cement copper. Since the precipitant (detinned cans) can contain from 8 0 to 9 5 percent metallic iron and the cement copper can contain from 60 to 90 percent metallic copper, the figures quoted for the iron consumption can vary considerably. 32 to the concentrations of acid and ferric ion. .The acid and ferric ion in the pregnant solution is consumed in the process as in the other operations. The effect of the acid-metallic iron reaction can be reduced to some extent if stagnant areas within the launders are eliminated by changing the shape of the launders and in jecting the solution through pipes in the bottom of the launders^23)• A minimum surface to volume ratio will reduce excess iron consumption by diminishing solution contact with oxygen(24)« A graph presented by Taylor and Whelan in a report on the Rio Tinto operation in Spain can be explained using the principles revealed by the experiments of Nadkarni and Wadsworth. Figure 5 is a reproduction of the graph illustrating the variation, in the content of the leach liquors during their passage through the launder system. The graph shows that the ferric ion is consumed at a faster rate than the cupric ion; however, the rate of removal of the cupric ion from solution increases after the consumption of most of the ferric ion. In the presence of high concentrations of ferric ion, the overall removal of cupric ion from solution is retarded by the redissolution of copper according to equation (18). An increase in the rate of iron consumption due to the hydrogen ion-metallic iron re action occurs when the cupric ion concentration is decreased to below 0.2 grams per litre. Iron consumption by this re action becomes quite significant as- the solution approaches the end of the launder system. 10. 50 1 1 1 1 1 1 1 1 I I 9.25 0 10 20 30 40 50 60 70 80 90 100 Percentage of launders traversed 5 - Variation in the content of leach liquors during passage through the launder system at Rio Tinto, Spain. -34 In any launder system the ferric ion is responsible for the greatest consumption of excess iron. Monninger states that metallic iron is the most efficient material for the re moval of ferric ion. There are at least three methods of decreasing the concentration of the ferric ion in the pregnant solution prior to cementation, namely, precipitation of the ferric ion as Fe(OH)g with lime, reduction to ferrous ion with sulphur dioxide, and reduction to ferrous ion with pyrrho-tite, marcasite, or chalcopyrite. The use of one of these methods is a matter of economics primarily dependent on the locality of the plant. The cone-type precipitator developed by Kennecott takes advantage of all the known improvements to the cementa tion process with either detinned cans or scrap iron as the precipitant; however, iron consumption remains excessive and the grade of precipitant is comparable to that obtained in the launder system. Since very few operating data have been published by Kennecott it is difficult to establish the reasons for the high iron consumption and the low grade of cement copper produced in the cone-type precipitator. Detinned cans are used for the precipitant therefore the following reasons for the high iron consumption could be common to both the launder system and the cone-type precipitator: (1) Virtually all of the cupric ion is removed from solution which ensures that some excess iron consumption will result from the metallic iron-hydrogen ion reaction. (See equation 1.4). (2) The fact that most of the ferric ion is removed 35 from solution indicates that there is a pro longed contact between the metallic copper and the ferric ion. (3) The detinned cans contain ninety percent or less of their'weight in unoxidized iron. This ' fact not only accounts for an increase In iron consumption but also a low grade of cement copper. (4) A loss of metallic copper between the precipi tation process and the drying process would account for a considerable increase in the calculated figure for iron consumption. Such a loss has never been reported; however, the author has suggested this as a possible reason because iron consumption is usually reported as being much higher than that which can be attri buted to equations (12) , (13) and (14). A further study of the available literature yielded several reasons for the low grade of the cement copper produced. G. THE PRODUCT OF CEMENTATION PLANTS: CEMENT COPPER There are at least four reasons why the cement copper seldom contains more than eighty-five percent copper, namely: 1. hydrolysis of dissolved cations 2. solids in suspension 3. oxidation during drying 4. impurities in the precipitant. 36 1. Hydrolysis of Dissolved Cations Table 2 shows the analyses of several leach liquors and illustrates the wide range of concentrations of cupric ions and impurity ions which can result from acid leaching of copper ores. The pregnant solutions at Anaconda's Weed Heights operations contains up to eight grams per litre of aluminum oxide plus phosphate, silica, calcium, and magnesium . (23) Monninger has found from plant experience that the grade of the final copper precipitate is inversely proportional to the amount of soluble salts present as impurities in the pregnant solution. At a pH approaching 3.0, most of the salts of the Al + + +, +++ +++ Fe , and Cr ions have precipitated out of solution. During cementation, hydrolysis may even occur when the solution acidity is considerably below 3.0 due to local consumption of acid at the iron surfaces or to high initial concentrations of the ions. The salts, once precipitated, will not redissolve even in solutions with abnormally high sulphuric acid concentrations. Ferrous iron, although present in high concentrations, is rarely a source cf contamination as a pH approaching 5.0 is required before extensive hydrolysis and precipitation occur. Most of the precipitated salts are retained in the final product, thus lowering the grade and increasing the overall moisture content. The retention of salt-loaded solutions in the final copper precipitate also lowers the grade since these salts, unless removed by washing, will remain with the copper.. Wartman and Robertson investigated the contamination of the cement copper by the hydrolized iron salts. They found 3 7 Table 2 : Analyses of Leach Liquors Composition of Leach Liquor Leach Liquor A B C D E F Cu 2.0 *Cgpl) 1.14 (gpl) 7.10 (gpl) 5. 94 (gpl) 2.26 (gpl) n.a. ' (gpl) Fe 0.15 n.a. n.a. n. a. 18.0 n.a. Fe 0.3. n. a. n.a. n.a. 2.0 n.a. Total iron 0.45 0. 36 11. 3 13.02 20.0 n.a. H2S04 n. a. 0 . 35 16.4 6.8 11. 0 n. a. Zn + + n. a. 0.23 0.08 1.16 2.0 1. 37 As + + + n. a. <0.01 0 .017 <0 . 01 0. 16 0 . 81 Sb + + + n. a. <0 .005 <0. 01 <0.005 n.a. 0.10 Bi++ + n.a. n.a. n.a. n.a. n.a. 0.22 CaO 0.58 0 . 75 0. 76 0. 94 0.002 n.a. MgO 0.67 1. 16 0 . 66 1.16 n.a. 5. 74 Pb+ + n. a. 0 . 01 <0.002 0.-016 n.a. 0.05 Si02 n.a. 0. 18 0 .12 0.43 n.a. n.a. Al + + + n.a. 0 . 35 1. 24 1.19 n.a. 2.00 Mn + + n.a. n. a. n.a. n. a. n.a. 9.7 "gpl - grams per liter 'n.a. - figures not available 38 that at a pH of 4.0 most of vhe ferric iron had precipitated and most of the acid had been neutralized. The precipitated salts contained thirty percent iron and ten percent of the original copper in the pregnant solution. R.D. GrovesQ^') investigated the effect that certain solution impurities might have on the purity of the cement copper. He added the impurities listed as solution F in Table 2 to a solution containing 7.1 gpl copper, 16.4 gpl . sulphuric acid and 11.3 gpl total iron. He found that the presence of aluminum, magnesium, manganese, and lead did not appreciably affect the purity of the product if the powder was thoroughly washed before drying. The presence of zinc, antimony, and arsenic contaminated the powder to a minor degree while the contamination from bismuth was extensive. The impurity contents of the powders produced were, in percent, Zn 0.13, '' Sb 0.20, As 0.17, and Bi 1.19. Sintering of the product removed ninety-seven percent of the arsenic and fifty-three percent of the antimony but the bismuth and zinc were unaffected. 2. Solids in Suspension At Anaconda's 3utte operation the pregnant solution has contained suspended solids up to 7.5 grams per litre^g). Recently a method of clarification by flocculation and settle ment has been practiced which reduces the solids content of -makeup'water, originating from underground, to 0.04 grams per litre^2g)» Contamination of the cement copper by suspended solids entering the -precipitation plant can be reduced by in creasing the solution flow to prevent settling; however, 39 effective removal of the solids from the final product can only be obtained with proper washing methods. 3• Oxidation During Drying-Very few data have been published about the methods employed to dry the cement copper from the cementation plant. Since most of the cement copper produced assays less than ninety percent copper and must be smelted, little effort has been made to prevent further reduction In grade by oxidation. In most cases drying is carried out under atmospheric pressures either in the sun ^ ^) or on gas-fired hot plates^g^. Anderson and Cameron^j reported a drop in copper content of five percent during drying at the Ohio Copper Company's operations. Pressure briqueting of the cement copper.has been used with minor success. The present author dried several samples of cement copper in a hydrogen atmosphere at elevated temperatures and compared the grade of copper obtained with identical samples dried at atmospheric temperature and pressure by washing with alcohol and acetone respectively. Hydrogen drying was found to give the better results. Oxidation, during the alcohol and acetone drying procedure, resulted in a reduction of copper content by approximately one percent. On a commercial scale either drying In a reducing atmosphere at elevated temperature or drying in a vacuum would be effective in preventing extensive oxidation of the cement copper. 