SULFIDIC ORE MIXED/MULTILAYER FIXED BEDS

ABDELAAZIZ AZIZI

GOLD CYANIDATION REVISITED – KINETIC &

ELECTROCHEMICAL STUDIES OF GOLD –

SULFIDIC ORE MIXED/MULTILAYER FIXED BEDS

Thèse présentée

à la Faculté des études supérieures de l’Université Laval

dans le cadre du programme de doctorat en génie chimique

pour l’obtention du grade de Philosophie Doctor (Ph.D.)

DÉPARTEMENT DE GÉNIE CHIMIQUE

FACULTÉ DES SCIENCES ET DE GÉNIE

UNIVERSITÉ LAVAL

QUÉBEC

2011

© Abdelaaziz Azizi, 2011

i

Résumé

Pour déchiffrer le rôle de minerais sulfureux sur la cyanuration de l’or, une étude détaillée

sur l’importance relative des phénomènes de passivation (PP) et d’interactions galvaniques

(IG) a été menée dans le présent travail de thèse. Une électrode tournante à disque (RDE)

Au/Ag plongée successivement dans une série de pulpes préparées à partir d’une large

gamme de sulfures métalliques a révélé l’impact négatif de ces derniers sur la lixiviation de

l’or. Puisque les contacts galvaniques permanents entre électrode tournante à disque d’or

(RDE) et suspensions de minerais sulfureux sont difficiles à mettre en œuvre, les tests

standards de cyanuration réalisés en mode RDE/slurry ont plutôt tendance à surestimer

l’importance des PP sur l’effet aidant des IG, intrinsèquement présents dans les particules

minérales. Ainsi, un nouveau réacteur électrochimique à lit fixe (PBER) a été développé et

testé pour découpler et quantifier les contributions individuelles des PP et d’IG sur la

lixiviation de métaux précieux (or et argent, PM) lors de la cyanuration de minerais riches

en sulfures. Le réacteur a été chargé de mélanges homogénéisés, de sulfures et de poudres

d’or et d’argent, où y sont établis les contacts galvaniques permanents à l’échelle inter-

particulaire entre tous les constituants. Les IG améliorent, à des degrés variables, la

lixiviation des métaux précieux, particulièrement celles dues à la pyrite, à la chalcopyrite et

à un minerai industriel ont été tellement positives qu’elles l’emportent largement sur les

effets négatifs dus aux PP.

L’extension de la nouvelle approche du PBER pour prendre en considération les

caractéristiques minéralogiques de plusieurs systèmes multi-minéraux a montré que les IG

Au- sulfures ont été le paramètre le plus important contrôlant la lixiviation de l’or.

Finalement, plusieurs stratégies ont été testées pour améliorer la cinétique de cyanuration

de l’or en présence de sulfures inhibiteurs. La galène neutralise largement l’effet négatif de

la dissolution des minéraux sulfureux sur la lixiviation de l’or. Des tests de pré-oxydation

menés sur des sulfures individuels puis sur leurs mélanges associés ont montré que les IG

sulfure-sulfure se produisant lors de l’étape de pré-oxydation au sien du mélange peuvent

donner lieu à des résultats totalement différents pendant la cyanuration.

ii

Abstract

To elucidate the role of sulfide ores on gold cyanidation, a detailed study on the relative

importance of passivation phenomena (PP) and galvanic interactions (GI) was carried out

in the present thesis work. A Rotating Disc Electrode (RDE) Au/Ag disc immersed

successively in slurries of a wide range of sulfide rich ores emphasized the negative impact

of sulfide minerals on the gold leaching rate. Because permanent GI between gold RDE and

slurried sulfide-rich ores are uneasy to achieve, the standard gold RDE/slurry cyanidation

arrangement has a tendency to inflate overly the importance of PP over the corrective trend

of GI inherently present within the ore grains. A new packed-bed electrochemical reactor

(PBER) was thus developed and tested to decouple and quantify the individual

contributions of PP and GI on precious metal (gold and silver, PM) leaching rates during

cyanidation of sulfidic ores. The PBER was filled with mixtures of sulfide minerals, gold

and silver powders, where permanent (inter)particle-particle electrical contacts were

ensured among all constituents. GI were found to ameliorate, to various degrees, the

leaching of PM, particularly those due to pyrite, chalcopyrite and an industrial ore were so

positive that they largely outweighed the negative impact of PM passivation.

Extending the new PBER approach to consider the mineralogical characteristics of several

multi-mineral systems indicated that GI between gold and sulfide mineral particles were the

most important parameters affecting gold leaching. Several leveraging strategies were

tested to increase gold cyanidation kinetics in the presence of PM-leaching inhibiting

sulfide minerals. Galena was found to largely neutralize the negative effect of sulfide

minerals dissolution on gold leaching. Pre-oxidation tested on individual sulfide minerals

and on their associated mixtures revealed that GI occurring between conducting phases

present in the ore may give rise to totally different cyanidation responses.

iii

Avant-Propos

Cinq chapitres composent le présent mémoire de thèse. Le chapitre I (Introduction) est

dédié à une revue de la littérature sur les principaux facteurs ayant une influence sur la

cinétique de l’oxycyanolixiviation de l’or dans un milieu typique de sulfures alors que les

chapitres II-V portent sur les travaux expérimentaux expressément réalisés pour l’obtention

du doctorat. Ces travaux ont donné lieu à quatre publications scientifiques dont deux ont

déjà été acceptées dans le journal international hydrometallurgy. Deux autres publications

sont soumises au même journal.

[1] Azizi, A., Petre, C.F., Olsen, C., Larachi, F., 2010. Electrochemical behavior of gold

cyanidation in the presence of a sulfide-rich industrial ore versus its major constitutive

sulfide minerals. Hydrometallurgy 101, 108-119.

[2] Azizi, A., Petre, C.F., Olsen, C., Larachi, F., 2011a. Untangling galvanic and

passivation phenomena induced by sulfide minerals on precious metal leaching using a new

packed-bed electrochemical cyanidation reactor. Hydrometallurgy, accepted paper.

[3] Azizi, A., Petre, C.F., Olsen, C., Larachi, F., 2011b. The effect of gold mineralogical

associations on its recovery by cyanidation from multi-mineral systems: packed-bed

electrochemical reactor approach. Hydrometallurgy, submitted paper.

[4] Azizi, A., Petre, C.F., Olsen, C., Larachi, F., 2011c. Leveraging strategies to increase

gold cyanidation in the presence sulfide minerals: packed-bed electrochemical reactor

approach. Hydrometallurgy, submitted paper.

Chaque publication constitue un chapitre séparé du corps de la thèse. Cependant quelques

parties superflues ont été éliminées des introductions du troisième et du quatrième article

pour éviter certaines répétitions. Les publications ont été rédigées par moi-même et

corrigées par mon directeur et mon codirecteur.

iv

Remerciements

Ce travail a pu être à terme grâce aux orientations et au soutien de mon directeur

scientifique de thèse, Professeur Faïçal Larachi. Qu’il trouve ici la preuve de toute ma

reconnaissance et mes remerciements pour sa grande disponibilité et pour le temps qu’il a

consacré au bon déroulement des travaux de doctorat en me faisant bénéficier de ses vastes

connaissances scientifiques.

Je tiens à remercier Monsieur Catalin Florin Petre, qui a bien voulu être mon codirecteur de

thèse, à qui je dois une reconnaissance particulière pour sa grande contribution scientifique,

pour ses critiques et pour ses encouragements. Qu’il veuille bien trouver ici l’expression de

ma profonde reconnaissance de m’avoir aidé à la réalisation de ce travail.

Ce travail a bénéficié d’une collaboration fructueuse dans le cadre d’un projet de recherche

et développement coopératif entre l’Université Laval et le Consortium de Recherche

Minérale du Québec (COREM). Je tiens à remercier COREM pour le financement qui m’a

permis de réaliser cette thèse. Les interactions rapprochées et nombreuses avec les

chercheurs et techniciens de COREM et les rétroactions mutuelles ayant eu lieu ont été une

source d’enrichissement permanent. Je pense particulièrement à Caroline Olsen, merci pour

son intérêt, sa collaboration et sa chaleur humaine.

Je tiens à remercier également Monsieur Abdelaziz Baçaoui, Professeur à l’université Cadi

Ayyad au Maroc pour son aide et ses précieux conseils.

Un grand merci à tous mes amis, professeurs et corps administratif et technique du

département de génie chimique de l’Université Laval. Si par hasard, je venais d’oublier

certaines personnes, qu’elles sachent que ma reconnaissance va bien au-delà de ces

remerciements

Il est difficile de trouver des mots assez forts pour exprimer mon immense gratitude et ma

plus grande affection à deux personnes à qui je dois tout : mes parents. Pour leur amour,

leurs sacrifices, leur dévouement et leur encouragement inlassable; qu’ils acceptent que je

leur dédie ce travail.

v

Table des matières

Résumé ..................................................................................................................................... i Abstract .................................................................................................................................. ii Avant-Propos ........................................................................................................................ iii Remerciements ....................................................................................................................... iv

Table des matières .................................................................................................................. v Liste des tableaux ................................................................................................................ viii Liste des figures ..................................................................................................................... ix CHAPITRE I. Introduction et objectifs .............................................................................. 1

I.1 Généralités ..................................................................................................................... 1

I.2 Cyanuration de l’or ........................................................................................................ 6 I.2.1 Complexe Au(CN)2

- ............................................................................................... 6

I.2.2 Mécanismes de dissolution de l’or ......................................................................... 7 I.2.3 Cinétique de la réaction de dissolution de l’or ....................................................... 9

I.2.3.1 Cyanure et Oxygène Dissous ........................................................................... 9 I.2.3.2 Effet de la Température ................................................................................. 11

I.2.3.3 Effet du pH .................................................................................................... 13 I.2.3.4 Surface de contact .......................................................................................... 15 I.2.3.5 Agitation ........................................................................................................ 15

I.2.3.6 Effet des dopants métalliques ........................................................................ 16 I.2.3.7 Phénomènes de passivation & d’interactions Galvaniques ........................... 17

I.3.1 Dissolution des minéraux sulfureux dans les solutions de cyanure ..................... 20 I.3.1.1 Minéraux de Cuivre ....................................................................................... 21 I.3.1.2 Minéraux de fer ............................................................................................. 24

I.3.1.3 Minéraux de zinc ........................................................................................... 25

I.3.1.4 Minéraux d’antimoine ................................................................................... 26 I.3.2 Spéciation du soufre dans les solutions de cyanure et passivation ....................... 28 I.3.3 Phénomènes de passivation et d’interactions galvaniques ................................... 30

I.4 Objectifs du projet ....................................................................................................... 36 I.5 Bibliographie ............................................................................................................... 40

CHAPITRE II. Electrochemical behavior of gold cyanidation in the presence of a sulfide-

rich industrial ore versus its major constitutive sulfide minerals ......................................... 43 II.1. Introduction ............................................................................................................... 45 II.2. Experimental ............................................................................................................. 48

II.2.1. Reagents ............................................................................................................. 48 II.2.2. Materials ............................................................................................................. 48 II.2.3. Electrochemical Campaign ................................................................................ 50

II.2.3.1 Preparation of Disc Electrodes ..................................................................... 50

II.2.3.2 Anodic and Cathodic Behavior of Gold and Mineral Disc Electrodes ........ 51 II.2.3.3 Disc Electrode Galvanic Couples ................................................................. 51

II.2.4. Sulfide Minerals and Gold Dissolution in Slurry Reactor ................................. 52

II.3. Results and discussion ............................................................................................... 53 II.3.1. Anodic and Cathodic Behavior of Gold and Mineral Disc Electrodes .............. 53 II.3.2. Disc Electrode Galvanic Couples ....................................................................... 56

vi

II.3.3. Disc Electrodes Passivation and Galvanic Interactions Effects on Gold

Leaching ........................................................................................................................ 59

II.3.4. Electrochemical Pre-oxidation of Mineral Disc Electrodes and Gold Leaching

...................................................................................................................................... 61 II.3.5. Sulfide Minerals and Gold Dissolution in Slurry Reactor ................................. 63 II.3.6. Alkaline Pre-oxidation of Sulfide Ores and Gold Leaching .............................. 65 II.3.7. Gold leaching, NaOH vs. Ca(OH)2 and pre-oxidation by DO2 vs. H2O2 ........... 77

II.4. Conclusion ................................................................................................................. 79 II.5. References ................................................................................................................. 81

CHAPITRE III. Untangling galvanic and passivation phenomena induced by sulfide

minerals on precious metal leaching using a new packed-bed electrochemical cyanidation

reactor 84

III.1 Introduction ............................................................................................................... 86 III.2. Experimental ............................................................................................................ 88

III.2.1. Materials and reagents ...................................................................................... 88

III.2.2. Equipment and procedures ................................................................................ 90

III.2.2.1. Sulfide minerals and PM dissolution in a packed bed electrochemical

reactor ....................................................................................................................... 90

III.2.2.2. Electrochemical campaign ......................................................................... 93 III.3. Results and Discussion ............................................................................................ 94

III.3.1. Effect of Pyrite (MRI-2) on Gold and Silver Leaching .................................... 94

III.3.2. Effect of Chalcopyrite (MRI-3) on Gold and Silver Leaching ......................... 98 III.3.3. Effect of Sphalerite (MRI-4) on Gold and Silver Leaching ........................... 101

III.3.4. Effect of Chalcocite (MRI-5) on Gold and Silver Leaching .......................... 106 III.3.5. Effect of galena (MRI-6) on gold and silver leaching .................................... 109 III.3.6. Effect of stibnite (MRI-7) on gold and silver leaching ................................... 111

III.3.7. Effect of industrial gold-containing ore (MRI-1) on gold leaching ................ 113

III.3.8. Effect of silver on gold leaching in the presence of sulfides .......................... 116 III.4. Concluding remarks ............................................................................................... 118 III.5. References .............................................................................................................. 119

CHAPITRE IV. The role of multi-sulfidic mineral binary and ternary galvanic

interactions in gold cyanidation in a multi-layer packed-bed electrochemical reactor ...... 122

IV.1. Introduction ........................................................................................................... 123 IV.2. Experimental .......................................................................................................... 126

IV.2.1. Materials and Reagents ................................................................................... 126 IV.2.2. Equipment and procedures ............................................................................. 127 IV.2.3. Electrochemical campaign .............................................................................. 129

IV.3. Results and discussion ........................................................................................... 130 IV.3.1. Gold Cyanidation and the Pyrite-Chalcopyrite-Silica System ....................... 130 IV.3.2. Gold Cyanidation and the Pyrite-Sphalerite-Silica System ............................ 136

IV.3.3. Gold Cyanidation and the Pyrite-Chalcocite-Silica System ........................... 139

IV.3.4. Effect of - -Au Areal Ratio on Gold Cyanidation in the Pyrite-Chalcocite-

Silica System ............................................................................................................... 142 IV.4. Conclusions ........................................................................................................... 145 IV.5. References ............................................................................................................. 146

CHAPITRE V. Leveraging strategies to increase gold cyanidation in the presence of

sulfide minerals - Packed-bed electrochemical reactor approach ....................................... 148

vii

V.1. Introduction ............................................................................................................. 149

V.2. Experimental ........................................................................................................... 151

V.2.1. Materials and Reagents .................................................................................... 151 V.2.2. Equipment and procedures ............................................................................... 152 V.2.3. Electrochemical campaign ............................................................................... 153 V.2.4. Alkaline pre-oxidation of sulfide ores in PBER .............................................. 154

V.3. Results and Discussion ........................................................................................... 155

V.3.1. Effect of Galena on the Leaching of Free Gold in the Presence of Pyrite ....... 155 V.3.2. Effect of Galena on the Leaching of Free Gold in the Presence of Chalcopyrite

.................................................................................................................................... 158 V.3.3. Effect of Galena on the Leaching of Free Gold in the Presence of Sphalerite 161 V.3.4. Effect of Galena on the Leaching of Free Gold in the Presence of Chalcocite164

V.3.5. Effect of Alkaline pre-Oxidation of Pyrite, Chalcopyrite and Sphalerite on Gold

Leaching ...................................................................................................................... 167

V.3.6. Combined pre-Oxidation and Lead Nitrate Addition on Gold Leaching in

Presence of Pyrite ....................................................................................................... 171

V.4. Concluding remarks ................................................................................................ 174 V.5. References ............................................................................................................... 175

CONCLUSION GENERALE ............................................................................................. 177 PERSPECTIVES ................................................................................................................ 180 APPENDIX A. Capillary electrophoretic analysis of sulfur and cyanicides speciation

during cyanidation of gold complex sulfidic ores .............................................................. 182 APPENDIX B. Untangling galvanic and passivation phenomena induced by sulfide

minerals on precious metal leaching using a new packed-bed electrochemical cyanidation

reactor - Supplementary Data SD-1 .................................................................................... 191 APPENDIX C. PBER: time-on-stream vs. physical time .................................................. 195

viii

Liste des tableaux

Tableau I-1 Constantes de stabilité et potentiels standards de réduction des complexes d’or

ayant une importance hydro métallurgiques (Zhang, 1997) ........................................... 3

Tableau I-2 Solubilité de minéraux de cuivre dans une solution 0.1 M CN- (Hedley et

Tabachnick, 1958) ........................................................................................................ 21

Tableau I-3 Solubilité de minéraux de zinc dans une solution de cyanure (Hedley &

Tabachnik, 1958) .......................................................................................................... 25

Tableau I-4 Constantes d’équilibres (K) (Marsden et House, 2006) .................................... 29

Table II-1 Mineralogical composition of the four sulfide ore samples. ............................... 49

Table II-2 Chemical composition of the four sulfide ore samples. ...................................... 49

Table III-1 Mineralogical composition of sulfide ore samples investigated in this study. ... 89

Table III-2 Elemental composition of sulfide ore samples investigated in this study. ......... 89

Table III-3 Sulfide/PM areal ratios implemented in cyanidation experiments. .................... 92

Table III-4 Relative gold recoveries (normalized with respect to silica base cases) from

Au/Ag disc immersed in slurries of minerals and from PM-mineral mixtures (case B)

in PBER. ..................................................................................................................... 116

ix

Liste des figures

Figure I-1 Schéma simplifié d’un circuit de cyanuration avec charbon actif en pulpe

(Jeffrey, 1997) ................................................................................................................. 5

Figure I-2 Diagramme potentiel-pH du système Au-CN-H2O à 25ºC (Osseo-Assare et al.,

1984) ............................................................................................................................... 6

Figure I-3 Schématisation du mécanisme de dissolution de l’or (Habashi, 1967) ................. 8

Figure I-4 Courbes de polarisation linéaire d’oxydation de l’or et de réduction de

l’oxygène, Conditions : 400ppm NaCN et 16ppm d’O2 (Heath & Rumball, 1998) ..... 10

Figure I-5 Courbes de polarisation anodique et cathodique de l’or à différentes

concentrations de cyanure et d’oxygène (Heath & Rumball, 1998) ............................. 11

Figure I-6 Effet de la température sur la lixiviation de l’or (Cathro & Koch, 1964) ........... 12

Figure I-7 Spéciation du cyanure libre et du cyanure d’hydrogène en fonction du pH

(Marsden & House, 2006) ............................................................................................ 13

Figure I-8 Effet du pH sur la consommation du cyanure (Ling et al., 1996) ....................... 14

Figure I-9 Courbe de polarisation montrant l’effet de faible concentration de plomb sur

l’oxydation anodique de l’or (Jeffrey & Ritchie, 2000) ............................................... 17

Figure I-10 Spéciation du cuivre et du cyanure en fonction de la concentration total du

cyanure dans une solution contenant 2mM de Cu(I) (Huang & Young, 1996) ............ 22

Figure I-11 Diagramme E-pH du système Cu-CN-H2O à 25ºC (Osseo-Assare, et al., 1984)

...................................................................................................................................... 23

Figure I-12 Diagramme E-pH du système Fe-CN-S-H2O à 25ºC (Osseo-Assare et al., 1984)

...................................................................................................................................... 24

Figure I-13 (a) produits d’oxydation des hydrosulfures dans une solution de cyanure aérée

(b) effet du pH sur la spéciation des polysulfures à 25 ºC (Senanayake, 2008) ........... 28

Figure I-14 Effets des interactions galvaniques Au-Sulfures sur la dissolution de l’or, pH=

10.5, [CN-] =10mM, T=25ºC (Aghamirian & Yen, 2005) ........................................... 33

Figure I-15 (a) Courbes de polarisation anodiques de l’électrode d’or et minérales (b) Effet

du contact galvanique Au-Chalcocite sur la lixiviation de l’or et potentiel mixte, [CN-]

=10mM (Dai & Jeffrey, 2006) ...................................................................................... 34

Figure II-1 Optical microscopic images of MRI-1: (a) chalcopyrite associated with pyrite,

500X and (b) sphalerite associated with pyrite, 500X. ................................................. 50

Figure II-2 Anodic voltammograms of Au/Ag, MRI-1 and MRI-2 electrodes; CN- = 10

mmol/L, DO2 = 0 mmol/L, pH = 11, ΔE/Δt = 0.5 mV/s, electrode rotation rate = 600

rpm. ............................................................................................................................... 54

Figure II-3 Cathodic voltammograms of Au/Ag, MRI-1 and MRI-2 electrodes; CN- = 0

mmol/L, DO2 = 0.25 mmol/L, ΔE/Δt = 0.5 mV/s, electrode rotation rate = 500 rpm,

pH = 11. ........................................................................................................................ 54

Figure II-4 Effect of coupling Au/Ag and MRI-1 electrodes in one electrochemical cell

(OEC) and in two electrochemical cells (TEC) on gold leaching rate. Reaction

conditions: CN- = 10 mmol/L, DO2 = 0.25 mmol/L, electrode rotation rate = 500 rpm,

pH = 11. ........................................................................................................................ 57

Figure II-5 Effect of galvanic interactions in MRI-1 on the evolution of galvanic potential

and galvanic current vs. time in one electrochemical cell (OEC), CN- = 10 mmol/L,

DO2 = 0.25 mmol/L, electrode rotation rate = 500 rpm, pH = 11. ............................... 58

x

Figure II-6 Effect of galvanic interaction of MRI-1 on the evolution of galvanic potential

and galvanic current vs. time in two electrochemical cells (TEC), CN- = 10 mmol/L,

DO2 = 0.25 mmol/L, electrode rotation rate = 500 rpm, pH = 11. ............................... 58

Figure II-7 Effect of electrode rotational speed on molar flux densities for gold leaching in

the presence of passivation and galvanic interactions. Also shown are the calculated

lines representing the diffusion of oxygen with the reduction of oxygen to (line a)

hydroxide and (line b) peroxide, respectively. Reaction conditions: CN- = 10 mmol/L,

DO2 = 0.25 mmol/L, pH = 11. ...................................................................................... 60

Figure II-8 Coupling Au/Ag and MRI-1 electrodes in one electrochemical cell (OEC):

effect of mineral electrode electro-oxidation on gold leaching. Reaction conditions:

CN- = 10 mmol/L, DO2 = 0.25 mmol/L, pH = 11, electrode(s) rotation rate = 500 rpm.

...................................................................................................................................... 62

Figure II-9 Effect of sulfide minerals on gold dissolution in slurry reactor: industrial ore

(MRI-1) and the equivalent of its major sulfide constituents (MRI-2, MRI-3 and MRI-

4). Reaction conditions CN- = 10 mM, DO2 = 0.25 mM, pH = 11, electrode rotation

rate = 500 rpm. .............................................................................................................. 64

Figure II-10 Cyanidation of Au/Ag electrode in the presence of MRI-1 slurry. Pre-oxidation

effect on: (a) gold leaching rate, (b) sulfur speciation and (c) metals speciation. Full

symbols: tests without pre-oxidation; empty symbols: tests with pre-oxidation.

Reaction conditions: CN- = 10 mM, DO2 = 0.25 mM, pH = 11.5, electrode rotation

rate = 500 rpm. .............................................................................................................. 68





rate = 500 rpm. .............................................................................................................. 70





rate = 500 rpm. .............................................................................................................. 73


effect on: (a) gold leaching rate, (b) sulfur speciation and metals speciation. Full



rate = 500 rpm. .............................................................................................................. 75

Figure II-14 Effect of lime (Ca(OH)2) on the gold leaching rate. Reaction conditions: 50%

solids by mass, CN- = 20 mM, DO2 = 0.25 mM, pH = 11.5, electrode rotation rate =

500 rpm. ........................................................................................................................ 78

Figure II-15 Cyanidation of Au/Ag electrode in the presence of MRI-1 slurry: effect of the

type of pre-oxidant (O2 or H2O2) on gold leaching rate. Reaction conditions: 50%

solids by mass, CN- = 20 mM, DO2 = 0.25 mM, pH = 11.5, electrode rotation rate =

500 rpm, 3.2% w H2O2. ................................................................................................ 79

Figure III-1 Sketch of a new packed-bed electrochemical reactor (PBER). Legend as

follows: 1, inlet section; 2, working section (Teflon); 3, outlet section; 4, stainless-

steel cylinder; 5, sintered-glass distributor; 6, peristaltic pump; 7, magnetic stirrer; 8,

oxygen probe electrode; 9, pH-meter electrode; 10, air bubbling system. ................... 91

xi

Figure III-2 Effect of pyrite (MRI-2) on: (a and a`) gold and silver dissolution, (b) the

evolution of galvanic potential and galvanic current vs. time between Au and MRI-2.

Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11. .................................. 96

Figure III-3 Effect of chalcopyrite (MRI-3) on: (a) gold and silver dissolution, (b) the


Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11. ................................ 100

Figure III-4 Effect of sphalerite (MRI-4) on: (a) gold and silver dissolution, (b) the



Figure III-5 Effect of galvanic interactions between gold and MRI-4 on: (a) oxy-sulfur ions

speciation and (b) cyanicides formation. Reaction conditions: CN- = 30 mM, DO2 =

0.25 mM, pH = 11. ...................................................................................................... 105

Figure III-6 Effect of chalcocite (MRI-5) on: (a) gold and silver dissolution and (b) the



Figure III-7 Effect of galvanic interactions between gold and MRI-5 on the reactivity of

MRI-5. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11. ................... 108

Figure III-8 Effect of galena (MRI-6) on: (a) gold and silver dissolution and (b) the



Figure III-9 Effect of stibnite (MRI-7) on: (a) gold dissolution, (b) the evolution of galvanic

potential and galvanic current vs. time between Au and MRI-7. Reaction conditions:

CN- = 30 mM, DO2 = 0.25 mM, pH = 11. .................................................................. 112

Figure III-10 Effect of the industrial gold ore (MRI-1) on: (a) gold dissolution and (b) the



Figure III-11 Effect of silver on gold leaching in the presence of pyrite (MRI-2). Reaction

conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11. ............................................... 118

Figure IV-1 Sketch of packed-bed electrochemical reactor (PBER) strategies used to study

the effect of gold distribution and mineralogical associations on gold recovery. (a)

pyrite mineral phase: ; chalcopyrite, sphalerite or chalcocite mineral phase: X; silica

layer; insulator sintered-glass disc filter; (b) homogenized mixture of mineral phases and X. .......................................................................................................................... 128

Figure IV-2. Effect on gold leaching of Au distribution within (a) electrically-disconnected

pyrite, chalcopyrite and silica layers: Au within pyrite (A), Au within chalcopyrite (B),

Au within silica (C,D); (b) electrically-connected pyrite-chalcopyrite//silica systems:

Au within pyrite-chalcopyrite (A’), Au within silica (B’). Inset: Effect of pyrite-

chalcopyrite galvanic association on free-gold leaching from inert silica layer.


Figure IV-3 Measured galvanic current and potential vs. time for pyrite-chalcopyrite

galvanic couple in the PBER. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM,

pH = 11. ...................................................................................................................... 133

Figure IV-4 Effect of galvanic interactions between pyrite and chalcopyrite on: (a) Fe-

cyanide and Cu-cyanide evolution and (b) sulfur-bearing anions speciation. Reaction

conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11. .............................................. 135

xii


pyrite, sphalerite and silica layers: Au within pyrite (A), Au within sphalerite (B), Au

within silica (C,D); (b) electrically-connected pyrite-sphalerite//silica systems: Au

within pyrite-sphalerite (A’), Au within silica (B’). Inset: Effect of pyrite-sphalerite

galvanic association on free-gold leaching from inert silica layer. Reaction conditions:

CN- = 30 mM, DO2 = 0.25 mM, pH = 11. ................................................................. 137 Figure IV-6. Measured galvanic current and potential vs. time pyrite-sphalerite galvanic

couple in the PBER. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11.

.................................................................................................................................... 138 Figure IV-7. Effect of galvanic interactions between pyrite and sphalerite on SO4

2- and

SCN- evolution. Reaction conditions: CN

- = 30 mM, DO2 = 0.25 mM, pH = 11. ..... 139


pyrite, chalcocite and silica layers: Au within pyrite (A), Au within chalcocite (B), Au

within silica (C,D); (b) electrically-connected pyrite-chalcocite//silica systems: Au

within pyrite-chalcocite (A’), Au within silica (B’). Reaction conditions: CN- = 30

mM, DO2 = 0.25 mM, pH = 11. .................................................................................. 140

Figure IV-9 Measured galvanic current and potential vs. time for pyrite and chalcocite

galvanic couple in the PBER. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM,

pH = 11. ...................................................................................................................... 141 Figure IV-10 . Effect of galvanic interactions between pyrite and chalcocite on: Fe-cyanide

and Cu-cyanide evolution. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH =

11. ............................................................................................................................... 142 Figure IV-11. Effect of (a) pyrite-Au and (b) pyrite-chalcocite-Au areal ratios on gold

leaching from multi-mineral systems containing pyrite, chalcocite, gold and silica.


Figure V-1. Sketch of packed-bed electrochemical reactor (PBER) strategies used to study

the effect of gold distribution and galena mineralogical associations on gold recovery

with/without pre-oxidations. (a) mineral phase: X (= pyrite; chalcopyrite, sphalerite or

chalcocite) or homogenized mixture of mineral phase X + galena: X + ; silica layer;

insulator sintered-glass disc filter; (b) mineral phase: X (= pyrite; chalcopyrite,

sphalerite or chalcocite); galena: ; silica layer; insulator sintered-glass disc filter. .. 153

Figure V-2. Effect of galena ( ) in the presence of pyrite ( ) on the evolution of: (a and a`)

free gold dissolution and (b) -Au galvanic current. Reaction conditions: CN- = 30

mM, DO2 = 0.25 mM, pH = 11. .................................................................................. 157

Figure V-3. Effect of galena ( ) in the presence of chalcopyrite ( ) on the evolution of: (a)

free gold dissolution, (b) -Au galvanic current, (c) sulfur anions speciation. Reaction


Figure V-4. Effect of galena ( ) in the presence of sphalerite ( ) on the evolution of: (a)



Figure V-5. Effect of galena ( ) in the presence of chalcocite ( ) on the evolution of: (a)



Figure V-6. Effect of pre-oxidation on gold leaching in presence of pyrite. Full symbols:

tests without pre-oxidation; empty symbols: tests with pre-oxidation. Reaction


xiii

Figure V-7. Effect of pre-oxidation on gold leaching in presence of chalcopyrite. Full


Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11. ............................... 170

Figure V-8. Effect of pre-oxidation on gold leaching in presence of sphalerite. Full


Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11. ............................... 171

Figure V-9. Effect of pre-oxidation of mixed mineral systems containing pyrite and

chalcopyrite or pyrite and sphalerite on free gold leaching. Full symbols: tests without

pre-oxidation; empty symbols: tests with pre-oxidation. Reaction conditions: CN- = 30

mM, DO2 = 0.25 mM, pH = 11. .................................................................................. 171

Figure V-10 Effect of lead nitrate addition strategy in presence of pyrite, on: (a) gold

leaching and (b) pyrite oxidation products speciation. Reaction conditions: CN- = 30

mM, DO2 = 0.25 mM, 5 mg/L Pb-Pb(NO3)2, pH = 11. ............................................ 173

1

CHAPITRE I. Introduction et objectifs

I.1 Généralités

L’or est un métal précieux. Il est inaltérable, malléable, recyclable mais aussi rare. Il

demeure indéniablement l’un des repères économiques importants dans nos sociétés. En

raison de sa rareté et de ses propriétés chimiques et physiques, l’or est depuis longtemps un

des matériaux les plus convoités. Aujourd’hui, en plus de ses applications dans les sociétés

traditionnelles et dans les systèmes financiers internationaux où il constitue une valeur

refuge, sa vocation grandissante de métal de haute technologie dans divers secteurs

industriels n’est plus à prouver.

L’or et/ou ses alliages sont de plus en plus utilisé(s) au cours des vingt dernières années en

électronique (Maxey et al., 1997) pour sa conductivité électrique élevée, dans les industries

spatiales et aéronautique (Jeffrey, 1997) pour sa réflectivité élevée aux infrarouges, dans

l’industrie chimique comme catalyseur des réactions d’hydrogénation, d’isomérisation, de

craquage des hydrocarbures (Chandler et al., 2000) et en médecine (Jeffrey, 1997) pour le

traitement des arthrites, la radiothérapie de certains cancers et pour plusieurs diagnostics

comme l’observation de la moelle des os ou la délimitation du foie et du poumon.

Aujourd'hui la production mondiale de l'or est d'environ 2 300 tonnes métriques (tm) et est

en progression constante avec les nouvelles technologies d'extraction (Gavin, 2007). Les

principaux gisements d’or se trouvent en Afrique du sud (500 tm/an), aux États-Unis (350

tm/année) et au Canada (150 tm/année) (Gavin, 2007). On en trouve aussi en Indonésie et

Nouvelle-Zélande (200 tm/an), en Russie (80 tm/an) et au Ghana (75 tm/année). L’or se

trouve dans la nature sous différentes formes : or natif, sulfure d’or, électrum. Il est

habituellement associes à des veines de quartz et de sulfures métalliques.

Le procédé le plus utilisé pour la production de l’or est l’extraction hydrométallurgique par

cyanuration (80% de la production mondiale) rendue possible suite à la découverte de la

grande solubilité de ce métal dans une solution de cyanure aérée par Elsner en 1846

(Marsden & House, 2006). Avant cette découverte, l’or était récupéré par gravimétrie, une

2

méthode qui ne permettait pas de récupérer les fines particules d’intérêt. Le procédé par

cyanuration est celui le plus répandu dans l’industrie minière aurifère. Malgré les coûts

attachés à la cyanuration, à la sécurité liée à la toxicité du cyanure et aux effets néfastes sur

l’environnement exigeant plusieurs traitements, cette méthode est la plus efficace jusqu'à ce

jour pour satisfaire la demande en or trop élevée dans le marché international.

De tous les métaux l’or est le métal le plus noble (Nicol et al., 1987) et existe naturellement

à l’état métallique. C’est là une conséquence de la stabilité de l’électron 6 S qui engendre

une faible activité chimique de l’or en solution aqueuse qui est fonction de son potentiel

standard de réduction.

Les états d’oxydation les plus communs de l’or sont Au(I) et Au(III) avec des potentiels

standards de réduction qui sont respectivement 1690 mV et 1500 mV (Nicol et al., 1987).

Ces valeurs de potentiel sont nettement supérieures au potentiel de réduction de l’eau, 1230

mV (Bard, 1973), indiquant ainsi que les formes aureuse (Au(I)) et aurique (Au(III)) sont

thermodynamiquement instables sous forme de composés ioniques en solution aqueuse.

Cependant, les cations aureux et auriques peuvent être stabilisés par un certain nombre de

ligands fortement donneurs permettant ainsi la formation de complexes peu dissociés

(Marsden & House, 2006).

La réaction de complexation des ions aureux par un ligand Ln, avec n charges positives ou

négatives, peut être représentée par la réaction I.1 dont la constante de stabilité est calculée

selon l’équation I.2.

Au+

+ xLn

→ 1 nx

xAuL (I.1)

Ks =

1 nx

X

xn

AuL

Au L (I.2)

Le potentiel standard de réduction du complexe peut être exprimé en fonction du potentiel

de réduction des ions aureux E0 Au

+/Au et de la constante de stabilité, Ks, selon

l’équation I.3,

3

0 0

/ln( )complexe Au Au

RTE E Ks

F (I.3)

Où ;

R : Constante des gaz parfait (J K-1

mol-1

)

T : Température absolue (K)

F : Nombre de Faraday (C mol-1

)

Il est clair que pour des valeurs élevées de la constante de stabilité Ks, les complexes d’or

sont thermodynamiquement stable (E0 < 1230 mV). Cependant la chimie aqueuse de l’or

est dominée par la formation d’un nombre très limité de complexes stables des ions aureux

et auriques. Ceux jouissant d’une importance hydro-métallurgique sont listés dans le

tableau ci-dessous.

Tableau I-1 Constantes de stabilité et potentiels standards de réduction des complexes d’or

ayant une importance hydro métallurgiques (Zhang, 1997)

Complexe Ks Réaction E0/mV

Au(CN)2-

2.1038

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

- -570

AuS-

2.1036

AuS- + e = Au + S -460

Au(HS)2-

1,3.1030

Au(HS)2- + e = Au + 2HS -90

Au(S2O3)23-

5.1028

Au(S2O3)23-

+ e = Au + S2O32-

150

Au(thiourea)2+

2.1023

AuTu2+

+ e = Au + 2 Tu+

352

Au(CN)43-

1056

Au(CN)43-

+ e = Au + 4CN-

400

AuI4-

5.1047

AuI4- + 3e = Au + 4I

- 560

Au(SCN)43-

1042

Au(SCN)43-

+ 3e = Au + 4SCN-

623

AuBr4-

1032

AuBr4- + 3e = Au + 4Br

- 870

AuCl4-

1026

AuCl4- + 3e = Au + 4Cl

- 1002

Bien que des recherches nombreuses aient été effectuées sur les lixiviants autres que le

cyanure, la cyanuration reste la technique la plus utilisée dans l’industrie de traitement de

minerais aurifères. L’or est connu pour être soluble dans une solution de cyanure aérée en

4

raison de l’abaissement du potentiel de réduction du métal en-deçà de celui de l’oxygène

dissous, mais aussi grâce à la grande stabilité en solution du complexe dicyanoaureux. L’or

métallique est également connu pour développer dans une solution de cyanure un potentiel

moindre que certain sulfures métalliques, électriquement conducteurs, avec lesquels il est

naturellement en contact au sein du minerai. C’est pour cela que le cyanure continue d’être

l’agent de lixiviation par excellence utilisé en industrie pour la dissolution oxydative et la

complexation de l’or (présent à l’état de traces, 1-5ppm) à partir des minerais aurifères

sulfureux (Hilson & Monhemius, 2006).

Le procédé d’extraction de l’or par cyanuration a été breveté par MacArthur & Forrest en

1887 (Marsden & House, 2006), et depuis, ce procédé reste le plus largement utilisé en

hydrométallurgie extractive de l’or. Le procédé d’extraction de l’or est effectué

généralement en trois étapes principales (voir Figure I.1), auxquelles peuvent s’ajouter des

étapes supplémentaires selon le type de minerai à traiter :

1. Broyage du minerai en milieu humide et généralement alcalin afin de libérer les

particules d’or et rendre leurs surfaces plus accessibles au cyanure. On ajoute usuellement

du cyanure à cette étape afin d’augmenter le temps de contact entre le minerai et la solution

lixiviante. Le degré de broyage est souvent élevé en raison de la finesse des particules d’or.

2. Le minerai broyé est ensuite transféré vers les cuves de cyanuration. Ces cuves, dont un

modèle simple est montré à la Figure I.1, fonctionnent à la pression atmosphérique avec

agitation. Les dimensions et le nombre de cuves sont choisis pour assurer un temps de

séjour moyen de la pulpe dans le milieu réactionnel de 20 à 48 heures. De l’air et une

solution de cyanure sont ajoutés dans les cuves pour permettre la lixiviation de l’or. L’air

fournit l’oxygène soluble nécessaire à l’oxydation en solution de l’or. La complexation des

cations aureux par le cyanure et la formation du complexe dicyanoaureux peuvent être

représentées selon la réaction globale classique d’Elsner :

4 Au +8CN-

+ O2 + 2H2O → 4Au(CN)2- + 4OH

- (I.4)

3. Une fois que l’étape de l’oxycyanolixiviation est achevée, et que le métal visé est mis en

solution sous forme de complexe dicyanoaurate, Au(CN)2-, l’or peut être récupéré des

solutions de cyanure. Il existe actuellement deux méthodes qui sont mises en œuvre à

5

l’échelle industrielle : la méthode du charbon actif en pulpe et celle de la précipitation par

cémentation avec la poudre de zinc. Bien que les nouvelles usines préfèrent utiliser la

première (Figure I.1), la majorité des concentrateurs actuels utilisent encore la seconde. La

première méthode consiste à faire adsorber l’or dissous sur le charbon actif en grain, et à

séparer celui-ci du reste de la pulpe par des tamis. On procède ensuite à l’élution du

charbon actif et on récupère l’or par électrolyse de la solution résultante sur un treillis de

fer. La deuxième méthode consiste à séparer les solides de la solution par des filtres ou par

des épaississeurs à contre-courant. L’or de la solution traitée et ensuite précipité par l’ajout

de la poudre de zinc et récupéré par filtration.

Figure I-1 Schéma simplifié d’un circuit de cyanuration avec charbon actif en pulpe

(Jeffrey, 1997)

6

I.2 Cyanuration de l’or

I.2.1 Complexe Au(CN)2-

Le complexe Au(CN)2- est le plus stable de tous les complexes que peut former l’or avec

les ligands ioniques, d’où sa grande importance en hydrométallurgie de l’or (Nicol, 1980a).

La grande affinité du cyanure pour l’or est attribuée à sa charge négative et à sa capacité à

entrer dans une liaison Π avec l’or (Wang & Forssberg, 1990). Dans ce complexe l’état

d’oxydation de l’or est (+1) et le nombre de ligands préféré est deux (Nicol et al., 1987).

La grande valeur de la constante de stabilité, Ks = 2.1038

M (Nicol, 1980a), du complexe

Au(CN)2- est d’une importance capitale dans le procédé d’extraction d’or dans la mesure où

la concentration en cyanure pourrait diminuer dans les cuves de cyanuration sans pour

autant engendrer une dissociation conséquente du complexe dicyanoaurate.

L’application des diagrammes de Pourbaix dans les procédés hydro métallurgiques a été

largement documentée par Osseo-Assare et al. (1984), Figure I.2. Ces diagrammes

indiquent les régions de stabilité thermodynamique des différentes espèces chimiques d’un

système donné et sont utilisés pour établir des prévisions sur les possibilités/impossibilités

de corroder un métal.

Figure I-2 Diagramme potentiel-pH du système Au-CN-H2O à 25ºC (Osseo-Assare et al., 1984)

7

Le diagramme potentiel-pH du système Au-CN-H2O à 25ºC (Figure I.2) fait apparaître la

prédominance du complexe Au(CN)2- dans de larges conditions de pH et de potentiels,

indiquant ainsi la grande stabilité de ce complexe en solution aqueuse. La superposition des

régions de stabilité de l’or métallique et de l’eau dans de larges plages de pH montre bien

l’importance de l’utilisation d’un oxydant plus fort pour la dissolution de l’or dans une

solution de cyanure. Elsner (1846) fut le premier à reconnaitre la nécessité de l’oxygène

dans le processus de dissolution de l’or.

I.2.2 Mécanismes de dissolution de l’or

En présence d’un agent lixiviant, à savoir, le cyanure de potassium, de sodium ou de

calcium, l’or se dissout en milieu oxygéné sous la forme d’un sel double soluble

MAu(CN)2-. Cette propriété est à la base du procédé de cyanuration.

Elsner (1846) est l’un des premiers chercheurs ayant mis en évidence cette réactivité. Il

propose l’équation suivante :

4Au + 8CN- +O2 + 2H2O → 4Au(CN)2

- + 4OH

- (I.5)

Bodlander (1896) a par la suite suggéré que la réaction procède en deux étapes :

2Au + 4NaCN + O2 + 2H2O → 2NaAu(CN)2- +2NaOH +H2O2 (I.6)

H2O2 + 2Au + 4NaCN → 2NaAu(CN)2- + 2NaOH (I.7)

L’équation globale reste la même que celle d’Elsner. Cependant, il n’existe pas de

consensus dans la littérature sur les mécanismes de réduction du peroxyde d’hydrogène à

la surface de l’or bien que sa présence ait été prouvée dans de nombreux cas.

Boostera (1943) a émis l’hypothèse que la cyanuration de l’or n’est rien d’autre qu’un

processus de corrosion mettant en œuvre l’oxydation anodique de l’or métallique, la

réduction cathodique d’oxygène et la complexation par le cyanure des ions aureux. Les ions

cyanure libres attaquent la surface des sites anodiques de l’or (oxydation), les électrons

libérés sont dirigés vers les sites cathodiques de l’or et l’oxygène réagit avec la surface de

ces derniers pour éliminer les électrons en surplus (réduction). Il s’agirait alors d’un

phénomène de surface dont la vitesse serait fonction de la diffusion des réactifs de la

8

solution vers la surface de l’or (Figure I.3). Cette interprétation a été approuvée par

plusieurs chercheurs (Habashi, 1967, Clem, 1982 et Worstell, 1987). Pour qu’il y ait

réaction dans un tel processus électrochimique, il faut que le réactif soit transporté vers la

surface de l’électrode où le transfert d’électrons aura lieu et que le produit de réaction soit

transporté de cette interface vers le cœur de la solution. Les réactions électrochimiques,

pour lesquelles l’une de ces deux étapes de transfert constitue l’étape limitative de la

cinétique, correspondent à des réactions caractérisées par un transfert de charges

particulièrement rapide. Pour ces réactions, la cinétique relative au transfert de charge est si

rapide que le potentiel et les concentrations à la surface de l’électrode sont toujours en

équilibre. Dans ce cas, l’équation thermodynamique de Nernst est vérifiée, il s’agit d’un

système appelé communément Nernstien.

Figure I-3 Schématisation du mécanisme de dissolution de l’or (Habashi, 1967)

Il se forme un film à la surface des particules d’or (couche limite de Nernst) à travers lequel

des échanges prennent place par diffusion (le mouvement des réactifs est sous l’influence

d’un gradient de potentiel chimique). L’oxygène dissous et le cyanure doivent diffuser à

travers ce film afin d’atteindre la surface de la particule d’or et y réagir. L’épaisseur de ce

film dépend également du niveau d’agitation de la solution.

9

La cinétique de la réaction de lixiviation de l’or est alors contrôlée par la vitesse de

diffusion des réactifs au travers de la couche limite, de la concentration des réactifs, du

degré d’agitation, de la surface d’or disponible et des phénomènes de passivation de

surface.

I.2.3 Cinétique de la réaction de dissolution de l’or

Le contrôle du procédé de cyanuration et l’amélioration de ses performances ne seraient

possibles qu’à travers une compréhension adéquate des mécanismes d’action de tous les

facteurs pouvant influencer la cinétique de dissolution de l’or.

I.2.3.1 Cyanure et Oxygène Dissous

En se basant sur la réaction de dissolution de l’or (two-electron process), il apparait qu’une

mole d’or nécessite 0.5 mole d’oxygène et 2 moles de cyanure dépendamment de

l’efficacité de réduction du peroxyde d’hydrogène en anion hydroxyde. Les conditions de

diffusion limite s’établissent quand la vitesse de diffusion du cyanure et/ou de l’oxygène est

proportionnelle à la vitesse de lixiviation de l’or. Ce flux à la surface de l’or est maximal

lorsque les conditions de potentiel à l’interface sont telles que les concentrations [CN-]

et/ou [O2] sont (est) nulle(s) (les réactifs sont instantanément consommés à la surface du

métal). De tels flux maximaux sont alors proportionnels à la concentration de réactif(s) au

sien de la solution. Le flux de diffusion limite pour une électrode à disque rotatif est calculé

selon l’équation de Levich (Heath & Rumball, 1998) :

Jx = 0.62D2/3

V-1/6 1/2

C(X) (I.8)

Avec,

J : Flux de diffusion (mol m2 s

-1) de l’espèce X

D : Coefficient de diffusion de l’espèce X en solution (m2 s

-1)

V : Viscosité cinématique de la solution (m2

s-1

)

: Vitesse angulaire de rotation de l’électrode en tours par seconde (s-1

)

C(X) : Concentration de l’espèce X (mol m-3

)

10

En utilisant la loi de Faraday et à partir de l’équation de Levich, le courant de diffusion

théorique peut être calculé selon l’équation suivante :

Id = 0.62nFAD2/3

V-1/6 1/2

C(X) (I.9)

Avec,

n : Nombre d’électrons échangés

F : Nombre de Faraday

A : Surface de l’électrode

Le potentiel auquel se croisent les courbes de polarisation anodique et cathodique est

appelé potentiel mixte ou potentiel de corrosion. À un tel potentiel, le courant cathodique

est égal en valeur absolue au courant anodique (Figure I.4) et la cinétique de la réaction est

contrôlée par l’espèce chimique présentant un plateau de courant de diffusion.

Figure I-4 Courbes de polarisation linéaire d’oxydation de l’or et de réduction de

l’oxygène, Conditions : 400ppm NaCN et 16ppm d’O2 (Heath & Rumball, 1998)

Il est généralement admis qu’à faibles concentrations en cyanure par rapport à l’oxygène, la

vitesse de dissolution de l’or est proportionnelle à la concentration en cyanure. Cette

relation est linéaire jusqu’à atteindre une vitesse limite qui correspond à l’activité de

11

l’oxygène. Inversement, à faible concentration en oxygène, la vitesse de dissolution est

proportionnelle à la concentration en oxygène (Figure I.5).

Figure I-5 Courbes de polarisation anodique et cathodique de l’or à différentes

concentrations de cyanure et d’oxygène (Heath & Rumball, 1998)

On aurait donc avantage à optimiser plutôt le rapport cyanure/oxygène afin de rationaliser

l’utilisation des réactifs. Des investigations faites sur des électrodes en alliage Au/Ag ont

montré des rapports molaires optimaux variant de 4:1 à 7:1 selon les auteurs (Habashi,

1966; Lorenzen & van Deventer, 1992; Heath & Rumball, 1998).

Cependant, pour des raisons pratiques, les concentrations de cyanure utilisées sont

beaucoup plus élevées pour éviter des conditions propices aux limitations en cyanure en

raison de l’action de certains sulfures de métaux qui peuvent consommer de grandes

quantités de cyanure.

