Recovery Of Noble Metals From Jewellery Wastes

RECOVERY OF NOBLE METALS FROM JEWELLERY WASTES

By Richard Kadiambuji Kady Mbaya

Submitted in partial fulfilment of the requirements for the degree

MAGISTER TECHNOLOGIAE: CHEMICAL ENGINEERING

In the Department of Chemical and Metallurgical Engineering FACULTY OF ENGINEERING TSHWANE UNIVERSITY OF TECHNOLOGY

Supervisor: Prof J.H. POTGIETER Co-Supervisor: Dr S.S. POTGIETER

March 2004

DECLARATION BY CANDIDATE

“I hereby declare that the dissertation/thesis submitted for M Tech: Chemical Engineering, at Tshwane University of Technology, is my own original work and has not previously been submitted to any other institution of higher learning. I further declare that all sources cited or quoted are indicated and acknowledged by means of a comprehensive list of references”.

-----------------------------------------R.K.K. Mbaya

Copyright© Tshwane University of Technology 2004

i

ACKNOWLEDGEMENTS

I would first like to thank God the Almighty, without his grace all this would be impossible. I would also like to express my sincere gratitude and appreciation to: •

Prof. J.H.Potgieter and Dr. S.S.Potgieter for their guidance, encouragement and positive criticism.

•

Mr. D. Newman and Mrs N. Newman of the Department of Fine Arts for supplying the material used in the investigation.

•

Mrs M. Loubser of Pretoria University for XRF analysis.

•

Dr. A.Teodorovic, Dr. L.Marjanovic, S.Mokgalaka, Dr. C.T.Mutale, T. Mukongo. And K. Lonji for their contribution.

•

Dr. H. Chikwanda head of Department of Chemical and Metallurgical Engineering for her support

•

Tshwane University of Technology for funding and making the materials and facilities for the completion of this work available.

ii

DEDICATION

This study is dedicated to: •

My wife Clarisse Ntumba Ngalula

•

My children Tresor Kabongo Mbaya, Papin Lukusa Mbaya, Rodine Ngalula Mbaya, Laurette Nseya Mbaya and Job Kady Mbaya for their support.

iii

ABSTRACT

Jewellery has attracted humans of all cultures and civilization since times immemorial. Gold for example, has a beautiful colour, excellent corrosion and oxidation resistance, superior malleability, and ease of fabrication. Its limited

availability

together

with

these

other

characteristics

are

collectively responsible for the unique place gold has in the family of metals. Silver and platinum play a similar the role. As a consequence much thought and extensive research have been devoted to methods of recovering these noble metals from native ores in general and from wastes in particular.

Various methods for treating noble metals have been reviewed and discussed by numerous experts in the field. Although cyanide leaching remains the overwhelming option for the treatment of gold ores because of its economy and simplicity, it suffers from certain inherent drawbacks such as toxicity and slow leaching.

Despite the impressive safety record of cyanide to date, it is conceivable that environmental concerns could fuel trends to alternative lixiviants such as halogen-based systems, nitric acid systems or thiourea for leaching. While difficulties remain to be overcome, leaching with solutions other than cyanide has considerable potential as an effective and less hazardous procedure for gold, silver and platinum recovery.

The present research work deals with leaching of jewellery wastes using nitric acid to dissolve silver and aqua regia (one part nitric acid, by volume, to three parts hydrochloric acid) to dissolve gold and platinum.

iv

The wastes were dissolved in nitric acid. Metallic silver with a purity of 98.3% was recovered by using HCl to precipitate AgNO3 as AgCl, and then cementing it with zinc before melting it at 1000 ºC with K2CO3

.

Gold and platinum were dissolved in aqua-regia. Metallic gold with a purity of 99% was recovered by precipitation with iron sulphate (FeSO4) and melting it at 1100 ºC.

Platinum with a purity of 99% was recovered by precipitation with ammonium chloride (NH4Cl) as ammonium hexachloroplatinate complex (NH4)2PtCl6 and converted to black platinum powder using hydrazine. All analyses were performed by titration, AAS, ICP, XRD and XRF.

v

INDEX

PAGE

DECLARATION OF CANDIDATE

i

ACKNOWLEDGEMENTS

ii

DEDICATION

iii

ABSTRACT

iv

LIST OF FIGURES

xi

LIST OF PLATES

xiii

LIST OF TABLES

xiv

LIST OF ABBREVIATIONS USED

xvi

NOMENCLATURE

xvii

GLOSSARY

xix

CHAPTER 1

INTRODUCTION

1

1.1 BACKGROUND AND MOTIVATION

1

1.2 PROBLEM STATEMENT

5

1.3 HYPOTHESIS

6

1.4 OBJECTIVES

6

vi

CHAPTER 2

LITERATURE REVIEW 2.1 HISTORY

7

2.2 OCCURRENCE, EXTRACTION AND REFINING

9

2.2.1 Gold

9

2.2.2 Silver

11

2.2.2.1 Smelting

12

2.2.2.2 Amalgation

12

2.2.2.3 Cyanidation

12

2.2.3 Platinum-group metals (PGM)

13

2.3 USES OF NOBLE METALS IN JEWELLERY

14

2.3.1 Gold

14

2.3.2 Silver

18

2.3.3 Platinum

19

2.4 PHYSICAL AND CHEMICAL PROPERTIES

20

2.4.1 Gold

20

2.4.1.1 Physical properties of gold and gold alloys

20

2.4.1.2 Chemical properties

22

2.4.2 Silver

24

2.4.2.1 Physical properties

24

2.4.2.2 Chemical properties

25

2.4.3 Platinum

28 vii

2.4.3.1 Physical properties

28

2.4.3.2 Chemical behavior of the platinum-group metals

28

2.5 CURRENT PREFERRED REFINING PROCESSES

29

2.5.1 Introduction

29

2.5.2 Current developments in extractive metallurgy of gold

30

2.5.2.1 Conventional methods

30

2.5.2.2 Direct leaching methods

39

2.5.2.3 Direct electrowining

44

2.5.2.4 Use of unconventional lixiviants

45

CHAPTER 3

EXPERIMENTAL APPROACHES

3.1 INTRODUCTION

61

3.2

62

MATERIALS

3.2.1 Chemicals and reagents

62

3.2.2 Instrumentation

62

3.2.3 Particle size analysis

64

3.3 METHODS

65

3.3.1 Silver

67

3.3.1.1 Dissolution of silver and other metals in nitric acid

67

3.3.1.2

68

Precipitation of silver chloride

viii

3.3.1.3 Cementation process

69

3.3.1.4 Refining process of silver

70

3.3.2 Gold

71

3.3.2.1 Gold dissolution in aqua regia

71

3.3.2.2

72

Precipitation of gold from aqua regia solution

3.3.3 Platinum

73

CHAPTER 4 RESULTS AND DISCUSSIONS 4.1 INTRODUCTION

74

4.2 SILVER

74

4.2.1 Effect of HNO3 concentration

75

4.2.2 Effect of stirring speed

78

4.2.3 Effect of temperature

79

4.2.4 Effect of solid/liquid ratio

80

4.2.5 Cementation results

81

4.3 GOLD

82

4.3.1 Effect of temperature on precipitation of gold

84

4.3.2 Effect of stirring speed on the precipitation of gold

86

4.4 PLATINUM

87

4.4.1 Effect of temperature on precipitation of platinum

89

4.4.2 Effect of stirring speed on precipitation of platinum

90

4.5 FINAL RESULTS

92

4.6 COST ANALYSIS OF THE PROCESS

95

ix

CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS

99

5.1 Conclusions

99

5.2 Recommendations

100

REFERENCES

102

APPENDIX

116

x

LIST OF FIGURES

Figure 2.1

Weight per cent composition ranges for various

17

colours of Au/Ag – Cu alloys Figure 2.2

Potential-pH equilibrium diagram for the gold – water 24 system for dissolved gold species of 10-4 M.

Figure 2.3

Potential-pH equilibrium diagram for the silver–water 27 System for dissolved silver species of 10-4 M.

Figure 2.4

Block diagram of hydrometallurgical Au reclamation

56

Process Figure 3.1

Flow Sheet of the process

Figure 4.1

Representation of Ag extraction as a function

Figure 4.2

66

of time for different HNO3 concentrations

76

Representation of Ag extraction as a function

78

of time for various stirring speeds Figure 4.3

Representation of Ag extraction as a function

80

of time for different temperature Figure 4.4

Representation of Ag extraction as a function

81

of time for various solid/liquid ratio Figure 4.5

Recovery of silver cemented from silver chloride

82

as a function of time Figure 4.6

Recovery of gold from pregnant aqua regia solution as a function of time

xi

84

Figure 4.7

Precipitation of gold from pregnant aqua regia

85

solution at different temperatures as a function of time Figure 4.8

Precipitation of gold from pregnant aqua regia

86

solution at different stirring speeds as a function of time Figure 4.9

Recovery of platinum from the filtrate after

88

removing gold as a function of time Figure 4.10 Concentration of platinum at different temperatures

89

as a function of time during recovery with hydrazine Figure 4.11 Concentration of platinum at different stirring speeds 91 as a function of time

xii

LIST OF PLATES

Plate 4.1

Metallic silver photo

93

Plate 4.2

Metallic gold photo

93

Plate 4.3

Platinum black powder photo

94

xiii

LIST OF TABLES

Table 2.1

The purity of gold alloys

15

Table 2.2

Gold usage in various applications

15

Table 2.3

Silver usage in various applications

18

Table 2.4

Platinum usage in various applications

19

Table 2.5

Physical properties of gold

21

Table 2.6

Physical properties of silver

25

Table 2.7

Physical properties of platinum

28

Table 3.1

Chemical analysis of jewellery waste sample

64

Table 3.2

Screen size analysis of jewellery waste

65

Table 4.1

Dissolution of jewellery waste sample by nitric acid

74

Table 4.7

Analytical lines used by industry for determination

83

of gold by ICP Table 4.10

Analytical lines used by industry for determination

87

of platinum by ICP Table 4.13

Final results

92

Table 4.14

Calculation of income

97

Table 4.15

Profit for the global process

97

Table 4.2

The result of effect of HNO3 concentration

119

on the dissolution of Ag from the jewellery wastes

Table 4.3

Effect of stirring speed on the dissolution of silver from jewellery wastes xiv

120

Table 4.4

The results of the effect of temperature on

121

the dissolution silver from jewellery wastes Table 4.5

Results on the effect of solid/liquid ratio on

122

the dissolution of silver from jewellery wastes Table 4.6

Results of cementation of silver chloride

123

Table 4.8

Effect of temperature on precipitation of gold

124

Table 4.9

Results of precipitation of gold using different

125

stirring speeds Table 4.11

Effect of temperature on precipitation of platinum

126

Table 4.12

Effect of stirring speed on precipitation of platinum

127

xv

LIST OF ABBREVIATIONS USED

AAS

Atomic Absorption Spectrometer

ICP-OES

Inductively Coupled Plasma Optical Emission Spectroscopy

XRD

X- ray Diffractometry

CSTR

Continuously stirred tank reactor

CIC

Carbon in columns

SX

Solvent extraction

CILO

Carbon in leach with oxygen

CIP

Carbon in pulp

CIL

Carbon in leach

ESA

Extended surface area

HMS

Heavy medium separation

BNH

Brinell hardness

CCD

Counter-current-decantation

SAG

Semi-autogenous grinding

RIC

Resin-in-column

LD50

Lethal dose to 50% of the population

xvi

NOMENCLATURE

Symbol

Explanation

K

degrees Kelvin

ºC

degrees Celsius

cm

centimeter

g/dm3

grams per cubic decimoles

h

hour

kg

kilogram

mg

milligram

ml

milliter

nm

nanometer

Oz

troy ounce = 31.104 grams

pH

negative logarithm of the concentration of hydrogen ions

s

seconds

SCE

Standard Calomel Electrode

SHE

Standard Hydrogen Electrode

UK

United Kingdom

USA

United States of America

USD

United States Dollar

t

Time

R

South African Rand - currency

xvii

MPa

mega Pascal

fcc

face-centered cubic

xviii

GLOSSARY

Aqua regia

a mixture of acids containing 3 parts hydrochloric acid and one part nitric acid. Some authors include one part water.

PGM

Platinum Group Metal, namely platinum, palladium, rhodium, ruthenium, iridium and osmium.

xix

CHAPTER 1 INTRODUCTION

1.1 BACKGROUND AND MOTIVATION

Jewellery production is one of the main uses of gold, and accounts for more than 80% of the total yearly market for the metal consumption. In small jewellery workshops, were jewellery is measured, finished, cleaned and polished, the waste is collected and transferred to foundries for extraction and final treatment. Jewellery workshops generate three different kinds of wastes (Ferrini, Manni and Massacci, 1998b: 529-534):

1. Handwashing waste 2. Floor sweepings waste 3. Jewellery polishing waste

Handwashing waste is generated by the rubbish coming from operators and clothes for laboratory cleaning; it is characterized by an organic matrix from which gold particles are segregated. It contains some other components, such as soap and even exhausted coffee powder that makes its beneficiation difficult. Jewellery polishing waste is generated when the artifacts are cleaned and polished using bristles of different hardness. It is characterized by a mixture of plastic and metal bristles, abrasive paste and 1

metal dust, the latter mainly composed of gold alloys. The floor sweepings waste contains large amounts of waste dust and debris, known commonly as sweeps.