40 4. Impurities in the Precipitant Since the grade of cement copper which is produced with detinned cans as the precipitant rarely exceeds eighvy-five percent copper at least ten to fifteen percent of the weight of the cans must consist of rust and miscellaneous residue from burned garbage. Jacky, reported a contamination of four \ I 8 ) percent, in the cement copper, from the oxidation of detinned cans that had been stored for four years. A copper powder containing more than ninety-nine percent copper was produced in the laboratory by R.D. GrovesQ4) using several forms of unoxidized scrap iron; however, it is unlikely that such a grade will be obtained under field con ditions. It should be feasible to produce a cement copper approaching ninety-nine percent purity with the use of pH control to minimize hydrolysis, washing to eliminate contami nation, suitable drying to reduce oxidation and a high purity iron to reduce contamination from impurities in the precipitant. The ideal precipitant, for the recovery of copper from solution with iron, should contain at least ninety-five percent unoxidized iron, have a large surface area to weight ratio, and be available in large quantity at a cost competitive with detinned cans. Several iron powders satisfy these conditions. ' ' 41 CHAPTER V IRON POWDER: A SOURCE OF IRON FOR CEMENTATION A. SUPPLY The former argument that iron powders were too expensive and not available in sufficient quantities was investi gated in a survey of several large producers of these products in the United States. Iron powders ranging in purity from ninety-two to ninety-eight percent metallic iron are available at a cost slightly above that of tin cans. Several producers now have facilities for large daily productions. The supply could be increased in proportion to the demand. The analyses of several readily available iron pow'ders are listed In Table 3. grinding scrap cast iron; however, some chemical iron Is available. Chemical iron can be produced from a basic ferric sulphate solution by reduction with hydrogen or carbon monoxide, according to the equations : Most of the iron powder available is produced by 4FeS04 + 4H20 + 02 -> ( Fe ( OH ) ) 2 . Fe2 ( S04 ) 3 + H0 SO (24) (Fe(OH)3)2.Fe (S04) -> .: 2Fe203+ 3S02 + 3H20 + l 1/2 02 (25 ) Fe203 + 3H2 -»• 2Fe+3H20 (26) Feo03+3C0 -v 2F.e+3C02 (27) Table 3: Chemical Analyses of Iron Powders Iron Powder Percent Composition Sample A B C D E F Elements Total Fe 9 8. 7 9 7.-9 3 98.5 93 91. 9 91. 5. Metallic 95 + 91 88.5 iron C . •;0. 02 0.02 3.9 3.1 3.0-3.5 Si 0.08 2.0-2.5 Mn 0. 25 0.4 0. 55 0. 7 0.65 0.5-0.7 S 0.02 5 0 .005 0.005 0.01-0.04 0. 01 0.01 P 0 . 014 0.012 0.012 0.02 0.1-0.3 Ni 0.0 4 Cr 0.05 V 0. 01 As 0. 001 • Pb 0 .002 • R"2 (loss) 0 . 57 Acid Insol 0.3 3 , 0 4.5 Type of ground ground ground iron cast hydro ge n hydroge n cast chemical cast powder iron reduced reduce d iron iron iron 43 Chemical iron is also-produced from limonite according to the reactions: 'Fe203.H20 + 6HCl •* 2FeCl3+4H20 (28) 2FeCl3+F{2 •*• FeCl2 + 2HCl (29FeCl2+H2 + Fe + 2HC1 (30) The hydrochloric acid produced in the process is recovered and used for leaching. Higher costs of production eliminate most chemical irons as precipitants of copper from dilute leach solutions at the present time. The prospect of producing iron powder from leach solutions, which normally contain high con centrations -of the ferrous ion, is'presently under investigation, by. at least one of the major copper producers in the south western United States. Sponge Iron is produced from either magnetite or roasted pyrite concentrates. The sponge, iron made from pyrite concentrates has considerably greater exposed surface area and is preferred, for this reason. Iron powder, throughout this report ..refers to high purity ground- cast iron and chemical iron as described previously. Table 3 is a summary of the chemical analyses of iron powder from several sources. The analyses were supplied by the manufacturer. Custom screening is available from most suppliers. Although high purity sponge iron would be as satisfactory as the powders tested, none was available at the time the experiments were conducted.-44 " B. ADVANTAGES: Iron powder can be shipped in air-tight bags contain ing one hundred pounds of powder, or in bulk by railroad end highway tankers. The powder weighs 12 0-15 0 pounds per cubic foot whereas detinned cans weigh 16-17 pounds per cubic foot, A large tonnage can be stored in enclosed buildings thus pre venting oxidation by exposure to the elements. A high purity iron powder will provide, maximum surface area of metallic iron and can be easily, agitated in a simple reaction vessel to ensure maximum contact between the metallic iron and the cupric ion s. The powder can be handled mechanically with conveyors, chutes, and hoppers or forced through piping by air or water. With a large scale operation the actual cost of the iron powder could be less than that of cans per pound delivered'to the precipitator units. C. IRON POWDER PRECIPITATORS An effective precipitator which utilizes sponge or •'. powdered iron to recover copper from dilute leach liquors was developed by A.E. Back et al on behalf of the Kennecott Copper Corp. The precipitator has been described in a previous section of this paper. The "burpers" used to remove hydrogen from copper particles and to funnel the hydrogen to the centre of the precipitator where it is used to control the batch process, are the main feature, of the precipitator. The copper is remov ed through the botto.m of the cone when valves are activated by 4 5 the hydrogen flow meter. A thorough appraisal of Back's precipitator is impossible with the limited information that has been published. The use of an inverted cone allows maximum contact between the metallic iron and the copper-bearing solutions. Control of the process with hydrogen is unique and apparently effective; how ever, it is a potential source of disruption to the operation. Although the loss'of.metallic copper over the rim of the pre cipitator was reduced from 2.7 to 0.4 percent with the "burners" in small scale treatment vessels a greater loss is inevitable with a large scale vessel due to a decrease in the degree of control. Since Back's precipitator allows the precipitated copper to remain in the cone until most of the iron has reacted there is a maximum contact between metallic copper and ferric ion which contributes to excessive iron consumption. T.D. Smyth^g) developed and carried out experiments with a cone-column precipitator based on the same principles as the Kennecott precipitator but without the "burpers". A small-scale acrylic model was tested at a leaching operation in the Highland Valley of British Columbia. The precipitator consisted of a seven and one-half inch diameter, five foot high column mounted on a thirty degree angle cone. The powdered iron was fed through the top of the column and the solutions were injected into the cone through a nozzle designed to produce a vigorous stirring action in the cone. Smyth's precipitator was capable of producing a high purity copper (96.0 percent reported); however, it had several disadvantages. The end point of the reaction was not obvious and there was no method outlined in the report which would allow automatic control of the batch operation as with Back's precipivator. A'considerable amount of the precipitated copper was forced over the top of the column requiring the use of a settling tank to recover this material. The excessive loss of precipitated copper was illustrated by the iron to copper ratio of 1.25 to 1.0 obtained during test runs with a synthetic solution containing cupric sulphate and sulphuric acid but no ferrous or ferric iron. No experiments involving variations in the concentrations of ferric or ferrous ions and free sul phuric acid were reported. The series of experiments conducted by the present author were prompted by the lack of available literature on the subject of recovering copper from solution with iron powder and the implications of previous investigators that iron powders offered several advantages over precipitants new in use. CHAPTER VI EXPERIMENTAL PROCEDURES. The experiments of Nadkarni and Wadsworth revealed the importance of minimizing contact between ferric ions' and metallic copper. In the batch precipitators used by Back et al and Smyth, the solids were contained in the precipitator until sufficient iron had reacted to produce a satisfactory cement copper product. . In both cases the process was not con tinuous and excess iron was consumed as a result of the ferric ion-metallic copper reaction. Back et al reported a reduction in ferric ion concentration from 0.75 to 0.01 grams per liter during precipitation of copper with particulate iron in a precipitator with a volume capacity of 125 gallons per minute •and a charge of 500 pounds of precipitant. A. APPARATUS AND METHODS EMPLOYED TO CONTROL VARIABLES. The precipitator used in the experiments described herein was similar In design to that used by Smyth. It con sisted of a vertical glass column, 1.22 meters long with an inside diameter of 7,52 centimeters, mounted on an inverted sixty degree cone as shown in Figure 6. The iron powder was added through a tube (h) inserted in the top of the column. The pregnant solution was injected through the apex of the cone (f). Suitable agitation of the iron powder was achieved V. J LEGEND Pregnant solution tank Jabsco pump Precipitator valve Drain valve By-pass valve Precipitator cone Precipitator column Iron feed tube Settling tank drain Barren solution tank Settling tank overflow Settling tank Precipitator overflow Sketch of precipitator used in the experiments. 