I.2.3.2 Effet de la Température

Le processus électrochimique global de lixiviation de l’or est limité, entre autres, par le

transfert de masse par diffusion du cyanure et/ou d’oxygène au sien de la couche de Nernst

(énergie d’activation de la réaction de dissolution de l’or de 1 à 5 kcal/mol). On pourrait

12

ainsi, supposer que la réaction est contrôlée par les flux de transport des réactifs à la surface

de l’or. Les coefficients de diffusion des réactifs sont directement proportionnels à la

température, DCN = K T; DO2 = K T. Cependant, la solubilité de l’oxygène diminue avec un

accroissement de la température. Il y’a donc moins d’oxygène potentiellement disponible

pour l’oxydation de l’or à température élevée. Ces tendances opposées laissent à penser que

le choix de la température est un compromis entre flux de réaction, qui augmente avec la

température, et disponibilité d’oxygène, qui diminue avec l’augmentation de celle-ci.

Des expériences réalisées à différentes températures (Figure I.6), montrent clairement

qu’une augmentation de celle-ci ne semble pas améliorer significativement la cinétique de

la réaction globale étant donné que toutes les courbes convergent vers un rendement final

pratiquement similaire après un temps de lixiviation de 5 h.

Figure I-6 Effet de la température sur la lixiviation de l’or (Cathro & Koch, 1964)

Les faibles sur-lixiviations obtenues durant les premières heures en augmentant la

température, par rapport à la température ambiante, ne peuvent que rarement justifier les

coûts d’énergie excédentaires (Marsden & House, 1992). Ceci n’est pas surprenant, vu que

13

le cyanure peut également se décomposer à des températures supérieures à 40 °C. Selon W.

McQuiston (1973), la température optimale de cyanuration se situe entre 15,6°C et 21,1°C.

I.2.3.3 Effet du pH

Lorsque le cyanure de sodium est mis en solution, un équilibre s’établit entre l’ion cyanure

(CN-) et le cyanure d’hydrogène (HCN) :

CN- + H2O ↔ HCN + OH

- Ka(25 ºC) = 6.2 10

-10 , pKa = 9.31 (I.10)

Le sens de déplacement de l’équilibre HCN/CN- dépend du pH de la solution (Figure I.7).

À pH = 9.31, 50% du cyanure total existe sous forme CN- et 50% sous forme de HCN. À

pH < 9.31, l’équilibre se déplace dans le sens de formation du cyanure d’hydrogène (plus

de 90% de cyanure total est sous forme de HCN à pH = 8.4). Cependant à pH > 9.31,

l’équilibre se déplace dans le sens de formation des ions CN-. C’est d’ailleurs la principale

raison pour laquelle les circuits de cyanuration de l’or sont toujours opérés à des pH

alcalins (souvent supérieure à 10).

Figure I-7 Spéciation du cyanure libre et du cyanure d’hydrogène en fonction du pH

(Marsden & House, 2006)

Selon Marsden & House (2006), la vitesse de dissolution de l’or est peu affectée par le pH

lorsqu’il est supérieur à 9.5. Cependant cette observation n’a été que très peu confirmée

dans la littérature. Ling et al. (1996) ont montré, lors de tests de lixiviation à partir de

14

minerais aurifères sulfureux, une diminution importante de la consommation de cyanure en

passant de pH = 10 à pH = 12 (Figure I.8).

Figure I-8 Effet du pH sur la consommation du cyanure (Ling et al., 1996)

Dans ce cas, il possible que les économies réalisées sur la consommation de cyanure

seraient dues à la formation d’une couche protectrice, à pH élevés, d’hydroxydes de métaux

à la surface de minéraux générateurs de cyanicides. Cependant cette tendance n’est pas

toujours en faveur de la cinétique de lixiviation de l’or. La forte alcalinité pourrait nuire à la

dissolution de l’or dans le cas où elle favorise la formation d’un film résistif d’hydroxydes

de métaux ou de sels d’agents alcalins à la surface du métal visé. Un tel film empêcherait

d’avantage l’accès par le cyanure à la surface active de l’or et ferait basculer la cyanuration

d’un contrôle externe par le cyanure à un contrôle plus sévère à travers le film solide.

Selon Hedley & Tabachnick, (1958), la décomposition des minéraux à base de métaux

d’arsenic (As) et d’antimoine (Sb), est fortement dépendante du pH. Par conséquent, leurs

effets inhibiteurs (formation d’une couche d’oxyde d’antimoine ou d’arsenic à la surface de

l’or) sont aussi fortement dépendants du pH du milieu.

Pour conclure, le pH de la cyanuration est dicté par plusieurs facteurs :

15

1. La vitesse de dissolution des différentes phases minérales entrant la composition

des minerais industriels. Ces dernières, peuvent affecter soit directement ou

indirectement le rendement global de la cyanuration;

2. Les coûts engendrés par l’ajustement du pH;

3. La précipitation des espèces en solution (CaSO4, Fe(OH)3, etc.)

De ce fait, le pH optimal de la cyanuration devrait être déterminé dépendamment de la

composition minéralogique du minerai à traiter et de la nature chimique des eaux de

procédé industriel.

I.2.3.4 Surface de contact

La cyanuration de l’or est un processus électrochimique (Boostra, 1943) de corrosion dont

la vitesse serait fonction de la diffusion des réactifs de la solution à la surface des particules

ciblées. La cinétique de la lixiviation est alors proportionnelle à la surface d’or exposée à la

solution. Un broyage poussé augmente considérablement les surfaces granulaires et facilite

théoriquement le contact des grains d’or avec la solution de cyanure.

Il est généralement admis que la vitesse de lixiviation est inversement proportionnelle à la

taille des particules minérales à cause de l’augmentation du degré de libération de l’or par

broyage de plus en plus fin. Cependant ce n’est pas toujours le cas, car des broyages

extrêmement fins font augmenter aussi la surface des métaux et des minéraux

consommateurs de cyanure et générateurs des anions sulfures (HS-), ce qui peut réduire les

bénéfices d’une plus grande surface de contact avec la solution de cyanure (Cornejo, 1984).

Pour l’optimisation du circuit de broyage, on aurait donc avantage à chercher plutôt un

compromis entre le pourcentage de l’or lixivié et le cyanure consommé.

I.2.3.5 Agitation

La lixiviation de l’or correspond à un phénomène de surface dont la vitesse est contrôlée

par la diffusion des réactifs au travers de la couche limite qui se forme autour des particules

d’or (couche de Nernst). L’épaisseur de cette couche est fonction du degré d’agitation de la

solution par rapport aux grains d’or. Plus l’agitation est importante, moins la couche de

16

Nernst est épaisse, plus la diffusion des réactifs à la surface de l’or est rapide. Cette

hypothèse a été validée lors de tests de lixiviation réalisés avec une électrode d’or tournante

(Heath & Rumball, 1998). Cependant, le contrôle de la cinétique de la lixiviation d’or à

partir de minerais réels peut être mixte, et les lois de diffusion ne déterminent que

partiellement la vitesse de dissolution de l’or dans la mesure où les phénomènes de

passivation bloquent la surface de l’or par un film résistif. Auquel cas, l’agitation

mécanique et/ou l’augmentation du flux volumétrique d’aération dans le réacteur ne

seraient susceptibles de compenser qu’en partie le déclin de la lixiviation d’or.

I.2.3.6 Effet des dopants métalliques

La passivation de l’or concernée dans la pratique industrielle de la cyanuration nous oblige

à nous intéresser au premier pic de polarisation survenant dans les régions des potentiels

négatifs (Figure I.9). Il existe un consensus dans la littérature à l’effet que la passivation

serait due à la formation d’un film ou d’un adsorbat du complexe neutre monocyanoaurate,

AuCN, (Jeffrey & Ritchie, 2000) à la surface de l’or. Des études ont montré que l’or pur

(ca. 100%) est plus vulnérable à la passivation et que la présence de certains additifs ou

cations bivalents peut, par l’entremise de mécanismes complexes de surface, accélérer la

cinétique de dissolution de l’or (Senanayake, 2008).

Selon Marsden et House (2006), la présence de faibles concentrations de plomb, de

mercure, de thallium ou de bismuth améliore nettement la vitesse de lixiviation de l’or en

évitant ou en réduisant l’effet de passivation survenant à -0.4 V. D’autres auteurs

(Deschênes et al., 2000; Jeffrey & Ritchie, 2000) rapportent que des concentrations de

plomb variant de 10-6

à 10-5

M sont bénéfiques alors qu’à des concentrations dépassant 10-4

M, les effets de Pb(II) sur la vitesse de dissolution de l’or sont plutôt négatifs.

Il existerait alors des conditions propices pour que ces additifs puissent s’adsorber à la

surface de l’or sous forme d’un dépôt (par cémentation par exemple), activer les atomes

d’or (entre autres, en catalysant la réaction cathodique de réduction de l’oxygène) et

empêcher ainsi la formation de la couche de passivation d’AuCN.

17

Figure I-9 Courbe de polarisation montrant l’effet de faible concentration de plomb sur

l’oxydation anodique de l’or (Jeffrey & Ritchie, 2000)

La manifestation d’un pic de polarisation de l’or (Figure I.9) à la place d’un plateau de

courant de diffusion signifie que l’étape d’activation de l’or, par le dopant métallique

utilisé, est relayé par une étape de passivation associée, non pas à AuCN, mais à la

formation d’une couche d’additifs eux-mêmes. Ceci expliquerait les tendances

contradictoires rapportées dans la littérature et consolide l’hypothèse liée à l’existence de

conditions propices pour lesquelles la présence d’additifs est en faveur du comportement

anodique de l’or. Les ions plombeux (II) sont également largement utilisés pour réduire les

effets néfastes, sur la dissolution de l’or, dus à la forte réactivité de certains sulfures

métalliques dans les solutions de cyanure alcalines (Senanayake, 2008). Cet aspect sera

abordé en détail dans le cinquième chapitre.

I.2.3.7 Phénomènes de passivation & d’interactions Galvaniques

À l’état naturel, l’or se trouve sous forme métallique soit pur, soit sous forme d’alliages

avec d’autres métaux. Il est également largement dispersé dans les minerais aurifères qui

n’en contiennent que de très faibles quantités. Ces minerais sont classés en fonction des

composés qui sont associés à l’or. La présence de ces composés inflige un manque à gagner

18

important aux industries durant la lixiviation de l’or par cyanuration. L’effet inhibiteur de

ces composés est essentiellement lié à leur nature chimique. On distingue l’or associé à des

sulfures de fer (pyrite : FeS2, pyrrhotite : Fe1-xS), de l’or associé à des sulfures d’arsenic

(arsénopyrite : FeAsS) ou d’antimoine (stibine : Sb2S3) et de l’or associé à des sulfures de

cuivre (chalcopyrite : CuFeS2, chalcocite : CuS2).

Lorsque l’or et les sulfures métallique sont en contact, ces derniers pourront influencer sa

vitesse de lixiviation selon deux mécanismes :

1. Phénomènes de passivation qui résultent de la formation d’une couche protectrice à

la surface de l’or entravant sévèrement sa dissolution.

2. Phénomènes d’interactions galvaniques entre l’or et les sulfures métalliques,

électriquement conducteurs. Dans ce cas, la dissolution de l’or peut être soit

favorisée, soit ralentie, dépendamment des conditions de potentiels aux interfaces

Au-sulfures.

Il s’agit alors d’un éventail de processus redox complexes mis en jeu au cours de la

cyanuration dont l’ampleur, de l’effet négatif ou positif, est étroitement liée aux spécificités

du minerai à traiter.

Ces phénomènes de passivation et d’interactions galvaniques seront largement détaillés

dans la partie suivante.

I.3 Effet des sulfures métalliques sur la cyanuration de l’or (passivation &

Interactions galvaniques)

Étant donné que de larges proportions de l’or sont étroitement associées à d’autres sulfures

métalliques, l’effet de ces derniers sur l’extraction de l’or par cyanuration a stimulé l’intérêt

de nombreux chercheurs en hydrométallurgie de l’or. Des travaux de recherche sur la

dissolution de l’or par cyanuration ont montré que certains métaux, en particulier le cuivre,

le zinc et le fer, provenant de la dissolution de sulfures métalliques, peuvent se complexer

avec le cyanure pour former des cyanocomplexes, et ainsi ralentir la vitesse de dissolution

de l’or en réduisant la concentration des réactifs disponibles pour l’oxycyanolixiviation de

19

l’or (Habashi, 1967). Vu la forte réactivité des sulfures métalliques associés à l’or dans les

solutions alcalines de cyanure, des réactions secondaires impliquant le cyanure, les

hydrosulfures et les oxysulfures instables peuvent se produire. Il est généralement accepté

que ces réactions parasites nuisent aux performances du procédé de cyanuration via les

surconsommations de cyanure et d’oxygène engendrées. Cependant, la mise en œuvre de

cinétiques de dissolution de l’or en présence des anions sulfures (HS-) a permis de suggérer

une incidence directe de ces derniers sur la vitesse de lixiviation du métal. Weichselbaum et

al. (1989) ont montré que des traces de sulfures de sodium inhibent dramatiquement la

cyanuration de l’or. Cet effet serait imputable à la formation d’une couche protectrice

d’Au/Sx à la surface du métal visé. Des résultats similaires ont été trouvés par Lorenzen et

van Deventer (1992) ayant étudié l’effet de sulfures modèles, très réactifs, à base de cuivre,

de fer et de zinc sur la cinétique lixiviation de l’or. Dans le même sens, Jeffrey et Breuer

(2000) ont utilisé une électrode en alliage Au(96%)-Ag(4%) pour investiguer l’effet des

sulfures. Les résultats obtenus montrent que les anions sulfures sont fortement incriminés

dans la passivation de la surface de l’or. Une telle passivation pencherait en faveur d’un

blocage de la surface de l’or par un film résistif d’Au/Sx.

Les phénomènes d’interactions galvaniques entre l’or et les sulfures métalliques,

électriquement conducteurs, sont également soupçonnés d’influencer directement la vitesse

de lixiviation de l’or, bien qu’il n’existe pas un consensus dans la littérature sur leurs effets.

Lorenzen et van Deventer (1992) ont tenté de séparer l’effet des interactions galvaniques de

celui de la passivation de la surface l’or en conduisant des tests de cyanuration sur un

montage électrochimique à une et à deux cellules. Les résultats obtenus montrent que

lorsque l’électrode d’or et l’électrode minérale sont électriquement en contact,

l’accentuation du courant cathodique occasionne indirectement un manque à gagner sur la

vitesse de dissolution de l’or en raison du déplacement du potentiel mixte dans les régions

de potentiels où l’or est passif. Cependant, d’autres chercheurs (Aghamirian & Yen, 2005;

Dai & Jeffrey, 2006) rapportent plutôt des effets positifs sur la lixiviation de l’or de telles

interactions galvaniques.

Les tendances contradictoires rapportées dans la littérature sur l’incidence de nombreux

sulfures métalliques sur les profils de dissolution de l’or pourraient être expliquées, en

20

partie, par le fait que les tests de cyanuration n’ont pas été réalisés dans les mêmes

conditions.

Une meilleure évaluation des phénomènes de passivation et d’interactions galvaniques ne

serait alors possible qu’à travers une meilleure compréhension des comportements des

différents minéraux sulfureux dans les solutions de cyanures en fonction des conditions

physicochimiques du milieu.

I.3.1 Dissolution des minéraux sulfureux dans les solutions de cyanure

Plusieurs minéraux de métaux de base tels que le cuivre, le zinc et le fer, plus abondants

dans les minerais industriels, notamment sous forme de sulfures, peuvent se dissoudre dans

les cuves de cyanuration. Ces réactions indésirables sont responsables de surconsommation

importantes de cyanure et d’oxygène, potentiellement disponibles pour

l’oxycyanolixiviation de l’or.

Au contact des solutions de cyanure aérées, les cinétiques de lixiviation de minéraux

sulfureux sont très rapides, et le milieu réactionnel se charge en produits d’oxydation qui

sont généralement des ions métalliques, des complexes cyanométalliques, des oxydes de

métaux ainsi que d’autres espèces de soufre incluant des thiocyanates, des sulfures, des

polysulfures, des thiosulfates, des sulfites, des sulfates, etc.

La lixiviation des minéraux sulfureux, contenant des métaux bivalents, peut avoir lieu selon

la réaction globale suivante (Marsden & House, 2006) :

2MS + 2(x + 1)CN- + O2 +2H2O ↔ 2M(CN)x

(x-2) + 2SCN

- + 4OH

- (I.10)

Mis à part les minéraux sulfureux, la plupart des complexes de métaux de transition, à

savoir les hydroxydes, les sulfates et les carbonates, sont solubles dans les solutions

alcalines de cyanure (voir Tableaux I.2, I.3). Cependant, les réactions de décomposition de

ces derniers ont moins d’effet sur les profils de dissolution de l’or en raison d’une part, du

fait qu’ils consomment moins de cyanure par apport aux sulfures métalliques, et d’autre

part, de l’absence du soufre pouvant passiver la surface de l’or.

21

I.3.1.1 Minéraux de Cuivre

La présence du cuivre dans les minerais aurifères est très problématique à cause de la forte

réactivité de celui-ci dans les solutions de cyanure alcalines.

Tableau I-2 Solubilité de minéraux de cuivre dans une solution 0.1 M CN- (Hedley et

Tabachnick, 1958)

Pourcentage de cuivre total dissous

Minéral 23 ºC 45 ºC

Azurite 2CuCO3Cu(OH)2 94.5 100.0

Malachite CuCO3Cu(OH)2 90.2 100.0

Chalcocite Cu2S 90.2 100.0

Cuivre métallique Cu 90.0 100.0

Cuprite Cu2O 85.5 100.0

Bronite Cu5FeS4 70.0 100.0

Enargite Cu3AsS4 65.8 75.1

Tetrahedrite 4Cu2S Sb2S3 21.9 43.7

Chrysocolla CuSiO3 11.8 15.7

Chalcopyrite CuFeS2 5.6 8.2

La dissolution du cuivre est généralement indésirable pendant l’extraction de l’or par

cyanuration en raison, entre autres, de :

1. La surconsommation de cyanure et d’oxygène;

2. L’oxydation catalytique du cyanure en cyanate (Breuer & Jeffrey, 2005) par le

Cu(II);

3. Le ralentissement de la vitesse de lixiviation de l’or;

4. La contamination du produit final.

22

Figure I-10 Spéciation du cuivre et du cyanure en fonction de la concentration total du

cyanure dans une solution contenant 2mM de Cu(I) (Huang & Young, 1996)

La Figure I.10 montre la spéciation de 2 mM de cuivre (I) avec un balayage en

concentration de cyanure en solution. À partir de cette figure, on constate qu’à des

concentrations en cyanure inferieures à 6 mM, le cuivre coexiste essentiellement sous

forme de Cu(CN)2- et de Cu(CN)3

2- et que la proportion du cyanure libre en solution est très

faible. Une première analyse de cette figure nous permettrait aussi de suggérer qu’à des

rapports molaires cuivre cyanure 1:2, le cuivre existe essentiellement sous forme de

Cu(CN)2-, cependant à des rapports 1:3 la forme Cu(CN)3

2- est prédominante. Étant donné

que la dissolution de l’or requiert que du cyanure libre soit présent en solution,

l’oxycyanolixiviation de l’or devrait être conduite à des concentrations en cyanure trois fois

supérieures à la concentration théorique nécessaire pour la formation du complexe

tricyanocupreux à partir du cuivre contenu dans les minerais à traiter.

23

Figure I-11 Diagramme E-pH du système Cu-CN-H2O à 25ºC (Osseo-Assare, et al., 1984)

Comme le diagramme E-pH du système Cu-CN-H2O l’illustre (Figure I.11), les minéraux

de cuivre réagissent avec le cyanure pour former une variété de complexes

cyanométalliques, à savoir Cu(CN)2-, Cu(CN)3

2- et Cu(CN)4

3,- avec une prédominance du

complexe tricyanocupreux lors de la cyanuration de l’or. La coexistence de ces complexes

en solution est due à leurs constantes de stabilité qui sont comparables (Marsden & House,

2006). Les proportions de ces composés dans des conditions spécifiques de pH, de

température, de concentration de cyanure libre et de cuivre peuvent être calculées en se

basant sur leurs constantes de stabilité. Cependant, ces calculs ne peuvent être utilisés qu’à

titre indicatif seulement en raison de l’absence d’une méthode analytique fiable permettant

de corriger les constantes de stabilités tout en tenant compte des spécificités

physicochimiques du milieu réactionnel.

Breuer et al. (2005) ont rapporté que la lixiviation de l’or pourrait aussi avoir lieu à partir

des complexes Cu(CN)32-

et Cu(CN)43-

. Ces lixiviations ont été expliquées par la

décomplexation partielle de ces ions en Cu(CN)2- et le cyanure ainsi libéré est utilisé pour

la dissolution de l’or. Cependant, aucune étude expérimentale permettant de confirmer ce

mécanisme n’a été conduite jusqu’à nos jours et la spéciation du cuivre dans des conditions

reflétant la physicochimie des cuves de cyanuration industrielles reste très ambiguë en

24

raison, entre autres, de la forte dépendance des constantes de stabilités des complexes

Cu(CN)x1-x

à la force ionique de la solution (Lukey et al., 1999).

I.3.1.2 Minéraux de fer

La plupart des minerais d’or contiennent des sulfures de fer tels que la pyrite (FeS2), et

moins souvent la pyrrhotite (FeS) ou la marcasite (FeS2) (structure cristalline différente de

celle de la pyrite). Le fer métallique n’est que modérément soluble dans les solutions de

cyanure. Par contre certains sulfures de fer (Hedley et Tabachnick, 1958), à savoir la

pyrrhotite, la marcasite, l’arsénopyrite et la pyrite se décomposent dans les solutions de

cyanure alcalines pour former des cyanocomplexes de fer ainsi que plusieurs espèces de

soufre.

Figure I-12 Diagramme E-pH du système Fe-CN-S-H2O à 25ºC (Osseo-Assare et al., 1984)

À partir du digramme E-pH du système Fe-S-CN-H2O (Figure I.12), il est clair que dans les

conditions de dissolution de l’or, la forme Fe(CN)64-

est prédominante. Ceci nous

permettrait de présumer que les sulfures de fer influencent la cinétique de dissolution de

l’or via le même mécanisme que celui des sulfures de cuivre.

Selon Hedley et Tabachnick (1958), la réactivité des sulfures de fer dans une solution de

cyanure alcaline est dans l’ordre suivant :

25

Pyrrhotite >> marcasite >> arsénopyrite >> pyrite

Cette réactivité est toutefois contrôlée par plusieurs facteurs, comme la présence de défauts

dans le réseau cristallin du minéral, la présence d’impuretés et/ou d’autres sulfures

conducteurs (Cruz et al., 2005) qui sont électriquement en contact avec le minéral en

question.

L’oxydation de la pyrrhotite dans les conditions de lixiviation de l’or peut avoir lieu selon

la réaction suivante :

2FeS + 12CN- + 5O2 + 2H2O ↔ 2Fe(CN)6

4- + 2SO4

2- + 4OH

- (I.11)

D’après le classement mentionné précédemment, il semble que la pyrite est la moins

réactive par rapport aux autres sulfures de fer. Des études électrochimiques qui ont été

effectuées dans le même sens (Hamilton & Woods, 1984; Ahlberg & Broo, 1996) ont

montré que la pyrite est le sulfure de fer le plus noble en plus d’être un très bon catalyseur

pour la réaction de réduction de l’oxygène (Biegler et al., 1975). De ces résultats on

pourrait soupçonner le fait que la pyrite pourrait aussi influencer, indirectement, la réaction

de lixiviation de l’or en augmentant la réactivité, par contact galvanique, des autres sulfures

qui sont fortement incriminés dans la passivation de l’or.

I.3.1.3 Minéraux de zinc

Les minéraux de zinc ne posent généralement pas de problèmes majeurs lors de l’extraction

de l’or par cyanuration car ils ne sont que modérément solubles dans les solutions de

cyanure (Tableau I.3).

Tableau I-3 Solubilité de minéraux de zinc dans une solution de cyanure (Hedley &

Tabachnik, 1958)

Minéral Teneur en Zn (%)

Initial Final

Extraction en %

Sphalérite (ZnS) 1.36 1.11 18.4

Willemite (Zn2SiO4) 1.22 1.06 13.1

Hydrozinicite (3ZnCO3,2H2O) 1.36 0.87 35.1

Calamine (H2Zn2SiO5) 1.19 0.95 20.2

26

Zinicite (ZnO) 1.22 0.79 35.2

Smithsonite (ZnCO3) 1.22 0.73 40.2

La sphalérite est le sulfure de zinc le plus abondant dans les minerais d’or. Il réagit avec le

cyanure pour former le cyanocomplexe Zn(CN)42-

et l’ion sulfure S2-

(dépendamment des

conditions d’oxygénation du milieu) selon la réaction suivante (Hedley & Tabachnik,

1958):

ZnS + 4CN- → Zn(CN)4

2- + S

2- (I.12)

L’effet du zinc sur le procédé de cyanuration est limité en raison de la faible valeur de la

constante de stabilité du complexe Zn(CN)42-

(Log Ks = 19.62), comparativement à celle du

complexe dicyanoaurate (Log Ks = 39.3) (Senanayake, 2008). Des études ont montré que

les cyanocomplexes de zinc pourraient servir d’agent lixiviant de l’or dans des conditions

de très faibles concentrations en cyanure libre (Hedley & Tabachnik, 1968). Cependant, la

formation de ces complexes nuit fortement au contrôle du procédé d’extraction de l’or par

cyanuration à cause des interférences dues aux cyanocomplexes de zinc lors du dosage du

cyanure libre par titrimétrie au nitrate d’argent. En effet, le cation argent dans AgNO3, en

faisant précipiter le cyanure libre sous la forme de dicyanoargentate d’argent, déplace les

diverses réactions de décomplexation des complexes cyanométalliques présents en solution

avec des vitesses différentes qui sont, entre autres, déterminées par les valeurs numériques

des constantes de stabilité de ces complexes, e.g., pK(Zn(CN)42-

= 19.62, pK(Cu(CN)32-

=

21.66, pK(Fe(CN)63-

= 43.6 (Marsden et House, 2006). Un tel parasitage est beaucoup plus

prononcé pour le zinc en raison des constantes de stabilités très comparables entre

Zn(CN)42-

et le complexe Ag(CN)2- (Log Ks = 20.48) qui se forme pendant le dosage du

cyanure libre. Un nouvel équilibre thermodynamique s’établit ainsi, et les risques de

surévaluation de la concentration en cyanure libre sont presque inévitables.