Due to the importance and cost of noble metals and its extremely low concentration level in various matrices, suitable methods for reliable leaching, extraction, recovery and determination are required. The efficient recovery of carat gold scraps and wastes is a vital component of a profitable jewellery manufacturing business. Much of the scrap can be recycled, but refining contaminated scraps to pure gold is an operation not without problems. The easy option is to send all such scraps and wastes to a specialist refiner. This is not always a realistic or cost-effective option and so an in-house refining capability may be desirable (Corti, 2002). However, most of these methods are very laborious and not practical. In a major advance in refining technology, modern hydrometallurgical techniques have been adopted to produce high-purity gold from waste containing silver and platinum with a wide range of gold contents (Embleton, 1989: 315-319).

Various methods for treating noble metals have been reviewed and discussed by numerous experts in the field (Prassad et al., (1991)). Although cyanide leaching remains the overwhelming option for the treatment of gold ores because of its economical benefit and simplicity, it suffers from certain inherent drawbacks such as toxicity and slow leaching (Hiskey and Alturi, 1988:96). Swaminathan, Pyke and Johnston (1992:1)

2

mentioned that despite the impressive safety record of cyanide to date, it is conceivable that environmental concerns could fuel trends toward lixiviants such as halogen-based systems, nitric acid systems or thiourea for leaching. While difficulties remain, leaching with solutions other than cyanide has considerable potential as an effective and less hazardous procedure for gold, silver and platinum recovery.

The Department of Fine Arts at Tshwane University of Technology has a section of jewellery design and manufacture. During the manufacture of various jewellery pieces, the filings and other waste are collected and need to be recovered for re-use. Its recovery is costly and time consuming and is the main driving force for this investigation into the development and verification of an “in-house” method for successful recovery of such wastes.

There are several methods for the treatment of jewellery wastes, but three possibilities seems feasible on an “ in-house” scale:

1. Separating the waste by melting it. 2. Selective dissolution and subsequent recovery by electrochemical means. 3. Heavy medium separation (HMS), by selecting a medium with a corresponding density to that of the metal to be separated from collection.

3

It is generally accepted that hydrometallurgical processing requires less capital and is more efficient than pyrometallurgy if the metal concentration to be recovered is low (from the per cent range down to parts per million). The heavy medium separation (Sink – Float separation) approach uses a variety of heavy liquids, some of which are extremely poisonous and give off toxic fumes (Wills, 1988: 420-424). The yield is often less than what can be achieved with a selective dissolution method. Furthermore, there is nowadays a growing interest in replacing cyanide by non-toxic reagents that are environmentally more safe than cyanide.

A number of review papers describing gold processing developments (Haque, 1987; Jha, 1987; Bhappu, 1990; Stanley, 1990; Dayton, 1987 as cited by Prassad et al., (1991:1259)), on both refractory and non-refractory ores and the dissolution chemistry of gold and silver in different lixiviants (Hiskey, 1984:173-178) have been published. Dahne, in his paper “ Gold Refining and Gold Recycling” discussed the various well-established pyrometallurgical and wet chemical refining procedures, such as the Miller Process, Wohwill electrolysis, wet chemical refining with nitric acid, the cupellation process, the lead collecting process, for recovery of gold from the raw gold derived from gold ores and recycling materials (Dube, 2001:3).

The present research work deals with leaching of jewellery wastes using nitric acid to dissolve silver and aqua-regia (one part nitric acid, by volume, to three parts hydrochloric acid) to dissolve gold and platinum.

4

1.2 PROBLEM STATEMENT

The project will attempt to find an “in-house” solution to the recovery of noble metals from jewellery wastes collected during practical sessions of students being trained in jewellery design, in order to save time and financial costs (of sending the wastes to a commercial operation for recovery) and to supply a “technical” grade purity of material for training purposes in introductory courses.

1.3

HYPOTHESIS

The work carried out in this study will attempt to prove that: •

Wet chemical processes can be used to separate and recover noble metals from jewellery wastes.

•

The refining of those noble metals can successfully be done inhouse.

•

The entire process can be done in an effective and cost efficient manner.

5

1.4

OBJECTIVES

The main objective of this project is to separate, recover and refine jewellery wastes obtained from Fine Arts Department using a simple and economical method that is environmentally safe.

6

CHAPTER 2 LITERATURE REVIEW 2.1

HISTORY

Gold, silver and platinum are all regarded as precious metals. In particular, gold and silver have been highly prized by man for thousands of years. Early civilizations associated the yellow colour of gold with the sun, and for the Egyptians, gold was the symbol of their sun god Ra. Just as gold was associated with the fire of the sun, so the brilliant white colour of silver was associated with the moon, “Luna”, meaning white and shining (Grimwade, 1985:1). In ancient literature gold is the universal symbol of the highest purity and value (cf. passages in Old Testament, e.g. Ps 19 verse 10). The luster and fine colour of gold have given rise to most of the words which are used to denote it in different languages. The word “gold” is probably connected with the Sanscrit word “jvalita” which is derived from the verb “jval”, to shine. An alternative explanation is that it stems from the old English word, “geola” meaning yellow (Grimwade, 1985:1).

The discovery of platinum is much more recent. Although it was used by the pre-Columbian Indians of South America to make small ornaments, it was not until the 18th century that its true worth was recognized (Grimwade,

7

1985:2). It was the Spaniards who first called this white metal “ platina”, a derogatory diminutive of their word for silver ‘plata’ and it was considered by many to be a worthless nuisance, up to as recently as 1780.

There are a number of reasons which make these metals so precious. Firstly, they have a remarkable resistance to attack by their environment. Whereas the base metals readily combine to form oxides, sulphides and other minerals, the precious metals, or “noble metals” as they are sometimes called, can occur naturally in the uncombined state as lode deposits in rocks, or as placer deposits in the gravel beds of rivers and streams and in alluvial sands. Secondly, gold and silver, different to platinum, are widely distributed throughout the world, which explains why so many different civilizations discovered these naturally occurring “native metals”. It was relatively easy to extract the gold and silver, and the early metalworkers soon discovered that they could be easily fabricated into shapes for decorative purposes to display their aesthetic qualities and beauty (Grimwade, 1985:1).

Today we find that beautiful works of art in gold, silver and platinum are being made not only using the traditional skills of the goldsmith, but also incorporating modern technology in their manufacture. South Africa is a major producer of gold, accounting for about 68% of total global production (Mohibe, 1981; Thomas and Boyle, 1986;

Lutley,

1997). For platinum,

South Africa produces 70% of total global production (Mintek, 2002).

8

2.2

OCCURRENCE, EXTRACTION AND REFINING

The precious metals are found in the earth’s crust, as are all other metallic and non-metallic elements, but in very small quantities. In 1924, Clark and Washington published an average composition of the earth’s crust based on the analysis of a vast collection of rock samples from all over the world and estimated that gold was present at between 1 and 9 mg.ton -1, with platinum at a similar level, and silver some ten times higher. This partly explains why these metals and the cost of extraction is very high. Even so, extraction would be extremely costly, if not uneconomic, if their distribution was uniform around the world. Fortunately, as with other important minerals, localized concentrations are found in certain regions as ore deposits and mining and extraction becomes economically feasible (Grimwade, 1985:6).

2.2.1

Gold

Gold is often found combined with silver. A small amount of copper may also be present in native gold. Generally, the gold content is 85-92%, but can range from about 50% in the very pale-yellow alloy with silver, known as “electrum” to almost pure gold (Grimwade, 1985:7). The biggest high-quality gold nugget ever discovered was about 60 cm long and 30 cm wide and assayed at 98.66% gold. The only mineral found in any quantity is gold

9

telluride (calaverite and sylvanite). Four types of gold deposits are found (Grimwade, 1985: 7):

1. Quartz veins and lodes within rock. 2. Massive deposits in which the fine native gold particles are associated with sulphides and fine-grained quartz. 3. Disseminated copper deposits in which gold is recovered as a by-product from copper–sulphide concentrates. 4. Placer deposits in streams or former stream gravel beds, alluvial and beach deposits. These deposits tend to occur in folded sedimentary rocks such as those found in the Americas, Australia and South Africa. The richest deposits are found in Gaunteng (Witwatersrand) and the Free State in South Africa in a 482,7 km semi-circular arc (Weston, 1983). It was discovered by George Harrison in 1886.

There are several methods of extracting gold but the leading ones are (http://37.19//encyclopedia.org/G/GO/GOKCHA.htm):

Simple washing, i.e. dressing auriferous sands, gravels, etc.; Amalgamation, i.e. forming a gold amalgam, afterwards removing the mercury by distillation; this process is employed to extract gold from both alluvial and reef deposit.

10

Chlorination or the Plattner Process, i.e. forming soluble gold chloride and then precipitating the metal using ferrous sulfate, charcoal and sulfuretted hydrogen, either alone or mixed with sulfur dioxide. Cyanide process, i.e. dissolving the gold in potassium cyanide solution,

and

then

precipitating

the

metal

using metallic zinc or by electrolysis. Electrolytic process, which is applied to the solutions, obtained from a washing process.

2.2.2 Silver

The occurrence of native silver is now much rarer than native gold since most of the silver mines of Asia Minor, Greece, Spain and the Americas have been exhausted. Only a small fraction of the world’s production comes from the amalgamation and cyanidation process as a by-product of gold extraction. Silver is mainly extracted from sulfide minerals such as argentite, tetrahedrite, polybasite, pyroargyrite and stephanite (Grimwade, 1985:9). Of these, only argentite is a simple sulfide, Ag2S, the others being complexes with antimony and copper. In addition, the silver sulphides are usually associated with other minerals such as lead, zinc and copper sulphides so that extraction is a complicated business and silver is a co-or by-product in the production of these other metals (Grimwade, 1985:9).

11

The following leading extraction methods are simply explained:

2.2.2.1 Smelting In the early days, considerable quantities of silver were extracted simply by smelting high percentage silver ores under oxidizing or reducing conditions with conversion of accompanying materials. They were almost exclusively ores containing elemental silver, silver sulfide, or silver halides (Habashi, 1997:1224).

2.2.2.2 Amalgamation

For many centuries, amalgamation was the most important process for the treatment of silver ores. Today, it is rarely used because of high costs, poor yields and the toxicity of mercury (Habashi, 1997:1224).

2.2.2.3 Cyanidation The aqueous cyanide process is the dominant process for extraction of silver from silver ores. The conditions are essentially the same as those for the gold cyanide process (Habashi, 1997:1225).

12

2.2.3 Platinum

The Spanish conquest of South America revealed that the pre-Columbian Indians were well versed in the art of making platinum trinkets. Analysis of a “platinum ingot” found in the esmeraldas in Columbia gave 84.95% platinum, 4.64% palladium, rhodium and iridium, 6.94% iron, and 1% copper (Grimwade, 1985:10). The platinum occurred as small grains of white native metal together with gold grains in placer deposits in streams.

Russia produced 93% of the world’s supply of PGMS, until the geologist Dr. Hans Merensky discovered the world’s largest platinum resources in an outcrop 136 km long at Rustenburg, north-west of Johannesburg, in South Africa. This outcrop, named in his honour as the Merensky Reef, now forms part of a much larger complex (the Bushveld Igneous Complex). The platinum content is partly in the form of native metal alloyed with iron, and partly as sulphides and arsenides e.g. cooperite (PtS) and sperrylite (PtAs2). Together with sulphides of iron, copper and nickel, are also the other PGMS associated with the platinum. The reef is about 50 cm thick with a maximum depth of about 900 m (Grimwade, 1985:11).

Global platinum demand reached a record level of 6.15 million ounces in 2001, driven mainly by a one-third increase in usage by the autocatalyst

13

industry. Total platinum supplies rose by 570 000 ounces to 5.86 million ounces, and South Africa’s output increased to 4.1 million ounces, compared with 3.8 million ounces in the previous year. If all the planned projects come to fruition, SA’s platinum output will exceed 6 million ounces per annum in the second half of the decade (Mintek,2002).

2.3

USES OF NOBLE METALS IN JEWELLERY

2.3.1

Gold

The gold value of jewellery alloys is determined by their gold content (fineness). In most countries, laws govern the terms used in designating the fineness of gold jewellery for manufacturers, processors and dealers. Pure gold has a fineness of 1000 and is much too soft to be used for jewellery manufacture (Grimwade, 1985:47). It is designated as 24 carat. A comparison of the different units in which the purity of a gold alloy is expressed, is given in table 2.1.

14

Table 2.1. The purity of gold alloys (Grimwade, 1985:47)

Caratage

Fineness

% Gold

24 22 20 18 15 14 12 10 9 8

1000 916.6 833.3 750 622 585 500 416 375 333

100 91.7 83.3 75.0 62.2 58.5 50.0 41.6 37.5 33.3

The usage of noble metals in various applications is given in the tables 2.2, 2.3 and 2.4.

Table 2.2 Gold usage in various applications (Grimwade, 1985:47)

Sectors

Distribution (%)

Carat jewellery Coins Electronics Industrial and decorative uses Dental Investment bars Other uses

15

57 21 6.5 5.0 4.7 3.7 2.1

In order to appreciate the development and use of the carat-gold alloys it is necessary to briefly examine the alloying behaviour of gold. To do this it is convenient to divide the gold into two classes, namely the yellow coloured carat gold and the white carat gold. The coloured carat golds are based on the gold-silver-copper (Au-Ag-Cu) alloy system although other additions, notably zinc (Zn), may be present in the lower-caratage alloys (Grimwade, 1985:49).