49 . by the turbulence resulting from the high input velocity of solution to the cone through the 0.5 cm.(I.D.) input pipe at the apev of the cone, The input velocity was approximately 390 feet per minute. Back^^ recommended an input velocity of 4 30 feet per minute for a precipitator with a capacity of 12 5 gallons per minute.. In all of the experiments described in this paper, the velocity of flow through the precipitator was adjusted to force precipitated copper over the top of the column, A settling tank (1) was used to collect the copper in the effluent solutions. All of the iron used in each experiment was added as a batch at the beginning of the test ; however, the results indicated that continuous addition of iron would be more appropriate. Since the iron was added and removed from the top of the precipitator rather than through the apex of the cone there would be no need to interrupt the input of solution. The process could be operated continuously. The experiments described and discussed herein were conducted to determine the relative importance of the para meters involved in the precipitation of copper with powdered iron or at least to establish a basis for further research which would serve this purpose. The factors investigated were the velocity of solution flow through the column, the particle size of the iron powder, the free acid concentration of the pregnant solution, the ferrous and ferric ion concentration of the pregnant solution, and the temperature of the pregnant solution. The results of variations in the above factors are discussed in Chapter VII. 50 • The temperature of the pregnant solution was ad justed to 15.6°C (60°F) at the beginning of each test, except where noted, but no attempt was made to maintain the. tempera ture at this figure throughout the duration of the experiment. A rise of one to two degrees centigrade was observed with solutions initially at 15.6°C. The solutions were forced through the column by a Jabsco plastic pump. Solution flow was controlled by the pre cipitator valve (c) and the by-pass valve (e) shown in Figure 6. The flow rate through the column was calculated from the time required to fill the settling tank and the dimensions of the column. Reproducible flow rates were obtained with a variation of less than two percent. Iron powder "A" in Table 3 was used for all tests, except where noted. The powder was divided into the two size fractions of minus 100 plus 200 and minus 200 mesh (Tyler) to gain some knowledge of the size requirements for precipitation. . The author realized at the start of the experiments that several more size ranges of iron powder should have been tested; how ever, the number of experiments Involved would have been prohibitive. Reagent grade cupric sulphate, ferrous sulphate, ferric sulphate and sulphuric acid were used in the preparation of the pregnant solutions. All the chemicals were weighed out and dissolved in hot tap water prior to each experiment with the exception of the ferric sulphate which was allowed to stand at least twenty-four hours to ensure that it dissolved. The hot solutions were mixed in the pregnant solution tank by the Jabsco pump, with the precipitator valve closed, after dilution to the required volume and adjustment of the temperature with tap water. The. iron powder was added as a batch as soon as solution started to enter the cone. A sample of the pregnant solution was taken prior to each teet and analyzed for the salts and acid added. The duration of each test, except where noted, was ten minutes and fifteen seconds. At the end of each experiment the barren solution was analyzed for iron salts and free acid. The solids which overflowed the column and collected in the settling tank (overflow) were analyzed for metallic copper and iron as were the solids which remained in the pre cipitator (underflow). Before analysis the samples were dried in alcohol and acetone. A portion of the dried sample was taken from each quarter and the portions were combined to obtain a sample for assaying. The weight of iron consumed per weight of copper produced was determined from the weights of the solids in the underflow and overflow and their respective assays. Throughout the thesis the ratio"of iron consumed to copper produced is called the iron consumption and denoted as Fe:Cu. Since the weights of the iron powder, underflow and overflow were rounded off to the nearest 0.05 grams the experimental error in the Fe:Cu ratio could be two percent. The above procedure did not yield satisfactory results with ferric ion in the pregnant solutions as the calcu lated iron consumptions were too erratic. Part of the error could have resulted from a delay in analysis which allowed the 5 2 ferrous ion to oxidize to the ferric ion. Since the ferric ion reactions warranted further .investigation part of the experimental procedure was altered. The same procedures as described for the previous tests were followed, with the ex ception that larger samples of iron were added to the precipitator and samples of the.effluent were taken at specific intervals throughout the duration of the test. Iron samples from three different sources were used in the tests with ferric ion. They are denoted as iron powders A, B, and C in Table 3. 1.Analyses _ The following procedures were followed- to determine the concentrations of salts and acid present In the pregnant and barren solutions. Copper determinations were made colorimetrically using a Fisher Electrophotometer. The solutions were prepared for colorimetry by oxidizing the ferrous iron to ferric iron with sodium hypochloride and precipitating all the iron present with excess ammonium hydroxide. The concentration of copper in the cupric ammonium complex (formed by the addition of 'excess ammonia) was compared against a standard, one thousand parts per million, cupric ion solution supplied by Fisher Scientific Co. Ltd. The possible error with the equipment used was five percent. The total amount of ferric ion and cupric ion in solution was determined by titration with sodium thiosulphate in the presence of potassium iodide and a starch indicator. The ferric ion concentration was obtained by sub traction of the amount of cupric ion (determined colorimetrical-53 ly) from the above figure. An error of two percent could have resulted from titration. Free sulphuric acid concentration was determined by titration with sodium carbonate in the presence of excess sodium thiosulphate. The concentration of ferrous ion in solution was determined by titration with a standard potassium permanganate solution after acidification with sulphuric acid. Errors of two percent and one and one half percent could be attributed to titration for the acid and ferrous ion determinations respectively. 2. Assays ~ The samples of the underflow'and overflow were assay ed for copper and iron by the following procedures. The solids were dissolved in nitric and sulphuric acids (hydrochloric acid was also added if an appreciable amount of metallic iron was present) and the solution was evaporated.to near dryness. The copper was dissolved in excess ammonium hydroxide and separated from the ferric hydroxide precipitate by filtration. Copper was determined by titrating the filtrate with sodium thiosulphate in the presence of potassium iodide with a starch indicator after acidification with sulphuric acid. The ferric hydroxide precipitate was redissolved with dilute hydrochloric acid and the amount of iron present was determined by titration with sodium thiosulphate in the presence of potassium iodide with starch as the indicator (22). Titration errors in the assay •method could account for variations of one percent of the figure quoted for copper and five percent of the figure quoted for iron. 54; CHAPTER VII RESULTS The results of the experimental work are presented in graph form and discussed in this chapter under the following headings: A) Velocity of Solution Through the Column .' B) Acid Concentration of the Pregnant Solution C) Temperature of the Pregnant Solution D) Ferrous Ion Concentration of the Pi^egnant Solution E) Ferric Ion Concentration of the Pregnant Solution. Data used in the graphs are tabulated in the appendix to this thesis. The results of the experiments were obtained before the author received papers describing the work of Nadkarni and Wadsworth^ 214) from which- most of the theory was procured. Nadkarni and Wadsworth used sheets of .1020 mild steel (area = 2.0 sq. in.) in their experiments; however, thei results and theories have been used to analyze the results of the experiments with iron powder precipitation described here in, wherever applicable. (A) Velocity of Solution Through the Column Figures 7 and 8 are plots of the velocity of the solution through the column versus the percent of copper in the overflow and the Fe:Cu ratio for -100+200 and -200 mesh iron powders respectively. The graphs show that iron consumption is not affected to any significant extent, by the variations 100 0.5 0.8 1.0 1.2 1.4 Velocity of solution (cm/sec) 1.6 Fig. 7 - Velocity of solution vs. percent copper in. overflow and Fe:Cu ratio with -100+20 0 mesh iron powder "A" (see Table 5-in Appendix). 