I.3.1.4 Minéraux d’antimoine

La stibine est un minéral (appelé antimonite ou encore sulfure d’antimoine) dont la formule

Sb2S3. C’est le principal minéral de l’antimoine. La présence de ce minéral, fréquemment

associé à d’autres sulfures dans les gisements d’or, bien que non-systématique inflige un

27

manque à gagner très important aux industries durant la lixiviation de minerais aurifères par

cyanuration.

Guo et al. (2005) ont montré que lors de tests de lixiviation de minerais synthétiques

contenant de très faibles concentrations en stibine (0.002% w/w), les taux de récupération

de l’or ne dépassent pas 38% au lieu de 90% en absence de celle-ci. Cet effet inhibiteur est

accentué en augmentant la concentration en stibine dans le minerai; seulement 22% d’or

disponible a été récupéré en présence de 0.01% de stibine. Il existe un consensus dans la

littérature à l’effet que les faibles taux de récupération d’or enregistrés en présence de

stibine seraient dus à la formation d’une couche d’oxyde d’antimoine (probablement

Sb2O5) insoluble à la surface de l’or (Senanayake, 2008). Cette explication pourrait être

supportée par le fait que l’antimoine ne forme pas des complexes stables avec le cyanure

libre en solution (Marsden & House, 2006). De ce fait, cette passivation pourrait être le

résultat d’un phénomène de surface incluant l’oxydation de la stibine, la réduction de

l’oxygène et la précipitation d’oxyde d’antimoine à la surface de l’or. Pour décrire ce

mécanisme, Guo et al. (2005) proposent les réactions suivantes :

Sb2S3 + 8 OH- → 2SbO2

- + 3S

o + 4H2O + 6e

- (I.13)

2SbO2- +2OH

- → Sb2O5 + H2O + 4e

- (I.14)

La réaction (I.13) aurait lieu préférablement à la surface de l’or en raison de la forte

conductivité électrique de ce métal favorisant ainsi le transfert d’électrons à l’oxygène.

L’effet du pH sur la solubilité de la stibine a été étudié afin d’appréhender l’influence de la

dissolution de ce minéral sur la lixiviation de l’or. Il est montré que la cinétique de

dissolution de la stibine est fortement accélérée en augmentant le pH. À pH = 12, la vitesse

de dissolution est trop élevée par contre à pH = 10, elle est très ralentie (Hedley &

Tabachnick, 1958). En conséquence, l’effet inhibiteur de la stibine sur l’extraction de l’or

dépend à son tour du pH. Pour alléger l’effet inhibiteur des oxydes d’antimoine résultant de

la dissolution de la stibine, des tests de lixiviation en réacteur slurry ont été réalisés par Guo

et al. (2005) en présence de dopants métalliques, en l’occurrence le nitrate de plomb, à

différentes concentrations. Les résultats obtenus montrent que la combinaison de faibles

28

valeurs de pH (9.8 < pH < 10.2) et des concentrations en nitrates de plomb supérieures à

100g/t améliorent nettement les profils de dissolutions de l’or. Cependant les ions CN-

peuvent se transformer en HCN qui à son tour se dissipe par désorption à faibles valeurs de

pH entrainant une baisse de performance du procédé de cyanuration. La recherche de

conditions optimales, pour que ces additifs puissent accélérer la vitesse de lixiviation de

l’or tout en évitant aussi bien les pertes en cyanure que la formation d’une couche d’oxydes

d’additifs eux-mêmes à surface de l’or, devrait être menée en tenant compte de la

minéralogie des minerais alimentant les réacteurs de cyanuration.

I.3.2 Spéciation du soufre dans les solutions de cyanure et passivation

La place du soufre dans le traitement des minerais aurifères sulfureux est très particulière.

En effet, le soufre élémentaire qui peut être formé pendant l’étape de prétraitement de

minerais sulfureux et/ou durant sa lixiviation par cyanuration, réagit rapidement avec le

cyanure pour former les thiocyanates, les sulfates, ainsi que d’autres espèces ioniques

incluant, entre autres, les anions sulfures, les polysulfures et les thiosulfates dépendamment

de la concentration en cyanure et du degrés d’oxygénation du milieu.

Figure I-13 (a) produits d’oxydation des hydrosulfures dans une solution de cyanure aérée

(b) effet du pH sur la spéciation des polysulfures à 25 ºC (Senanayake, 2008)

Le soufre élémentaire pourrait également réagir avec les anions sulfures en milieu alcalin

(voir Tableau I.4) ce qui augmente la teneur en ions de soufre dissous dans les cuves de

(b) (a)

29

cyanuration via la formation d’une série de polysulfures et accentuerait la surconsommation

de cyanure via les réactions indésirables (Figure I.13 (b)) de formation de thiocyanate.

Tableau I-4 Constantes d’équilibres (K) (Marsden et House, 2006)

Réactions K

HS- + H

+ = H2S 10

7

2S + HS- + OH

- = S2S

2- + H2O 10

2.2

3S + HS- + OH

- = S3S

2- + H2O 10

3.9

4S + HS- + OH

- = S4S

2- + H2O 10

4.6

5S + HS- + OH

- = S5S

2- + H2O 10

4.6

6S + HS- + OH

- = S6S

2- +H2O 10

2.3

La présence de sulfures stables sous forme d’hydrosulfure HS- vers pH =10 (Figure I.13

(b)) est connue pour inhiber la dissolution de l’or dans les solutions de cyanure aérées.

Cette observation a été justifiée par une étude expérimentale réalisée par Jeffrey et Breuer

(2000) sur une électrode d’or (rotating electrochemical quartz crystal microbalance) avec

un balayage en concentration de sulfure de sodium à pH = 10. Cette étude a montré que

l’effet inhibiteur est décelable à partir d’une concentration de 0.01mM en ion HS-.

L’analyse des profils de dissolution montre que le déclin de la lixiviation de l’or s’accentue

avec l’augmentation de la concentration utilisée en sulfure de sodium. Certaines évidences

pencheraient en faveur d’un blocage de la surface d’or, sans que le mécanisme ne soit

suffisamment élucidé, par un film résistif de type Au/Sx (Weichselbaum et al., 1989;

Jeffrey & Breuer, 2000) au détriment du complexe Au(CN)2-. Un tel film empêcherait

davantage l’accès au cyanure à la surface active et ferait basculer la cyanuration d’un

contrôle diffusionnel externe par CN- vers un contrôle diffusionnel à travers le film de

sulfure, plus sévère. Auquel cas, l’augmentation de la concentration en cyanure pourrait

être susceptible de compenser plus au moins le déclin de la lixiviation de l’or. Ceci serait

dû à une remise en solution des dépôts de sulfures à la surface de l’or sous forme de

30

thiocyanate (voir réaction I.15) et à la formation du complexe Au(CN)2- au détriment de la

couche passivante d’Au/Sx (Jeffrey & Breuer, 2000).

S + CN- SCN

- (I.15)

Cependant, pour une meilleure compréhension de l’effet de sulfures sur la lixiviation de

l’or et afin d’identifier les espèces soufrées incriminées dans la passivation, il serait crucial

de compléter ces études par des études cinétiques de lixiviation de l’or en présence des

anions sulfures (HS-) à différentes concentrations d’oxygène dissous, de cyanure libre et

dans une gamme de pH entre 10 et 12 pour avoir des conditions représentatives des

solutions industrielles où le soufre s’oxyde graduellement en sulfate. Parallèlement à ces

études, le dépistage de la surface de l’or, par les techniques de caractérisation de surface,

serait très utile pour la mise en évidence de la passivation ainsi que la détermination de la

nature chimique du film passivant.

I.3.3 Phénomènes de passivation et d’interactions galvaniques

Lorsque l’or et les sulfures métalliques, électriquement conducteurs, sont mis en contact,

une pile galvanique s’établit et la différence de potentiel entre les deux matériaux indique le

sens de la corrosion. Un des deux matériaux (le moins noble) s’oxyde et se dissout (anode),

tandis que sur l’autre (le plus noble) a lieu la réduction d’oxygène (cathode). Les

interactions galvaniques peuvent alors empêcher la dissolution de l’or si le potentiel des

sulfures métalliques est inférieur à celui de l’or, favorisant la réduction de l’oxygène au lieu

de la dissolution du métal. Idéalement, lorsque le potentiel de ce dernier est inférieur à celui

des sulfures métalliques, les interfaces Au-sulfures représentent autant de lieu propices aux

interactions galvaniques en faveur d’un comportement anodique de l’or est donc conduisant

à sa dissolution accrue. Étant donné que l’or est connu pour développer dans une solution

de cyanure un potentiel moindre que certains sulfures métalliques, e.g. pyrite, pyrrhotite et

chalcopyrite (Aghamirian & Yen, 2005; Dai & Jeffrey, 2006), avec lesquels il est

naturellement en contact, un accroissement de la vitesse de l’oxycyanolixiviation de l’or

résultant de la corrosion galvanique est attendu. Cependant, il n’existe pas un large

consensus autour des hypothèses fondamentales qui sous-tendent cette tendance à la

lumière des résultats contradictoires rapportés dans la littérature. Filmer (1982) et Paul

31

(1984), ont mis en évidence l’effet inhibiteur du contact galvanique direct, Au-pyrite, sur le

comportement anodique de l’or. Ils suggèrent que la réduction de l’oxygène pourrait avoir

lieu sur toute la surface des 2 matériaux (Au + pyrite). En interférant de la sorte avec la

réduction de l’oxygène, l’accentuation du courant cathodique occasionne indirectement un

ralentissement de la vitesse de lixiviation de l’or en raison du déplacement du potentiel

mixte dans la zone de passivation d’Au.

Lorenzen et van Deventer (1992) ont étudié l’effet des interactions galvaniques entre l’or et

les sulfures métalliques sur la vitesse de lixiviation de l’or. Pour se faire, des couples

galvaniques macroscopiques ont été formés an associant les électrodes d’or et minérales par

paires. Les résultats de cette étude ont montré que la mise en contact direct de l’or avec la

pyrite, la pyrrhotite et la chalcopyrite occasionne un déclin significatif des profils de

dissolution du métal d’intérêt. L’action inhibitrice de la chalcopyrite est particulière (50%

d’or seulement a été récupéré par rapport à l’électrode d’or toute seule). Ils ont également

suggéré que les ions en solution provenant de la dissolution des sulfures métalliques

entravent sévèrement la lixiviation de l’or à travers la formation d’une couche passivante à

la surface du métal. Cependant, l’étude n’a pas rapporté de suivi des potentiels et des

courants galvaniques parallèlement à la vitesse de lixiviation de l’or ainsi que les courbes

de polarisation anodiques et cathodiques des électrodes d’or et minérales, bien que les

potentiels et les courants de corrosions aient été rapportés dans certains cas. Ceci rend très

difficile l’interprétation des mesures potentiométriques en raison de l’absence d’une

relation directe entre le potentiel et le courant de corrosion (Mills, 1959). De plus,

l’évaluation de l’importance relative des phénomènes de passivation et d’interactions

galvaniques dans cette étude a été faite par rapport à l’électrode d’or toute seule en ignorant

tout effet d’interaction entre ces deux phénomènes sur la vitesse de dissolution de l’or. Ce

qui n’est pas nécessairement vrai.

Aghamirian & Yen (2005) ont également tenté d’évaluer l’effet des phénomènes

d’interactions galvaniques entre l’or et les sulfures métalliques sur la cinétique de

dissolution de l’or en formant des couples galvaniques dans une seule cellule

électrochimique. Les résultats de cette étude montrent que les interactions Au-galène, Au-

pyrite et Au-pyrrhotite sont en faveur du comportement anodique de l’or tandis que celles

32

Au-chalcopyrite et Au-chalcocite occasionnent un manque à gagner sur la vitesse de

dissolution de l’or. Les potentiels en circuit ouvert dans une solution aérée de cyanure à

10mM de l’or, de la chalcopyrite, de la pyrite, de la pyrrhotite, de la galène et de la

chalcocite sont respectivement -450mV, -264mV, -108mV, -224mV, -85mV et -687mV

(Aghamirian & Yen, 2005). Une première analyse de ces potentiels permettrait de suggérer

que seule la chalcocite pourrait être à l’origine d’une pile électrochimique qui impose un

sens de parcours aux électrons pouvant entraver la réaction de corrosion de l’or, la

chalcocite se dissolvant (elle est sacrifiée) et l’électrode d’or restant pratiquement intact

(Figure I.14). Les courants galvaniques les plus élevés qui ont été enregistrés dans le cas

des couples Au-pyrite et Au-pyrrhotite ainsi que les activités élevées de ces deux sulfures

vis-à-vis de la réduction d’oxygène expliqueraient leurs effets promoteurs sur la lixiviation

de l’or. Dans le cas de la chalcopyrite et de la galène, des faibles courants galvaniques ont

été enregistrés. Ceci serait attribué à la faible différence de potentiels entre l’or et la

chalcopyrite tandis que la faible activité de la galène vis-à-vis de l’oxygène serait

responsable de l’atténuation du courant cathodique et par conséquent de celui du couple

galvanique.

Les tendances contradictoires enregistrées entre courants et potentiels galvaniques dans le

cas de la chalcopyrite et de la galène, par rapport aux autres sulfures, permet de conclure

que la différence de potentiels indique le sens des réactions mais pas leur ampleur. La

solubilité du sulfure métallique en question est aussi un facteur potentiellement très influant

à prendre en compte. Ceci n’est surprenant à la lumière des scénarios de passivation ou

d’activation de l’électrode d’or qui pourrait résulter de la dissolution non-électrochimique

de certains sulfures métalliques. La dissolution non-oxydative de la chalcocite (Fisher,

1994) en est un exemple :

Cu2S + 8CN- → 2Cu(CN)4

3- +S

2- (I.16)

33

Figure I-14 Effets des interactions galvaniques Au-Sulfures sur la dissolution de l’or, pH=

10.5, [CN-] =10mM, T=25ºC (Aghamirian & Yen, 2005)

Dans l’étude réalisée par Aghamirian et Yen (2005), les effets observés sur les profils de

dissolution de l’or résultent plutôt de l’action conjuguée des interactions galvaniques et de

la passivation, induite sur l’électrode d’or et/ou minérale par la dissolution du sulfure

associé à l’électrode d’or, et pas seulement des interactions galvaniques Au-sulfure. Le

découplage de ces deux phénomènes ne serait possible qu’à travers une étude

complémentaire dans un montage électrochimique à deux cellules. Dans ce cas, les effets

dus aux interactions galvaniques et les passivations induites devraient être évalués,

relativement, lorsque les deux électrodes sont plongées dans la même cellule, puis,

lorsqu’elles sont physiquement séparées.

34

Figure I-15 (a) Courbes de polarisation anodiques de l’électrode d’or et minérales (b) Effet

du contact galvanique Au-Chalcocite sur la lixiviation de l’or et potentiel mixte, [CN-]

=10mM (Dai & Jeffrey, 2006)

Dai et Jeffrey (2006) ont également réalisé une étude systématique portant sur l’effet

galvanique de certains sulfures métalliques (pyrite, chalcopyrite, pyrrhotite, galène,

chalcocite, etc.) sur la cinétique de lixiviation de l’or à partir d’une électrode en alliage

Au(95%)-Ag(5%). Les résultats de cette étude montrent que tous les sulfures étudiés

provoquent une sur-lixiviation de l’or lorsqu’ils sont en contact direct avec ce dernier. Ces

résultats ne corroborent pas ceux trouvés par Lorenzen et van Deventer (1992). Ces

derniers suggèrent que l’accentuation du courant cathodique, suite à la plus grande surface

disponible pour la réduction de l’oxygène (Au + sulfure), engendre un ralentissement de la

vitesse de lixiviation de l’or en raison du déplacement du potentiel de corrosion dans la

zone de passivation de l’or. Cependant, la courbe de polarisation de l’électrode d’or (en

alliage Au-Ag(95%-5%)) réalisé par Dai et Jeffrey (2006) ne montre aucune région de

passivation (Figure I.15 (a)), et l’augmentation du potentiel mixte ne pourrait dans ce cas

qu’accentuer la vitesse de lixiviation de l’or. Les différences entre les résultats de ces deux

études pourraient être expliquées par le fait que Lorenzen et Van Deventer (1992) ont

utilisé une électrode d’or pur. Cette dernière est connue pour être très vulnérable à la

passivation (Jeffrey & Ritchie, 2000).

Une analyse comparative des résultats obtenus par Dai et Jeffrey (2006) et ceux obtenus par

Aghamirian et Yen (2005), (Figures I.14 et I.15), nous permet de mettre en relief les

(a) (b)

35

tendances contradictoires rapportés dans le cas de la chalcocite. Contrairement à la

cinétique de dissolution de l’or pratiquement nulle dans le cas de la Figure I.14, le contact

galvanique Au-chalcocite (Figure I.15) occasionne une nette amélioration de la vitesse de

lixiviation de l’or. Ceci pourrait être expliqué en partie par le fait que l’expérience

présentée sur la Figure I.15 a été effectuée pendant une courte durée (3 min) où le

mécanisme de passivation de la surface de l’or par dissolution de la chalcocite est resté très

limité.

La majorité des études ayant été réalisées jusqu'à présent ont porté, seulement, sur la

corrosion de l’or, soit seul (cible monométallique) soit dans le cas bimétallique (Au-Ag,

Au-Cu), soit en contact galvanique avec un seul sulfure métallique. Cependant, l’influence

de l’association galvanique entre les différents sulfures, au sein des minerais aurifères

sulfureux, sur leurs réactivités dans des solutions de cyanure n’a pas été prise en

considération. Par conséquent, l’étude de l’effet des contacts galvaniques sulfures-sulfures

sur la cinétique de dissolution de l’or serait cruciale pour une compréhension intégrée des

phénomènes de passivation et d’interactions galvaniques. Des études réalisées par Cruz et

al. (2005) ont monté que la réactivité de la pyrite, dans une solution de nitrate de sodium

0.1M, est fortement influencée par la présence de traces de sphalérite et/ou de galène dans

la matrice pyritique.

La problématique de dissolution de l’or en présence de sulfures métalliques est très

complexe, en raison de l’interaction d’un grand nombre de facteurs : (i) passivation de la

surface de l’or due à la formation d’une couche protectrice de type Au2S, Ag2S ou d’un

dépôt de soufre, (ii) formation de couches résistives d’oxydes et/ou d’hydroxydes de

métaux, résultants de la dissolution des sulfures, à la surface des particules d’or et

minérales, (iii) oxydation des minéraux sulfureux et surconsommation de cyanure et

d’oxygène via les réactions parasites de formation de complexes cyanométalliques, de

thiocyanate et des oxysulfures et (iv) phénomène d’interactions galvaniques. Très souvent,

l’intrication entre ces différents facteurs contraint à déployer une stratégie ad hoc devant

tenir compte de la spécificité des minerais à traiter afin de maximiser la récupération de

l’or.

36

I.4 Objectifs du projet

Bien que plusieurs études aient porté sur les phénomènes de passivation et d’interactions

galvaniques en présence de sulfures métalliques, celles-ci suscitent, néanmoins, quelques

interrogations. Premièrement, peu d’études ont évalué l’importance relative sur la

dissolution de l’or des phénomènes de passivation et d’interactions galvaniques

corollairement, avec des études cinétiques détaillées d’oxydation des dits sulfures

métalliques. Deuxièmement, la plupart des études ont porté sur l’effet des interactions

galvaniques Au-sulfure, sans tenir compte de l’effet des interactions galvaniques sulfure-

sulfure sur la réactivité de ces derniers, et par conséquent sur la lixiviation de l’or.

Troisièmement, aucune étude de corrosion de l’or n’a été faite sur des cibles minérales

reflétant de façon réaliste la topochimie de surface des minerais industriels en terme de

contacts galvaniques permanents à l’échelle intra-particulaire ainsi que de rapports de

surfaces cathode/anode.

Le présent travail de thèse vise les objectifs suivant :

1. Évaluer l’importance relative sur la dissolution de l’or : (i) des interactions

galvaniques entre l’or et les sulfures métalliques, ii) des interactions galvaniques

permanentes entre les différents sulfures, iii) de la passivation induite sur la surface

de l’or par les réactions indésirables de dissolution des sulfures métalliques. Et

parallèlement à cet objectif, contribuer à une compréhension fiable et plus fine des

aspects fondamentaux de la cyanuration via un suivi méticuleux des cinétiques de

lixiviation de minéraux sulfureux dans les solutions de cyanure alcalines.

2. Rechercher des moyens de réduction de la forte réactivité des sulfures métalliques

et/ou de prévention de la passivation de l’or durant la cyanuration à travers des

études cinétiques et électrochimiques originales.

Au deuxième chapitre de ce mémoire, des études électrochimiques et cinétiques sont

mises en œuvre sur un montage de type Rotating Disk Electrode (RDE). Ce dernier est

constitué de cellules munies d’électrodes tournantes permettant, en sus d’un suivi en ligne

des concentrations d’espèces mises en jeu, de mesurer les réponses électrochimiques des

37

électrodes Au-Ag et minérales étudiées. Pour tenter d’apporter des indices quant au

découplage et au chiffrage de la contribution des phénomènes de passivations versus celles

reliées aux phénomènes d’interactions galvaniques, l’électrode d’Au-Ag a été connectée à

l’électrode minérale (préparée à partir d’un minerai industriel) par une jonction sans

résistance (ZRA) permettant de mesurer, courant et potentiel galvaniques en fonction du

temps. Les effets dus aux interactions galvaniques et à la passivation sont comparés lorsque

les deux électrodes sont plongées dans la même cellule, puis, lorsqu’elles sont

physiquement séparées par immersion dans deux cellules distinctes. La cinétique de

dissolution de l’or est suivie par absorption atomique simultanément avec les signaux de

potentiel et courant galvaniques. Dans le but de comprendre le comportement du minerai

industriel étudié, une étude systématique des effets des sulfures majeurs entrant dans sa

composition a été entreprise. Plus spécifiquement, nous allons présenter des mesures de

cinétique de lixiviation de l’or à partir de l’électrode Au-Ag que nous avons plongée dans

une série de pulpes formés successivement par les sulfures majeurs du dit minerai

industriel, pris séparément. Afin de rechercher des moyens d’atténuation de l’effet

inhibiteur des sulfures majoritaires présents dans le minerai industriel, sur la vitesse de

lixiviation de l’or, des tests de pré-oxydation de ce dernier ainsi que de ses constituants ont

été réalisés en réacteur slurry.

Au troisième chapitre, pour répondre à certaines questions ouvertes sur la dissolution de

l’or en réacteur slurry et en couples galvaniques macroscopiques, il a été décidé

d’investiguer les effets dus aux interactions galvaniques permanentes Au-sulfure, et aux

phénomènes de passivation dans un nouveau réacteur électrochimique à lit fixe. Dans ce

dernier, les minerais (sulfures métalliques commerciaux et industriels) et de la poudre d’or

sont disposés en lit fixe sous forme de mélange alimenté par une solution de cyanure aérée

recirculée en boucle fermée à l’aide d’une pompe péristaltique. Contrairement aux couples

galvaniques utilisés précédemment en mode RDE, la disposition de la poudre d’or et des

sulfures métalliques de manière à refléter de façon réaliste la topochimie de surface des

minerais industriels en termes de contacts galvaniques permanents à l’échelle intra-

particulaire a été possible. Ceci nous a permis de découpler les phénomènes de passivation

des interactions galvaniques et d’évaluer leur importance relative sur la dissolution de l’or

et des sulfures, tout en utilisant des rapports de surfaces cathode/anode plus réalistes, dont

38

la mise en œuvre a été impossible que ce soit en couples galvaniques macroscopiques ou

bien en réacteur slurry.

Dans le quatrième chapitre de ce travail, l’application du réacteur électrochimique à lit

fixe en cyanuration a été étendue pour tenir compte, en plus des interactions galvaniques

Au-sulfure, de l’effet des associations galvaniques sulfure-sulfure sur la réactivité de ces

derniers est par conséquent sur la lixiviation de l’or. Tout en ayant l’avantage de contrôler

les associations galvaniques, au moyen du réacteur à lit fixe, entre les poudres de métaux

précieux (Au & Ag) et celles de sulfures métalliques à l’échelle inter-particulaire, nous

étions capable de démontrer que le comportement électrochimique des différents minéraux

est fortement influencé par leur statut minéralogique dans les systèmes qui les renferment.