The 18 carat Au-Ag-Cu alloys are yellow in colour and have very good mechanical properties. As such, it is the most popular alloy combination in the production of jewellery pieces. A standard 18 carat Au-Ag-Cu alloy contains 75% Au by weight, with a formula of Au75Ag12.5Cu12.5 (Zhang et al., 1995:603-609). In an 18 carat alloy the remaining alloy content of 25% consists of silver and copper in varying proportions, as shown in the ternary equilibrium diagram in figure 2.1.

16

Figure 2.1 Weight per cent composition ranges for various colours of Au/Ag – Cu alloys (Zhang et al., 1995:603)

Coloured gold alloys used in the jewellery industry are mostly based on the ternary alloy system Au - Ag - Cu, allowing a wide variety of colours that can be produced. The workability of an alloy and its resistance to wear depend on its mechanical properties; these, and its resistance to corrosion, can be controlled by adding zinc to the basic formulation (Zhang et al., 1995:603-609).

White gold alloys were first developed in the early 1900s, in an effort to replace platinum by a cheaper material with identical properties (Grimwade, 1985:53). It is possible to produce a 9 carat white gold in the 17

Au-Cu-Ag system with up to 6% Cu such that the Ag + Cu content is 62.5%. Such alloys are used for jewellery manufacture but they are rather unsatisfactory as they are too soft. Nickel and palladium are the only suitable additives to give gold a colour approaching whitish-grey. The demand for white gold has fallen in the last few years in favour of coloured gold and platinum (Grimwade, 1985:54).

2.3.2 Silver

Table 2.3 Silver usage in various applications (Grimwade, 1985:47)

Sectors

Distribution (%)

Photography Electrical and electronics Silverware and jewellery Brazing alloys Other uses

32 31 17 13 7.0

In earlier times, the manufacture of silver articles for jewellery and general use was the main outlet for silver, apart from coinage manufacture, and it is still very important. Today, silver-copper alloys with a fineness of 925, 835, and 800 are used in all forms of jewellery because of their favourable properties for both manufacture and use. The standard materials, Ag925, Ag835, Ag800 are supplied in the form of sheets, bands, wires, rods, and profiles (Habashi, 1997:1252).

18

2.3.3 Platinum

Table 2.4 Platinum usage in various applications (www. Platinuminfo.net/pgm-pt.htm)

Sectors

Distribution (%)

Chemical Electrical Glass Petroleum Other uses Jewellery Autocatalyst

6 7 4 2 9 51 21

The platinum standard of 950 fineness has been accepted in the United Kingdom for hallmarking purposes, and this standard is likely to be adopted in many other countries. Platinum has been in such demand for jewellery and high-quality watchmaking that more than one-third of all the platinum consumed has been used for this purpose (Grimwade, 1985:65). Of this amount, Japan uses about 90%, with the United States taking the next largest share. Most of the alloys employed contain 95% platinum and are stamped PLAT 950. Pure platinum is too soft for use in jewelry and so it is hardened by alloying with cobalt or rubidium (Habashi, 1997:1279).

19

2.4 PHYSICAL AND CHEMICAL PROPERTIES

2.4.1

2.4.1.1

Gold

Physical properties of gold and gold alloys

Gold, atomic number 79, atomic mass 196.96654 g.mol-1, has only one naturally occurring isotope,

197

Au. Its most important radioisotope, which is

used in medicine, is 195Au. It emits α and γ rays and has a half-life of 183 d. The electronic configuration of gold is [Xe] 4f145d106s1 and its atomic radius is 0.1439 nm. The ionic radius for coordination number 6 is 0.1379 nm for Au+, and 0.085 nm for Au3 + (Habashi, 1997:1186). Some physical parameters of gold is given in table 2.5:

20

Table 2.5 Physical properties of gold (Habashi,1997:1186)

Properties

Values

Melting point Boiling point Density at 20 °C Density at 1065 °C Vapor pressure at 1064 °C Atomic volume at 20 °C Electrical resistivity at 0 °C Thermal conductivity at 0 °C Specific heat Enthalpy of fusion Enthalpy of vaporization Tensile strength

1064.4 °C 2808 °C 19.32 g cm-3 17.32 g cm-3 0.002 Pa 10.21 cm3 mol-1 2.06 x 10-6 Ω cm 3.14 W cm-1 K-1 0.138 Jg-1 K-1 12.77 kJ mol-1 324.4 kJ mol-1 127.5 MPa

The unit cell of gold is face-centered cubic (fcc), with a lattice constant (ao) of 0.40781 nm. Gold as it occurs in nature usually does not have a very crystalline appearance. It exists in thread-like, leaf-shaped, and spherical forms, on which cubic, octahedral, and dodecahedral surfaces can sometimes be seen. When large amounts of molten gold solidify, a characteristic pattern of concentric rings appears on the surface. Pure gold that has not been mechanically pre-treated is very soft. Its hardness on the Mhos’ scale proved that it is the most ductile of all metals and can be cold drawn to give wires of less than 10 µm diameter, and beaten into gold foil with a thickness of 0.2 µm (Habashi, 1997:1186). Due to its softness, gold can be highly polished. The colour of utility gold is less rich and varies considerably according to its alloy composition. Very thin gold foil is

21

translucent;

and

transmitted

light

appears

blue-green

(Habashi,

1997:1186).

2.4.1.2 Chemical properties The dissolution behaviour of gold and silver in the presence of various complexing ligands can be determined with the aid of Pourbaix diagrams. The development and use of these diagrams in hydrometallurgical systems is well documented (Hiskey and Alturi, 1998:95-134). Various sources of thermodynamic information provided data to calculate standard potentials and equilibrium constants. Latimer (1952) and Sillen and Martell (1950) represented the basic reference sources for most of the species considered. Additional data were obtained from Jorgensen and Pourodier (1970:124-127) and then from Hancock, Finkelstein and Evers (1974:25392543). The thermodynamic values used throughout these figures are for a temperature of 25 ºC.

Gold-Water System From the Pourbaix diagram showed in figure 2.2 the Eh–pH relationships for the Au–H2O system conveniently illustrate the extremely noble nature of gold in the absence of coordinating ligands. The diagram was constructed for dissolved gold species having concentrations equal to 10-4 M. The dashed lines delineate the stability limits of water. Metallic gold is

22

predominant in a very large area, including the entire domain of water stability. The aurous ion (Au+) does not appear on the diagram because it disproportionates spontaneously according to the following reaction for the aqueous ions

3 Au+ (aq) →

2 Au (s) + Au3+ (aq)

(1)

The value of the equilibrium constant for this reaction at 25 °C is K = 6.7x 109. The auric ion (Au3+) and other oxidized forms of gold only occur at potentials above the upper stability limit of water. Oxygen and the various oxidized gold species reduced simultaneously to the metallic state. This means that gold cannot be oxidized in strong acids or strong alkalis in the absence of complexing ligands. Due to gold being completely and perfectly stable in the presence of water, it is found in nature mainly in the native form of atomic gold.

23

Figure 2.2 potential-pH equilibrium diagram for the gold-water system for dissolved gold species of concentration 10-4 M (Hiskey and Alturi, 1987).

2.4.2 Silver

2.4.2.1 Physical properties Silver has an atomic number of 47 and a relative atomic mass of 107.8682. Natural silver consists of the stable isotopes Ag107 (51.8%) and Ag109 (48.2%). The electronic configuration of silver is [Kr] 4d105s1 (Habashi, 1997:1218).

24

Some physical parameters for silver is given in table 2.6:

Table 2.6 Physical properties of silver (Habashi, 1997:1218)

Properties

Values

Lattice constant at 20 °C Atomic radius Melting point Boiling point Specific heat capacity at 25 °C Thermal conductivity Vapor pressure at 1030 °C Density at 20 °C Density of liquid at melting point Brinell hardness Tensile strength Resistivity at 0 °C

0.4077 nm 0.144 nm 961.9 °C 2210 °C 0.23 J kg-1K-1 418 Wm-1K-1 1.33 Pa 10.49 gcm-3 9.30 gcm-3 26 BHN 140 MPa 150 µΩcm

Like gold, silver crystallizes in a face-centered cubic structure in which each metal atom is surrounded by 12 neighbours. This high degree of symmetry results in a structure with many slip planes and is the reason for its good mechanical formability. Silver has the highest electrical conductivity, the highest thermal conductivity, and the lowest electrical contact resistance of all metals (Habashi, 1997:1219).

2.4.2.2 Chemical properties Silver is generally monovalent, although other valencies are known to exist. One remarkable property of silver is the fact that in the molten state it can

25

dissolve vast quantities of oxygen, e.g. 100 g of silver will dissolve 213.5 ml of oxygen, which is equivalent to 0.305% by weight at 973 °C (Grimwade, 1985:36). Unlike gold, silver will dissolve in nitric acid and in hot concentrated sulphuric and hydrochloric acids. This fact is made use of during gold assaying when silver is ‘parted’ from gold by dissolution in HNO3 (Grimwade, 1985:36).

Silver – Water System Figure 2.3 shows the Eh-pH behaviour for pure silver at 25 °C. The diagram was constructed for dissolved silver species having concentrations equal to 10-4 M. Silver, like gold, is a highly noble metal covering a large portion of the domain of water stability. Metallic silver is fully stable in the absence of oxidizing agents and complexing ligands at all pH values.

However, silver can be dissolved under oxidizing conditions in acidic to moderately alkaline solutions to yield Ag+ and in strongly alkaline solutions to yield AgO-. The shaded areas in figure 2.3 indicate the respective regions for these species at a 10-4 M concentration level.

26

Figure 2.3 Potential –pH equilibrium diagram for the silver- water system for dissolved silver species of 10-4 M. Shaded regions show where silver is soluble in the domain of water stability (Hiskey and Alturi, 1987).

In natural waters Ag+ is stable under certain conditions and its existence in aqueous environments allows for the formation of a great many silverbearing minerals. The silver ion can conceivably precipitate from solution to form sulfides, halogenides, selenides. Furthermore, Ag+ can undergo secondary enrichment type reactions with primary minerals to form silverbearing minerals.

As a result of these factors, silver occurs in

approximately 200 different minerals and is characterized by complex mineralogy (Gasparini, 1984:99-110).

27

2.4.3 Platinum

2.4.3.1

Physical properties

Some physical parameters for platinum is given in table 2.7:

Table 2.7 Physical properties of platinum (Habashi, 1997:1268)

Properties Atomic number Relative atomic mass Crystal structure Lattice constants at 20 °C Atomic radius Melting point Boiling point Specific heat at 25 °C (Cp) Thermal conductivity Density at 20 °C Brinell hardness Tensile strength Specific electrical resistance at 0 °C

2.4.3.2

Values 78 195.08 fcc 0.3923 nm 0.139 nm 1772 °C 4170 °C 0.13 Jg-1K-1 73 Wm-1K-1 21.45 g cm-3 50 BHN 137.3 MPa 9.85 µΩ cm

Chemical behaviour of platinum

Platinum and palladium dissolve in a mix of hydrochloric and nitric acid in ratio 3:1, the other platinum-group metals are largely insoluble in it (Lowen, 1995).

Platinum has a very valuable function in chemical engineering;

many chemical reactions can proceed at a very rapid rate in the presence of a platinum surface whereas they would be uneconomically slow in its

28

absence (Letowsky and Distin, 1985, 1987, 1989). This behaviour is known as catalysis and platinum stands supreme among all metals as a catalyst (Krystyna, 1998; Grimwade, 1985:67).

2.5 CURRENT PREFERRED REFINING PROCESSES

2.5.1 Introduction

Processes for the extraction of gold have improved dramatically over the years. The latest developments in gold beneficiation technology have not only reflected the economic aspects in terms of increased efficiency and reduced costs, but also addressed the environmental aspects, particularly with regard to gaseous emissions and liquid effluents discharged from gold plants (Prassad et al., 1991:1260). The dissolution of gold by alkaline cyanide remains the most common method of extraction of gold from its ores. However, many gold ores do not respond very well to the conventional cyanidation process. Such ores are referred to as refractory ores, and are characterized by low gold recoveries and high cyanide consumptions when subjected to direct cyanide leaching (Prassad et al., 1991:1260).

These ores are subjected to oxidation pretreatment methods such as roasting, chemical oxidation, pressure oxidation and bio-oxidation prior to

29

cyanidation. Furthermore, there is general interest in replacing cyanide by lixiviants which are non-toxic and environmentally safe.

2.5.2 Current developments in the extractive metallurgy of gold

2.5.2.1 Conventional methods

Free milling of gold ores Cyanidation More than 87% of gold extracted in the United States in 1988 employed classical, direct cyanidation (Lucas, 1987:416; Jha, 1984). This process is ideal for free milling, non-refractory ores. Ore is first ground by conventional multistage crushing and ball milling, or by semi-autogenously grinding (SAG), then leached in agitated or pachuca-type tanks, using lime, dilute cyanide solution and oxygen. The leached pulp then undergoes counter-current-decantation (CCD) or filtration, using drum or belt filters. The precious metal values are finally precipitated from the pregnant leach liquor using zinc dust (the Merrill – Crowe process), or are electrodeposited on a steel wool cathode. The overall chemical reactions are (Prassad et al., 1991:1262): Leaching: 4 Au(s) + 8 NaCN(aq) + O2 (g) + 2 H2O (aq) = 4 NaAu(CN)2 (aq) + 4 NaOH (aq)

(2)

30

Merrill-Crowe precipitation: 2 NaAu(CN)2 (aq) + Zn(s) = Na2Zn(CN )4 (aq ) + 2 Au0(s)

(3)

Electrodeposition: 4 OH- = O2 + 2H2O + 4 e-

(4)

2 e- + 2 Au(CN)2 = 2 Au0 + 4 CN-

(5)

Direct alkaline cyanidation is applicable only to free milling ores. If the ore is refractory, a pretreatment step is necessary to make it amenable to cyanide leaching.