100 3 O £ 90 cu > o •ri 8 8.0 o . o % Cu • Fe:Cu o G"— — *» 1 1 e « 1 1 1 1 0.6 0.8 1.0 1.2 1.4 Velocity of solution (cm/sec) 1.6 1.3:1.0 1.2 1.1 1.0 0.9 0 . 8 3 o Fig. 8 - Velocity of solution vs. percent copper in overflow and Fe:Cu ratio with -200 mesh iron powder "A" (see Table Sin Appendix). in the velocity of the solution through the column. Agitation appeared to be sufficient, over the range of velocities tested, to assume that the experiments were carried out with a degree of agitation comparable to the high stirring speeds used by Nadkarni et al. Scaling-up of the cone-column precipitator would require a series of tests to determine the most suitable angle of the cone for the capacity required. The slight in crease in the iron consumption at a velocity of 1.70 centimeters per second in Figure 7 could be attributed to a loss of col loidal-sized copper particles in decantation. The fine particle are produced by attrition and Impact from violent agitation of the coarse particles in the -100+200 mesh size fraction. The percent copper in the overflow decreased above a"crltical velocity", the velocity above which the contact time of the precipitant with the cupric ion is insufficient for complete cementation. The critical velocity is a function of the copper concentration of the pregnant solution, the particle size of the iron powder (surface area), and the purity of ;the iron powder (the percentage of unoxidized iron). The maximum percentage of copper is attained at. velocities of solution flow below the critical velocity with the -100+200 mesh iron powder. The maximum grade of copper which can be produced with the iron used, in the experiments is 9 7+_l percent. The variation in the grades of copper obtained below the critical velocity is due primarily to oxidation during the drying process. The grade is slightly lower with the finer powder since some of the finest size fractions are forced out of the column before cementation is complete and oxidation is 5 7 greater during drying due to the larger available surface area. At a cupric ion concentration of 2.2 + 0.1 gpl the critical velocities are in the order of J. .35 and 0.90 centimeters per second for the -100+200 and -100 mesh size fractions respective ly. In practice the velocity of solution would be plotted against the percentage of copper in the overflow and the velocity would be chosen according to market requirements. A comparison of iron consumptions in Figures 7 and 8 shows that a slightly higher iron consumption is common to the finer powder. The difference is due to a loss of colloidal-size copper powder during decantation of the solution from the solids. Although the finest size fractions of iron powder have the highest surface area to weight ratio they are not the most suitable precipitants of copper in the cone-column precipitator. The decision to separate the fine particles must be based on economic considerations at a specific time and location. The mechanics of copper precipitation with powdered iron in the cone-column precipitator are quite simple to describe non-mathematically. At the apex of the cone the precipitant is subjected to violent agitation. As the copper forms on the iron particles attrition and impact free the copper from the iron, exposing fresh iron surfaces or, at the least, maintaining the diffusion layer at a minimum. As the cross-sectional area of the cone increases, agitation de creases until the particles attain' fluidized bed conditions. The particles are thus in a state of dynamic equilibrium where the largest particles, being nearest the apex of the cone, are 58 in contact with the strongest solutions and are. subjected to the greatest agitation. The velocity of flow through the column and the length of the column must be such that particles small enough to escape from the fluidized bed and to be carried to the overflow contain a minimal amount of iron. The column serves as an elutriator since particles are carried up through the 'column upon reaching a combination of shape, size and specific gravity. Due to the sponge-like nature of the iron powder and the precipitated copper, the specific gravity of any particle in the precipitator is established by a combination of copper, iron, solution, salts, hydrogen, oxygen, impurities in the precipitant and impurities, in the pregnant solution. According to the experiments of Nadkarni et al the rate of the cementation reaction is dependent upon the degree of agitation up to a limit above which the reaction rate is independent of agitation. The degree of agitation can be governed by altering the vertex angle of. the cone or by using a suitable nozzle. The amount of the Iron charge in the precipi tator will be considered in the section dealing with the Influenc of ferric Ion on the cementation process. 59 (B) Acid Concentration of the Pregnant Solution Data from operating cementation plants show that most of the acid present in the pregnant solution is consumed if copper precipitation is conducted in a launder system using scrap iron or detinned cans. Back stated that "due to the in timate contact in the cone (precipitator) between the metallic iron and the solution, copper precipitation is complete with less neutralization of the acid by the iron than in a conven tional launder plant", but he did not publish data to substanti ate his statement. Figures 9 and 10 are plots of acid concentration of the pregnant solution against percent copper in the overflow and the iron to copper ratio for -100+200 and -200 mesh iron powders respectively. Both size ranges of the iron powder show no substantial increase in iron consumption due to the presence of excess free sulphuric acid even in the pH range of 1.0 to 1.8 which is well below the "economical limit". The effect .of excess acidity is not noticeable up to a concentration of five grams per litre of free sulphuric acid. Above this concentra tion the influence of the. hydrogen ion-metallic iron reaction is indicated by a slight increase in iron consumption. The decrease in the percent copper in the overflow above an acid concentration of five grams per litre is less than one percent with the -100+200 mesh iron; however, it is quite significant with the finer material.- Without "burpers" or some other means of removing hydrogen from the surface of the particles, the finer material is floated up through the lOOr 5 o rH LM fc cu 90 > e c •rl O 8 0 o % Cu . Fe:Cu _ 4 8 12 16 Acid concentration (gpl) 20 1.3:1.0 1.2 1.1 CJ 0 . 9 0 . 8 Figure 9 - Acid concentration of pregnant solution vs. percent copper in overflow and Fe:Cu ratio with -100+200 mesh iron powder "A" (see Table 7 in Appendix). 100 1.3:1.0 Figure 10 -4 8 12 16 20 Acid concentration (gpl) Acid concentration of pregnant solution vs. percent copper in overflow and Fe:Cu ratio with -100+200 mesh iron powder "A" (see Table 8 in Appendix). 61 column and into the overflow before cementation has been completed. The suppression of the hydrogen ion-metallic iron reaction is due to the following: 1. The rapid rate of the precipitation reaction with the lcrge surface area of the iron. 2. The control of the amount of copper removed from solution by controlling the velocity of solution through the column. 3. The vigorous agitation of the powder in the solution which results in the rapid removal of the copper from the iron as soon as precipitation occurs. 4. The high purity of the iron precipitant which enables rapid cementation to occur with a minimum of hindrance from inactive impurities. The high purity iron reacts more slowly.with acid solutions as its purity gives less galvanic action than an impure iron. The experiments of Nadkarni and Wadsworth have indi cated that an additional series of experiments is warranted to determine the influence of different acid concentrations on the iron consumption at low cupric ion concentrations. (C) Temperature of the Pregnant Solution Figures 11 and 12 are plots of the pregnant solution against the percent copper in the overflow and the Fe:Cu ratio for -100+200 and -200 mesh iron powders respectively. The curves are shown as linear plots because the variations are within the experimental error discussed in Chapter VI. The 61a graphs show that the temperature of the pregnant soluvion has little influence on either the percentage of copper in the product or the iron consumption over the range of temperatures normally encountered under field conditions. (At the time the experiments were conducted, the author was concerned with operat ing temperatures in Canada and did not realize that solution temperatures approaching 45°C were common to several operations in the south-western United States). Nadkarni and Wadsworth established the dependence of the specific rate of reaction, K upon the temperature. On an Arrhenius plot of KQ vs. temperature the specific rates were approximately 6.5xl0~6, lO.OxlO-5, lO.lxlO-6, 21xl0~6 and 42xl0~6 cm.-2 sec.-1 for temperatures of 5, 25, 30 , 60 and 90°C respectively under conditions of maximum agitation. The re-^ / _ . lationship between the specific rate and the surface area., are--'-aS^ "defined by"the equation resulting from the integration of equation (16) viz. In S- - -KnAT C where -KQA is the slope of the straight line which results from c a plot of In — vs. time since the reaction is of the first order. Co _If it is assumed that the particles of iron used in the cementa tion process are spherical with diameter x and specific gravity \ 6 •/ -jr G there are ——3- particles in a unit weight and the surface • /\ to s 1 v 6 C area per unit weight is . Since the value of In — is Gsx CQ directly proportional to the particle size of the iron powder-used and is dependent on the temperature as represented by the 6 2 : oo 3 o H cu > o c •H O 80 o 0 o 5 o % Cu _ — • Fe:Cu e 0 1 • . 