Ainsi, la nature de la phase minérale associée à l’Au a été démontrée pour être, souvent, le

principal facteur contrôlant l’oxycyalixiviation de l’or dans les systèmes multi-minéraux

synthétiques étudiés. L’effet du rapport de surface anode/cathode sur la lixiviation de l’Au

a également été pris en considération dans ce chapitre.

Le cinquième et dernier chapitre est dédié à la recherche de moyens d’atténuation de

l’effet inhibiteur des sulfures métalliques sur la vitesse de lixiviation de l’or. Les résultats

trouvés au niveau du deuxième, du troisième et du quatrième chapitre, nous ont permis de

démontrer que le comportement des différents minéraux sulfureux (étudiés séparément)

dans les solutions de cyanure et leur incidence inhibitrice sur la vitesse de lixiviation de

l’or, ne permettaient pas une explication aisée aussi bien du comportement du minerai

mélange natif (minerai industriel) que de celui de plusieurs systèmes synthétiques multi-

minéraux, constitués des mêmes sulfures. Par conséquent, les mesures de prévention de la

passivation investiguées dans le présent chapitre ont été mises en œuvre sur le réacteur

électrochimique à lit fixe. Premièrement, pour ne pas choisir des cations exotiques, l’effet

de la galène, utilisé comme source des ions Pb (II), sur la vitesse de lixiviation de l’or en

présence d’une large gamme de minéraux sulfureux a été exploré aussi bien en présence

qu’en absences des interactions galvaniques galène-sulfure. Deuxièmement, des tests de

pré-oxydation ont été menés en tenant compte des interactions galvaniques Au-sulfure et

sulfure-sulfure à l’échelle inter-particulaire. Une combinaison d’études électrochimiques et

de spéciation chimique des produits de la dissolution des sulfures a été utilisée pour

39

contribuer à une interprétation plus étayée des mécanismes mises en jeu lors des différentes

tentatives d’amélioration de la dissolution de l’or.

Étant donné qu’ une meilleure compréhension des comportements des différents minéraux

sulfureux dans les solutions de cyanure est capitale pour une contribution effective à

l’explication des mécanismes d’extraction de l’or par cyanuration, un protocole d’analyse

par électrophorèse capillaire permettant la séparation et la quantification des principaux

ions résultant de la dissolution des minerais aurifères sulfureux durant la lixiviation de l’or

par cyanuration a été développé. La séparation et la quantification de 11 ions en,

l’occurrence : S2O32-

, Cu(CN)32-

, Fe(CN)64-

, Fe(CN)63-

, SCN-, Au(CN)2

-, Ag(CN)2

-, SO4

2-,

SO32-

, HS- et OCN

- a été possible au moyen d’un protocole indirect (UV-Visible). Pour

plus de détails sur cette étude, voir article publié dans Journal of Separation Science

présenté en annexe A.

40

I.5 Bibliographie

Aghamirian, M.M., Yen, W.T., 2005. Mechanisms of galvanic interactions between gold

and sulfide minerals in cyanide solution. Miner. Eng. 18, 393-407.

Ahlberg, E., Broo, A.E., 1996. Oxygen reduction at sulphide minerals. Int. J. Miner.

Process. 46, 73-89.

Bard, A.J., 1973. Encyclopedia of Electrochemistry of the Elements. Marcel Dekker, New

York.

Biegler, T., Rand, D.A.J., Woods, R., 1975. Oxygen reduction on sulphide minerals. Part 1.

Kinetics and mechanism at rotated pyrite electrodes. J. Electroanal. Chem. 60, 151-162.

Bodlander, G.,1896. Die Chemie des Cyanideverfahrens. Z. Angew. Chem. 9, 583-587.

Cathro, K.J., and Koch, D.F.A., 1964. The anodic dissolution of gold in cyanide solutions.

Journal of Electrochemical Society 111: 1416-1420.

Chandler, B.D., Alexander, B., Schabel, A.B., Pignolet, L.H., 2000. Preparation and

Characterization of Supported Bimetallic Pt–Au and Pt–Cu Catalysts from Bimetallic

Molecular Precursors. Journal of Catalysis. 193, 186-198.

Cruz, R., Luna-Sanchez, R.M., Lapidus, G.T., Gonzalez, I., Monroy, M., 2005. An

experimental strategy to determine galvanic interactions affecting the reactivity of sulfide

mineral concentrates. Hydrometallurgy 78, 198-208

Dai, X., Jeffrey, M.I., 2006. The effect of sulfide minerals on the leaching of gold in

aerated cyanide solutions. Hydrometallurgy 82, 118-125.

Deschênes, G., Lastra, R., Brown, J.R., Jin, S., May, O., Ghali, E., 2000. Effect of lead

nitrate on cyanidation of gold ores: progress on the study of the mechanisms. Miner. Eng.

13, 1263-1279.

Elsner, L., 1846. Über Das Verhallen Verschiedener Metalle İn Einer Wassrigen Lôsung

Von Cyanakalium, J. Prakt. Chem. 37, 441-446

Ficher, W.W., 1994. Comparison of chalcocite dissolution in the sulfate, perchlorate

nitrate, chloride, ammonia, and cyanide system. Miner. Eng. 7, 99-103.

Filmer, A., 1982. The dissolution of gold from roasted pyrite concentrations. J. S. Afr. Inst.

(March), 90-94.

Guo, H., Deschênes, G., Pratt, A., Fulton, M., Lastra, R., 2005. Leaching kinetics and

mechanisms of surface reactions during cyanidation of gold in the presence of pyrite and

stibnite. Miner. Metall. Process. 22, 89-95.

Gavin, M.M., 2007. Global trends in gold mining: Towards quantifying environmental and

resource sustainability. Resources Policy 32, 42–56

Habashi, F. 1967. Kinetics and mechanism of gold and silver dissolution in cyanide

solution. Montana Bureau of Mines and Geology Bulletin. vol. 59.

Habashi, F., 1966. The theory of cyanidation. Transactions of the Mineralogical Society of

AIME 235:236-239.

41

Hamilton, I.C., Woods, R., 1984. A volumetric study of the surface oxidation of sulfide

minerals. In: Richardson, P.E. et al., (Eds.), Proceedings of Int. Sym. On Electrochemistry

in Mineral and Metal Processing. Electrochemical Society Inc., pp. 259-285.

Hedley, N., Tabachnick. H., 1958. Chemistry of cyanidation. Mineral Dressing Note 23.

New York: American Cyanamid Company.

Heath, A.R. and Rumball, J.A., 1998. Optimising cyanide : oxygen ratios in gold CIP/CIL

circuits. Miner. Eng. 11, pp. 999-1010

Huang, H.H., Young, C.A., 1996. Mass-balanced calculations of Eh-pH diagrams using

STABCAL. In: Woods, R., Richardson, P., Doyle, F.M. (Eds), Electrochemistry in

minerals and Metals Processing IV. The Electrochemical Society, Pennington, NJ, pp. 227-

238.

Jeffrey, M.I., Breuer, P.L., 2000. The cyanide leaching of gold in solutions containing

sulfide. Miner. Eng. 13, 1097-1106.

Jeffrey, M.I., Ritchie, I.M., 2000. The leaching of gold in cyanide solutions in the presence

of presence of impurities: I. The effect of silver. J. Electrochem. Soc. 147, 3272-3276.

Jeffrey, M., 1997. PhD thesis: A kinetic and electrochemical study of the dissolution of

gold in aerated cyanide solutions: the role of solid and solution phase purity. Curtin

University of Technology, Western Australia.

Ling, P., Papangelakis, V.G, Argyropoulos, S.A., Kondos, P.D., 1996. An Improved Rate

Equation for Cyanidation of a Gold Ore. Can. Metal. Quaterly. pp. 225-234.

Lorenzen, L., van Deventer, J.S.J., 1992. Electrochemical interactions between gold and its

associated minerals during cyanidation. Hydrometallurgy 30, 177-194.

Lukey, G.C., van Deventer, J.S.J., Huntington, S.T., Chowdhury, R.L., Shallcross. D.C.,

1999. Raman study on the speciation of copper cyanide complexes in highly saline

solutions. Hydrometallurgy 53 (3), 233-244.

MacArthur, J.S., Forrest, R.W., Forrest, W., 1888. Improvements in obtaining gold and

silver from ores and other compounds. British Patent 14174

Marsden, J.O., House, C.I., 2006. The Chemistry of Gold Extraction. 2nd Edition. Society

for Mining, Metallurgy, and Exploration (SME), Littleton. CO, USA.

Medley, C.D., Smith, J.E., Tang, Z., Wu, Y., Bamrungsap, S., Tan, W., 2008. Gold

nanoparticle-based colorimetric assay for the direct detection of cancerous cells. Anal.

Chem. 4, 1067-72

Maxey, A., Gonnella, P., Ball, Y., Herkenhoff, P., 1997. Gold. In Register of Australian

Mining 1996/1997, Ed. Louthean, R., Resource Information Unit Ltd. Bassendean, WA.

Nicol, M., Fleming, C. and Paul, R., 1987. The chemistry of the extraction of gold, in The

Extractive metallurgy of gold, vol. 2, Ed. Stanley, G.G., South African Institute of Mining

and Metallurgy, Johannesburg, S. Africa, pp. 831-905.

Nicol, M.J., 1980. The anodic behaviour of gold, Gold Bulletin. Vol. 13, pp. 105-111

42

Osseo-Assare, K., Xue, T. and Ciminelli, V.S.T., 1984. Solution chemistry of cyanide

leaching systems, in Precious Metals: Mining Extraction and Processing, The Metallurgical

Society/AIME., Warrendale, Pennsylvania, pp. 173-197.

Paul, R.L., 1984. The role of electrochemistry in the extraction of gold. J. Electroanal.

Chem. 168, 147-162.

Senanayake, G., 2008. A review of effects of silver, lead, sulfide, and carbonaceous matter

on gold cyanidation and mechanistic interpretation. Hydrometallurgy. 90, 46-73.

Wang, X. and Forssberg, K.S.E., 1990. The chemistry of cyanide-metal complexes in

relation to hydrometallurgical processes of precious metals, Mineral Processing and

Extractive Metallurgy Review, vol. 6, pp. 81-125.

Weichselbaum, J., Tumilty, J.A., Schmidt, C.G., 1989. The effects of sulfide and lead on

the rate of gold cyanidation. Proceedings Aus. I.M.M. Annual conference Perth/Kalgoorlie.

Australasian Institute of Mining and metallurgy, Melbourne, pp. 221-224.

Zhang, H.G., 1997. PhD thesis: some aspects of the use of thiourea in gold processing,

Murdoch University, Murdoch. WA.

43

CHAPITRE II. Electrochemical behavior of gold

cyanidation in the presence of a sulfide-rich industrial ore

versus its major constitutive sulfide minerals

Abdelaaziz Azizi, Catalin Florin Petre,1 Caroline Olsen,

1 Faïçal Larachi

Department of Chemical Engineering, Laval University, Québec, Canada, G1V 0A6

1- COREM Research Center, 1180 Rue de la Minéralogie, Québec, Canada, G1N 1X7

Abstract/Résumé

A detailed study on the relative importance of passivation phenomena and galvanic

interactions during gold cyanidation was carried out. Mineral disc electrodes consisting of a

sulfide-rich industrial ore and major sulfide components therefrom were prepared along

with an Au electrode (gold/silver alloy) in use for gold leaching rate tests. These leaching

tests conducted by hyphenating gold and mineral disc electrodes conjointly in one

electrochemical cell or in two separate electrochemical cells objectified both passivation-

induced setbacks as well as boosts by Au/mineral galvanic interactions on gold dissolution.

To decipher the role of sulfide ores on gold cyanidation, a systematic study was performed

by monitoring the leaching behavior of an Au disc electrode successively immersed in

slurries of industrial ore and its major sulfide constituents, i.e., pyrite, sphalerite and

chalcopyrite. The tested mineral constituents and ore exhibited an inhibiting effect on gold

leaching, decreasing in the following order: chalcopyrite > sphalerite > industrial ore >

pyrite. Pre-oxidation of the industrial sulfide ore prior to cyanidation improved the gold

leaching rate. However, in spite of noticeable reductions in cyanide consumption, no

beneficial effect of pre-oxidation on gold leaching was observed for the major sulfide (ore)

constituents when tested separately. Although cooperating permanent galvanic interactions

between gold and main constitutive minerals in the industrial ore prompted higher gold

leaching rates, predictability of the latter from lab-controlled leach tests of the nearly pure

constitutive sulfide minerals still remain premature.

44

Une étude détaillée sur l’importance relative des phénomènes de passivation et

d’interactions galvaniques lors de l’extraction de l’or par cyanuration a été menée. Des

électrodes d’or (en alliage Au/Ag) et minérales (un minerai industriel et ses constituants

majeurs) ont été préparées et utilisées pour les expériences de cyanuration. Les tests de

lixiviation réalisés en appariant les électrodes d’or et minérales dans un banc

électrochimique à une et à deux cellules séparées ont montré que les interactions

galvaniques entre l’or et les sulfures étudiés présentent globalement un effet positif sur la

vitesse de dissolution de l’or. Cependant, la passivation de ce dernier inflige un manque à

gagner sur sa dissolution. Pour déchiffrer le rôle de minerais sulfureux sur la lixiviation de

l’or, une étude systématique a été conduites en suivant la cinétique de lixiviation d’une

électrode d’or plongée successivement dans une série de slurries préparés à partir du

minerai industriel étudié et de ses constituants majeurs. Les résultats obtenus montrent que

tous les minerais testés présentent un effet négatif sur la vitesse de lixiviation de l’or. La

gradation de la répression de la plus sévère vers la moins sévère suit l’ordre suivant :

chalcopyrite > sphalérite > minerai industriel > pyrite. La pré-oxydation du minerai

industriel avant sa cyanuration améliore nettement la vitesse de dissolution de l’or.

Cependant, malgré une réduction notable sur la consommation du cyanure, la pré-oxydation

des sulfures métalliques majeurs, pris isolément, ne semble pas améliorer la lixiviation de

l’or durant l’étape de cyanuration. Bien que les interactions galvaniques permanentes entre

l’or et les constituants majeurs dans le minerai industriel favorisent la vitesse de

cyanuration de l’or, la prédiction de cette dernière à partir de tests contrôlés réalisés au

laboratoire sur les sulfures constitutifs reste encore prématurée.

45

II.1. Introduction

Since the discovery of gold dissolution in cyanide solutions in 1783 by Carl Wilhelm

Scheele (Marsden and House, 2006), myriad studies followed to elucidate the mechanisms

in play during this reaction. Elsner (1846) was the first to shed light on the mechanism of

gold dissolution in aerated cyanide solutions while cyanidation evolved coevally into a

commercial process after a patent filed by MacArthur et al. (1887). Nowadays, with

virtually depleted gold-bearing oxide deposits around the globe, gold extraction from less

friendly sulfide deposits is stretching the limits of cyanidation into uncharted territories.

Unlike oxides, which divert marginal quantities of cyanide and have generally minor

impact on gold leaching (Marsden and House, 2006), most of the corresponding metal

sulfides display a range of reactivity in alkaline gold leaching cyanidation (Marsden and

House, 1992; 2006). Due to poor cyanide selectivity for gold over the enclosing sulfide

mineral matrix, such parasitic reactions inflict utility extra costs and profit losses resulting

from high levels of cyanide consumption and/or some side-reaction products which can

directly reduce the gold leaching rate. Consequently, gold processing has become more

complex and early mechanisms, as described by Elsner in simple cyanide solutions, cannot

be relied upon to anticipate the behavior of gold dissolution in the presence of metal

sulfides.

Since most of the sulfide minerals are endowed with significant electric conductivity

allowing charge transfer at their surface, galvanic interactions, by virtue of gold-sulfide

contacts, could substantially increase or decrease the leaching of gold in aerated cyanide

solutions. In principle when gold open circuit potential is lesser than that of sulfides, gold

anodic dissolution can be prompted (Aghamirian and Yen, 2005). Conversely, galvanic

phenomena can also impede gold dissolution when the adjoining sulfide mineral acquires

anodic potential prompting on gold surfaces oxygen reduction rather than gold dissolution

(Aghamirian and Yen, 2005). However, there is still some controversy regarding the role of

galvanic interactions on gold dissolution. Filmer (1982), Paul (1984) and Lorenzen and van

Deventer (1992) postulated that the presence of pyrite and pyrrhotite substantially

decreases gold leaching rate that they ascribed to galvanic effects. Aghamirian and Yen

(2005) and Dai and Jeffrey (2006) interrogated systematically the interrelation between

46

galvanic effects by some sulfides (i.e., pyrite, chalcopyrite, pyrrhotite, galena, chalcocite)

and the kinetics of gold leaching. Their findings concluded that pyrite, pyrrhotite and

galena induce positive galvanic interactions promoting gold dissolution when a gold

electrode was paired with one of those sulfides; unlike chalcopyrite and chalcocite for

which contradictory results were reported in the literature.

Meaningful comparisons of the findings among above studies are difficult to establish since

various experimental strategies were used by different authors. For example, Lorenzen and

van Deventer (1992) used a high-purity gold electrode that is now accepted to be highly

vulnerable to passivation (Jeffrey and Ritchie, 2000) and not representative of the naturally-

occurring gold/silver alloys. In the same study, while a correct approach was devised to

study macroscopic galvanic couples by pairing the gold and mineral electrodes first in one

cell and then in two separate cells, in contrast, the galvanic current and potential as a

function of time were not reported. In addition, no anodic and/or cathodic behaviors of

minerals were considered. In Aghamirian and Yen’s (2005) study, galvanic couples were

tested only in one single electrochemical cell unveiling only global outcomes on gold

leaching from combined galvanic interactions and surface passivation. In Dai and Jeffrey’s

work (2006), the galvanic current/potential conditions relating to the Au-minerals contacts

were barely detailed. In addition, in the majority of the relevant studies, indications about

the mineralogy of the studied minerals were seldom provided. For example, pyrite often

occurs naturally in association with other minor sulfides which, even if present in minute

quantities, tend to play an important role in shaping pyrite reactivity (Cruz et al., 2001) and

as a consequence in affecting the extent of galvanic interactions on gold leaching.

A number of sulfide minerals readily dissolve in aerated cyanide solutions making gold

cyanidation a process difficult to optimize. Hence, numerous investigations were carried

out to understand how the resulting dissolution products would interfere with gold leaching

(Hedley and Tabachnick, 1968; Liu and Yen, 1995; Deschenes et al., 1998; Guo et al.,

2005; Dai and Jeffrey, 2006; Breuer et al., 2008). It is generally accepted that sulfides

interfere with gold dissolution through: (1) excess cyanide and oxygen consumptions as a

result of transition metals (such as Cu, Fe and Zn) dissolution (Habashi, 1967), (2) gold

surface passivation by some reaction products such as Fe(OH)3 (Guo et al., 2005) or sulfur

47

(Lorenzen and van Deventer, 1992; Dai and Jeffrey, 2006), which are suspected to form a

passivating layer on the surface of gold, (3) galvanic interaction between gold and sulfide

phase (Lorenzen and van Deventer, 1992)

Atmospheric pre-oxidation was shown to be an effective tool to knock-down the

undesirable reactivity of certain sulfide minerals such as marcasite and pyrrhotite (Marsden

and House, 2006). Ore pre-oxidation strategies are thus investigated as preventive measures

to turn their sulfide cyanicides into barren entities vis-à-vis cyanide consumption. However,

pre-oxidation mitigation measures were often tested on individual sulfide minerals. Multi-

factorial galvanic interactions stemming from the different mineralogical phases present in

the ore during pre-oxidation were disregarded which may give rise to totally different

subsequent cyanidation responses. In addition, the effect of these interactions on the

speciation of the reaction products was only poorly addressed in the open literature. Thus,

developing pre-oxidation strategies that will consider the mineralogical specifications of

the industrial gold ore still remains an opportunity where a contribution could be made.

In consideration of the above interrogations, this work aims at several targets:

1. To investigate the effects of sulfide ores and minerals on gold leaching during

cyanidation;

2. To attempt sorting out the individual contributions during cyanidation as effected by

galvanic interactions between gold and sulfide minerals and by the formation of

passivating layers on gold surface;

3. To investigate the effects on gold leaching rate of pre-oxidation of sulfide ore and its

sulfide minerals components prior to cyanidation.

48

II.2. Experimental

II.2.1. Reagents

Distilled water was used for all our cyanidation/pre-oxidation experiments. Sodium

cyanide, NaCN (98%, Sigma-Aldrich Canada), sodium hydroxide, NaOH (Fisher Scientific

Canada), calcium hydroxide Ca(OH)2 (95%, Sigma-Aldrich Canada), boric acid H3BO3

(99.5, Sigma-Aldrich Canada), KAu(CN)2 (98%, Sigma-Aldrich Canada) used in this study

were all certified analytical grade.

II.2.2. Materials

Four sulfide-rich ore samples were used in the present work. One sample (MRI-1) came

from a mine in the northwest of Québec province, Canada, while the other three samples

were provided by Ward’s Natural Science and referred to as pyrite (MRI-2), chalcopyrite

(MRI-3) and sphalerite (MRI-4) by virtue of the dominant proportion of the named sulfide

mineral therein (see Table II.1). The samples had a controlled particle size distribution and

were stored in the freezer to minimize surface alterations.

Tables II.1 and II.2 describe, respectively, the mineralogical composition and the elemental

analysis of each mineral. All minerals were characterized using a combination of atomic

absorption spectrometry, scanning electron microscopy (SEM) and X-ray diffraction

patterns (XRD). The first mineral (MRI-1) is clearly more complex than the others, having

more metallo-species and also a more complicated mineralogy. The mineralogical weight

distribution of sulfides for MRI-1 ore was 44.15% pyrite, 0.11% chalcopyrite, 0.31%

sphalerite and some traces of galena and arsenopyrite (Table II.1). Figures II.1a and II.1b

show the mineralogical associations of the main sulfide phases for MRI-1; as will be shown

later, these associations are believed to have a direct influence on the behavior of the ore

during cyanidation.

49

Table II-1 Mineralogical composition of the four sulfide ore samples.

Ore / Phase Pyrite

(FeS2)

Chalcopyrite

(CuFeS)

Sphalerite

(ZnFeS2)

Galena

(PbS)

Arsenopyrite

(FeAsS)

Gangue

% w % w % w % w % w % w

MRI-1

(Industrial ore)

44.15 0.11 0.31 0.05 0.08 55.3

MRI-2 (Pyrite) 96.6 tr. tr. tr. tr. 3.4

MRI-3

(Chalcopyrite)

0.8 63.2 18.3 tr. tr. 17.7

MRI-4

(Sphalerite)

4.7 tr. 90 tr. tr. 5.3

tr.: trace ( 0.05 %w); Gangue: SiO2 (maj.), Al2O3, CaO, MgO

Table II-2 Chemical composition of the four sulfide ore samples.

Ore /

Element

Au Ag Cu Zn Fe S total As Al Si

mg/kg mg/kg % % w % w % w % % w % w

MRI-1 1.02 17.75 0.03 0.18 20.6 26.8 0.02 7.6 15

MRI-2 tr. tr. 0.05 tr. 43 51 tr. tr. 0.54

MRI-3 tr. tr. 20.4 4.63 22.2 23.4 tr. 0.22 4.1

MRI-4 tr. tr. 0.07 58 0.28 31 tr. tr. 0.91 tr.: trace ( 0.01 %w)

a

Pyrite

Chalcopyrite

50

Sphalerite

Pyrite

b

Figure II-1 Optical microscopic images of MRI-1: (a) chalcopyrite associated with pyrite,

500X and (b) sphalerite associated with pyrite, 500X.

II.2.3. Electrochemical Campaign

II.2.3.1 Preparation of Disc Electrodes

The experiments were carried out using a Rotating Disk Electrode (RDE) system, as this

technique was shown to ensure reproducible experimental conditions (Churchill and Laxen,

1966). The electrochemical behavior of gold was studied using a gold/silver alloy (96%-4%

by weight) disc, mimicking industrial situations where gold usually occurs alloyed with

silver. The active surface area of the Au/Ag disc, in contact with the cyanidation medium,

was 1.09 cm2 and was embedded in a Teflon disk holder so only its lower surface was

exposed to the solution. This disk holder was then threaded and attached to a Teflon rod

itself threaded and attached to the main steel shaft. One extremity of the shaft was

electrically connected with the Au/Ag disc using a spring, while a carbon brush was used to

maintain electrical contact between the second extremity of the shaft and the potentiostat,

thus making allowance for current and potential measurements while the working electrode

is rotating.

51

The mineral electrodes were prepared by mixing in an agar mortar together, in a 3:1 mass

ratio, the target ore (Table II.1) with graphite powder (used to increase the electric

conductivity of the mineral specimens) and 5-6 drops of silicone oil (used as binding agent)

until a homogeneous paste was obtained. The resulting specimens were mechanically

pressed inside a Teflon holder (similar with the one used for the Au/Ag electrode) and then

dried overnight under N2 at room temperature.