Developments in the treatment of refractory ores. Refractory ores

A detailed analysis of various causes of refractoriness of gold ores is described by Jha (1987: 331 – 352). If the refractoriness is due to the presence of sulfides, an oxidation pretreatment is given to render the ore amenable for cyanidation. The conventional method of treating these ores and concentrates is by oxidative roasting followed by cyanidation.

31

Pretreatment methods Roasting of sulfides

This approach is both expensive and environmentally undesirable and has therefore rendered many ores uneconomical (Hake, 1987). Gold roasting plants are currently seeking new technology to reduce environmentally undesirable emissions of SO2, particulates, mercury and arsenic. The removal of SO2 is affected by use of lime scrubbing of effluent gases in which up to 95% of the SO2 is removed. The major disadvantage of this process is that the production of sulfuric acid from the SO2 generated is only feasible under special conditions where the concentration of SO2 in emissions is high (Hake, 1987).

However, many of the refractory gold ores do not meet this special condition. Several innovations in roasting are in developmental stages and the trend is towards fluid-bed roasting with dry ore grinding or with highdensity slurry feed of flotation concentrates (Prassad et al.,1991:1261). Subsequent cyanidation leach recoveries from roasting ores generally range from 75 to 90%, whereas recoveries from concentrate roasting range from 85 to 95%. Arsenopyrite often requires a two-stage roast. The first stage roast is at lower temperature and oxygen deficient to produce As2O3 and avoid the formation of FeAsO4. The second stage roast is at a higher temperature with excess oxygen to produce Fe2O3 and complete

32

conversion of the sulfides to oxides (Coleman and Robert, 1989). The relevant roasting reactions are: Stage 1 roast: FeAsS (s)

→

Fe S (s) + As (g)

2 As (g) + 3 /2 O2 (g)

→

As2 O3 (g)

(6) (7)

Stage 2 roasts: As2O3 (g) + O2

→

As2O5 (s)

(8)

Fe2O3 (s) + As2O5 (s)

→

2FeAs O4 (s)

(9)

4 FeS2 (s) + 11O2 (g)

→

2 Fe2O3 (s) + 8 SO2 (g) (10)

Alternative commercially viable methods of oxidation have now been developed to eliminate the problems associated with roasting. These include chemical oxidation and bio-oxidation processes (Prassad et al., 1991: 1261).

Chemical oxidation Commercial application of chemical oxidation at ambient pressure was first applied at the Carlin Mines (USA) for the oxidation of carbonaceous sulfide ores (Prassad et al., 1991:1261). The major purpose of the Carlin type process is to oxidize the carbonaceous material and humic acid using chlorine as oxidizing agent. The Carlin process could use either sodium hypochlorite generated in situ by electrolysis of brine containing pulp or use chlorine directly. The relevant chlorine reactions are:

33

Cl2 (g) + NaOH (aq) = NaOCl (aq) + HCl (aq)

(11)

HCl (aq) + NaOH (aq) = NaCl (aq) + H2O(l)

(12)

Au (s) + 3/2Cl2 (aq) = AuCl3 (aq)

(13)

FeS2 (s) + 7NaOCl (aq) + 2NaOH(aq) = FeCl2 (aq) + 2Na2SO4 (aq) + 5NaCl (aq) + H2O(l)

(14)

An improvement of the Carlin process can be achieved by the ”Double Oxidation” process (Prassad et al., 1991:1261). It involves the pre-oxidizing of the pyrite before chlorination was applied to conserve chlorine. In this process, the slurry is aerated at 80-86 °C until a considerable portion of the pyrite and some of the carbonaceous materials are oxidized. The preoxidation is then followed by the chlorination method to oxidize the carbonaceous material and the rest of the pyrite. The success of this double oxidation process for the pyrite at Carlin is due to the porosity and large surface area of the spheroidal pyrite.

Further developments in the area of chemical oxidation on gold ores include Caro’s acid oxidation developed by Ontario Research Foundation (Prassad et al., 1991:1261) and the Nitrox Process (Van Weert et al., 1986:84-85). The Caro’s acid oxidation was attempted on gold bearing arsenopyrite ores or concentrate and was shown to improve gold extraction from less than 65% by direct cyanidation to 85% during the development

34

stage. The Nitrox process, which treats the ore for 1 to 2 h in nitric acid in the presence of air at atmospheric pressure to oxidize pyrites and arsenopyrites prior to cyanidation, claims to increase the gold recoveries from 30% to 90% (Prassad et al., 1991:1262).

Pressure oxidation

Pressure oxidation or autoclaving is an aggressive pretreatment method for highly refractory gold ores and concentrates (Wall et al., 1987:393-401). This chemical oxidation process uses high-pressure oxygen and heat to break down the sulfides and carbonaceous ores to free the gold, which is usually recovered by a subsequent cyanidation leach. The ore may be treated directly or concentrated by flotation or gravity methods prior to pressure oxidation. Preconcentration of the ore may be preferable since it limits the size of the plant. The exact pressure and temperatures required depend on the ore, and are held as low as possible to minimize the plant cost.

Generally pressure oxidation is conducted at temperatures in the range of 180 – 210 ºC and pressures from 200 – 300 psi. The use of pressure oxidation was first employed at Homestake’s McLaughlin Mine in California, USA (Prassad et al., 1991:1262). The process was based on the use of autoclaves operating at pressures in excess of 188 psi (13 bar)

35

and temperatures of 160 – 180 ºC to affect the oxidation of sulfides and carbonaceous materials. The slurry enters the autoclave at a temperature of 90 – 120 °C and is acidified by the addition of sulfuric acid to a pH of 1.8 – 1.9. Oxygen is sparged into the autoclave at 34 – 45 kg/t slurry. The oxidized ore exits at a temperature of about 175 ºC and flows into thickeners where the acid is washed out. Lime is then added to neutralize the acid and also raise the pH to 10.8, with a subsequent cyanidation leach. This process is claimed to achieve an overall gold recovery of about 93%.

Another interesting process is the “high pressure low alkalinity cyanidation” which was tested on pilot scale in South Africa on stibnite concentrates (Prassad et al., 1991:1262). The process involves batch cyanide leaching of concentrates at high pressure (8.8 MPa) at 7.0 pH in a tube reactor with oxygen over a pressure of 12.0 MPa. It was reported that gold extractions improved from low values of 1% for direct cyanidation to about 85%.

Bio-oxidation

Bio-oxidation is a process of using bacteria to aid in the chemical break down of some sulfide ores (Prassad et al., 1991:1262). Thiobacillus ferroxidans is the most common bacteria used for this purpose. Other bacteria used include thiooxidans and sulfolobus acidocaldarious which

36

thrive at higher temperatures. These bacteria are found in mine drainage and natural hot springs. The typical bacteria such as thiobacillus ferrooxidans gets its energy by oxidizing sulfur via the reaction:

S2- = S6+ + 8 e-

(15)

Thus the sulfur in FeS2 becomes sulfate anions. The other important reaction is the oxidation of the ferrous to the ferric iron: 7 Fe2 (SO4 )3 (aq) + Fe S2 (s) + 8 H2O (l) → 15 Fe SO4 (s) + 8 H2SO4(aq)

(16)

This reaction is important because it breaks down the pyrite, allowing the encapsulated material such as gold to be leached by other extraction methods. The original bio-oxidation process used fixed bed reactors such as dumps and heaps. However, recent developmental work (Prassad et al., 1991:1262) has demonstrated that a continuously stirred tank reactor (CSTR) will give the optimum results. The use of a CSTR for bio-oxidation had been constrained by the need for long residence times, low pulp densities, high power costs for agitation and efficient heat removal from the system (Prassad et al., 1991:1262).

The basic requirements for bacteria/organisms include enough substrate, air/oxygen, water and acidity. The key factors controlling the biological leaching process include:

37

Substrate - if material is not acid-soluble the disseminated bacterial action is impossible.

Oxygen

- oxygen is only available to bacteria if it is dissolved in water.

The amount of dissolved oxygen can be improved by as much as 40% by increasing the pressure in the vessel.

Temperature

- bacteria can withstand temperatures up to 35 ºC. Low

temperatures will slow down bacterial activity.

Water - water is the transport medium for nutrients and transfer of metal values from the solid sites. The bacteria cannot live without water. It is estimated that bio-oxidation followed by cyanidation can improve gold recovery from 68% to 98% (Prassad et al., 1991:1262).

Canales, Acevedo and Gentina (2002:1051-1055) used bio-oxidation of gold concentrates as a pretreatment stage prior to cyanidation in gold mining. Bacterial attack removes part or the entire sulfide layer that covers the gold microparticles, thus facilitating the gold extraction. Results show that the bacterial attack (thiobacillus ferrooxidans) was directed mainly to pyrite as suggested by extractions of 32% for iron (pyrite) and only 2.4%

38

for arsenic (enargite). Increasing the residence times had a positive effect on bio-oxidation, while high pulp densities had an adverse effect.

2.5.2.2 Direct leaching methods

Heap Leaching

The primary reason for utilizing this technique is that the gold/silver values are located in the fracture filling and that the cyanide solution is able to contact them at coarser sizes ranging from run of mine ore down to three stage crushed product (Prassad et al., 1991:1263). In practice, the crushed ore is placed on an impervious surface and leached by percolation over a period ranging from 30 to 150 days or more, depending on the ore size, height, and mineralogy of the valuable minerals. In general, the recovery of gold by the heap leaching technique range from 60 to 80%. In spite of its applicability to low-grade ores, heap leaching suffers from drawbacks such as long leaching time, low permeability due to the presence of clayey material or a mixture of coarse and fine particles in the ore and in some cases seasonal operational efficiency due to cold climates. The heap leaching process with its modifications such as agglomeration, solution heating, air injection and staged heap leaching (Jha, 1987:331-352) feature very low capital and operation costs, although heap recoveries are somewhat lower than by agitation leach methods. Moreover, the heap

39

processes are very flexible, can be operated under snowy conditions and are environmentally very attractive (Lewis, 1983:48-56).

Activated carbon-based technology

Adsorption of gold and silver onto activated charcoal from pregnant solutions has found worldwide acceptance in the last decade. The preference of the carbon adsorption over the conventional Merrill – Crowe zinc dust precipitation system is mainly due to more effective recovery of precious metals from lower grade leach solution along with lower capital and operating costs. A majority of new plants designed in the last decade in the USA, Australia and South Africa adopt carbon adsorption systems (Prassad et al., 1991:1263). The various forms of carbon adsorption techniques that are now in use in the precious metals industry are discussed below.

Carbon-in-pulp (CIP) technology is generally used to treat low grade ore feed where a granular carbon as coarse as 6 mesh is moved countercurrent to the cyanided pulp in the adsorption contactors (Davidson, 1974:67-76). The loaded carbon is then eluted and the gold and silver values in the eluate are recovered by the conventional Merrill – Crowe zinc dust precipitation or by electrowinning on steel cathodes using 2.5 to 3.5 V per cell and amperage consistent with 30 to 40 current efficiency. The

40

process includes cyanide agitation leaching, counter-current carbon-pulp contact and separation of carbon from the pulp for desorption of gold and silver values (Prassad et al., 1991:1263). The ore is normally ground to 100 meshes or finer and leached in a thick slurry of about 45 – 50% solids to which lime is added to ensure the alkalinity of sodium cyanide. The pulp is vigorously agitated and aerated for 2 – 24 hours for complete dissolution of the precious metal values. Carbon in columns (CIC) are predominantly used to recover gold from heap leaching pregnant solutions. Activated carbon is made from wood, nut shells, coal, petroleum coke. Coconut shell is

preferred due to its high adsorption capability and commendable

durability for gold and silver cyanides. An improved technology for the carbon-in-pulp (CIP) process has been developed by Davy McKee Corporation, based on a new contactor design (Prassad et al., 1991:1264). This new contactor uses a modification of the air swept interchange screen design which is able to handle high carbon concentrations and resulting in reduced contactor size. An increase in the carbon concentration from 5% for conventional CIP to 25% for the Davy McKee contactor has reduced the stage volume by 80%. This new design is claimed to have resulted in the reduction in the adsorption plant capital cost of up to 45% or 20% of the total carbon plant from adsorption to electrowinning.

Carbon in leach with oxygen (CILO) is another new development which was developed by Hazen Research Inc. (Prassad et al., 1991:1265). The

41

process consists of utilizing oxygen rather than air in the carbon leach process. The test results on different ores indicate that the leaching rate can be significantly increased by a factor of 4.8 when using oxygen in the process. Contrary to expectation, the cyanide consumption for the CILO process was less than that in a comparative CIL test. The major advantage of the CILO process is the reduction in the retention time. Thus both leaching and gold adsorption can be carried out in about the same number and size of tanks as conventionally used for CIP alone. The process can be utilized to reduce capital and operating cost or to increase production or improve gold recovery at existing plants (Prassad et al., 1991:1265).