1 1 1.3:1.0 1.2 1.1 3 o —11.0 v. 10 15^ 20 Temperature (°C) 25 0 . 9 0.8 Figure 11 - Temperature of.pregnant solution vs. percent copper in overflow and Fe:Cu ratio with -100+200 mesh iron powder "A" (see Table 9 in Appendix). x 100 o\° 6 o 90 > o c •H CJ 80 — 0 o % Cu • Fe:Cu 1 1 • 1 — 5 10 15 20 25 Temperature (°C) 1.3:1.0 1. 2 1.1 1.0 0.9 0.8 Figure 12 - Temperature of pregnant solution vs. percent copper in overflow and Fe:Cu ratio with -200 mesh iron powder "A" (see Table 10 in Appendix). 63 specific rate determined from the Arrhenius plot it is obvious that decreasing the particle size of the precipitant is more significant, and practical, than Increasing the tem perature of the pregnant solution. As the particle size is decreased over a specific range of temperature, such as the range used in the experiments of the present author (5-25°C) it would be reasonable to assume that the effect of temperature would decrease to a point where it becomes almost insignificant. This hypothesis is substantiated by the results listed in Tables 9 and 10. With the -100+200 mesh powder more copper has been removed from solution at the higher temperatures within the range tested; however, with the finer powder the amount of copper removed from solution is relatively constant over the temperature range tested. (D) Ferrous Ion Concentration of the Pregnant Solution Figures 13 and 14 are plots of ferrous ion concentra tion in the pregnant solution against percent copper in the overflow and the iron to copper ratio for -10 0+2 0 0 and -2 00 mesh iron powder respectively. The ferrous ion in the pregnant solution did not affect either the iron consumption or the purity of the overflow as was also predicted and shown by other experimenters ^ ) • Precipitated salts of the ferrous Ion are not a source of contamination of the cement copper as hydrolysis and precipitation is not significant within the pH range of cementation and with the concentrations of ferrous ion normally encountered in leaching operations. In the presence of the iron oxidizing bacteria Thiobacillus ferrooxidans and 64 Ferrous ion cone, (gpl) Figure 13 - Ferrous ion cone, of pregnant solution vs. percent copper in overflow and Fe:Cu ratio with -100+200 mesh iron powder "A" (see Table 11 in Appendix). 1 0Q •4 6 8 10 Ferrous ion cone, (gpl) 12 1.3:1.0 14 Figure 14 - Ferrous ion cone, of pregnant solution vs. percent copper in overflow and Fe:Cu ratio with -200 mesh iron powder "A" (see Table 12 in ApDendix). .65 Ferrobacillus ferrooxidans oxidation of tha ferrous ion to ferric ion can occur. A high concentration of ferric.ion in solution is desirable for the leaching of sulphides, but, it is a definite disadvantage to the cementation process as is shown in the next section. (E) Ferric Ion Concentration of the Pregnant Solution. Figures 15 and 16 are plots of the concentration of ferric ion in the pregnant solution against percent copper in the overflow and the iron to copper ratio for -10 0+20 0 and -200 mesh iron powders respectively. The grade of copper in the overflow was not affected by increasing the ferric ion con centration; however, iron consumption increases rapidly beyond a concentration of 0.5 grams per litre of ferric ion. Tests with -200 mesh iron powder yielded erratic results with very high iron to copper ratios. The higher iron consumption with the finer powder contradicts the statement by Kennecott re searchers^ ^ that "if powdered iron of high surface area is used, • the copper precipitation reaction is found to be predominant and to be essentially completed before excessive amounts of iron have been consumed by the other two reactors". The other two reactions referred to are the metallic iron-hydrogen ion reaction and the metallic iron-ferric ion reaction. Apparently, at the time of publication, the writers were not aware of the ferric ion-metallic copper reaction which was illustrated later by the experiments of Nadkarni and Wadsworth. Since the methods used by the present author for 66 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Ferric ion cone, (gpl) Figure 15 - Ferric ion cone, of pregnant solution vs. percent copper in overflow and Fe:Cu ratio with -100+200 mesh iron powder "A" (see Table 13 in Appendix). I I | I I LJo.o 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Ferric ion cone, (gpl) Figure 16 - Ferric ion cone, of pregnant solution vs. percent copper in overflow and Fe:Cu ratio' with -200 mesh iron powder "A" (see Table 14 in Appendix). 6 7 previous tests did not provide enough information about either the ferric ion reaction or the above quotation, a series of tests were conducted in which samples of the overflow wera taken for analysis at specific times throughout the experiment, Larger samples of iron were used and the tests were continued until most of the solids had been forced to overflow the column. Samples of iron powder from three different sources were used in the experiments. The chemical analyses of the powders used are shown in Table 3 as powders A, B, and C. Figures 17, 18 and 19 are plots of the cupric and ferric ion concentrations of the effluent solutions against time for the three different iron samples tested. Table 4 lists the screen analyses, as given by the respective suppliers, for' each sample and pertinent data for each test. In Figures 17, 18 and 19 the consumption in ferric ion in solution corresponds approximately to the increase in copper surface area in the reaction vessel reaching a maximum prior to the removal of the precipitated copper from the column, when the surface area is also at a maximum. The cupric ion-metallic iron reaction is obviously predominant. The reaction between metallic iron and the hydrogen ion was previously shown to be negligible and the-reaction between ferric ions and metallic Iron must also be negligible as very little ferric ion was removed in the first few minutes of precipitation when the surface area of metallic iron was at a maximum. The statement quoted above from the paper by Spedden,Malouf and Prater-is basically true, but very misleading, since excess acid consumption is'..due primarily to th reaction between metallic copper and ferric ions. TABLE 4 SCREEN ANALYSES FOR IRON POWDERS A, B AND C AND DATA TO ACCOMPANY FIGURES 15, 16 and 17 IRON POWDER SAMPLE Screen Analysis "A" "B" "C" (Tyler) mesh % +100 i1. -100+200 35. -200 63. 2 5 3 (ASTM Standard) mesh % -100+200 100.0 (ASTM Standard) mesh % Supp. Actual - + .8 0.5 0.4 _ 8+40 70-90 82.5 -40+80 10-25 12.9 -80 5.0 4.2 Composition of Pregnant Solution Cu++ (gpl) Fe++ (gpl) Fe+++ (gpl) • Free H2S04 (gpl) 1.90 2.0 0 . 82 5.0 1.70 2.0 0.96 5.0 1. 90 2.0 1.08 5.0 Data Used for Calculations -Wt. of Iron Powder Charge (gms) 80.00 80.00 80.00 Wt. of Overflow (gms) 57.9 46 . 8 47.7 Wt. of Underflow (gms) 3. 3 12 . 9 7.1 Copper in Overflow (%) 9 8.1 96 . 3 98.1 Iron in Overflow (%) 0 . 9 1.6 0 , 9 Copper in Underflow (%) 95 .9 80 . 9 83.1 - Iron in Underflow (%) 2 . 1 18.1 15.9 Fe:Cu 1.31 :1 1.39:1 l.',-9 :1 0.0 H OH b0 10 20 30 40 Time (minutes) + + + to -++ . . . c Figure 17 - Cu and Fe ion cone. vs. time with 4-> us fc -f-> c cu o c o o c o •H o •H fc fc <i) XJ C £ O •rH o •H fc O, o iron powder "A" 1.2 0.6 0.0 Time (minutes) Figure 13 + + + + + Cu and Fe ion cone, iron Dowder "B" time with 50 Figure 19 + + Time .(minutes) +.++ Cu'' and Fe ion cone. vs. time with iron powder "C" 70 The variation in the iron consumption for the three samples tested is due to the variation of the concentration of ferric ion in the pregnant solutions, the difference in the size ranges of the iron tested and the variation in the per centage of the metallic iron in the powders. Iron consumption increased almost directly with an increase in ferric ion concen tration of the pregnant solution. Larger particle sizes of iron increase the contact time of the copper and iron with the ferric ion thus increasing the iron consumption. Although the initial surface areas of the copper and iron are smaller with larger particles, the surface area must approach the same figure per unit weight of charge before the material can rise in the column. Iron consumption is increased in direct proportion to the Impurities in the precipitant with the other factors constant. "The reaction sequence of the cementation process , in the apparatus employed for the experiments, is illustrated in Photographs 1 to 7. The photographs were taken during the test plotted in Figure 17. Photograph 1 was taken one minute after the addition of iron, powder to the precipitator. The iron was immediately coated with copper on addition to. the cone. The cupric ion-metallic iron reaction has priority at this stage as signified by the small amount of ferric ion removed from solution. An absence of solids in the column at the start of the experiment, when the surface area of metallic iron is at a maximum, indi cates that there has been a negligible amount of hydrogen re leased by the metallic iron-hydrogen ion reaction. 