II.2.3.2 Anodic and Cathodic Behavior of Gold and Mineral Disc Electrodes

Linear polarization experiments for gold/silver and/or mineral disc electrodes were

performed using 0.01 M boric acid as a background electrolyte in a conventional three-

electrode cell. The pH of electrolyte was adjusted to 11 using NaOH and was measured

using an Oakton 1000 series pH-meter with a precision of ±0.01. A platinum wire was used

as the counter-electrode and a standard calomel electrode (SCE) as the reference. All

potentials were measured relative to the SCE and converted to the standard hydrogen

electrode (+242 mV vs. SHE). The electrochemical studies were carried out, using a

multichannel potentiostat-galvanostat (model VSP-27 from Bio-Logic SA) and the EC-Lab

software. A series of preliminary tests revealed that a scan rate of 0.5mV s-1

was low

enough to minimize physical retardation phenomena (Dai and Jeffrey, 2006; Aghamirian

and Yen, 2005). Before each experiment, the Au/Ag electrode was routinely polished with

sandpaper, followed by polishing with of 5.0 µm alumina particles and then finally rinsed

with distilled water. The mineral electrodes were washed with distilled water and

transferred immediately into the electrochemical cell to prevent further alterations. Fresh

mineral surface was created before each experiment by removing 0.1 cm of the mineral

electrode disc. For the anodic polarization experiments, nitrogen was continuously bubbled

through the solution, while for the cathodic polarization experiments air was sparged as

oxygen source.

II.2.3.3 Disc Electrode Galvanic Couples

Gold dissolution pattern of a gold electrode in electrical contact with another electrode may

be influenced by galvanic interactions, or by the dissolving species liberated from the

contiguous electrode, or by combinations thereof (Lorenzen and Van Deventer, 1992). In

targeting specifically the effects by galvanic interactions, the electrodes will have to be

52

immersed in two separate electrochemical cells of the same starting cyanidation solution,

whereas in gathering the combined effects of galvanic interaction and dissolved species, the

electrodes will have to be placed in the same cell. This strategy enabled evaluating the

effect of combining the two electrodes in the same cell on both gold leaching rate and

electrochemical dynamic responses of galvanic potential and current. In all experiments,

the circuits were completed by platinum electrode and calomel reference electrode, both of

which were brought into close contact with the working electrode by a Luggin capillary. In

each test the electrodes were rotating at 500 rpm while air was continuously bubbled into

the solution. The Au/Ag and sulfide mineral electrodes were electrically connected to each

other through a zero resistance ammeter enabling acquisition of the galvanic current and

potential signals as time elapses. The solution was stirred with a magnetic stirrer to enhance

mixing and homogenization of the solution as well as diffusion of reactants and products

nearby the electrodes.

II.2.4. Sulfide Minerals and Gold Dissolution in Slurry Reactor

To investigate the effects of minerals dissolution on gold leaching rate, a systematic

cyanidation study has been performed using the RDE equipment and the gold/silver

electrode. The Au/Ag electrode was immersed in slurries containing one of the four ores

used in this study (MRI-1 to MRI-4) and both gold dissolution and solution chemistry were

monitored as a function of time. Gold cyanidation experiments were carried out at room

temperature in a 0.40 L magnetically-stirred gas-slurry glass reactor. Combining 150 g of

distilled water and 50 g (crushed, P80 =75µm) of the targeted ore resulted in slurry

containing 25% solid by weight. Air was bubbled through the reactor to maintain constant

dissolved oxygen level (DO2 ~ 8.5 mg/L at 25ºC), which was monitored by means of a

dissolved oxygen probe (DOB-930 model from Omega, accuracy ±0.1 mg/L). The pH of

the slurry was controlled to 11.5 using NaOH. Once pH and dissolved oxygen were

adjusted, NaCN was added to the slurry to initiate reaction. During the course of reaction,

small aliquots were periodically withdrawn from the reactor and filtered on 0.45µm

Millipore membrane filters to remove solid particles prior to analysis.

In all experiments, the concentrations of dissolved metals (Au, Ag, Cu, Zn, Fe, etc.) were

measured using a Perkin-Elmer AA-800 Atomic Absorption Spectrometer (AAS), while a

53

capillary electrophoresis instrument (CE, Agilent Technologies) was used to analyze the

dissolved sulfur and cyanicides and where the experimental analytical errors were

estimated to vary between 5% and 10% (Petre et al., 2008). Residual (unreacted or free)

cyanide was quantified using a silver nitrate titration method with rhodamine as color-

change indicator. This method has some inconveniences, as the presence of copper and zinc

cyanide complexes is known to interfere with the quantification of free cyanide (Marsden

and House, 2006). Nevertheless, the error was not quantified as the method is largely used

in gold plants and remains a simple and quick method for free cyanide quantification.

II.3. Results and discussion

II.3.1. Anodic and Cathodic Behavior of Gold and Mineral Disc

Electrodes

Linear sweep voltammograms for Au/Ag, MRI-1 and MRI-2 discs in a 10 mmol/L sodium

cyanide solution at pH 11 are presented in Figure II.2. MRI-1 and MRI-2 materials were

chosen for these tests as they correspond to a representative sulfide-rich industrial ore

(MRI-1) along with its major sulfide constituent (pyrite). The figure shows that gold

oxidation commences around ca. -350 mV, immediately displaying a steeply rising current

density as the imposed potential increased. A diffusion-limited current plateauing near 10

mA m-2

was reached at about +400 mV. In contrast, oxidation currents of the industrial ore

(MRI-1) and pyrite (MRI-2) broke through after the imposed potential exceeded +200 mV

and +400 mV, respectively. At first sight, these results suggest that galvanic interactions at

the Au-sulfide interface will tend to promote a gold anodic behavior, and thus gold

leaching, for both MRI-1 and MRI-2.

54

Figure II-2 Anodic voltammograms of Au/Ag, MRI-1 and MRI-2 electrodes; CN- = 10

mmol/L, DO2 = 0 mmol/L, pH = 11, ΔE/Δt = 0.5 mV/s, electrode rotation rate = 600 rpm.

Figure II-3 Cathodic voltammograms of Au/Ag, MRI-1 and MRI-2 electrodes; CN- = 0

mmol/L, DO2 = 0.25 mmol/L, ΔE/Δt = 0.5 mV/s, electrode rotation rate = 500 rpm,

pH = 11.

55

Furthermore, the different patterns revealed by the anodic polarization curves of MRI-1 and

MRI-2 suggest that other sulfides (i.e., sphalerite and chalcopyrite, Figure II.1a,b)

coexisting among with the most abundant one (i.e., pyrite in MRI-1) may alter the

magnitude of galvanic interactions between gold and ore, and so likely their reactivity.

Oxidation of MRI-1 ore is triggered at a lower potential (beginning at +200 mV) while

showcasing higher current density compared to pyrite (MRI-2) oxidation (lower current

density and higher break-through potential, +400 mV). Such differences are ascribed to

coexistent minority sulfides, e.g., chalcopyrite and sphalerite, associated with pyrite in

MRI-1. Cruz et al. (2005), using cyclic voltammetry, studied the influence of sphalerite

impurities on pyrite reactivity. This latter was found to decrease upon association even with

small quantities of sphalerite due to sphalerite less positive equilibrium potential in favor of

galvanic protection of pyrite. Other studies have shown that when pyrite and chalcopyrite

are in contact in solutions, chalcopyrite is preferentially oxidized protecting in return the

pyrite (Madhuchhanda et al., 2000). The higher Tafel slope in oxidation current density for

MRI-1 in comparison to that of MRI-2 seen in Figure II.2 may be attributed to

simultaneous oxidations of pyrite, chalcopyrite and/or sphalerite. Such reactions could

influence the equilibrium potential of pyrite, thus making the galvanic interactions between

gold and pyrite lesser in the case of MRI-1 than for MRI-2.

Oxygen reduction voltammograms for gold/silver, pyrite and industrial ore electrodes are

illustrated in Figure II.3. Since gold is soluble in cyanide solutions, oxygen reduction

experiments on the gold surface were conducted in cyanide-free aqueous media. Oxygen

reduction over both MRI-1 and MRI-2 electrodes is far more active than on gold electrode

irrespective of the applied potential. This finding is in disagreement with Aghamirian and

Yen (2005) who found pyrite was less active for oxygen reduction than gold at potentials

lower than -300mV and trend reversal at higher potentials. Pyrite (both in MRI-1 and MRI-

2) seems to offer suitable surface sites for oxygen reduction across the whole potential

range interrogated in Figure II.3. In support of our results, pyrite was previously shown to

possess good electro-catalytic ability for oxygen reduction (Rand, 1977) whose mechanism

and kinetics are largely dependent on the mineral mineralogy and surface properties

(Ahlberg and Broo, 1996b; Cruz et al., 2001). Ahlberg and Broo (1996a,b) pointed out that

oxygen reduction reaction proceeds through the formation of hydrogen peroxide and

56

therefore pyrite may develop a surface layer consisting of ferric hydroxide in aerated

alkaline solutions (Buckely and Woods, 1987). These considerations are in support with the

appearance of passivation peaks at ca. –75mV for MRI-2 and at ca. –350mV for MRI-1

(Figure II.3) to be attributed to the oxidation of pyrite surface by hydrogen peroxide formed

during the cathodic potential sweep.

Since gold oxidation can occur at potentials more positive than –400mV (Figure II.2) and

considering the enhanced magnitude of cathodic currents (Figure II.3), it can be speculated

that galvanic interactions would drive favorably the gold anodic behavior. These

interpretations must, however, be confirmed by further electrochemical studies such as

pairing gold, MRI-1 and MRI-2 electrodes in electrochemical cells to enable simultaneous

monitoring of the galvanic potential and current, and gold leaching rate as a function of

time.

II.3.2. Disc Electrode Galvanic Couples

In an attempt to decouple and assess the contribution of passivation phenomena versus

galvanic interactions on gold leaching rate, the Au/Ag and MRI-1 electrodes were first

electrically connected in one electrochemical cell and then in two separated cells.

Electrically connected MRI-1 and Au/Ag electrodes, either in one (OEC) or in two (TEC)

separated electrochemical cells promoted higher gold dissolutions in comparison to a

solitary Au/Ag electrode immersed in an aerated cyanide solution (Figure II.4). These

results come in support of the previous observations from the anodic and cathodic

polarization curves (Figures II.2, II.3). Nonetheless, the gold dissolution rate achieved in

OEC arrangement was lesser than in the TEC one. This difference could be explained by

partial obstruction of the gold surface by resistive films of Au2S and/or metal hydroxides

originating from the leaching sulfide ore electrode. This is not unlikely, bearing in mind

that sulfide minerals are soluble to some extent in cyanide solutions and therefore some

sulfur and metals (particularly Fe, Cu and Zn) can be leached off into the solution.

Coherent with the formation of inhibiting Au2S surface layers (Weichselbaum et al., 1989;

Jeffrey and Breuer, 2000), addition of sodium sulfide traces to cyanide solutions has been

shown to significantly retard gold dissolution. Our findings are also in line with those of

57

Guo et al. (2005) who reported important slowdown in gold dissolution in the presence of a

20% pyrite slurry which was rationalized by the authors as caused by pyrite leaching and

subsequent formation of Fe(OH)3 and sulfur passivating films on gold surfaces.

Nevertheless, as illustrated in Figure II.4, for our case the positive effect of galvanic

interaction outweighs the negative effect of gold surface passivation resulting in a net

acceleration of gold leaching.

Figure II-4 Effect of coupling Au/Ag and MRI-1 electrodes in one electrochemical cell

(OEC) and in two electrochemical cells (TEC) on gold leaching rate. Reaction conditions:

CN- = 10 mmol/L, DO2 = 0.25 mmol/L, electrode rotation rate = 500 rpm, pH = 11.

0

5

10

15

20

25

0 10 20 30 40 50 60 70

Time (minutes)

Au

le

ac

he

d (

g/m

2)

Au/Ag+MRI-1; OEC

Au/Ag+MRI-1; TEC

Au/Ag; OEC

0

5

10

15

20

25

0 10 20 30 40 50 60 70

Time (minutes)

Au

le

ac

he

d (

g/m

2)

Au/Ag+MRI-1; OEC

Au/Ag+MRI-1; TEC

Au/Ag; OEC

58

Figure II-5 Effect of galvanic interactions in MRI-1 on the evolution of galvanic potential

and galvanic current vs. time in one electrochemical cell (OEC), CN- = 10 mmol/L, DO2 =

0.25 mmol/L, electrode rotation rate = 500 rpm, pH = 11.

Figure II-6 Effect of galvanic interaction of MRI-1 on the evolution of galvanic potential

and galvanic current vs. time in two electrochemical cells (TEC), CN- = 10 mmol/L, DO2 =

0.25 mmol/L, electrode rotation rate = 500 rpm, pH = 11.

-250

-230

-210

-190

-170

-150

0 10 20 30 40 50 60

Time (minutes)

E (

mV

) vs S

HE

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

I (m

A/m

2)

E

I

-250

-230

-210

-190

-170

-150

0 10 20 30 40 50 60

Time (minutes)

E (

mV

) vs S

HE

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

I (m

A/m

2)

E

I

-210

-200

-190

-180

-170

-160

-150

-140

0 10 20 30 40 50 60

Time (minutes)

E (

mV

) vs S

HE

-5

0

5

10

15

20

25

30

35

40

45

I (m

A/m

2)

E

I

-210

-200

-190

-180

-170

-160

-150

-140

0 10 20 30 40 50 60

Time (minutes)

E (

mV

) vs S

HE

-5

0

5

10

15

20

25

30

35

40

45

I (m

A/m

2)

E

I

59

Lorenzen and van Deventer (1992) suggested that when gold is in electrical contact with

conductive sulfide minerals such as pyrite and chalcopyrite (also present in MRI-1), the

reduction of oxygen could take place over the entire exposed surfaces, gold + mineral. The

magnitude of oxygen reduction would exceed that of gold oxidation in the region of

cathodic potentials, and, therefore, passivation of gold would occur. Our findings in Figure

II.4, on the contrary, do not show any decrease in gold leaching and, as stated by Dai and

Jeffrey (2006), the higher the surface area available for oxygen reduction (Figure II.3), the

higher the is leaching rate for gold. As alluded to in the introduction, the fact that a non-

alloyed high-purity gold electrode was used by Lorenzen and van Deventer (1992) in their

electrochemical cell tests could be at the origin of the discrepancies with our Figure II.4

results. Gold leaching out of a pure gold electrode was shown by Jeffrey and Ritchie (2000)

to be disrupted by AuCN insoluble films that form on the surface. Figures II.5 and II.6

present the evolution of the galvanic current and galvanic potential as a function of time for

Au/Ag and MRI-1 electrodes paired in one (OEC) and in two electrochemical cells (TEC),

respectively. In OEC arrangement, the evolution of the corrosion potential rapidly peaks

near 3.5 mA/m2 after ca. 6 min, and then gradually decreases to a value of 1.8mA/m

2 after

60 min, Figure II.5. TEC arrangement, in contrast, displayed monotonically increasing

galvanic currents as a function of time which then plateaued around 40 mA/m2 after 60

min, Figure II.6. The lower galvanic current obtained in OEC arrangement are likely to be

attributed to a drop in anodic current as a result of the formation of a passivating Au2S film

on gold surface following the release of several sulfide species from the MRI-1 ore during

cyanidation (Petre et al., 2008). This appears also corroborated by Jeffrey and Breuer,

(2000) who showed that when gold was exposed to a hydrosulfide-containing solution, low

anodic currents were registered along with depressed gold leaching rates presumably

hindered by instantaneous surface passivation.

II.3.3. Disc Electrodes Passivation and Galvanic Interactions Effects on

Gold Leaching

Gold leaching rates generally improve with increasing electrodes rotational speed (Jeffrey

and Ritchie, 2000). Up to now, no one investigated the effect of electrode rotational speed

60

on Au leaching rate under Au/Ag and mineral electrodes coupling. As oxygen reduction

takes place on mineral and gold surfaces alike, Figure II.4, increasing rotational speeds may

improve oxygen diffusion and replenishment on gold and mineral electrode surfaces.

Hence, provided the circumstances of positive galvanic interactions prevail as portrayed in

Figures II.4-6, any increase in rotational speeds would result in improved gold leaching rate

had the system been operating under oxygen-diffusion control. Figure II.7 presents a

comparative study between measured and theoretical (Levich, 1962) molar flux densities

for gold leaching.

Figure II-7 Effect of electrode rotational speed on molar flux densities for gold leaching in

the presence of passivation and galvanic interactions. Also shown are the calculated lines

representing the diffusion of oxygen with the reduction of oxygen to (line a) hydroxide and

(line b) peroxide, respectively. Reaction conditions: CN- = 10 mmol/L, DO2 = 0.25

mmol/L, pH = 11.

As shown in Figure II.7, regardless of whether the Au/Ag electrode was tested in solitary

conditions or in association with MRI-1 ore in OEC and/or in TEC arrangement, the gold

leaching rate is suggested to be oxygen-diffusion limited. The two electrodes in OEC and

0.0

1.0

2.0

3.0

4.0

5.0

6.0

0 2 4 6 8 10 12

0.5 / s-0.5

Leach

ing

rate

x 1

05 (

mo

l/m

2/s

)

Au/Ag+MRI-1; OEC

Au/Ag+MRI-1; TEC

(b)

Au/Ag; OEC

(a)

0.0

1.0

2.0

3.0

4.0

5.0

6.0

0 2 4 6 8 10 12

0.5 / s-0.5

Leach

ing

rate

x 1

05 (

mo

l/m

2/s

)

Au/Ag+MRI-1; OEC

Au/Ag+MRI-1; TEC

(b)

Au/Ag; OEC

(a)

61

TEC arrangements were operated with the same rotational speed. Figure II.7 shows that for

all cases, the leaching rate of gold is linearly increasing with the square-root of gold

electrode rotation speed. However, for the Au/Ag and MRI-1 electrodes in OEC

arrangement (empty triangle symbols), a significant drop in slope took place. Dissolution

indicates progressive decrease in the leaching rate of gold above 400 rpm ( 6.4 (rad/s)0.5

)

for OEC arrangement, sliding below that of TEC arrangement. Such results could be

explained by an enhanced sulfide mineral leaching at electrode rotating speeds in excess of

400 rpm. Consequently, promoting the release in solution of further species from the

dissolution of the mineral electrode will result in further occlusion of the gold electrode

surfaces. Hence, reactions may transit from oxygen diffusion control to a mixed diffusion-

chemical control as suggested by the lowered slope in Figure II.7.

II.3.4. Electrochemical Pre-oxidation of Mineral Disc Electrodes and

Gold Leaching

The surface of the (MRI-1) mineral electrode was electrochemically pretreated by applying

a constant oxidizing potential to the electrode while being immersed in an aerated cyanide-

free alkaline solution. The mineral electrode (MRI) was installed in conventional three-

electrode glass cell where platinum wire was used as counter electrode and saturated

calomel electrode (SCE) as reference. The experiment started near the open circuit potential

and a constant oxidizing potential of +1000 mV was applied to the MRI-1 electrode for

varying time intervals (from 1 to 3 h). At the end of this electrochemical pretreatment, the

oxidized MRI-1 electrode and the Au/Ag electrode were electrically connected in OEC

arrangement before cyanidation tests were resumed.

The results summarized in Figure II.8 show a significant increase of gold leaching after

electro-oxidation of the mineral surface. It can also be seen that the anodic behavior of gold

is strongly promoted by the increase of pre-oxidation time. Such improvement is coherent

with a likely passivation of the surface of the mineral electrode resulting from the

formation of a resistant layer protecting from the dissolution of the sulfide minerals during

cyanidation. Several authors (Zhu et al., 1993 and references therein; Cruz et al., 2005),

62

suggested that the reaction products of pyrite alkaline electrochemical oxidation, major

sulfide in MRI-1, are Fe(III) hydroxides, SO42-

, and partially oxidized sulfur species:

FeS2 + 2H2O → FeOOH + 2S0 + 3H

+ + 3e

- (II.1)

FeS2 + 10H2O → FeOOH + 2SO42-

+ 19H+ + 15e

- (II.2)

Such Fe(III) hydroxide protective layers forming on the mineral surface limits MRI-1

dissolution and its effects on gold passivation (S and/or Fe(III) hydroxide deposits).

Figure II-8 Coupling Au/Ag and MRI-1 electrodes in one electrochemical cell (OEC):

effect of mineral electrode electro-oxidation on gold leaching. Reaction conditions: CN- =

10 mmol/L, DO2 = 0.25 mmol/L, pH = 11, electrode(s) rotation rate = 500 rpm.

Gold leaching rates obtained with pre-oxidized MRI-1 electrode in OEC arrangement even

surpassed those in TEC arrangement without ore pre-oxidation, see Figures II.4 and II.8.

Aghamirian and Yen (2005) suggested that the presence of iron hydroxide layer on the

mineral electrode would impair the process of oxygen reduction. This would entrain a

decrease in the cathodic current due to oxidation of the most active species at the mineral

0

5

10

15

20

25

30

35

40

45

0 10 20 30 40 50 60 70

Time (minutes)

Au

le

ac

he

d (

g/m

2)

Au/Ag+MRI-1; OEC ; Ox 1h

Au/Ag+MRI-1; OEC; Ox 3h

Au/Ag; OEC

Au/Ag+MRI-1; OEC

0

5

10

15

20

25

30

35

40

45

0 10 20 30 40 50 60 70

Time (minutes)

Au

le

ac

he

d (

g/m

2)

Au/Ag+MRI-1; OEC ; Ox 1h

Au/Ag+MRI-1; OEC; Ox 3h

Au/Ag; OEC

Au/Ag+MRI-1; OEC

63

surface, the consequence of which would be a decrease in gold leaching rate. This

expectation is unlike the trend revealed in our study. One possible explanation put forth

being despite deeply oxidized surfaces, it is virtually impossible to fully cover them with

hydroxides and thus oxygen reduction could still be permitted provided some porosity is

left across the deposited hydroxide. This explanation is corroborated with Ahlberg and

Broo (1996a,b) findings who demonstrated that oxygen reduction at the oxidized surface of

pyrite proceeds with the formation of important amounts of hydrogen peroxide due to the

porosity of the hydroxide layer. It can be speculated that in the presence of a porous iron

hydroxide layer, the catalytic activities towards oxygen reduction may increase at mineral-

iron hydroxide interface, as this surface could actually provide more active sites where

oxygen reduction would occur. This aspect appears to be very complex and further

investigations might be needed in the future.

II.3.5. Sulfide Minerals and Gold Dissolution in Slurry Reactor

Figure II.9 shows the time evolution of gold leaching from the rotating Au/Ag electrode

during cyanidation tests in a slurry reactor. Four slurries consisting of the industrial ore

(MRI-1), pyrite (MRI-2), chalcopyrite (MRI-3) and sphalerite (MRI-4) minerals were

tested in the proportions and properties described in sections II.2.2 and II.2.4. Unlike in the

mineral disc configurations, the results shown in Figure II.9 for the slurries reveal that all

tested minerals brought important reductions in gold leaching rates with respect to the

standard test of a solitary gold electrode immersed in clear aerated cyanide solution. The

following gradation takes place from worst and upwards: chalcopyrite > sphalerite >

industrial ore > pyrite.

Drastic setback in gold leaching rate in the presence of chalcopyrite is to be imputed in part

to massive copper leaching as monitored by capillary electrophoresis and atomic absorption

that will be discussed later. Substantial copper dissolution occurs mainly in the early

reaction stages and barely lasts for five minutes during which more than 70% of initial free

cyanide is diverted to form copper cyanide complexes (Cu(CN)x(x-1)-

). Accumulation of

small concentrations of hydrosulfide, 4 to 5 mg/L, in the presence of chalcopyrite is worthy

of notice. This leans towards conjugated negative effects prevailing during cyanidation of

gold in chalcopyrite where, in addition to rapid depletion of free cyanide through copper

64

complexation, gold could also have been passivated as indicated by the presence of

hydrosulfide ions in the leaching solution in accordance with literature findings

(Weichselbaum et al., 1989; Jeffrey and Breuer, 2000).

0,0

5,0

10,0

15,0

20,0

25,0

0 10 20 30 40 50 60 70

Au

leach

ed

(g

/m2)

Au/Ag

Au/Ag-MRI-2

Au/Ag-MRI-1

Au/Ag-MRI-3

Au/Ag-MRI-4

Time (minutes)

Figure II-9 Effect of sulfide minerals on gold dissolution in slurry reactor: industrial ore

(MRI-1) and the equivalent of its major sulfide constituents (MRI-2, MRI-3 and MRI-4).

Reaction conditions CN- = 10 mM, DO2 = 0.25 mM, pH = 11, electrode rotation rate = 500

rpm.

Regarding gold leaching profiles for the tests with MRI-1 and MRI-2, a slight decline of

the dissolution rate was noted in the presence of MRI-1 ore in comparison with MRI-2

mineral. This reflects the inter-dependence between ore reactivity and its mineralogical

composition. Indeed, the measured concentrations of S2O32-

and SO42-

after 60 min of

cyanidation were, respectively, 13 mg/L and 512 mg/L for MRI-2, and 53 mg/L and 1000

mg/L for MRI-1. The galvanic effects between the main associated mineralogical phases

(pyrite, chalcopyrite and sphalerite) in the industrial ore speak for themselves where despite

a lesser content of total sulfur in MRI-1 (total sulfur = 26.8%w, Table II.2), more soluble

65

oxy-sulfur species were generated from MRI-1 comparatively to MRI-2 (total sulfur =

51%w, Table II.2).