Ion-exchange/solvent extraction technique

One of the technologies deserving attention involves the recovery of precious metals from pregnant solutions by ion-exchange resins using the resin-in-column (RIC) technique (Gilmore, 1967:63-65). The flow sheet for this system is similar to that of carbon-in-column (CIC) plant. One major exception involves the operation of the elution (stripping) stage, which does not require elevated temperatures and pressures for removal of precious metals from loaded resins. Either strong or weak base resins may be used. In general, the stripping of the loaded gold/silver values from the resins has proved to be somewhat difficult. Weak base resins may be eluted by dilute caustic solution at ambient temperature. On the other hand, the strong

42

base resins require concentrated NH4SCN or Zn(CN)4 solutions as reported by researchers from MINTEK (Paul et al., 1983). In this case, the resin needs to be regenerated using acid, which breaks down the zinc cyanide complex producing HCN gas which must then be collected and readsorbed into caustic soda for recycling.

The alternate technique of Solvent-Extraction (SX) of gold/silver from pregnant solutions is also worthy of consideration. Mooiman and Miller (1983) identified a suitable extractant with an organic phase consisting of a weak base amine in which the basicity has been increased by the addition of an organic phosphorus oxide modifier, all in an appropriate carrier. Maximum loading is achieved at about pH 9.5, even for a very weak gold bearing solution. Stripping of the loaded solvent is effected at pH 12 – 13 by deprotonating the amine or with a 0.1 to 0.5% caustic solution. Both the loading and stripping reaction rates are relatively fast, making the technique very attractive from a design and cost consideration point of view. Moreover, the proposed system is very selective for gold/silver complexes over other metal cyano-anions such as ferric and cupric complexes which should result in obtaining a relatively pure gold/ silver product (Prassad et al., 1991:1265).

43

Desorption of gold

The desorption of gold from activated carbon has generally been undertaken using either the Zadra or AARL Elution process (Bailey, 1987). Both techniques utilize a low ionic strength aqueous eluent and produce low-grade eluate requiring electrodeposition techniques for gold recovery.

A new technique has been developed based on alcohol (usually methanol) called the Micron Alcohol Desorption Process. This process operates in a reflux/distillation mode with carbon acting as a fractionating medium. The main advantage of the Micron System is the production of a small concentrated volume of gold eluate, with a high Au concentration of typically 5000 – 15,000 ppm Au. The gold from the commercial Micron eluates are currently recovered by electrodeposition on to aluminum foil, although chemical methods of gold recovery are under development (Prassad et al., 1991:1263).

2.5.2.3 Direct electrowinning

Direct electrowinning of gold from dilute aurocyanide solutions containing 1 to 10 ppm Au has been proposed for treating pregnant solutions generated in heap leaching operations (Paul et al., 1983). In such a system the electrodes must have extended surface area (ESA), which is provided by a

44

flow through a supported fiber electrode or as a particulate carbon (packbed) electrode. In this case, the mass transfer limiting currents and thus the rate of electrodeposition is increased by the porous nature and the large active electrode surface area.

2.5.2.4

Use of unconventional lixiviants

The dominant chemical used in the gold industry for many decades has been cyanide in aqueous solution.

This has been used for the initial leaching process to solubilise the gold and throughout the subsequent processes of purification to the elution of the gold from loaded activated carbon and its subsequent electrowinning. Concerns about cyanide toxicity, its relatively slow kinetics, and its lack of selectivity in some refractory ores have prompted research into alternative lixiviants (Swaminathan, Pyke and Johnston, 1992). The application of lixiviants other than cyanide to the leaching of precious metals is dependent on the apparent ability of these other ligands to form stable complexes with gold and silver. The following discussion will focus on the dissolution of gold and silver with halogens, thiourea, and thiosulfate.

45

Halogens Bromine Although bromine has been recognized for many years as a powerful gold extractant, it is only recently that it has been applied. Sergent (1988:167-169) have proposed the dissolution of gold in bromine as follows:

Br2 (DMH) + 2 H2O = 2 HOBr + H2 (DMH)

(17)

2 Au0 + 3 HOBr + 3 NaBr = 2 Au Br3 + 3 NaOH

(18)

Au Br3 + NaBr = Na+ ( Au Br4 )-

(19)

where DMH indicates dibromodimethyl hydation. The overall reaction can thus be represented as follows:

4 Au0 + 3 Br2 (DMH) + 10 NaBr + 6 H2O

+ 6 NaOH + 3 H2(DMH)

=

4 Na (Au Br4 )

(20)

The natural advantages claimed for bromine are rapid extraction, and adaptability to a wide range of pH values (Dadgar et al., 1989). Dadgar et al. (1989) have studied the leaching characteristics of bromine on a refractory gold ore and compared the results obtained with conventional cyanidation. The high dissolution rate and low recovery costs of the bromine process are expected to give a definite economic advantage over

46

cyanidation even though preliminary laboratory data indicated that the cost of reagents for both the processes is about the same. Disadvantages on the use of bromine are likely to be high consumption and interference in Atomic Absorption (AA) and Inductively Coupled Plasma (ICP) techniques during gold analysis.

Iodine Iodine forms the most stable gold complexes of all the halogens. It leaches gold from its ore at low concentrations and can penetrate rocks particularly well. Jacobson and Murphy (1987) reported that iodine does not absorb to any great extent on gangue minerals, which results in excellent recoveries of the reagent that will ultimately reduce the cost of the process. The application of iodine for leaching gold was patented by McGraw and Murphy (1979). The excessive consumption and the high cost aspect of iodine are the main reasons that it could not compete with cyanide.

Thiourea

Thiourea, CS(NH2)2, is an organic compound, the crystals of which dissolve in water to yield an aqueous solution stable under acidic conditions. The complexation reaction for gold and silver are shown below with respective formation constants (Hiskey and Alturi, 1988: 114) Au+ + 2 CS(NH2)2

→

Au [CS(NH2)2]+2

47

β

= 9.1x1021 (21)

Ag+ + 3 CS(NH2)2

→

Ag [CS(NH2)2]+3

β

= 1.3x 1013 (22)

Kazakov et al. (1964) described the gold dissolution reaction as follows:

Au(s) + 2 CS(NH2)2 (aq)

→

Au [CS (NH2)2]+2 (aq) + e-

(23)

The high consumption and low rate of dissolution are attributed to an inhibiting coating of sulfur on the surface of gold particles formed due to the degradation of thiourea. Schulze (1985) discussed how thiourea consumption could be reduced by stabilizing the reagent. The procedure involves adding SO2 to the leaching solution, thereby controlling the redox couple between thiourea and its first decomposition product formamidine sulfide, and in doing so keep the formation of sulfur to an absolute minimum.

Leaching of gold from a refractory aurostibnite flotation concentrate using acidified thiourea was described by Hisshion and Walter (1984:243-273). The initial extraction of gold is around 50–60% in 10 to 15 minutes from the concentrate and, after that, as much free gold is recovered by gravity separation. The critical parameters for the thiourea leach are as follows:

pH

: 1.4 adjusted by H2SO4

Redox potential

: Max 250 V : Min 150 V

Thiourea Conc.

: 1%

48

Thiourea Consump.

: 2 kg mt-1

Leach Time

: 10 – 15 min

The thiourea process offers an attractive potential for treating refractory ores and flotation concentrates.

Gaspar et al.(1994: 369-381) studied the possibility of gold and silver extraction from products obtained during treatment of copper anodic slimes through the construction of potential-pH diagrams for Au-CS(NH2)2-H2O and Ag-CS(NH2)2-H2O systems. The effect of leach parameters such as thiourea and ferric ion concentrations in the solution and level of the redox potential were considered. Optimum conditions for gold extraction of 99.8% are a leach time of 1 h at the concentration of 10 g L-1 thiourea and a concentration of 5.0 g L-1 ferric ion as oxidant .

Thiosulfate Thiosulfates are compounds containing the group S2O32- which is a structural analog of sulfate with one oxygen atom replaced by a sulfur atom. Gold and silver dissolve in thiosulfate solutions forming stable complexes according to the following reactions:

Au+ + 2 S2O32-

→

Au (S2O3)23-

Kc

= 1.0 x 1026

(24)

Ag+ + 2S2O32-

→

Ag (S2O3)23-

Kc

= 2.2x 1013

(25)

49

Kakovskii and Tyvrin (1957) have studied the fundamentals of gold and silver dissolution in thiosulfate solutions (Prassad et al., 1991:1268). In the presence of ammonium thiosulfate, gold is solubilized solely as the thiosulfate complex:

4 Au (s) + 8S202-3 + 2 H2O + O2 (g) → 4 Au (S2O3)23- (aq) + 4OH- (26) Silver reacts in the same way : 4 Ag (s) + 8S2O32- + 2 H2O + O2 (g) →

4 Ag (S2O3)23- (aq)+ 4OH-

(27) Tozawa et al. (1981:25) have investigated the effect of copper concentration on the dissolution of pure gold plates in ammonical thiosulfate solutions. Gold dissolution was found to be very sensitive to both thiosulfate and copper concentration, and exhibit a complex dependency on temperature. Berezowsky and Sefton (1979:17) reported that oxidative degradation of thiosulfate to tetrathionate occurred in the presence of air, particularly when both soluble copper and sulfides were present and this reduced the efficiency of the lixiviant.

The prelimary evaluation of the feasibility of thiosulphate leaching for the extraction of gold from precious metals ores was done on a laboratory scale by Abbruzzese et al.(1995:265-276). The experimental work has allowed the authors to point out the influence of temperature, thiosulphate concentration, ammonia concentration and copper sulphate concentration

50

on the gold dissolution from an ore (51.6 g/t Au) that originated from the Dominican Republic. Gold was recovered from the leach liquors by adsorption onto activated carbon or by electrowinning. With the optimization of the process parameters about 80% gold recovery was achieved, similar to the case with conventional cyanidation. This process has advantages over cyanide in that a decreased interference from foreign cations and a lower environmental impact is achieved.

An alternative non-cyanide lixiviant, namely sodium hypochlorite, has been used for the selective extraction of gold and silver from a copper concentrate (Puvvada and Murthy, 2000:185-191). Hypochlorite leaching of the as-received, as well as aqueous pressure-oxidized copper concentrate, was carried out. Direct hypochlorite leaching yielded gold and silver recoveries of 42.7% and 45%, respectively. With aqueous pressure oxidation followed by hypochlorite leaching, it was possible to selectively recover 90.0% of gold and 92.5% of silver from the copper concentrate.

Fourie (1986) investigated the following ligands for the leaching of gold and silver: cyanide, thiourea, thiocyanate, chloride, polysulfide and thiosulfate. Very good leaching (>90%) was obtained with nearly all the ligands, except chloride which gave a maximum silver leaching of only 30% and polysulfide which gave a maximum gold leaching of 35%. When chloride was used as a ligand, the maximum leaching was obtained within 15 min for both gold

51

and silver. With thiocyanate, polysulfide and cyanide, maximum leaching was achieved within 15 min for silver and 1 h for gold. With thiourea and thiosulfate, maximum leaching was obtained within 1 h for both gold and silver.

Gold recovery from arsenopyrite concentrates obtained by ore flotation concentration has been studied by Kononova et al.(2001:115-123). The thiosulphate leaching of the products obtained after the chemical preparation of the concentrates was carried out in the system Na2S2O3 NH3 – H2O at 20 – 22 ºC. The gold recovery on ion-exchangers and carbon adsorbents from the thiosulphate solutions was investigated. It was found that the use of strong basic or polyfunctional anion-exchangers with a macroporous or porous structure worked very well.

The study done by Ubaldini et al.(1998: 113-124) at laboratory scale on a gold-bearing ore with 4 g.ton

-1

Au, has permitted identification of the

influence of temperature, leaching agent concentration, pulp density and leaching time on the gold dissolution. Gold was recovered from the leach liquors

by

adsorption

onto

activated

carbon

and

subsequent

electrodeposition. After the optimization of the experimental conditions of the process, about 80% gold recovery was obtained. For the complete cycle of treatment the consumption of reagents has been low (5 g.kg-1of sulphuric acid and 0.5 g.kg-1 of ferric sulphate).

52

Kudryk et al. (1984:145) studied the kinetics of leaching of gold from pyrite concentrate using acidified thiourea. The results show that a very short time is needed to leach 95% of gold as compared with 24 h

in the

cyanidation process. The consumption of thiourea was comparable to that of sodium cyanide.

Gold can be effectively extracted from various sources such as refractory ores and jewellery scrap by dissolution in ammoniacal solutions with the aid of appropriate oxidants (Prassad et al., 1991:1266). Electrochemical and kinetic aspects of the dissolution of gold from the elemental state and various ores in ammoniacal solutions were studied. The role of a number of oxidants including oxygen, cupric amine, cobaltic amine, and iodine has been examined. The effects of concentration of free ammonia, ammonium ions and other salts on the overall rate of dissolution of gold were investigated. Excellent dissolution yielding better than 95% gold and silver recovery from sulfidic and carbonaceous refractory ores after 1 to 2 h of leaching at about 160 to 190 ºC was possible in a single stage operation. Gold dissolution was affected greatly by the reaction temperature and concentration of ammonia and oxidants.

A study was conducted to determine optimum conditions for the chlorination of gold in decopperized anode slime with chlorine gas in

53

aqueous medium by the Taguchi experimental design method (Bunyamin, 1999:81-90). It showed that:

•

The important parameters are reaction temperature, reaction time, stirring speed and solid:liquid ratio, respectively. The solubility of gold increases with increasing reaction temperature, reaction time and stirring speed, and with decreasing solid-liquid ratio.