71 PHOTOGRAPH 3: 11 minutes PHOTOGRAPH 4: 15 minutes afterstart of experiment after start of experiment 72 PHOTOGRAPH 7: 25 minutes after start of experiment PHOTOGRAPH 8: Settling tank 73 When Photograph 2 was taken, at five minutes after the start of the experiment, the ferric ion concentration in the effJuent had decreased. The decrease in ferric ion corres ponded to an increase in the available metallic copper surface area in the reaction vessel. The bulk of the solids in the reaction zone had increased, at this point, to approximately one arid one-half times the initial volume of the charge.. The increase in volume is due in part to an increase in the number of particles by attrition and impact in the turbulent zone of the precipitator and in part to the spongy nature of the preci pitated copper. At eleven minutes, the solids in the reaction zone approached their maximum volume as shown in Photograph 3. This phenomenon is in agreement with the findings of A.E. Back who reported an increase in the bulk of solids at the height of the cementation reaction of from two, to two and one-half times the initial volume. In Photograph 4, taken fifteen minutes after the addi tion of the precipitant, the surface area of metallic copper is at a maximum as indicated by the high consumption of ferric ion At this point some of the particles started to rise in the column,which increased their contact time with the ferric ion. At eighteen minutes cementation approached completion as most of the copper was in motion towards the overflow as sho in Photograph 5. The ferric ion consumption decreased in pro portion to "the decrease in the available metallic copper surfac area. Photograph 6 shows that at twenty minutes most of the part-'cles have passed through the bottom half of the column. The cupric ion and ferric ion have increased with the decrease in the available metallic iron and copper respectively. When Photograph 7 was taken, twenty-five minutes after the iron was added, cementation had been completed. The results of the experiments plotted in Figures 17, 18 and 19 indicate that some control of the consumption of ferric ion can be exercised during cementation with powdered iron. Ferric ion is consumed according to the following react ions: +++ ++ ++ 2Fe +Cu(s) 2Fe +Cu (6) Cu+ + + Fe, . -> Cu, N + Fe + + (1) the sum of which is + + +  + 2Fe + Fe, . -> 3Fe (2) ( s) The influence of the ferric ion on excess iron consumption is expressed by equation (2) regardless of whether the ferric ion reacts directly with metallic iron or with metallic copper. Each 0.1 grams per liter of ferric ion remaining in solution after precipitation represents a saving of 0.5 pounds of iron per 1,00 0 gallons of solution pumped through the precipitator. Experimental work has shown that the ferric ion consumes most of the excess iron by equations (6) and (1), and that the cupric ion-metallic iron reaction occurs faster than the ferric ion-metallic copper reaction. Minimum contact between the ferric ion and the metallic copper could be obtained by removing the 75 metallic copper from the reaction zone immediately after pre cipitation. This could be achieved by adding the iron et a rate equal to its consumption and removing the copper from the top of the column. (F) Ideal Conditions for the Systems Employed. The ideal conditions for cementation with the high purity iron powder, the reaction vessel and the synthetic copper leach solutions used in the experiments by the present author and described herein are as follows : a) Iron particles in the size fraction of -100+200 mesh. b) A velocity of solution flow of approximately 1.3 5 centimeters per second. c) A concentration of free sulphuric acid up to 5 gpl, although, concentrations up to 20 gpl did not increase the iron consumption or lower the grade of product significantly, . . d) Normal operating temperatures encountered in the field. e) Normal concentrations•of ferrous ion encountered in heap leaching operations. f) A concentration of ferric ion in the pregnant solution of..less than .0.5 gpl but preferably a very low con centration of ferric ion with respect to the concen tration of cupric ion of approximately 2 gpl. g) A concentration of cupric ion of approximately 2 gpl. h) Constant concentrations of ions in solution, and continuous operation with,a minimum charge of preci pitant in the reaction vessel. 75a The above conditions apply only to the- systems tested. Methods of evaluating the parameters under conditions which vary from those tested are presented in Chapter VIII. 76 CHAPTER VIII EVALUATION OF THE PARAMETERS IN THE CONE-COLUMN PRECIPITATOR. Although the experimental work left several un answered questions the results did establish a basis for further work and a basis upon which the important parameters can be evaluated. The chemical and physical properties of the iron powder used as the precipitant and the cupric and ferric ion concentrations in the pregnant solution are the parameters which govern the velocity of solution flow in the column, the length of the column, the rate of feed to the precipitator and the amount of precipitant in the reaction vessel. The effect of the ferric ion on the cementation re action was defined by Nadkarni and Wadsworth with the following equation ^ip--- WFe++t]Ko' - AFeCCu++3Ko <") o S The equation was applied to a closed system in which there was no additional input of pregnant solution after the beginning of the reactions. The equation applies only if A^, Ape, [Fe+++] and [Cu++] are treated as variables. In the continuous precipitator, the initial concen trations of cupric and ferric ions are constants which can be 7 7 controlled by proper circular.ion and dilution of the solution. If iron powder is added at a rate equal to consumption, and copper is removed.at a constant rate, the areas of copper and iron in the precipitator may be considered as constants. The specific rates, K and K ', may also be treated as constants c o o ' . at a specific pH and temperature. In a commercial operation the pH could be maintained at a constant figure and temperature would be the only variable. Temperature variations would be small and distributed over several weeks or months depending on the locality of the operation. The experiments of Nadkarni et al and the experiments by the present author have indicated that the ideal iron powder should have a particle size within the range of minus 60 plus 200 mesh. Material coarser than sixty mesh does not react completely, according to Nadkarni et al, whereas particles finer than 200 mesh require a low velocity of solution flow in the column, which decreases the capacity of the reaction vessel, and produce colloidal-size particles which are difficult to recover from the effluent solution. The experiments discussed in this thesis eliminate the temperature of the pregnant solution and the concentration of the ferrous ion in the pregnant solution as important para meters in the cementation of copper 'with iron powder in the cone-column precipitator. The influence of variations in acid concentrations on the iron consumption is not significant, at concentrations up to twenty grams per liter, if the cupric ion is not stripped from solution. 78 The common practice of most cementation plant opera tors is to recover the maximum amount of copper from solution rather than an "optimum" amount. The optimum recovery is that which retains the maximum concentration of ferric ion and sulphuric acid in the effluent solution after precipitation'. With a launder system there is little advantage to retaining cupric ion in the effluent solution as only thirty percent of the copper is precipitated before most of the ferric ion is removed according to Figure 5. The experiments described and discussed herein have indicated that with a cone-column pre cipitator using powdered iron as the precipitant, the removal of cupric and ferric ions from solution can be controlled by varying the velocity of the solution through the precipitator and the charge in the precipitator. For example, in Figure 17, at three minutes, seventy-five percent of the cupric ion has been removed with only a twenty-five percent reduction in the concentration of ferric ion. If the charge in the preci pitator was minimized by the addition of iron powder at a rate equal to the iron consumption, the ferric ion reaction with metallic copper would be kept to a'minimum. There would be several advantages to recirculating a dilute copper solution to the leaching area: ,1. By allowing most of the ferric ion to remain in solution the iron consumption would be reduced considerably. 2. The ferric ion in the barren solution is required as a solvent for the leaching of sulphides; therefore, a high concentration of ferric ion in the barren solution is desirable. 3. Bacteria present in the ore and contributing to the leaching process would not be subjected to high fluctuations in the concentration of copper detrimental to their growth rate and activity. 