On the contrary, more thiocyanate was generated (63 mg/L) from MRI-2 dissolution than

from MRI-1 (45 mg/L) with no hydrosulfide ions formation with both slurry materials.

Some studies reported that autogenous hydrosulfide produced in cyanide solutions in the 8-

12 pH range can be oxidized by O2 into polysulfides, thiocyanate, thiosulfate and sulfate

ions (Luthy and Bruce, 1979; Marsden and House, 2006; Bard and Startmann, 2006).

Recently, Breuer et al (2007, 2008) reported that pyrite is able to catalyze the oxidation of

dissolved sulfide ions via electron transfer from the sulfide ion to dissolved oxygen on the

pyrite surface, with polysulfides formed as intermediate oxidation products. Then,

polysulfides undergo homogeneous oxidation in the presence of cyanide to form mainly

thiocyanate, along with some thiosulfate and sulfite. This explanation is partially in

agreement with our results, as no hydrosulfide ions were detected while thiocyanate

concentration constantly increased with the increase of pyrite (MRI-2) cyanidation time

(data shown later). However, unlike Breuer et al (2007, 2008) findings, high sulfate

concentrations were detected whereas sulfite remained below instrumental detection limits.

The absence of sulfite in this study can be explained by its possible reaction with

polysulfides to form thiosulfate as explained by the same authors, Breuer et al. (2007,

2008).

Finally, the inhibiting effect on gold leaching rate of sphalerite (MRI-4) was found to be

more pronounced than for MRI-1 ore or MRI-2 mineral, Figure II.9. This result appears

odd in light of sphalerite low reactivity (Hedley and Tabachnik, 1968) in aerated cyanide

solutions as confirmed by the low amounts of produced SO42-

(20 mg/L) and SCN- (6

mg/L). Further investigations are needed to decipher the behavior of this mineral.

II.3.6. Alkaline Pre-oxidation of Sulfide Ores and Gold Leaching

The industrial gold ore (MRI-1) as well as the three mineral constituents (MRI-2, MRI-3

and MRI-4) were pre-oxidized prior to cyanidation tests. Room-temperature oxidative pre-

treatments were conducted by sparging, for varying periods of time, pure oxygen through

each of the designated slurries prior to cyanidation. Note that the Au/Ag electrode was not

66

immersed into the slurry. After pre-oxidation was completed, a sample aliquot was

withdrawn from the reactor for analysis to set the initial solution concentrations;

cyanidation was then initiated by immersing the Au/Ag electrode into the oxidized slurry.

A series of preliminary tests was carried out using MRI-1 to identify the optimal pre-

oxidation time. This optimal time corresponds to the duration that maximizes cyanidation

efficiency which is defined as the ratio of leached gold divided by the consumed free

cyanide. The optimal pre-oxidation time was found to be 4 h. Hence, high cyanidation

efficiency combines the highest gold leaching achievable for minimum cyanide

consumption. The benchmark gold leaching curve is the usual solitary Au/Ag electrode in a

clear O2-NaCN solution and is shown in Figures II.10a to II.13a as filled diamond symbols.

The most remarkable pre-oxidation effects are obtained with the industrial ore (MRI-1,

Figure II.10a) where the gold leaching rate increased from 6.6x10-6

mol/m2/s for non-

pretreated ore to 2.16x10-5

mol/m2/s after 4h ore pre-oxidation. Considering the pre-

oxidized slurries for the individual sulfide minerals, Figure II.11a identifies pyrite (MRI-2)

pre-oxidation as counter-productive for leveraging gold leaching rate, whereas pre-

oxidation of chalcopyrite (MRI-3, Figure II.12a) or sphalerite (MRI-4, Figure II.13a) led to

marginal gold leaching improvements on already low performances.

To assess the impact of any deposits and/or alteration of the gold surface during pre-

oxidation on the gold leaching by cyanidation, further pre-oxidation of MRI-1 has been

carried out with the gold electrode being immersed in the slurry during the pre-treatment

phase. Although immersing the gold electrode in the MRI-1 pulp during pre-oxidation may

lead to some losses in gold cyanidation rate comparatively to the normal pre-oxidation

procedure described above, yet a significant improvement of gold leaching rate was

obtained (two times, Figure II.10a) in comparison with cyanidation without slurry

pretreatment. In addition, the difference in gold leaching rates between the two pre-

oxidation tests stems from likely passivating layers of mineral oxidation products forming

on the gold surface during pre-treatment.

67

0

5

10

15

20

25

0 10 20 30 40 50 60 70

Time (minutes)

Au

leach

ed

(g

/m2)

Au/Ag

Au/Ag+MRI-1; Ox 4h

without immersion

Au/Ag+MRI-1(a)

Au/Ag+MRI-1; Ox 4h

with immersion

0

200

400

600

800

1000

1200

1400

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70

Lea

ched

sp

ecie

s (m

g/L

)

Time (minutes)

SO42- ; Ox 4h

SO42-

S2O32-

S2O32-; Ox 4h

SCN-; Ox 4hSCN-

(b)

68

0

10

20

30

40

50

60

0 10 20 30 40 50 60 70 80 90

Leach

ed

sp

ecie

s (

mg

/L)

Time (minutes)

Zn(CN)42- ; Ox 4h

Zn(CN)42-

Cu(CN)32-

Cu(CN)32- ; Ox 4h

(c)



symbols: tests without pre-oxidation; empty symbols: tests with pre-oxidation. Reaction

conditions: CN- = 10 mM, DO2 = 0.25 mM, pH = 11.5, electrode rotation rate = 500 rpm.

The pre-oxidation of pyrite and chalcopyrite did not appear to restrict their reactivity during

the cyanidation. Monitoring the evolution of released sulfur-bearing anions as a function of

time (Figures II.11b and II.12b) indeed revealed pretty active pre-oxidation exacerbating

the formation of sulfur oxyanions and thiocyanate species during cyanidation. For the case

of pyrite (Figure II.11b), the total sulfur formed during the cyanidation step, calculated as

the sum of sulfur from S2O32-

, SO42-

and SCN-, climbed from 212 mg/L without pre-

oxidation to 320 mg/L for the case with pre-oxidation. For the case of chalcopyrite (Figure

II.12b), the total sulfur slightly increased from 95mg/L without pre-oxidation to 120 mg/L

with pre-oxidation.

The generation of thiocyanate during cyanidation increased remarkably in the case of pre-

oxidized pyrite (Figure II.11b). Pre-oxidation is believed to lead to an accumulation of

polysulfides in high concentrations as the intermediate products of sulfide ore oxidation in

alkaline solutions. Therefore, once cyanide was added to trigger gold leaching, polysulfides

69

could divert substantial amounts of free cyanide (up to 25%) to form thiocyanate (Figure

II.11b). This is in agreement with Breuer and Jeffrey (2008) who stated that the main

mechanism for thiocyanate formation is a homogenous pathway between free cyanide and

polysulfides:

Sx2-

+ CN- → SCN

- +Sx-1

2- (3)

This explanation is also coherent with the observed adverse effect of pyrite pre-oxidation

on gold leaching rate, as shown in Figure II.11a. Thus, it can be speculated that the

formation of polysulfides during sulfide ores dissolution may be doubly negative on gold

leaching rate because of:

polysulfides consumption of free cyanide via homogeneous pathway

producing thiocyanate,

polysulfides chemisorption on gold surface as demonstrated by Lustemberg

et al. (2008).

0,0

5,0

10,0

15,0

20,0

25,0

0 10 20 30 40 50 60 70

Au

leach

ed

(g

/m2)

Time (minutes)

Au/Ag

Au/Ag+MRI-2;

Ox 4hAu/Ag+MRI-2;

Ox 4h

(a)

Au/Ag+MRI-2

Au/Ag

70

0

500

1000

1500

2000

2500

0

20

40

60

80

100

120

140

160

0 10 20 30 40 50 60 70

Leach

ed

sp

ecie

s (

mg

/L)

Time (minutes)

SCN-; Ox 4h

SO42-; Ox 4h

S2O32-- ; Ox 4h

SCN-

SO42-

S2O32--

(b)

0

2

4

6

8

10

12

0 10 20 30 40 50 60 70

Leach

ed

sp

ecie

s (

mg

/L)

Time (minutes)

Fe(CN)64

-

Fe(CN)64-; Ox 4h

OCN-

OCN-; Ox 4h

Cu(CN)32-; Ox 4h

Cu(CN)32-

(c)





71

Regarding the formation of metallo-cyanide complexes during cyanidation, it is clear that

the galvanic effects occurring between the main sulfide constituents of the industrial ore

(MRI-1) play an important role in their leaching kinetics (Figure II.10c). For example, it

can be noted that most of the dissolution of copper and zinc occurs during the first 5 min of

leaching where the leached metal concentrations represent up to 80% of the metal

concentrations to be reached after 60 min of cyanidation (Figure II.10c). Hence sphalerite

and chalcopyrite are rapidly leached out from the ore. From the data obtained it can be

suggested that galvanic chalcopyrite-pyrite and sphalerite-pyrite associations, with

chalcopyrite and sphalerite having lesser equilibrium potentials than pyrite, lead to the

electrochemical activation of chalcopyrite and sphalerite and to repression of pyrite

leaching. This is also confirmed by the absence of dissolved iron in the leaching solution

(Figure II.10c), despite abundant iron in MRI-1 ore (20.6% w/w). In contrast, when pyrite

mineral (MRI-2) was used, 10 mg/L of Fe(CN)64-

were detected after 60 min in the

leaching solution (Figure II.11c), thus suggesting that galvanic protection by sphalerite

and/or chalcopyrite for the MRI-1 ore are responsible for pyrite passivation (Figure II.10).

The same galvanic interactions between sulfides in the industrial ore (MRI-1) could also be

responsible for the important leaching of Zn, Figure II.10c. Despite the fact that sphalerite

mineral (MRI-4) is far richer in zinc (58% w) than MRI-1 ore (0.18% w), the amount of

Zn(CN)42-

leached out during MRI-1 cyanidation (Figure II.10c) notably exceeds that

released from MRI-4 leaching (Figure II.13b). Finally, all these results are consistent with

those of Cruz et al (2005) and Qing You et al. (2007) who noted that even small quantities

of sphalerite and/or chalcopyrite induce a significant decrease on pyrite reactivity.

Figures II.11c and II.12c show that the leaching of copper and iron from pyrite and

chalcopyrite was differently affected by pre-oxidation. Indeed, pre-oxidation went from

having no effect on copper dissolution from MRI-2 (Figure II.11c) to being detrimental in

reducing the leaching of this metal from MRI-3, as the leaching of Cu increased by 40%

after pre-oxidation (Figure II.12c). In contrast, iron leaching was greatly repressed by pre-

oxidation for both MRI-2 and MRI-3. Fe(CN)64-

concentration decreased after 60 min of

cyanidation from 10 mg/L without pre-oxidation to 2.8mg/L with pre-oxidation for MRI-2

mineral and from 1.5mg/L to 0 mg/L for MRI-3 mineral (Figures II.11c and II.12c). In

72

addition, Fe(CN)64-

concentrations measured by capillary electrophoresis were far less than

those for total leached sulfur and reflect neither pyrite nor chalcopyrite ores stoichiometry.

These results suggest that pyrite and chalcopyrite break down in aerated cyanide solution to

form sulfur oxyanions, thiocyanate, copper cyanide and solid iron hydroxide rather than

dissolved iron cyanide. This explanation seems to support the thermodynamic calculations

by Zhang et al. (1997) who showed that competition between Fe and Cu to yield their

cyanide complex counterparts results in precipitation of FeOOH prior to Cu(OH)2 with free

cyanide being consumed almost exclusively by Cu. In our case, this competition is more

visible with MRI-3 ore (Figure II.12c), where only very low concentrations of Fe(CN)64-

(1.5mg/L) were detected in solution. Unlike Fe(CN)64-

, large amounts of Cu(CN)32-

were

released into the solution, which, for the case of pre-oxidized MRI-3, consumed up to 90%

of the available free cyanide (Figure II.12c).

0,0

5,0

10,0

15,0

20,0

25,0

0 10 20 30 40 50 60 70

Au

leach

ed

(g

/m2)

Time (minutes)

Au/Ag+MRI-3;

Ox 4h

Au/Ag

Au/Ag+MRI 3

(a)

73

0

50

100

150

200

250

300

350

400

450

0 10 20 30 40 50 60 70

Leach

ed

sp

ecie

s (

mg

/L)

Time (minutes)

S2O32-; Ox 4h

SO42-; Ox 4h

SO42-; Ox 4h

S2O32-; Ox 4h

SCN- and SCN-, Ox 4h

(b)

0

0,5

1

1,5

2

2,5

3

3,5

4

4,5

5

0

50

100

150

200

250

0 10 20 30 40 50 60 70

Leach

ed

sp

ecie

s (

mg

/L)

Time (minutes)

Cu(CN)x(x-1)

Cu(CN)x(x-1) - ;Ox 4h

Fe(CN)64- ; Ox 4h

Fe(CN)64-

(c)





74

In the case of sphalerite mineral (MRI-4), despite its remarkable inhibitory effect on the

dissolution rate of gold (Figure II.13a), only extremely low concentrations of leached ions

were detected (Figure II.13b) when compared with the other cases (Figures II.11b, II.11c,

II.12b and II.12c). Moreover, pre-oxidation of this mineral significantly reduced its

reactivity during cyanidation as observed from Figure II.13b. As a matter of fact, the total

leached sulfur during cyanidation dropped from 14.5 mg/L without pre-oxidation to less

than 2 mg/L with pre-oxidation. MRI-4 pre-oxidation leads to significant decrease in

sulfate formation and also prevents the formation of thiocyanate and thiosulfate during

cyanidation (Figure II.13b). However, this had no beneficial effects on the gold leaching

rate as seen from Figure II.13a and, therefore, the effect of sphalerite on gold dissolution

kinetics will need further investigation.

0,0

5,0

10,0

15,0

20,0

25,0

0 10 20 30 40 50 60 70

Au

leach

ed

(g

/m2)

Time (minutes)

Au/Ag+MRI 4; Ox 4h

Au/Ag+MRI 4

Au/Ag

(a)

Au/Ag+MRI-4

Ox 4h

Au/Ag+MRI-4

75

0

5

10

15

20

25

0 10 20 30 40 50 60 70

Leach

ed

sp

ecie

s

(mg

/L)

Time (minutes)

SO42-

SO42- ; Ox 4h

Zn(CN)42-; Ox 4h

Zn(CN)42-

S2O32-

SCN-

(b)


effect on: (a) gold leaching rate, (b) sulfur speciation and metals speciation. Full symbols:

tests without pre-oxidation; empty symbols: tests with pre-oxidation. Reaction conditions:

CN- = 10 mM, DO2 = 0.25 mM, pH = 11.5, electrode rotation rate = 500 rpm.

Pre-oxidation of the industrial gold ore (MRI-1) drastically decreased its reactivity during

cyanidation when compared to the benchmark cyanidation test. Correlatively, the total

sulfur leached after 60 min of cyanidation decreased from 396 mg/L without pre-oxidation

to 26mg/L with pre-oxidation (Figure II.10b). Moreover, it can be seen that pre-oxidation

virtually stopped thiosulfate and sulfate leaching during the subsequent cyanidation step.

Pre-oxidation had also a significant effect on reducing the formation of zinc cyanide and, to

a lesser extent, copper cyanide, which decreased from 44 mg/L to 21 mg/L and from 33

mg/L to 28 mg/L, respectively. These results appear to corroborate previous interpretations

made in Section II.3.4 where partial passivation was suggested for the mineral surface

during pre-oxidation due to iron hydroxides (FeOOH) deposition (Marsden and House,

2006).

Interesting observations can be made when the response of MRI-1 ore to pre-oxidation is

compared with the response of the other three minerals (Figures II.10 vs. Figures II.11-13).

76

The results suggest that the presence of sphalerite and chalcopyrite in close contact with

pyrite within the industrial gold ore could be the most important parameter affecting the

efficiency of the oxidative pre-treatment. Peters (1976) proposed that chalcopyrite and

sphalerite oxidize as follows in alkaline media:

CuFeS2 + 10H2O ↔ Cu(OH)2 + Fe2+

+2SO42-

+ 18H+ + 16e

- (II.4)

ZnS + 6H2O ↔ Zn(OH)2 + SO42-

+ 10H+ + 8e

- (II.5)

Due to high alkalinity (pH =11) and strong oxidation condition during pre-treatment

procedure, Fe2+

species are further oxidized into Fe3+

, which will precipitate as FeOOH

(Osseo-Asare et al., 1984). Moreover, the presence of pyrite in the industrial ore lattice acts

as a charge acceptor and theoretically promotes higher kinetics for reactions II.4 and II.5,

thus favoring accumulation of precipitates (i.e., FeOOH, Cu(OH)2 and Zn(OH)2) on the

MRI-1 surface. This accumulation can be held responsible for the reduction of MRI-1

reactivity and subsequent increase in gold leaching rate, as observed in Figures II.10a-c.

Thus, as the galvanically promoted dissolution of chalcopyrite and sphalerite advances

during pre-oxidation, the newly formed metal coatings which passivate the industrial ore

surface will play a crucial role in defining its reactivity during cyanidation.

The pre-oxidation treatment also had a beneficial effect on free cyanide consumption (i.e.,

lesser cyanide consumed) for both the industrial ore and its major sulfide constituents, with

an exception of chalcopyrite. Pre-oxidation of pyrite, sphalerite and industrial ore prior to

cyanidation leads to a decrease in cyanide consumption of 16%, 6% and, 25%, respectively

(data not shown). In the case of chalcopyrite, the results are less positive as pre-oxidation

increased copper leaching during cyanidation (see Figure II.12c) and, therefore, an over-

consumption of free cyanide.

Finally, it can be argued that studying the behavior of different sulfide minerals during

cyanidation cannot easily explain the behavior of an industrial sulfide-rich gold ore (MRI-

1). In other words, the global behavior of the industrial ore appears to be influenced by the

mineralogy and the topochemical environment of its constituent sulfides. The differences

obtained for the gold leaching rate, when the major sulfide components were tested

77

individually and in association within the industrial ore, can be assigned to the permanent

galvanic contacts between the major constituents in the industrial ore. However, further

investigations are needed to improve the fundamental understanding of the reactions and

mechanisms involved during pre-oxidation and/or cyanidation of the naturally associated

sulfides as in actual industrial ores.

II.3.7. Gold leaching, NaOH vs. Ca(OH)2 and pre-oxidation by DO2 vs.

H2O2

To mimic closely conditions similar to those encountered in gold plants, cyanidation

experiments were carried out in a slurry reactor using lime as an alkaline reagent and slurry

concentration at 50% solid by weight, everything else being kept identical as described

above (particle sizes (P80 <75µm), sodium cyanide and dissolved oxygen concentrations).

Cyanidation tests were performed to study the effect of lime (Ca(OH)2) on gold leaching

kinetics. The results of the experiments carried out using the Au/Ag RDE in aerated

cyanide solutions, show that the use of lime as alkaline reagent causes an important decline

in gold leaching when compared to cyanidation with sodium hydroxide (Figure II.14). In an

attempt to explain this phenomenon the Au/Ag electrode disc was analyzed by back scatted

electron imagining (SEM) coupled with X-ray energy dispersion spectroscopy (EDS). The

analysis showed that deposits containing calcium, oxygen and carbon were formed on the

gold surface, probably point to the formation of a passivating film of Ca(OH)2 and/or

CaCO3 (see also Li et al., 2006) on the gold surface, thus affecting gold leaching.

In a recent study, Davidson and sole (2007) suggested that the presence of lime causes the

formation of low-solubility calcium aurocyanide complex (Ca[Au(CN)2]2) which was held

responsible for the losses in gold recovery in plant circuits using lime to control pH.

Therefore, to check whether possible precipitation of aurocyanide complex with calcium

could also be involved in the observed decline in gold recovery (Figure II.14), a solution

containing 22.4 mg/L of Au(CN)2- was mixed during 1 h with 5 g/L of lime. These values

represented, respectively, the concentration of Au(CN)2- recovered from the gold electrode

in a clear solution in the presence of NaOH (black squares in Figure II.14) and the amount

of lime normally needed to maintain pH at 11.5 for 1 h in MRI-1 cyanidation tests. After 60

78

min of stirring, an aliquot was withdrawn and filtered on 0.45 m Millipore membrane filter

to remove solid particles, and then gold concentration was measured by atomic absorption.

The concentration of dissolved gold found (22.2 mg/L) indicated, in our case, that gold

surface passivation by gold-free calcium-bearing precipitates is the main cause for the

observed gold leaching losses in the presence of lime.

0,0

2,0

4,0

6,0

8,0

10,0

12,0

14,0

16,0

18,0

20,0

22,0

24,0

0 20 40 60 80

Au

leach

ed

(g

/m2)

Time (minutes)

Au/Ag-NaOH

Au/Ag-Ca(OH)2

Figure II-14 Effect of lime (Ca(OH)2) on the gold leaching rate. Reaction conditions: 50%

solids by mass, CN- = 20 mM, DO2 = 0.25 mM, pH = 11.5, electrode rotation rate = 500

rpm.

Figure II.15 shows the beneficial effect of the oxidative pre-treatment of MRI-1 ore, either

by dissolved oxygen or by hydrogen peroxide, on the gold leaching rate. Pre-oxidation tests

were conducted in the same manner as previously when dissolved oxygen was used as

oxidizing reagent. However, when hydrogen peroxide was used, 10 mL of H2O2 solution at

32% w was added to a suspension of 100 g of ore and 90 mL of distilled water and the pre-

oxidation time was varied from 1 h to 1.5 h. The results showed that the percentages of

gold recovery, calculated on the basis of gold leached in a clear solution after 60 min

(benchmark cyanidation test, black squares in Figure II.15) were 34%, 66%, 72% and 79%,

respectively, in direct cyanidation without pretreatment, cyanidation after pre-oxidation

79

with dissolved oxygen for 4 h, cyanidation after pre-oxidation with H2O2 for 1 h, and

cyanidation after pre-oxidation with H2O2 for 1.5 h. Pre-oxidation of industrial ore before

cyanidation may improve the dissolution rate of gold suggesting that this artifice can

become a useful tool to improve leaching of free and/or attached gold in sulfide-rich ores.

0,0

2,0

4,0

6,0

8,0

10,0

12,0

14,0

16,0

18,0

0 20 40 60 80

Au

leach

ed

(g

/m2)

Time (minutes)

Au/Ag

Au/Ag+MRI-1

Au/Ag+MRI-1;

Ox 4h (O2)

Au/Ag+MRI-1;

Ox 1h (H2O2)

Au/Ag+MRI-1;

Ox 1.5h (H2O2)

Figure II-15 Cyanidation of Au/Ag electrode in the presence of MRI-1 slurry: effect of the

type of pre-oxidant (O2 or H2O2) on gold leaching rate. Reaction conditions: 50% solids by

mass, CN- = 20 mM, DO2 = 0.25 mM, pH = 11.5, electrode rotation rate = 500 rpm, 3.2%

w H2O2.

II.4. Conclusion

An investigation aiming at an assessment of the relative importance of the phenomena of

passivation and galvanic interactions during the dissolution of gold in cyanidation

processes was conducted. Two mineral electrodes were prepared using a sulfide-rich

industrial gold ore (MRI-1) and its main sulfide component (pyrite), along with an Au/Ag

electrode. Cyanidation experiments were conducted by coupling the Au/Ag electrode with

one of the two mineral electrodes in one or two electrochemical cells. It was found that for

the case of tests in one electrochemical cell gold dissolution rate is controlled by positive

80

galvanic interactions rather than by passivation. However, when gold and mineral

electrodes are immersed in two distinct cells, gold leaching is controlled only by positive

galvanic effects giving rise to high galvanic currents.

To understand the influence of sulfide ores on gold cyanidation, a systematic study was

performed by immersing a gold/silver electrode disc successively in slurries containing the

industrial ore or one of its major sulfide constituents (pyrite, sphalerite and chalcopyrite). In

addition, solution speciation by capillary electrophoresis was used to quantify sulfur and

cyanicides for better depiction of the reactions taking place during cyanidation. The results

showed that all minerals had a negative effect on the leaching rate of gold. Moreover, the

mineralogy of the ore was found to directly influence its dissolution, both the nature and

the abundance of species released being affected by the galvanic interactions between the

main mineralogical phases present in the ore. This was also confirmed by the results

obtained when pre-oxidation was applied to the sulfide minerals. Pre-oxidation was

effective to improve gold leaching in the case of industrial ores while, despite a decrease in

cyanide consumption, no beneficial effect of pre-oxidation was observed for the major

sulfide constituents of MRI-1 ore.

Finally, it can be concluded that the behavior of an industrial sulfide-rich gold ore during

cyanidation cannot be easily explained only by studying the individual behavior of its

different sulfide mineral constituents. The global behavior of the industrial ore appears to

be influenced both by the mineralogy and the topochemical environment of its constituent

sulfides.

81

II.5. References

Aghamirian, M.M., Yen, W.T., 2005. Mechanism of galvanic interactions between gold

and sulfide minerals in cyanide solution. Minerals Engineering 18, 393-407.

Ahlberg, E., Broo, A.E., 1996a. Oxygen reduction at sulphide minerals. 1. A rotating ring

disc (RRDE) study at galena and pyrite. International Journal of Mineral Processing 46, 35-

52.