•

Fast dissolution and high-grade gold extraction permit the use of continuous processes in hydrometallurgical practice.

•

It is a non-toxic technique compared to the conventional cyanidation with low capital investment.

•

Under optimal leaching conditions 99.78% of the gold is solubilized.

Chmielewsky et al. (1997:334-344) developed a process of gold reclamation from jewellery production wastes. The problem was important because of significant losses of valuable gold in the production wastes. It is, however, difficult to solve because the physical form and chemical composition of the wastes vary considerably. On the basis of extensive laboratory studies they developed a process that suited the technical conditions of jewellery manufacturing and a production line for processing of gold-containing wastes. The process consists of the following steps, as shown in the figure 2.4:

54

1.

low-temperature carbonization and roasting of the wastes.

2.

a first step of leaching with nitric acid solution in order to remove silver and other metals.

3.

a second step of leaching with nitro-hydrochloric acid (aqua regia).

4.

selective solvent extraction of gold with diethyl malonate.

5.

separation of metallic gold from the organic phase by reduction.

In the first leaching step the impact of nitric acid concentration, mixing time, temperature and liquid/solid ratio was investigated. As a result, the following parameters of the process were chosen: a HNO3 concentration of about 8 M (1:1 dilution of concentrated HNO3), temperature 40 – 50 ºC, time of mixing about 7 h, using a liquid/solid ratio of 5:1. In the second step of leaching with aqua regia the influence of temperature and mixing time was studied. Optimum results were obtained at 40 – 60 ºC for 7 h mixing. In the solid residue the concentration of gold did not exceed 0.03 – 0.08 wt%. Besides gold, the solution contained other elements: B, Si, P, Fe, Mn, Pb, Bi, Sn, Ti, Ag, Al, Ca, Cu, Na and Zn. The gold obtained was 99.99% pure. The total efficiency of the gold recovery was about 97%.

55

Jewellery waste

Thermal degradation of waste (KS type furnace) HNO3

Thermally treated waste material Ag leaching at 40 °C

Aqua regia

Desilverized waste

AgNO3 for Ag recovery

Au leaching at 40 °C Diethyl malonate

Au solution

Solid waste disposal

Au extraction Waste-water 0.01 M HNO3

Au extract Extract washing Waste-water

Washed Au extract H2SO4 conc.+ H2O2 + (COOH)2

Au reduction re-extraction

Metallic gold

Waste water

Figure 2.4 Block diagram of hydrometallurgical Au reclamation Process (Chmielewsky et al.(1997:334-344))

56

Schreier and Edtmaier (2002) investigated separation of iridium, rhodium and palladium from secondary platinum scraps by precipitation and calcination. The scraps were dissolved in boiling aqua regia, diluting with water or hydrochloric acid and precipitated with ammonium chloride. The mother liquor gave a precipitate, where the accompanying platinum group metals (PGM) iridium, rhodium and palladium were separated from the platinum matrix. Calcination of the precipitate gave a platinum sponge of high purity (>99.90%). In Australia pilot plant test work on the Calcine-Leach-Metals Recovery Process yielded a recovery of over 90% for platinum + palladium +gold (PLA, 2001). The new PLA process also has the potential to be used on other PGM ores, not only in Australia to reduce costs and increase revenue. It will also allow economical production of smelter concentrates from ores that have previously proved difficult to treat. A patented process [patent (No. ZL9411118736.5), 1996] developed for recovering platinum-group metals from spent catalyst materials has yielded high-purity Pt and Pd (purity > 99.95%). Resin was used for recovering Pt and Pd from spent catalyst materials. The resin has good selectivity and high adsorption capacity for Pt and Pd. Total recovery of Pt and Pd is claimed to be more than 96%. Mahmoud (2003) investigated a leaching process based on the ability of platinum-group metals to form stable chloro-complexes in acidic solutions.

57

Industrial catalyst losses were examined for the recovery of platinum, palladium, and rhodium by leaching with a mixture of sulfuric acid and sodium chloride to avoid using aqua regia or autoclave conditions. Extraction of platinum and rhodium in 60% H2SO4 at 135ºC steadily increased with increasing NaCl concentrations reaching 95% and 85%, respectively, at 0.1 M NaCl after 2 h. By comparison, palladium was dissolved more quickly but also reached 85% under the same conditions. More than 99% extraction of each metal was obtained after 10 h using 0.1 M NaCl and 60% H2SO4 at 125ºC. The use of gauzes containing palladium to recover platinum was first developed in 1969, using Pd/Au alloy. This technology recovered around 65% of the platinum . In 1980, Engelhard developed MTL technology (http://www.engelhard-clal.fr/dpi/gb/nitroprocess3.htm), which stands for Mass Transfer Limited, a new type of gold-free palladium gauze which raised platinum recovery efficiency to 85%. Catalytic converters as investigated by Duane (1994) are used to control hydrocarbon combustion emissions form gasoline powered vehicles. The converters contain small quantities of platinum group metals (PGM). These act as catalysts in chemical reactions, which convert carbon monoxide to carbon dioxide and water, and nitrous oxides to nitrogen. Although the concentrations are small, the amount of PGM’s in catalytic converters is still two orders of magnitude higher than any of the richest naturally

58

occurring ores. Researchers estimate that if a plant with the capacity to process 3,500 tons of catalyst per year could recover 80,000 oz. of palladium, and 8,000 oz. of rhodium, the market value would be $29.2 billion. The research by Duane (1994) has produced a suitable method for recovery of PGM’s . The method is currently being refined to improve the yield of individual PGM’s. Incineration is the traditional method used to recover precious metals from spent catalysts, but it has a number of drawbacks, particularly from an environmental point of view. A British-Swedish joint venture, involving Chematur Engineering AB and Johnson Matthey, has developed a process that uses supercritical water oxidation to recover the precious metals. The technology, called AqauCat, offers efficiencies of close to 100%, but without any of the problems associated with incineration (Grumett, 2003:163-166). The work presented in the rest of this thesis will utilize some of the approaches described in this literature review, or at least employ the general principles outlined by previous research workers. The next chapter will describe the experimental procedures applied to accomplish the recovery of the various noble metals from the jewellery waste used.

59

Literature on precious metals leaching discussed in this chapter, indicate that noble metals can be recovered from jewellery waste. The process used in this study followed a two stage procedure: first nitric acid was used to remove silver and other metals and then aqua regia was used to dissolve gold and platinum.

This process was chosen because

•

It is a non-toxic technique in contrast to the conventional cyanidation method.

•

All reagents used in this study are easily obtainable.

•

The process is easy to use and was found to be financially viable.

•

The process corroborates Pourbaix diagrams regarding dissolution of gold, silver and platinum in water.

•

Ammen (1997: 285-294) described different processes depending on the ratio of gold-silver in the sample. In this case, the chemical analysis of the sample showed that the gold-silver ratio is 1:4 and the process mentioned above was the only one required.

•

The process corroborates Scheieder and Edtmaier investigations regarding platinum separation

60

CHAPTER 3 EXPERIMENTAL APPROACHES 3.1

INTRODUCTION

The growing demand for highly pure precious metals to be used in hightech applications, the need for modern and clean processes and the increasing volume of low grade noble metals from secondary sources available for recycling today is having a direct impact on the industrial practice of recovering and refining precious metals (Ritchie, 1998:1). However, many issues remain to be resolved in connection with the final stage of precious metal refining to produce the individual metals from the overall process. According to Chimielewski et al. (1997) the problem is important because of significant losses of valuable gold in the recovery from wastes. The chemical composition of the waste normally varies widely and dictates the process to be used. In this study the focus will be on the development of a process which would allow jewellery waste to be safely recovered for reliable re-use in jewellery manufacturing.

61

3.2

MATERIALS

3.2.1 Chemicals and reagents

Wastes used in this study were jewellery wastes from the department of Fine Arts of Tshwane University of Technology, as well as an aqueous waste solution containing silver chloride residues (obtained from the department

of

Chemistry

and

Physics,

Tshwane

University

of

Technology). All the reagents used for the investigation were for analytical/laboratory grade, procured either from Merck, Unilab, Aldrich or Rochelle Chemicals.

3.2.2 Instrumentation

In this study, the following were used: •

Atomic Absorption Spectrometer (AAS) Perkin Elmer model 3110 with air-acetylene flame

•

Fume hood

•

Inductively coupled plasma spectrometer (ICP-OES SPECTRO CITROS)

•

Analytical balance ( Mono Bloc AB 104 –S )

•

Magnetic stirrer with hotplate

•

Muffle furnace

•

ARL9400XP + Spectrometer

•

Electronic Sieve Shaker, Model ES 200, Mark IV.

62

The chemical composition of the jewellery wastes used is presented in table 3.1. The sample was mounted on a PVA base and analysed without any sample preparation. Analysis were executed using the ARL9400XP + Spectrometer using the wide confidence limit program.

63

Table 3.1: Chemical analysis of the jewellery waste

Elements

Concentration (%)

Si

1.43

Ti

0.63

Al

0.50

Fe

0.35

Mn

0.16

Mg

0.17

Ca

1.20

Na

1.14

K

0.50

P

0.34

S

2.13

Cl

1.20

Cr

590 ppm

Ni

590 ppm

Cu

13.9

Zn

1.09

Ag

63.5

Pt

2.46

Au

9.11

As can be seen from table 3.1,quantitative chemical analysis by XRF indicates a gold content of 9.11%, silver content of 63.5% and platinum content of 2.46%.

3.2.3 Particle size analysis Screen analysis was performed by an electronic Sieve Shaker, Model ES 200, Mark IV. The data in table 3.2 showed that 85.41% of the material

64

used was below 212 µ m, which is favourable for precious metals leaching (Ritchie, 1999; Ubaldini et al., 1998).

Table 3.2 : Screen size analysis of jewellery waste Sieve size range

Weight % retained

Size distribution (%)

14.59

85.41

(µm) +212

3.3

-212

+150

18.23

67.18

-150

+106

18.35

48.83

-106 +75

26.52

22.31

-75

22.31

0

Total

100

EXPERIMENTAL METHODS

The approach used in this study was leaching-precipitation and melting. A flow sheet of the process is given in Figure 3.1

65

Sample (Au , Ag , Pt and impurities)

HNO3 + H2O [1:1] Soluble (Ag+, Cu2+)

Insoluble [Au(s), Pt (s)]

HNO3 + HCl [1:3]

HCl concentrated

Insoluble AgCl(s)

Soluble [AuCl4]-

Filtration Soluble CuCl2 (aq)

Filtration

Filtration Residue

and [PtCl6]2-

FeSO4.7H2O H SO (conc.)+ Zn (s) Insoluble 2 4 Ag (s)

Filtration

N2H4(hydrated conc.)

Heat at 1000 ºC Ag (s)

Insoluble Au (s)

Soluble [PtCl6]2-

Insoluble Pt Powder

Au (s) Heat at 1100 ºC

Figure 3.1 Flow Sheet of the process

66

3.3.1 Silver

3.3.1.1 Dissolution of silver and other metals in nitric acid.

After preliminary separation to remove unwanted pieces of wood, plastic and paper, 50 g of waste was weighed off. A magnet sandwiched inside a plastic sheet was used before weighing to remove magnetite and hematite. Experiments were performed in a 250 ml Pyrex beaker acting as a small reactor, using a stirroheater. A leaching solution was made up according to the literature (Ammen, 1997) from a mix of 52.8 ml HNO3 conc. and distilled water in the ratio [1:1]. The reaction between the leachant and jewellery wastes gave off reddish brown fumes, mostly NO2, which is the product of the interaction of HNO3 with base metals, in this case copper. The reaction was exothermic. The solution changed to a dark blue-green colour in the process. Due to copper dissolution when addition of fresh acid did not result in the evolution of any more visible brown fumes (NO2), the solution was cooled and diluted with distilled water in the proportion [1:3]. Separation was accomplished with a filter. Dissolved silver and other base metals remained in solution, while the insoluble materials like gold, platinum and impurities accumulated on the filter. The residue was dried and put aside for further processing. During the leaching stage, samples were taken every 30 min and analyzed by ICP for silver concentration. The dissolution can be described by the following chemical reactions:

67

4 Ag (s) + 6 HNO3 (aq) → 4 AgNO3 (aq) + NO (g)+ NO2 (g) + 3 H2O (aq) (1) 3 Ag (s)+ 4 HNO3 (l)→ 3 AgNO3 (aq) + NO (g) + 2 H2O

3.3.1.2

(2)

Precipitation of silver chloride from the filtrate containing

AgNO3, Cu(NO3)2 and other base metal nitrates.