4. The metallic iron-hydrogen ion reaction, which is responsible for a significant consumption of excess iron at low cupric ion concentra tions, would be eliminated or at least kept to a minimum. The following tests are proposed as methods of defining the velocity of solution flow in the column, the length of the column, the rate of addition of powder to the precipitator and the amount of precipitant in the reaction vessel. Velocity of Solution Flow in the Column The velocity of solution flow in the column must be as close to the critical velocity as possible to ensure maximum capacity of the reaction vessel but low enough to produce a -cement copper with a minimal percentage of iron. Initial experiments should be conducted with a. solution velocity of about 1.0 centimetres per second. An analysis of the cement copper in the overflow will determine whether the velocity should be increased or decreased to .provide the desired product. An iron consumption of 1.5 lbs of iron per pound of copper can 80 be used to calculate the approximate rate of iron feed for the velocity tests if ferric ion is present in the pregnant solution. Length of Column A plot of the critical velocity versus the length of .the column will establish figures for the most suitable column length. The contact time of precipitant with the pregnant solution increases in direct proportion to the length of the column; however, the column length will be limited by the costs of construction.. Rate of Feed to the-.Precipitator The input of precipitant to the reaction vessel must be balanced with the discharge of cement copper according to the following relationship: Fe,\ = VAC Cut4" _ Cut+) (Fe/Cu) (31) (s; i 1 where Fe(s) is t-he, weight of iron powder feed per unit time V is the velocity of solution flow A is the cross-sectional area of the reaction vessel Fe/Cu is the iron consumption ratio Cu^+ is the concentration of cupric ions in the influent solution Cu^ is the concentration of cupric ions m the effluent solution. A balance between the input and output can be calcu lated ' or established experimentally by measuring the weight of solids in the overflow and adjusting the feed rate until 81 equation (31) has been satisfied. Amount of Precipitant in the Reaction Vessel The required amount of precipitant in the reaction vessel can be established by plotting the weight of charge against the iron consumption and the concentrations of the cupric and ferric ions in the effluent solution. An excess amount of iron in the reaction vessel will not only consume excess ferric ion but will also strip the cupric ion from solution thus' allowing additional iron consumption by the hydrogen ion-metallic iron reaction. The hydrogen released by the reaction' between the hydrogen ions and the metallic iron will carry unreacted iron particles Into the overflow and lower the critical velocity.. A summary of the suggestions for further work is .in cluded in the final chapter. 82' CHAPTER IX . SUMMARY AND CONCLUSIONS The availability of iron powder at a cost slightly higher than that of scrap iron renders it a potential precipi tant of copper from dilute leach liquors. From a consideration of the results of the experiments with iro'n powders discussed herein,'the following observations were made: 1) The cone-column precipitator, with powdered iron as a precipitant is capable of continuous recovery of • copper from dilute copper sulphate solutions. 2) A velocity of solution flow through the column of approxi mately 1.0 centimeters per second is suitable for • particles retained on a 200 mesh Tyler screen and a cupric ion concentration of approximately two grams per liter. Particles finer than 2 00 mesh require a velocity of solution flow below 1.35 centimeters per second and produce colloidal-size copper which is difficult to recover from the effluent solution. Back et al recommended a solution velocity of approxi mately 0.25 centimeters per second at the overflow of their cone-type precipitator in order 'to prevent the escape of particles from the reaction vessel when using a particulate iron in the size fraction of -10+200 mesh. The volume capacity of the cone-column precipitator described herein is approximately 3-5 times that of 83 the vessel developed by Back et al for the same size fraction of iron and similar dimensions. The hydrogen ion-metallic iron reaction does not con tribute to the iron consumption to any significant extent up to concentrations of twenty grams per liter of free sulphuric acid, (the maximum concentration tested); however, no attempt was made to remove all of the cupric ion from solution. Variations in the temperature of the pregnant solution over the range commonly encountered in commercial operations was found to have no effect upon the ce mentation rate when iron powder is used as the precipitant The presence of ferrous ion in the pregnant solution has no influence on either the iron consumption or the grade of the cement copper produced. Ferric ion is the major consumer of excess iron. The iron consumption with a minus 100 mesh iron powder con-+ + taining approximately 2 gpl of Cu and 5 gpl of free H2S04 is in the order of 0.97:1.00, whereas, in a + + solution containing 1.90 gpl of Cu ,5.0 gpl of free + + + H^SC^ and 0.82 gpl of Fe the iron consumption is 1.31:1.00 - 25.6 percent of the iron consumption being due to the reduction of ferric ion. The ferric ion reacts with metallic copper and the cupric sulphate • reaction formed by thisVis responsible for the consumption of excess iron by the cementation reaction. The cupric ion-metallic iron reaction is significantly faster than the ferric ion-metallic iron reaction. The experiments conducted in the presence of ferri ion indicated that the consumption of excess iron by the ferric ion should be minimized if powdered iron is added to the precipitator at a rate 'equal its consumption thereby minimizing the surface are of metallic copper in the reaction vessel. 84 Suggestions for Further Experimentation As with most processes, the relationships between the variables must be redefined with, any change in one or more of the variables. The experimental work discussed herein has established that the following additional experiments would be required to study the feasibility of iron powder precipitation of copper with the cone-column precipitator at anv leaching operation. • 1) Continuous precipitation with various size fractions of iron powder must be conducted to establish the most economical size range and respective velocity for the grade of cement copper required. 2) Continuous precipitation must be conducted to obtain the rate of iron powder feed necessary to satisfy equation (31) and to establish the height of the precipitator column. 3) Continuous precipitation must also be employed to . establish the most economical concentration of cupric ion to be retained in the effluent solution which is governed by the ferric ion content of the pregnant solution. 4) If there is very little ferric ion jLn the pregnant • solution, the most economical amount of cupric ion to be retained in the effluent solution is governed by the acid content of the pregnant solution and must be established bv continuous precipitation. • •[ The recovery of copper from dilute leach solutions can 85 be practiced as a simple but profitable art. In order to obtain maximam efficiency with the process it must be practiced as-a science. The use of powdered iron as a precipitant of copper from dilute leach solution:.' can neither be accepted nor rejected.on the basis of the findings presented in this thesis without a consideration of the economics Involved at a specific operation. ( 86 LITERATURE CITED 1. AGERS, D.W. , HOUSE,.J.E.t SWAMSON, R.R. , and DROBNICK, "A New Reagent for Liquid Ion Exchange Recovery of Copper," Mining Eng., Vol. 17, No. 12, Dec. 1964, p. 76. -2. ALDRICH, H.W. and SCOTT, W.G., "The Inspiration Leaching Plant", Trans. AIME 106 , 1933 , pp. 65.0-77. 3. ANDERSON, A. E. and CAMERON, F.K., "Recovery of Copper by' Leaching, Ohio Copper Co. of Utah." Trans. AIME, vol. 73, 1926, pp. 31-57. 4. ARGALL, G.O. Jr., "How Leaching Recovers Copper from Waste and Leach Dumps in the Southwest", Mining World, Oct. 1963. .5. BACK, A.E., FISHER, K.E. and KOCHERHANS, J., "Process and Apparatus for the Precipitation of Copper from Dilute Acid Solutions", U.S. Patent No. 181,001, assigned to Kennecott Copper Corporation, New York, N.Y. Filed Mar. 20, 1962. 6. BACK, A.E., "Use of Particulate Iron in the Precipitation of Copper from Dilute Solutions", Paper presented at the Annual Meeting of the AIME, New York, N.Y., Feb. 2 7 to Mar. 3, 1966. 7. BEALL> J.V., "Kennecott Completes Four-year Expansion of Utah Copper Division", Min. Eng., June 1967, pp. 58-65. 87 .8. BEAN, J.J., "LPF at Miami - Analysis of Latest Operations", Min. Eng., Dec. 1960. 9. BIGLEY, A.C, FRICK, F.F. , KcCANNA, M. , and RYAN, J.P.. ".New Precipitation Plant at Butte." Mining Congress Journal, March 1951. 10. CADDICK, A.J., "Precipitation of Copper by Hydrogen Sulphide." Mining Magazine (London), 52, 20(1935).' 11. CAMPBELL, T.C , "A Brief Description of the Reduction Plant of the Chile Exploration Company of Chuquicamata, Chile, S.A." Trans. AIME 106, 1933, pp. 559-608. 12. EVANS, D.J.I. , RQMANCHUK, S., and MACKIVJ, Y.N., "Production of Copper Powder by Hydrogen Reduction Techniques." Can. Min. and Met. Bulletin, July 1961, pp. 530-538. 13. GREENWALT, W., "The Hydrometallurgy of Copper." McGraw-Hill Book Co., New York, N.Y., 1912, p. 256.x 14. GROVES, R.D., "Preparation of Copper'Powder from Leach Solutions after Precipitation with Iron." USBM, RI 6486, 1964. 15. GRUNENFELDER, J.G., 'The Hydrometallurgy of Copper, : Chap. 13, from "Copper, the Science and Technology of the Metal, its Alloys and Compounds", Ed. Butts, A., Reinhold, N.Y., 1954, p. 322-335. 16. HARDvJICK, W.R. , "Mining Methods and Costs', Inspiration Consolidated Copper Co. Open-pit Mine, Gila County, Arizona." USBM, IC 8154, 1963. 17. HUDSON, A.W. and VAN ARSDALE, CD., "Heap Leaching at Bisbee, Arizona", Trans. AIME V6 9, 192 3, pp. 13 7-154 . 18. JACKY, H.V/. , "Copper Precipitation Methods at Weed Heights", 88 18. JACKY, H.W. , Journal of Metals, April 19G7, pp. 22-2.7. 19. JACOBI, J.S., "Unit Process in Hydrometallurgy", Gordon •and Breach Science Publishers, New York, N.Y. 1963, p. 617. 20. KNOBLER, R.R. and JOSEPH, W., "The Mantos Blancos Operation", Mining Eng., Jan. 1962, pp. 40-45. 21. LOWER, G.O. and BOOTH, B., "Cyanidation Studies: Recovery of Copper by Cyanidation", presented at the Annual Meeting of the AIME, New York, N.Y., Feb. 2 7 to Mar. 3, 1966. 22. MAYLING, A.A., Personal Communication, Chief Chemist, Biometals Corporation Ltd., Vancouver, British Columbia. 23. MONNINGER, F.M., "Precipitation of Copper on Iron", Mining Congress Journal, Oct. 1963, pp. 4 8-51. 24. NADKARNI, R.M. and .WADSWORTH, M.E., "A Kinetic Study of the Cementation of Copper with Iron" (unpublished) PhD Thesis, Dept. of Metallurgy, University of Utah, 1966-. 25. NADKARNI, R.M., JELDEN, C.E., BOWLES, K.C., FLANDERS, H.E., and WADSWORTH, M.E., "A Kinetic Study of Copper Precipitation on Iron - Part I", Trans, of Metallurgical Soc. of AIME, Vol. 239, April 1967, pp. 581-585. 26. O'LEARY, V.D. , "Clarification and Treatment of Mine Drainage Water Prior to Use for Dump Leaching in Butte." Presented at the 95th Annual Meeting of AIME, New York, N.Y., Feb. 27-Mar. 3, 1966. 89 27. RAMSEY, R.H. (ed), "Anaconda's Nevada Project - New Approach to Copper Mining", Eng. and Min. Journal, Vol. 155, No. 8, Aug. 1954, pp. 74-92. 28. SMYTH, T.D., "Iron Powder as a Precipitant of Copper from Dilute Solution" (unpublished) summer report, Dept. of Mineral Engineering, The University of British Columbia, Oct. 1966.. 29. SPEDDEN, H. R. , MALOUF, E.D. , and PRATER, J.D., "Cone-type Precipitators for Improved Copper Recovery", Min. Eng., April 1966, pp. 57-62. 30. TAYLOR, J.H., and WHELAM, P.F., "Leaching of Cuprous Pyrites and the Precipitation of Copper at Rio Tinto, Spain." 3ull. Inst. Min. and Met,, Mo. 457, 1942. 31. TOBELMANN, H.A., "Handbook of Nonferrous Metallurgy", LIDDEL, D.M. (edd) McGraw-Hill Book Co., Inc., New York, N.Y., 1945, p. 346. 32. WARTMAN, F.S., and ROBERTSON, A.M., "Precipitation of Copper from an Acid Mine Water." ESBM, RI 3746, 1944 . 33. VAN ARSDALE, G.D., US Patent 1,232,080. 90 APPENDIX ( TABLE 5 - DATA FROM TESTS WITH VARIATIONS IN THE VELOCITY OF THE SOLUTION THROUGH THE COLUMN WITH IRON POWDER "A" (-100+2 00 MESH) CONSTANTS (PREGNANT SOLUTION) TEMPERATURE 15.6°C (60°F) Cu++ 2.2 + 0.1 gpl Fe++ 0 gpl Fe++  gpl H2S04 5.1 + 0.1 gpl Velocity of WT OF Fe :Cu Solution OVERFLOW UNDERFLOW IRON FEED (cm/sec) Wt(gms) % Cu % Fe Wt(gms) % Cu % Fe (gms) 0 . 51 •0.2 0 97.0 2.2 20. 25 71. 7 27. 7 20.00 0.98:1.00 0.73 1.30 97.8 1.0- 19.85 94. 9 3. 5 2 0.00 0.96 :1.00 0.93 2 . 35 95 . 8 3.0 18.65 31.0 18.2 20.00 0.96:1.00 1 .14 3.20 98.1 0 . 9 17.55 92.0 7.0 20.00 0.96:1.00 1. 34 2.75 97 . 5 1. 3 17.40 51. 0 38.0 20.00 1.00 :1.00 1.69 2 . 79 92.0 •6.9 17. 12 48. 0 50 . 7 20.00 1.0 3 :1.00 'TABLE 6 - DATA FROM TESTS WITH VARIATIONS IN THE VELOCITY OF THE SOLUTION THROUGH THE COLUMN WITH IRON POWDER "A" (-200 MESH) ' CONSTANTS (PREGNANT SOLUTION) TEMPERATURE 15.6°C (60°F) . ' Cu++ 2.2 + 0.1 gpl Fe++ 0 gpl Fe++ + 0 gpl H2SO,4 5.1 + 0.1 gpl Velocity of WT OF Solution OVERFLOW UNDERFLOW IRON FEED Fe: Cu (cm/sec) Wt(gms) % Cu % Fe Wt(gms) % Cu % Fe (gms) 0 .56 1.45 9 5.4 3. 5 19.18 97.5 2.2 20.00 0.98:1.00 0 . 75 2.50 96.0 3 . 0 17. 80 9 7.5 1. 8 20.00 0.99:1.00 0.37 2.50 1 9 7.0 2.6 18. 20 97.5 1. 3 20.00 0.97:1.00 1.07 • 2.60 9 3.0 3.0 17. 80 97.5 1. 8 20.00 0.98:1.00 1. 34 5 . 30 86.5 12. 3 13.55 91.0 8. 8 20.00 1.03:1.00 1.71 6.22 83.5 15 . 8 14. 55 87.0 12 . 2 20.00 0.98:1.09 TABLE 7 - DATA FROM TESTS WITH VARIATIONS IN THE ACID CONC. OF THE PREGNANT . SOLUTION WITH IRON POWDER "A" (-100+200 MESH) CONSTANTS (PREGNANT SOLUTION) 15.6°C (60°F) 2.0 + 0.1 gpl 0 gpl 0 gpl" 0.90 + 0.01 cm/sec. H2SO,4 PREGNANT SOLUTION (gpl) . Cone. . BARREN SOLUTION (SPD pH PREGNANT SOLUTION OVERFLOW UNDERFLOW WT. OF IRON FEED (gms) Fe-Cu Wt(gms) % Cu % Fe Wt(gms) % Cu % Fe 1.0 1.0 1.85 2.95 98.7 0 . 3 18. 35 87.0 12. 3 20.00 0.94 :1. 0 0 1.4 1.4 1.75 1.70 98. 7 0.4 19.60 83.0 16. 3 20 .00 0.94 :1.00 2 . 9 2 . 8 1.70 2.72 9 8.2 0 . 6 18. 15 71.0 28.1 20.00 0.96:1.00 5.1 5.0 1.55 2.35 98.2 9 . 6 18.20 94 . 0 4.8 20.00 0.96:1.00 10.3 10 .1 1.2 2. 55 98.6 .0.9 17. 60 92.0 '7.0 20.00 1.00 :1.00 20.2 20 . 0 1.1 4.90 98.0 1.5 15. 3 97.2 2.2 20.00 1.00:1.00 TEMPERATURE n + + Cu Fe + + Fe+++' Vel of Soln CD CO o ( TABLE 8 - DATA FROM TESTS WITH VARIATIONS IN THE ACID CONC. OF THE PREGNANT SOLUTION WITH IRON POWDER "A" (-200 MESH) CONSTANTS (PREGNANT SOLUTION) TEMPERATURE 15.6°C (60°F) Cu++ 2.0 + 0.1 gpl Fe++ 0 gpl Fe+++ 0 gpl Vel of Soln. 0.90 + 0.01 cm/sec. H2SOH . PREGNANT SOLUTION (SPD Gone. BARREN SOLUTION (gpl) pH PREGNANT SOLUTION OVERFLOW UNDERFLOW WT OF IRON FEED .(gms) Fe - Cu Wt(gms) % Cu % Fe Wt(gms) % Cu % Fe 0 . 4 0.2... 2 . 0 6.40 94 . 0 5. 3 14 . 80 96.0 3.5 20.00 0.95:1.00 0.9 0 . 7 1. 85 3.55 93.5 5. 3 17 . 55 95 . 5 2.6 20. 00 0 . 95 :1. 00 1.4 1. 3 1. 75 4 . 00 95.0 4.0 17.00 97.0 2.2. 20.00 0.96:1.00 3 . 3 2.9 1.65 3 . 80 95.0 4 . 2 15 .'9 5 9 7.0 •2 . 0 20.00 0.97:1.00 7.7 7.3 1.4 5 3. 85 •95. 5 5.3 17. 15 96 . 0 3.1 '20.00 0.92:1.00 10 . 2 9.9 1.25 4.-0 0 94 . 0 4.4 16.95 94 . 5 4.4 20.00 0.9 7 :.1.0 0 19 . 8 19 . 4 1.2 3.15 91. 0 7.9 17.25 97.0 1.8 20.00 1.00 :1.00 TABLE 9 - DAT^ FROM TESTS WITH VARIATIONS IN THE TEMPERATURE OF THE PREGNANT SOLUTION WITH IRON POWDER "A" (-10 0+20 0 MESH) CONSTANTS (PREGNANT SOLUTION) Cu++ 2.15 + 0.0 5 gpl • Fe++ 0 gpl Fe+++ 0 gpl H2SO 5.0 +0.1 gpl Vel of Soln. 0.90 + 0.01 cm/sec. WT OF TEMPERATURE OVERFLOW UNDERFLOW IRON FEED : • Fe:Cu ° C ' Wt (gms) % Cu % Fe Wt(gms) % Cu % Fe (gms) 6.1 12 . 2 15 .6 13.9 23.9 3.40 3.10 2. 35 3 .00 2.60 96 . 0 96.5 95. 8 95 . 4 97 . 0 3.1 2.6 3.0 •: 3 . 7 1.8 17.10 18.05 18.65 18.15 18.00 63.8 87. 6 81. 0 97.0 95.0. 35.6 11.9 18.2 2.2 4 . 2 20 .00 20.00 20.00 20.00 20.00 0.98:1.00 0 .95:1.00 0.96:1.00 0.95 :1.00 0.98:1.00 CO Cn o TABLE 10 - DATA FROM TESTS WITH VARIATIONS IN THE TEMPERATURE OF THE PREGNANT SOLUTION WITH IRON POWDER "A" (-20 0 MESH) CONSTANTS (PREGNANT SOLUTION) Cu++ 2.15 + 0.0 5 gpl Fe++ 0 gpl Fe+++ 0 gpl H2S04 5.0 + 0.1 gpl Vel of Soln. 0-.90 + 0.01 crn/sec. TEMPERATURE • 0 C OVERFLOW UNDERFLOW WT OF IRON FEED (gms) Fe: Cu  Wt(gms) % Cu % Fe Wt(gms) % Cu ' % Fe 7. 2 3.95 9 7.5 I. 3' 10 . 85 97.0 3.0 20.00 0.95:1.00 12.8 4 .00 96 . 5 3 . 0 16 .75 97.0 2.2 20 . 00 0.97:1.00 15.6 2.50 95.4 2.6 18.20 97.5 1.3 20.00 0.97:1.00 18.4 3. 30 96.5 2.4 17. 30 97.0 2.2 20.00 0.97:1.00 23.9 3. 85 9 5.4 3.7 16.40 96.5 2 . 6 20.00 1.00:1.00 CO CD TABLE 11 - DATA FROM TESTS WITH VARIATIONS IN THE FERROUS ION CONCV OF THE PREGNANT SOLUTION WITH IRON POWDER "A" (-100+200 MESH) CONSTANTS (PREGNANT SOLUTION) 15.G°C (60°F) 2.0 + 0.1 gpl 0 gpl. 5.0 + 0.1 gpl , 0-. 9 + 0.01 cm/sec. x, + + Fe of Preg Soln gpl OVERFLOW UNDERFLOW WT OF IRON FEED (gms). Fe:Cu Wt(gms) % Cu % Fe • Wt(gms) % Cu % Fe 0.00 2.95 97.2 1.6 18.2 .94 . 0 4. 8 20 ; 0'' ...0.96:1 1.00 2.90 97.4 1. T 18.0 9 2 . 0 6 . 7 20.0 0.97:1 1.93 2. 80 96.6 2.2 18.0 91. 2 7.9 20.0 0.97:1 5 . 80 3.0 9 7.2 1.8 17.7 6 8.0 30 . 8 20.0 0.97:1 8.10 2.3 96 . 8 1.7 17. 7 70. 6' 2 8.4 20.0 0.96:1 14.20 3 . 5 97.2 1.8 17.0 62 . 0 37.0 20 . 0 0.98:1 TEMPERATURE Cu+ + Fe + + + H2S04 Vel of Soln. c TABLE 12 - DATA FROM TESTS WITH VARIATIONS IN THE FERROUS ION CONC. OF THE PREGNANT SOLUTION WITH IRON POWDER "A" (-2 00 MESH) CONSTANTS (PREGNANT SOLUTION) TEMPERATURE 15.6°C (50°F) Cu++ 2.0 + 0.1 gpl Fe++ 0 gpl H2SOl4 5.0 + 0 .1 gpl Vel of Soln. 0.9 + 0.01 cm/sec. Fe of Preg Soln. SPl • • OVERFLOW UNDERFLOW WT OF IRON FEED (gms) Fe :Cu Wt(gms) % Cu % Fe Wt(gms) % Cu % Fe 0. 0 2.35 95.8 3.0 18.65 81.0 18 . 2 20.0 0.95 :1 1.00 2.80 95.9 3.1 17.90 95 .1 3.9 2 0.0 0.97:1 1.93 7 . 9 94 . 9 4.1 11. 8 94.7 ',4.2 20 . 0 0.97:1 3.00 '4.7 • 95.4 3.7 15 .8 97.2 2.2 20.0 0.98:1 5 . 00 4 . 2 95 . 0 4.2 16.2 95.8 3.1 20 . 0 0.99:1 5.000 9 . 5 95 . 8 3.1 10 . 6 97.7 1.3 20 . 0 1.00:1 13.90 4.6 96. 3 2 . 8 15 . 7 96.1 2 . 7 20.0 1.00:1 o TABLE 13 - DATA FROM TESTS WITH VARIATIONS IN THE FERRIC ION CONC. OF THE PREGNANT SOLUTION WITH IRON POWDER "A" (-100+200 MESH) CONSTANTS (PREGNANT SOLUTION) 15.6°C. (60°F) 2.0 + 0.1 gpl 0 gpl 5.0 +_ 0.1 gpl 0.9 + 0.01 gpl + + + Fe Cone.of Preg Soln. gpl OVERFLOW UNDERFLOW WT. OF IRON FEED (gms) Fe:Cu Wt(gms) % Cu % Fe Wt(gms) % Cu % Fe 0.00 2.95 97 . 2 1.6 18.20 94 . 0 4.8 20.00 0.96:1.00 0.25 1. 35 97.0 2.2 15 . 30 95 . 0 4 . 0 20 .00 1.08:1.00 0 . 50 1.50 9 8.0 1. 3 14. 30 93.0 1.1 20 .00 1.28:1.00 1.00 1.70 96 .1 3.1 10 .50 94 . 3 4. 8 20 . 00 1.67:1.00 1.25 1.02 95.8 2 .1 8. 80 95 . 8 2.1 20.00 2.08:1.00 1.25 0 .90 9 7.0 1.8 8.60 94 . 5 4.4 20 . 00 2.18:1.00 1.25 0 .70 97.6 1. 3 3.50 95 . 9 3.2 20.00 2.23 :1.00 2 . 50 0 .30 97.4 2.1 3. 30 •91. 3 7.9 20.00 6.0:1.00 TEMPERATURE n + + Cu TT + + Fe H2S04 Vel of Soln. CO CO (.) TABLE 14 - DATA FROM TESTS WITH VARIATIONS IN THE FERRIC ION CONC. OF THE PREGNANT SOLUTION WITH IRON POWDER "A" (-200 MESH) CONSTANTS (PREGNANT SOLUTION) TEMPERATURE 15.6°C (60°F) Cu++ 2.0 + 0.1 gpl Fe++ 0 gpl H2S04 5.0+0.1 gpl Vel .of Soln 0.9 + 0.01 cm/sec Fe+++Conc.of Preg Soln. gpl OVERFLOW UNDERFLOW WT OF IRON FEED (gms) Fe :Cu Wt(gms) % Cu % Fe Wt(gms) % Cu % Fe 0.00 2. 35 95 . 3 3. 0 13 . 65 81. 0 18. 2 20.00 0 .95 :1. 00 0.25 4.70 93. 3 5 . 7 11. 70 97.2 • 1.8 20 . 00 1. 24 :1.00 0.50 3.10 95 . 8 2.2 10 .10 97.7 1.3 20.00 1.54 :1. 00 1.0 0 0. 30 98. 7 0.4 5.80 9 8.7 0.4 20.00 3. 32 :1.00 1.00 . 3. 30 95 . 5 3.4 5.9 0 94.5 . 4.4 20.00 2 .25 :1. 00 3 .00 3.40 98.5 0.4 0 . 30 45 . 2 52.8 20 . 00 5.57 :1.00 3.00 2 . 30 94 . 7 4.1 0.20 83.9 15 . 2 20 .00 8.46 :1.00 o o 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Country Views Downloads
United States 25 2
China 9 1
Japan 6 0
Chile 4 1
United Arab Emirates 4 0
Germany 4 0
Russia 3 0
India 3 4
Canada 3 0
France 3 0
Zambia 2 0
Spain 1 0
City Views Downloads
Unknown 21 14
Mountain View 6 0
Beijing 5 0
Ashburn 5 0
Tokyo 5 0
Seattle 4 0
Shenzhen 4 1
Saint Petersburg 3 0
Mississauga 2 0
Nagpur 2 0
Santiago 2 0
Wilmington 2 0
Shizuoka 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0081087/manifest

Comment

Related Items