The filtrate used contained AgNO3 , Cu(NO3)2 and other trace amounts of base

metals soluble in nitric acid from the previous step. Selective

precipitation was performed by forming a chemical compound with silver ions, while leaving the copper ions dissolved in the solution (Ammen, 1997). The blue-green solution was diluted in the proportion 1:3 with HNO3/H2O according to the procedure described in the literature (Ammen, 1997). HCl was then added, which immediately precipitated silver chloride as a curdy white precipitate. The solution was allowed to settle for 24 hours. The process can be described by the following chemical reactions:

Ag NO3 (aq) + HCl (aq) → AgCl (s) + H+(aq)+ NO3- (aq)

(3)

Cu (NO3) 2 (aq) +2 HCl (aq) →CuCl2 (aq) + 2 H+(aq)+ 2 NO3- (aq) (4)

68

3.3.1.3 Cementation process

The AgCl precipitate from waste1 was washed free from any blue-green copper stain with hot water. It was placed in a shallow glass-evaporating dish and then put onto a hotplate. 1 g of AgCl was mixed with 60 ml sulfuric acid diluted in the proportion of [1:15] with water (Ammen, 1997) while stirring and heating it slightly. 0.52 g zinc powder was then added. It dissolved in the acid solution with evolution off H2 bubbles. As the zinc dissolved, the AgCl in contact with it was reduced to metallic silver. It is very important to keep stirring so that every particle of the silver chloride makes contact with the dissolving zinc and is reduced to metallic silver. When reduced, the silver chloride went from white to grey granular silver (called cemented silver). To free the cemented silver from any undissolved zinc, it was mixed again with a weak sulfuric acid solution prepared by diluting it with distilled water in the proportion [1:20] and stirred well (Ammen, 1997). When everything settled down and no bubbles evolved anymore, all the zinc was in solution. The silver cement was washed onto a filter and rinsed with plenty of hot water. The process can be described by the following chemical reactions:

2 AgCl (s)

+

Zn (s)

+ 2H+ (aq)

+ SO42- (aq)

=

2 Cl- (aq) + SO2-4 (aq) + Ag (s) + H2 (g)

2 AgCl (s) + Zn (aq) + 2H+ →

2 Ag (s) + ZnCl2 + H2 (aq)

69

Zn2+ (aq) + (5)

(6)

3.3.1.4

Refining process of Silver

Silver chloride can be converted to the metallic form by heat treatment of a mixture of silver chloride and sodium carbonate at 600 °C (Schneller, 1981: 22). The process works efficiently, except that the conversion is not entirely complete and the silver product has to be reprocessed.

The process used in this study for recovering elemental silver from silver residues was performed according to the procedure described by Paul Steed (1974). Three samples were selected: • Ag1 sample of silver chloride from jewellery waste of the Fine Arts Department • Ag2 sample of silver chloride from chemical waste of the Chemistry Department • Ag3 sample of silver cement from waste of Fine Arts Department Samples were dried in a furnace at 110 ºC and then mixed with an equivalent amount of potassium carbonate (K2CO3) in the ratio of [4:3] in a clay crucible and roasted in a furnace at 1000 ºC. The furnace was hooded because COCl2 could potentially form, which is poisonous. The crucible was removed, the slag discarded, and the molten silver was poured into another clay crucible. K2CO3 was used as a flux reducing agent for both waste Ag1 and Ag2 and its performance was compared to Ag3.

70

3.3.2 Gold

3.3.2.1 Gold dissolution in aqua regia

17 g of dried material originating from the dissolution of jewellery waste by nitric acid was dissolved using 102 ml of aqua regia solution (http://www.ishor.com/RefinAcidinstr.htm). The solution containing residue was washed off with hot water, filtered with paper filter and set aside. The residue was dried and sent for analysis. The filtrate was boiled to a syrupy constituency. Some drops of HCl were added. The operation was repeated three times. Although it was time consuming this operation of evaporation and re-evaporation with addition of HCl was the most important part of the entire refining process because of purity. When all excess nitric acid was driven off, the beaker containing the syrup-like solution was cooled down and diluted with hot water to four times its original bulk volume and left to stand for several days. Its total volume was approximately 1000 ml. The process can be described by the following chemical reactions:

3 HCl (aq) + HNO3 (aq) → Cl2 (g)+ 2 H2O (aq)+ NOCl (g)

(7)

Cl2 (g) + H2O (l) → HClO (aq) + H+(aq) + Cl- (aq)

(8)

HOCl(aq) → H+ (aq) + ClO- (aq)

(9)

71

Au (s) + ½ Cl2 (g) → AuCl2- (aq)

(10)

AuCl2- (aq) + Cl2 (g) → AuCl4- (aq)

(11)

The overall reaction can be written as Au (s) + 3/2 Cl2 (g) + Cl- (aq) → AuCl4- (aq)

3.3.2.2

(12)

Precipitation of gold from aqua regia solution using iron

sulphate solution.

A clear apple green solution obtained by mixing 120 g of FeSO4 7 H2O, 300 ml of distilled water and 40 ml of HCl was poured slowly into 400 ml of gold solution while stirring. The gold precipitated and was washed with hot water and dried. It was a brown powder. It was then mixed with borax fluxes and melted at 1100 °C in a muffle furnace to get metallic gold. The filtrate was sent for analysis to calculate gold recovery and assist with further platinum recovery. The process can be described by this chemical reaction:

3 FeSO4 (aq) + AuCl-4 (aq) → Au (s) + FeCl3 (aq) + Fe2(SO4)3 (aq) (13)

72

3.3.3 Platinum

The precipitation process was carried out by slowly pouring a solution of 214 mg.L-1 ammonium chloride (NH4Cl) to the filtrate remaining from the gold precipitation step, while stirring it at 80 to 90 ºC (Habashi, 1997; Narita, 1998; Schreier and Edtmaier, 2003). The reaction gave a complex orange precipitate called ammonium hexachloroplatinate (NH4)2PtCl6

.

The experiment was done in a fume hood because of NOCl production which is corrosive and dangerous for the lungs. The complex was reduced by hydrazine in aqueous solution at 80 ºC and yielded a black platinum powder. The process can be described by the following chemical reactions:

• 8 HCl (aq) + 2 HNO3 (aq) + Pt (s) → H2PtCl6 (aq) + 4 H2O (aq) +

2NOCl(g)

(14)

• H2PtCl6 (aq) + 2 NH4Cl (aq) → (NH4)2[PtCl6] (aq) + 2HCl (aq) (15)

• (NH4)2PtCl6 (aq) + N2H4 (l) + 6 NaOH (s) → Pt (s) + 6 NaCl (aq) + 2 NH3(g) + N2 (g) + 6H2O(aq)

73

(16)

CHAPTER 4 RESULTS AND DISCUSSION

4.1 INTRODUCTION

The aim of this experimental work was to assess the feasibility of leaching or extracting noble metals from jewellery wastes on a laboratory scale. The investigation evaluated the influence of HNO3 leaching agent concentration, temperature, solid/liquid ratio and stirring speed on the noble metal dissolution.

4.2 SILVER

Table 4.1 Dissolution of jewellery waste sample by nitric acid

Components

Mass (g)

Distribution (%)

Sample

50

100

Soluble noble metal

33

66

17

34

(Ag) Insoluble noble metals (Au, Pt)

74

Table 4.1 presents the results of the dissolution of jewellery waste by nitric acid. It is seen that 66% of the metals sample is soluble.

In this phase of the

examination different variables, like effect of nitric acid concentration, stirring speed, temperature and solid-liquid ratio were studied.

4.2.1 Effect of HNO3 Concentration The effect of HNO3 concentration on the leaching of jewellery waste was investigated. These series of experiments were carried out by varying the nitric acid concentration in the range from HNO3/H20 = 1:16 to 1:1 using the following working conditions: - Temperature 40 °C - Solid/liquid ratio 1:5 - Stirring speed 200 rpm - Mass sample 0.5 g

Figure 4.1 represents typical curves of silver recovery as function of time at different HNO3/H20 ratios.

75

Recovery(% )

100 90 80 70 60 50 40 30 20 10 0

Ratio 1:16 Ratio 1:8 Ratio 1:4 Ratio 1:2 Ratio 1:1

0

50

100

150

200

Time(min)

Figure 4.1 Representation of Ag extraction as a function of time for different HNO3 concentrations.

The silver dissolution process can be represented by reactions (1) and (2) while gold and platinum remain undissolved. Nitric acid is known to be effective oxidizing reagent and its function during the leaching was to promote oxidation of silver to water soluble silver nitrate. The by-products of the reaction were various nitrogen oxides, which reacted with water and regenerated the nitric acid (Kudryk et al., 1984). The separation becomes increasingly efficient as the composition of the waste approaches Ag75 Au25 (Habashi, 1997). It can be seen from the curves that the rate of leaching increased with increasing nitric acid concentration, which means that leaching at the highest HNO3 concentration (namely ratio (1:1)), the rate of silver extraction was considered complete. A recovery of about 97,1% was achieved after 1 h. The results in

76

table 4.2 (see Appendix B for details) also show that no big variation in rate of recovery occurs when increasing the time for longer than 1 h. The wastes had a fine grain size. The size distribution of the sample revealed that 85.4% was below 212 µm (see table 3.2). According to the literature, fine grinding not only increases the surface area for acid attack, but also creates a more reactive surface (Ritchie, 1998:4). As is so often the case in hydrometallurgy, an increase in surface area is equivalent to an increase in concentration of the metal to recover (Tshilombo,

2000: 12) and Kudryk et al.

(1984) also

mentioned that fine grain sizes reduced the cost of solvent consumption. The reaction is diffusion controlled and this model describes the dissolution reaction best. According to the literature, an increase in the diffusion coefficient value with temperature is relatively small and results in a low activation energy for dissolution of ~ 20 J.mol-1(~ 5 cal.mol-1) or less. This aids the identification of a diffusion controlled process, since a similar value would be obtained from an Arrhenius plot of the logarithm of the overall reaction rate against the reciprocal of absolute temperature (Dutrizac, 1986: 429).

The calculated magnitude of activation energy using an Arrhenius plot is 0.243 kcal.mol-1 less than 5 kcal.mol-1 and strongly supports a diffusion controlling mechanism (Dutrizac, 1986: 429). Its low value suggested that the extraction is controlled by a physical phenomenon (Theeraporn and Penpun, 2000). Therefore physical parameters like temperature and stirring speed/agitation variations can be expected to influence the leaching rate of the silver significantly.

77

4.2.2 Effect of stirring speed

The rate of silver leaching was measured as a function of stirring speed. The silver extracted from jewellery waste was monitored as a function of time for five different stirring speeds between 0 and 800 rpm. It can be seen from the data presented in figure 4.2 that there is a big increase in the dissolution of silver

Recovery(% )

from 0 to 200 rpm.

100 90 80 70 60 50 40 30 20 10 0

0 rpm 200 rpm 400 rpm 600 rpm 800 rpm

0

50

100

150

200

Time(min)

Figure 4.2 Representation of Ag extraction as a function of time for various stirring speeds.

A recovery of 97.1% in 1 h was reached with 200 rpm and for the increased stirring speeds of 400, 600 and 800 rpm no further change was observed. According to Kudryk et al. (1987) agitation leaching is usually the most economic choice for higher-grade ore or for those which need fine grinding for liberation of the desired metal. Subsequent leaching work was then carried out at a fixed stirring speed of about 200 rpm. The fact that an increased stirring 78

speed above 200 rpm did not increase the recovery of the silver, justifies the conclusion that sufficient leachant is delivered to the surface of the waste particles and that the rest of the process is diffusion controlled.

4.2.3 Effect of temperature

Temperature changes can effect the leaching reactions in two ways: it can influence the rate of the reaction, with the rate normally increasing with temperature, and it can also influence the overall efficiency of leaching by increasing or decreasing the solubility of sparingly soluble compounds (Kudryk et al., 1984). The influence of temperature was studied by carrying out leaching at 25 °C, 40 °C and 60 °C under the following conditions: -

Mass sample 0.5 g

-

Solid/liquid ratio 1:5

-

HNO3/H2O 1:1

-

Stirring speed 200 rpm

The data is presented in figure 4.3 and table 4.4(see appendix).

79

100

Recovery (% )

90 80 70 60

25 C

50 40

40 C 60 C

30 20 10 0 0

50

100

150

200

Time (min)

Figure 4.3 Representation of Ag extraction Curves as a function of time for different temperatures.

It can be seen that the recovery of silver significantly improves by increasing the temperature from 25 °C to 40 °C. For example only 61.7% of the silver is extracted after 60 min of leaching at 25 °C, while 97.1% is leached at 40 °C and 97,0% at 60 °C. It is observed that by increasing the contact time beyond 60 min the recovery rate is not affected to any greater extend. Increasing the temperature from 25 °C to 40 °C helped to start the reaction as it was exothermic (Ammen, 1997:118). The increase in temperature above 40 °C did not have any significant impact on the dissolution rate, confirming again the assumption that the reaction is at least to a certain degree diffusion controlled.

4.2.4 Effect of solid/liquid ratio

80

The effect of solid/liquid ratio has been evaluated by performing different runs at: - Mass sample 0.5 g - HNO3/H2O ratio 1:1 - Stirring speed 200 rpm - Temperature 40 °C

Recovery (% )

100 Ratio1:10

80 60

Ratio1:5

40

Ratio1:4

20

Ratio1:3

0 0

50

100

150

200

Time (min)

Figure 4.4 Representation of Ag extraction as a function of time for various solid/liquid ratio.

The results of these tests are represented by the plots in figure 4.4. It can be seen that a recovery of 97.1% was obtained after 1 h with solid/liquid ratio of 1:5. Increasing the solid/liquid ratio results in a decrease in the percentage recovery of silver.

4.2.5 Cementation results 81

The following conditions were used to precipitate metallic silver from the leach solution : 1 g AgCl, 0.52 g Zn powder, 60 ml of H2SO4/H2O (1:15), time 0-40 min. It can be seen from the figure 4.5 that a recovery of 80% was achieved within 5 min and then reached 99.7% after 30 min. After 30 min no change is observed in the recovery rate of the silver from the leach solution.

100

R ecovery (% )

80 60

Ag

40 20 0 0

10

20

30

40

Tim e (m in)

Figure 4.5 Recovery of silver cemented from silver chloride as a function of time.

4.3 GOLD

The solution of aqua regia was sent for analysis to ICP to evaluate the concentration of gold. It can be seen from the table 4.7 that the average concentration of the gold solution was 700 mg.L-1.

82

Table 4.7 Analytical lines used by industry for determination of gold by ICP (Mokgalaka, 2001:33)

Element 1 2 3

Wave length (nm) 208.209 242.795 267.595

Concentration (mg L-1) 719.8 698.6 681.4 Average: 700.0 mg L-1

After removing the gold from the solution, it was sent to ICP for analysis to determine the remaining amount of gold. It was found that only 0.7mg.L-1 remained after precipitation, which indicates a very efficient recovery of the gold.

The recovery was calculated using this formula

Ci - Cf R (%) =

------------------- x 100

(1)

Ci Where Ci is the concentration before precipitation in mg.L-1 Cf is the concentration after precipitation in mg.L-1 R is the recovery in % The recovery of gold from an aqua regia solution of 400 ml, treated with 300 ml copperas solution (FeSO4.7H2O) at a temperature 40 °C, and a stirring speed of 200 rpm, lead to a recovery of 99.9% within an 1 h and no increase was

83

observed with an increase in the treatment time. The results are graphically represented in figure 4.6.

100

Recovery (% )

90 80 70 60 50 40

Series1

30 20 10 0 0

50

100

150

200

Time (min)

Figure 4.6 Recovery of gold from pregnant aqua regia solution, as a function of Time.

The effect of temperature and the stirring speed on the precipitation of gold were subsequently investigated.

4.3.1 Effect of temperature on precipitation of gold

Tests were conducted using the following working conditions: - Aqua regia solution 400 ml -Copperas solution (FeSO4.7H2O ): 300 ml

84

-Stirring speed: 200 rpm -Temperature in the range 25 ºC, 40 ºC, 60 ºC and 80 ºC.

100 90

Recovery (% )

80 70

25 C

60

40 C

50

60 C

40

80 C

30 20 10 0

0

50

100

150

200

Time (min)

Figure 4.7 Precipitation of gold from pregnant aqua regia solution at different temperatures as a function of time.

It can be seen from table 4.8 ( appendix ) that at 25 °C, the concentration is 700 mg.L-1 at 0 min and decreased to 214 mg.L-1 only after 2 h, while at 40 °C it decreased to 0.7 mg.L-1 during the same time. The increase of the temperature above 40 °C didn’t increase the recovery rate of the gold as shown in figure 4.7. This is in agreement with other results in the literature, which indicate that increasing the temperature slightly, results in better recovery (Bunyamin et al., 1999). In this case it can be claimed that the precipitation was effective,

85

because a recovery of 99,9% was achieved. This also indicates that the reaction went to completion.

4.3.2 Effect of stirring speed on the precipitation of gold

The investigations on precipitation of gold were conducted using: - 400 ml aqua regia solution - 300 ml copperas solution (FeSO4. 7H2O ) - Temperature: 40 ºC -Stirring speed from 0 to 600 rpm

It can be seen from the table 4.9 (in appendix) and the curves of in figure 4.8 that the precipitation of gold from the solution is dependent on the stirring speed used during the recovery treatment.

Recovery (% )

100 80 0rpm

60

200rpm 400rpm

40

600rpm

20 0 0

50

100

150

200

Time (min)

Figure 4.8 Precipitation of gold from pregnant aqua regia solution at different stirring speeds as a function of time.

86

At 0 rpm the concentration decreased from 700 to 283,2 mg.L-1, which is a 59.5% recovery, while the concentration decreased from 700 to 0.7 mg.L-1 for the same time (2 h) when stirring at 200 rpm. This latter condition yields a recovery of 99.9%. The increase in stirring speed above 200 rpm did not have any further beneficial effect on the precipitation process This is important from a consumption of energy point of view when designing a final treatment process.

4.4 PLATINUM

After removing the gold from the aqua regia solution, the filtrate was sent for analysis to determine its platinum concentration. The results indicated that the remaining solution contained 110 mg.L-1 platinum (table 4.10).

Table 4.10 Analytical lines used by industry for determination of platinum by ICP (Mokgalaka, 2001:33).

Element 1 2

Wave length (nm) 177.709 191.170

Concentration (mg L-1) 110.3 110.8 Average: 110.5

After removing platinum from that solution, the filtrate was sent for ICP analysis to determine whether any platinum remained, and only 0.6 mg.L-1 remained in solution after treatment. It therefore seems that the precipitation process was successful and the reagent used was selective.

87

The following working conditions were employed during the investigation to study the recovery of platinum: hexachloroplatinate solution 100 ml, hydrazine solution 100 ml, temperature 80 °C, stirring speed 200 rpm, time ranging from 0-180 min. The curve in figure 4.9 indicates that the maximum recovery of 99.4% was reached in 1 h.

100

Recovery (% )

90 80 70 60 50 40

Pt

30 20 10 0 0

50

100

150

200

Time (min)

Figure 4.9 Recovery of platinum from solution as a function of time.

The effect of two parameters on platinum precipitation was investigated, namely temperature and stirring speed. According to Schreier and Edtmaier (2003), platinum precipitates completely in an optimum process leaving only minor impurities in the solution.

88

4.4.1 Effect of the temperature

The effect of temperature on precipitation of platinum from 100 ml hexachloroplatinate solution with 100 ml hydrazine solution at a stirring speed 200 rpm was studied in the temperature range from 25 °C, 40 °C, 60 °C, 80 °C and 100 °C.

Recovery (% )

100 80 25 C

60

60 C 80 C

40

100 C

20 0 0

50

100

150

200

Time (min)

Figure 4.10 Concentration of platinum at different temperatures as a function of time during recovery with hydrazine.

According to Harris and Stanley (1986) the selective recovery of platinum from chloride liquors originating from the treatment of various primary and secondary materials constitutes a complex and challenging problem in precious metals refining. In our case a suitable reagent was sought that would yield a good recovery within in 1 h. Table 4.11 showed that at 25 °C, the concentration with

89

hydrazine decreased from 110 to 69 mg.L-1 which is equivalent to a 37.9% recovery. At 60 °C, it decreased from 110 to 24 mg.L-1 which is equivalent to a 78.5% recovery. At 80 °C and 100 °C it decreased from 110 to 0.7 mg.L-1 which yields a recovery of 99.4%. The curve of figure 4.10 shows the results graphically. Once again the concentration decreases with increase of temperature.

4.4.2 Effect of stirring speed on precipitation of platinum

The investigation was done using the following working conditions: -Hexachloroplatinate: 100 ml - Hydrazine hydrate: 100 ml -Temperature: 80 °C, varying -Stirring speed of 0 to 600 rpm

It can be seen from the curve of the figure 4.11 that the precipitation of platinum from solution is dependent on the stirring speed used during the recovery process.

90

Recovery (% )

100 80 0rpm

60

200rpm 400rpm

40

600rpm

20 0 0

50

100

150

200

Time (min)

Figure 4.11 Concentration of platinum at different stirring speeds as a function of time.

Table 4.12 (in appendix) shows that at 0 rpm the concentration decreased from 110 to 31 mg.L-1, which is a 71.9% recovery, while for the same time (1 h) when stirring at 200 rpm, the concentration decreased from 110 to 0.7 mg.L-1, which yields a recovery of 99.4%. The increase in stirring speed above 200 rpm did not have any further beneficial effect on the precipitation process. Once again, this is important from a consumption of energy point of view when designing a final treatment process.

91

4.5 FINAL RESULTS

A summary of the optimum conditions for recovery of Ag, Au and Pt is given in the table 4.13.

Table 4.13 Optimum conditions for recovery of noble metals from jewellery wastes

Element

Optimum Temperature

Stirring

Solid/liquid

( °C)

speed

ratio

Recovery

Purity

yield (%)

(%)

(rpm) Ag

40

200

1:5

94.9

98.3

Au

40

200

1:5

99.0

99.1

Pt

80

200

1:5

99.4

99

As can be seen from the data in table 4.13, using the described procedure mentioned in paragraphs 3.3.1.4, 3.3.2.2 and 3.3.3 gold and silver were recovered and refined with a purity of 98.3% for metallic silver and 99% for metallic gold, and platinum powder to a purity of 99%. The final products are shown in plates 4.12, 4.13 and 4.14.

92

4.6 COST ANALYSIS OF THE PROCESS

The overall goal of this investigation was to develop a cost effective process, which would recover Au, Ag and Pt from jewellery wastes on a small scale. The economics of developing and operating a small pilot plant was a further consideration.

During investigations capital and operating costs were

considered. The capital cost did not include facilities provided by Tshwane University of Technology and wastes (solids and liquids) generated in each step of the operation, but only reagents and energy consumption were considered. According to guidelines of permissible utilization and disposal of sewage sludge Ed1 T.T. 85/95 August 1997(Department of Water Affairs and Forestry) the metal and inorganic content are limited to acceptably low levels. The analysis of waste solutions done by ICP after removing noble metals showed that the concentrations of elements present in solutions were negligible and LD50 for the soluble compounds was less than the limit. The wastes solutions generated do not pose a great threat to public health and the environment and therefore do not pose a major disposal problem because they can be evaporated to near dryness and encapsulated in cement. The hydrochloric and nitric acids used during the process could be reused within the process if the operation could be done at pilot scale, but eventually would require oxidation or neutralization with caustic soda or lime.

Calculations were performed assuming a unit mass of 1 kg of waste. Since the facilities of Tshwane University of Technology were used, the study did not take

93

in to consideration costs of apparatus. The energy consumption was minimized because agitation was restricted to 200 rpm for leaching and precipitation operations. The cost of stirring was calculated assuming the use of mechanic stirrer and extrapolated to laboratory scale where a magnetic stirrer was used. The cost of reagents was calculated referring to the Aldrich hand book, 20032004 where prices of all reagents are mentioned. The refining process needed high temperatures. The cost of energy consumption was therefore calculated assuming an electricity cost of R0.85/KWh (Masenya, 2003).

Table 4.14 shows the income calculated using data from the Financial Mail of November 14, 2003 and Aldrich, 2003- 2004. Currencies: Rand/Dollar $

- 6.97

Rand/Pound

- 11.64

Rand/Euro

- 8.01

Commodity prices: Precious Metals ( USS/oz ) Gold

383.1

Platinum

756

Silver

5.05

OZ = Troy ounce = 31.1 grams.

Dillon (1955: 574) as cited by Julsing

(1997:90)

Table 4.14 Income calculated for the global process assuming 1kg of wastes

94

Noble metal

Recovery

Purity (%)

yield (%)

Amount in

Monetary

g/kg of

value

waste

(Rands)

Silver

94.9

98.3

592

670

Gold

99.0

99.1

89

7676

Platinum

99.4

99.0

23.8

4102

Total

12448

Table 4.15 Profit for the global process assuming 1kg of wastes.

12448

Income (Rand) Cost for stirring (Rand)

796

Cost for refining (Rand)

3450

Cost for reagents

4125

Profit

4077

The table 4.15 gives an indication of the economic benefit of the proposed treatment process. The indications are that a reasonable amount of money can be saved by employing the proposed process in-house, because commercial refining companies normally charge a substantial percentage of the refined metals’ value to do this work. Furthermore, it has the advantage of being time

95

efficient too, as it can be accomplished within a week at the most, even if a mass of one kilogram of the waste needs to be processed.

96

CHAPTER 5

CONCLUSIONS AND RECOMMENDATIONS

5.1 Conclusions

The hypothesis was proven and the objectives given achieved. The investigations showed that it is possible to separate, recover and refine silver, gold and platinum from jewellery wastes.

The purpose of the

research was not only economical, but also to produce something that has an acceptable purity that can be re-used for daily practical work. As mentioned at the beginning, the study was devoted to the preliminary evaluation of the extraction of noble metals from jewellery wastes at a laboratory scale. The final refining of these precious metals needs to be investigated further.

The investigations determined of the influence of

temperature, leaching agent concentration and solid /liquid ratio for the dissolution of the noble metals. The choice of technique or process was driven by the materials used, their quantity, the impurities present, and the product made. It was therefore important to know the chemical composition, particle size distribution and reactions taking place.

The process followed a two-stage procedure: Nitric acid was used to separate silver and other base metals from gold and platinum. Then aqua

99

regia was employed to dissolve gold and platinum. In addition, the process has the following advantages: - it is non-toxic in contrast to the conventional cyanidation method - All reagents used in this study are easily obtainable. - The process is easy to use and seems to be financially viable according to the preliminary economic study done. -The process is suitable for the chemical composition and particle size distribution of the particular wastes investigated.

5.2 Recommendations

•

Design the correct way of sampling in order to obtain a product truly representative of the waste.

•

More studies are needed for the determination of the size distribution of noble metals in the wastes, because it will indicate if they are concentrated in the coarse or fine fractions.

•

It is necessary knowing the quantity of wastes produced per day or per month.

•

Optimization of parameters to minimize the consumption of reagents and energy, is needed.

•

More precise evaluation of the economic aspect of the process may be needed to establish the profitability.

100

•

Refining of the recovered metals to a purity of 99.99% needs to be investigated.

•

Safety rules must be respected (see appendix A and B).

101