Ahlberg, E., Broo, A.E., 1996b. Oxygen reduction at sulphide minerals. 3. The effect of

surface pre-treatment on the oxygen reduction at pyrite. International Journal of Mineral

Processing 47, 49-60.

Bard, A.J., Startmann, M., 2006. Encyclopedia of Electrochemistry. In: Fritz, S., Pickett,

C.J. (Eds.), Inorganic Chemistry, vol. 7a. Wiley-VCH Verlag GmbH & Co. KGaA,

Weinheim, Germany.

Breuer, P.L., Hewitt, D.M., Jeffrey, M.I., Rumball, J.A, 2007. The leaching and oxidation

of sulfide minerals in cyanide solutions - Quantification of reaction products and the effect

of lead and oxygen. World Gold Conference. 183-189.

Breuer, P.L., Jeffrey, M.I., Hewitt, D.M., 2008. Mechanisms of sulfide ion oxidation during

cyanidation. Part I: The effect of lead (II) ions. Minerals Engineering 21, 579-586.

Buckley, A.N., Woods, R., 1987. The surface oxidation of pyrite. Applied Surfcae Science

27, 437-452.

Churchill, M., Laxen, P.A., 1966. The rotating disc system and its applications in the

dissolution of gold. Nat. Inst. Metall. Johannesburg (Mintek), pp.11.

Cruz, R., Bertrand, V., Monroy, M., González, I., 2001. Effect of sulfide impurities on the

reactivity of pyrite and pyritic concentrates: a multi-tool approach. Applied Geochemistry

16, 803-819.

Cruz, R., Luna-Sánchez, R.M., Lapidus, G.T., González, I., Monroy, M., 2005. An


mineral concentrates. Hydrometallurgy 78, 198-208.



Davidson, R.J., Sole, M.J., 2007. The major role played by calcium in gold plant circuit.

The Journal of the Southern African Institute of Mining and Metallurgy 107, 463-467.

Deschenes, G., Rousseau, M., Tardif, J., Prud’homme, P.J.H., 1998. Effect of the

composition of some sulfide minerals on cyanidation and use of lead nitrate and oxygen to

alleviate their impact. Hydrometallurgy 50, 205-221.

Elsner, L., 1846. Über das Verhalten regulinischer Mettale in einer wässrigen Lösung von

Cyankalium. Journal für Praktische Chemie 37, 441-446 (DOI:

10.1002/prac.18460370167).

82

Filmer, A.O., 1982. The dissolution of gold from roasted pyrite concentrations. The Journal

of the Southern African Institute of Mining and Metallurgy 82, 90-94.

Guo, H., Deschenes, G., Pratt, A., Fulton, M., Lastra, R., 2005. Leaching kinetics and


stibnite. Minerals and Metallurgical Processing 22, 89-95.

Habashi, F., 1967. Kinetics and mechanism of gold and silver dissolution in cyanide

solution. Montana Bureau of Mines and Geology Bulletin 59, Montana Bureau of Mines

and Geology, Butte, Montana, USA.

Hedley, N., Tabachnick, H., 1968. Mineral Dressing Notes No 23, Chemistry of

Cyanidation. American Cyanamid Company, New Jersey, USA.

Jeffrey, M.I., Breuer, P.L., 2000. The cyanide leaching of gold in solution containing

sulfide. Minerals Engineering 13, 1097-1106.

Jeffrey, M.I., Ritchie, I.M., 2000. The leaching of gold in cyanide solution in the presence

of impurities II. The effect of silver. Journal of Electrochemical Society 147, 3272-3276.

Levich, V.G., 1962. Physicochemical Hydrodynamics, Prentice Hall, Englewood Cliffs,

NJ.

Li, J., Dabrowski, B., Miller, J.D., Acar, S., Dietrich, M., LeVier, K.M., Wan, R.Y., 2006.

The influence of pyrite pre-oxidation on gold recovery by cyanidation. Minerals

Engineering 19, 883-895.

Liu, G.Q., Yen, W.T., 1995. Effects of sulfide minerals and dissolved oxygen on the gold

and silver dissolution in cyanide solution. Minerals Engineering 8, 111-123.



Lustemberg, P.G., Vericat, C., Benitez, G.A., Vela, M.E., Tognalli, N., Fainstein, A.,

Martiarena, M.L., Salvarezza, R.C., 2008. Spontaneously formed sulfur layers on gold in

electrolyte solutions: Adsorbed sulfur or gold sulfide? The Journal of Physical Chemistry

C. 112, 11394-11402.

Luthy, R.G., Bruce, S.G., 1979. Kinetics of reaction of cyanide and reduced sulfur species

in aqueous solution. Environmental Science & Technology 13, 1481-1487.

MacArthur, J. S., Forrest, R. W., Forrest, W, 1887. Improvements in obtaining gold and

silver from ores and other compounds. British Patent 14 174.

Madhuchhanda, M., Devi, N.B., Rao, K.S., Rath, P.C., Paramguru, R.K., 2000. Galvanic

interaction between sulfide minerals and pyrolusite. Journal of Solid State Electrochemistry

4, 189-198.

Marsden, J.O., House, C.I., 1992. The Chemistry of Gold Extraction. Ellis & Horwood,

London.

Marsden, J.O., House, C.I., 2006. The Chemistry of Gold Extraction. 2nd Edition. Society

for Mining, Metallurgy and Exploration (SME), Littleton, Co, USA.

83

Osseo-Asare, K., Xue, T. and Ciminelli, V.S.T., 1984, Solution chemistry of cyanide

leaching systems, in Precious Metals: Mining Extraction and Processing, The Metallurgical

Society/AIME., Warrendale, Pennsylvania, pp. 173-197.

Paul, R.L., 1984. The role of electrochemistry in the extraction of gold. Journal of

Electroanalytical Chemistry 168, 147-162.

Peters, E., 1976. Direct leaching of sulphides: Chemistry and applications. Metallurgical

Transactions B 7, 505-517.

Petre, C.F., Azizi, A., Olsen, C., Baçaoui, A., Larachi, F., 2008. Capillary electrophoretic

analysis of sulfur and cyanicides speciation during cyanidation of gold complex sulfidic

ores. Journal of Separation Science 31, 3902-3910.

Qing You, L., Heping, L., Li, Z., 2007. Study of galvanic interactions between pyrite and

chalcopyrite in a flowing system: implications for the environment. Environmental

Geology 52, 11-18.

Rand, D.J.A., 1977. Oxygen reduction on sulfide minerals. Part III. Comparison of

activities of various copper, iron, lead, and nickel mineral electrodes. Journal of


Weichselbaum, J., Tumilty, J.A., Schmidt, C.G., 1989. The effect of sulphide and lead on


Australasian Institute of Mining and Metallurgy, Melbourne, pp. 221-224.

Zhang, Y., Fang, Z.H., Muhammed, M., 1997. On the solution chemistry of cyanidation of

gold and silver bearing sulphide ores. A critical evaluation of thermodynamic calculations.

Hydrometallurgy 46, 251-269.

Zhu, X., Li, J., Wastworth, M.E., 1993. Kinetics of transpassive oxidation of pyrite. In:

Hager, J.P. (Ed.), EPD Congress. The Minerals, Metals & Materials Society 355-368.

84

CHAPITRE III. Untangling galvanic and passivation

phenomena induced by sulfide minerals on precious

metal leaching using a new packed-bed electrochemical

cyanidation reactor

Abdelaaziz Azizi,1 Catalin Florin Petre

1, Caroline Olsen

2, Faïçal Larachi

1

1Department of Chemical Engineering, Laval University, Québec, Canada, G1V 0A6;

2COREM Research Center, 1180 Rue de la Minéralogie, Québec, Canada, G1N 1X7

Abstract/Résumé

Because permanent galvanic contacts between gold rotating disc electrodes and slurried

sulfide-rich ores are uneasy to achieve, the standard gold rotating-disc-electrode/slurry

cyanidation arrangement has a tendency to inflate overly the importance of passivation

phenomena over the corrective trend of galvanic interactions inherently present within the

ore grains. A new packed-bed electrochemical reactor (PBER) was thus developed and

tested to decouple and quantify the individual contributions of passivation phenomena (PP)

and galvanic interactions (GI) on precious metal (gold and silver, PM) leaching rates during

the cyanidation of sulfide-rich ores. The PBER was filled with homogenized mixtures of

minerals (pyrite, chalcopyrite, sphalerite, chalcocite, galena, stibnite, industrial ore), gold

and silver powders, which were arranged as electrically-isolated sulfide-quartz segregated

bilayer(s) with permanent (inter)particle-particle electrical contacts among all constituents

in each layer. Implantation of PM powders successively in the quartz layer and then in the

sulfide layer, enabled PP and GI to be singled out and quantified separately for each

sulfide. Galvanic interactions were found to ameliorate, to various degrees, the leaching of

PM. Locating the PM powders in the quartz layer emphasized the negative impact of PP of

the sulfide layer on Au and Ag dissolution, except for galena which enhanced gold leaching

kinetics. Galvanic interactions due to pyrite, chalcopyrite and an industrial ore were so

85

positive that they largely outweighed the negative impact of PM passivation. The galvanic

current measured between Au/Ag electrode and each of the above minerals confirmed the

positive effect of GI on gold leaching rate. In close agreement with PM recoveries, stronger

galvanic currents were measured for Au-pyrite, Au-chalcopyrite and Au-industrial ore

galvanic couples in comparison to those for Au-sphalerite, Au-chalcocite and Au-stibnite.

Puisque les contacts galvaniques permanents entre électrode tournante à disque d’or (RDE)

et suspensions de minerais sulfureux (slurries) sont difficiles à mettre en œuvre, les tests

standards de cyanuration réalisés en mode RDE/slurry ont plutôt tendance à surestimer

l’importance des phénomènes de passivations sur l’effet aidant des interactions galvanique,

intrinsèquement présents dans les particules minérales. Un nouveau réacteur

électrochimique à lit fixe (PBER) a été alors développé et testé pour découpler et quantifier

les contributions individuelles des phénomènes de passivations (PP) et d’interactions

galvaniques (IG) sur la lixiviation de métaux précieux (or et argent, PM) lors de la

cyanuration de minerais riches en sulfures. Le réacteur a été chargé de mélanges

homogénéisés de sulfures (pyrite, chalcopyrite, sphalérite, chalcocite, galène, stibine et

minerai industriel) et de poudres d’or et d’argent, lesquels ont été disposés en bicouche(s),

électriquement-séparées (sulfure-quartz), où le contact galvanique permanent à l’échelle

inter-particulaire a été possible entre tous les constituants de chaque couche. L’implantation

des poudres d’or et d’argent (PM) successivement dans la couche de quartz puis dans la

couche de sulfure a permis de découpler les phénomènes de passivations des interactions

galvaniques et de les quantifier à titre individuel pour chaque sulfure. Les interactions

galvaniques améliorent, à des degrés variables, la lixiviation des métaux précieux. La

localisation des poudres de métaux précieux dans la couche de quartz a permis de souligner

l’impact négatif des phénomènes de passivation induite par les couches de sulfures sur la

dissolution de l’or et de l’argent, à l’exception de la galène qui influence positivement la

cinétique de lixiviation de l’or. Les interactions galvaniques dues à la pyrite, à la

chalcopyrite et au minerai industriel ont été tellement positive qu’elles l’emportent

largement sur les effets négatifs dus aux phénomènes de passivation. Le courant galvanique

mesuré entre l’électrode d’Au/Ag et chacun des sulfures précités confirme l’effet positif des

86

interactions galvaniques sur la vitesse de dissolution de l’or. Corroborant les résultats

obtenus sur la lixiviation de l’or, le courant galvanique mesuré a été plus élevé dans le cas

des couples galvaniques Au-pyrite, Au-chalcopyrite et Au-minerai industriel qu’avec les

couples Au-sphalérite, Au-chalcocite et Au-stibine.

III.1 Introduction

Precious metals (PM), gold in particular, naturally occur in contact within a wide range of

refractory sulfide mineral systems. Hence, one of today’s most challenging R&D tasks

consists in turning to account the presence of sulfides on the cyanidation behavior of gold.

An important body of knowledge reveals that metal sulfides display a range of reactivities

and interferences in alkaline cyanide solutions (Marsden and House, 1992; 2006).

Refractory ores genuinely reflect in fine Au inclusions entwined within the metal sulfide

matrix calling for its heedful breakage prior to cyanidation. Gold in the form of coarse

inclusions in the so-called transition sulfide ores is exposed and thus subject to three

specific interferences: (1) waste of reagents due to (cyanide and dissolved oxygen)

consumption by the sulfide matrix components (Habashi, 1967) increasing reagent cost or

decreasing gold leaching activity, (2) gold surface passivation by solid reaction products

from the reacting sulfides, e.g., precipitating dissolved iron or sulfur species on the gold

surface (Guo et al., 2005; Jeffrey and Breuer, 2000), or loss of dissolved gold cyanide by

re-precipitation on sulfide surfaces driven by surface Fe(OH)2 oxidation to Fe(OH)3 (Linge,

1995 and references therein), (3) galvanic interaction between the sulfide phase and gold

surface (Lorenzen and van Deventer, 1992; Aghamirian and Yen, 2005a).

In principle, gold anodic dissolution can be prompted when gold open circuit potential is

lesser than that of the coexisting sulfides. Conversely, when the adjoining sulfide mineral

acquires an anodic potential, galvanic interactions impede gold dissolution and prompt,

instead of gold dissolution, oxygen reduction on the gold surfaces. Apart from such a

straightforward dichotomy, the outcome from galvanic phenomena related to real ores is

difficult to anticipate as numerous experiments led to ambiguous electrochemical behaviors

of gold and their sulfide mineral components. Despite several studies on the subject, there

is still some controversy regarding the role of galvanic interactions on gold dissolution

87

(Lorenzen and Van Deventer, 1992; Aghamirian and Yen, 2005a; Dai and Jeffrey, 2006;

Azizi et al., 2010). Thus far, experimental strategies to study such interactions on gold

cyanidation used galvanic cells in which mineral and gold monolithic electrodes, either

naturally occurring or fabricated wafers, were short-circuited or attached together through a

conductor wire (Lorenzen and Van Deventer, 1992; Aghamirian and Yen, 2005a; Azizi et

al., 2010); the anodic and cathodic behaviors of gold and sulfide minerals being compared

individually (Dai and Jeffrey, 2006). However, in such cells, the contact between gold and

mineral is made via an external circuit and/or an electrolyte solution (Lorenzen and van

Deventer, 1992; Aghamirian and Yen, 2005a; Dai and Jeffrey, 2006; Azizi et al., 2010).

This might not be representative of the micro-environments present within the industrial

sulfide-gold ores, in which intimate permanent (intra)particle-particle contacts exist

between all the components forming the galvanic cell. Therefore, these experimental

strategies might not reflect the actual surface topochemistry of industrial ores in terms of

permanent galvanic contacts at an intra-particle scale. In the case of fabricated wafer

mineral electrodes, another disadvantage relates to their preparation which requires mixing

hydrophobic (binding) mineral oils with mineral and graphite powders to increase electrode

electric conductivity powders (Azizi et al., 2010). The presence of such foreign additives

does not guaranty electrochemical representativeness of the behavior of mineral particles.

In most of the available electrochemical studies, the effects of galvanic interactions (GI)

and passivation phenomena (PP) via mineral dissolution were often studied separately

because of limitations inherent to the experimental strategies (Lorenzen and van Deventer,

1992; Dai and Jeffrey, 2006). First, when using mineral electrodes, the concentrations of

species liberated in the solution were insufficient to allow investigation of those species’

effects on gold dissolution kinetics. In addition, the anode/cathode areal ratio being often

close to unity didn’t reflect the usual industrial cyanidation conditions. Setting

representative ratios between mineral and gold surfaces is crucial to capture both PP and GI

effects. Second, when gold leaching is studied in slurries containing sulfide mineral

particles and gold/silver powder (Guo et al., 2005; Dai and Jeffrey, 2006) or an Au/Ag disc

immersed in slurry (Lorenzen and van Deventer, 1992; Liu and Yen, 1995), permanent

galvanic contacts between gold and mineral particles, as in real ores, were virtually

impossible to achieve. Therefore, the effects of galvanic interactions on the speciation of

88

dissolved species were ignored leaving unexplored instances in which certain sulfides

would benefit from PM-mediated galvanic protection. It is believed that current

electrochemical strategies exhibit limited reliability for both prediction and assessment of

the various reactivity scenarios of Au, Ag and minerals in cyanidation circuits.

In consideration of the above restrictions, this work aims at the following targets:

1. To present a new simple electrochemical technique and methodology that can be

applied to understand and evaluate the effects of GI and PP on PM cyanidation

kinetics via sulfide minerals dissolution. This system, consisting of a packed-bed

electrochemical reactor (PBER), was developed and tested to investigate the

effects of a wide range of sulfide-rich ores on the PM leaching rates in aerated

cyanide solutions;

2. To attempt, using this new PBER configuration during cyanidation, sorting out

the individual contributions of galvanic interactions between Au, Ag and several

sulfide minerals and/or the formation of passivating layers on PM surfaces.

III.2. Experimental

III.2.1. Materials and reagents

Seven sulfide-rich ore samples were tested. The first sample (MRI-1) came from a mine in

the northwest of Québec province (Canada), while the other six samples were provided by

Ward’s Natural Science and referred to as pyrite (MRI-2), chalcopyrite (MRI-3), sphalerite

(MRI-4), chalcocite (MRI-5), galena (MRI-6) and stibnite (MRI-7) by virtue of the

preponderant sulfide mineral therein (Table III.1). The samples were sieved to remove

particles coarser then 90 µm and finer than 45 µm. Consequently, the same granulometric

fraction was used for the different ores, providing nearly equal total area per unit mass in

all cyanidation experiments. Tables III.1 and III.2 describe, respectively, the mineralogical

and elemental composition of each mineral. Pure gold (P80 = 39 µm, Alfa Aesar, 99,998%

USA) and pure silver (P80 = 26 µm, Alfa Aesar, 99,9% USA) powders were also used in

the present work. All powders were characterized using a combination of atomic absorption

89

spectrometry (AAS), scanning electron microscopy (SEM) and X-ray diffraction patterns

(XRD).

Distilled water was used for all cyanidation experiments. The reagents used in this study,

sodium cyanide, NaCN (98%, Sigma-Aldrich Canada), sodium hydroxide, NaOH (Fisher

Scientific Canada) and boric acid, H3BO3 (99.5%, Sigma-Aldrich Canada) were all certified

analytical grade.

Table III-1 Mineralogical composition of sulfide ore samples investigated in this study.

Ore /

Phase

Pyrite

(FeS2),

w%

Chalcopyrite

(CuFeS),

w%

Sphalerite

(ZnFeS2),

w%

Chalcocite

(Cu2S),

w%

Galena

(PbS),

w%

Stibnite

(Sb2S3),

w%

Gangue,

w%

MRI-1 44.2 0.11 0.31 tr. 0.05 tr. 55.4

MRI-2 96.6 tr. tr. tr. tr. tr. 3.4

MRI-3 0.8 63.2 18.3 tr. tr. tr. 17.7

MRI-4 4.7 tr. 90 tr. tr. tr. 5.3

MRI-5 17.4 tr. 0.03 57.6 0.1 tr. 24.8

MRI-6 1.2 0.1 0.03 tr. 97.1 tr. 1.56

MRI-7 2.5 0.09 0.04 tr. 0.02 49.5 47.9

tr.: trace ( 0.03 %w); gangue: SiO2 (maj.), Al2O3, CaO, MgO

Table III-2 Elemental composition of sulfide ore samples investigated in this study.

Ore /

Element

Au

mg/kg

Ag

mg/kg

Cu

w%

Zn

w%

Fe

w%

S total

w%

As

w%

Sb

w%

Pb

w%

MRI-1 1.02 17.75 0.03 0.18 20.6 26.8 0.02 tr. 0.67

MRI-2 tr. tr. 0.05 tr. 43 51 tr. tr. 0.04

MRI-3 tr. tr. 20.4 4.63 22.2 23.4 tr. tr. 0.35

MRI-4 tr. tr. 0.07 58 0.28 31 tr. tr. tr.

MRI-5 tr. tr. 46 0.02 8.9 20.9 tr. tr. 0.09

MRI-6 tr. tr. 0.04 0.02 0.05 13.3 tr. tr. 84.1

MRI-7 tr. tr. 0.03 0.03 0.18 15.2 tr. 35.6 0.02

tr.: trace ( 0.01 %w)

90

III.2.2. Equipment and procedures

III.2.2.1. Sulfide minerals and PM dissolution in a packed bed electrochemical reactor

The schematic of a new packed-bed electrochemical reactor (PBER) is shown in Fig. III.1.

The reactor, built from stainless-steel with Teflon interior lining, consisted of three

sections: the inlet, working and outlet sections. The stainless-steel reactor body had contact

neither with the powder particles nor with the cyanide solution. The cylindrical working

section (25 mm H x 15 mm I.D) was filled with homogenized mixtures of minerals, gold

and silver powders, arranged as fixed bed layer(s) with the following features: a)

establishment of permanent (inter)particle-particle electrical contacts among all

constituents; b) procurance of a porous three-dimensional mineral electrode with a high

relative (anodic or cathodic) surface area per unit volume with respect to that of gold and

silver (see Table III.3). The working section (2) was connected to the inlet (1) and outlet (3)

sections, which in turn were connected to the liquid pumping circuit (Fig. III.1). The

aerated cyanide solution feed was continuously recirculated in a closed loop through the

fixed bed using a peristaltic pump. The time on stream against which all the concentration

profiles of this study were plotted corresponded to the time the aerated cyanide solution

sojourned inside the PBER excluding the remainder of the transit time of flight in the

external loop of the liquid circuit. This time was computed after the bed porosity was

determined for every powder loading in the PBER and the prevailing cyanide solution

volumetric flow rate. This latter was kept constant at 10.4 mL/min for all the experiments.

A schematic diagram illustrating the mechanism of galvanic interactions between gold,

silver and a conductive sulfide mineral in a cyanide medium inside the packed bed is shown

in Fig. III.1 (inset). An advantage of using the PBER in gold cyanidation experiments is to

achieve very short residence times of cyanide solution in the packed bed layer, unlike in

conventional slurry systems. This will allow more trustful capture of the fastest formation

steps regarding cyanicides as well as any rapid deposits formation on the gold surface.

To prevent development of pH and CN- axial profiles along the PBER, the feed solution

was prepared by adding fairly large amounts of free cyanide to 0.01 M boric acid solution

which was buffered to pH 11 with NaOH and stored in a 250 mL magnetically-stirred glass

91

reservoir. Typical cyanidation experiments were carried out at room temperature by

pumping upwardly through the column, 100 mL of the above solution at a constant flow

rate of 10.4 mL/min. This resulted in PBER operating in a gradientless mode at a maximum

per-pass residence time of cyanide solution equal ca. 5 s for the sulfide layer and 4 s for the

silica layer. Air was bubbled through the reservoir to maintain constant dissolved oxygen

level (DO2 ~ 8.5 mg/L at 25ºC), monitored by means of a dissolved oxygen probe (DOB-

930 model from Omega, accuracy ±0.1 mg/L), while pH was maintained at 11 ± 0.01 using

an Oakton 1000 series pH-meter. During the course of experiments, small aliquots were

collected for chemical analysis from the reservoir using a syringe equipped with a particles

filter (VWR, 0.45µm).

O2/H2O

OH-

e-

e-

Gold

CN-

CN-

Au(CN)2-

Silver

Ag(CN)2-

Mineral

Mineral

Mineral

3

1

2

4

67

5

8

9

10

Figure III-1 Sketch of a new packed-bed electrochemical reactor (PBER). Legend as

follows: 1, inlet section; 2, working section (Teflon); 3, outlet section; 4, stainless-steel

cylinder; 5, sintered-glass distributor; 6, peristaltic pump; 7, magnetic stirrer; 8, oxygen

probe electrode; 9, pH-meter electrode; 10, air bubbling system.

Dissolution of gold and silver when in electrical contact with a mineral may be influenced

by GI or by the dissolving species liberated from the contiguous mineral (PP), or by

combinations thereof (Lorenzen and van Deventer, 1992). In targeting specifically the PP

effects caused by the dissolved species, the mineral and PM powders were packed in two

separate layers referred to as the bilayer case A. The mineral powder (4 g of one of MRI-1

92

to MRI-7) was packed in the lower part of the PBER working section and was electrically

isolated from the upper PM-containing powder layer by a sintered-glass filter disc (VWR

International, USA). This upper PBER powder layer consisted of a known amount of inert

quartz particles (P80 ≤ 149 µm Sigma Aldrich, Canada) in which 50 mg Au and 20 mg Ag

were dispersed. It is also worthy of mention that the quartz particles are chemically and

electrochemically inert thus having no effect whatsoever on the leaching of gold in aerated

cyanide solutions (Aghamirian and Yen, 2005b). Alternatively, the combined GI and PP

effects were assessed after the mineral and PM powders were mix-packed altogether. The

same amounts of PM powders (as in case A) were mixed with 4 g of mineral powder to

form the PBER lower-layer whereas the quartz upper layer was left unloaded with PM, thus

kept chemically and electrochemically inactive. This was referred to as the bilayer case B.

The sulfide/PM areal ratios relevant to case B are listed in Table III.3. Finally, a benchmark

test in which both PP and GI were simultaneously incapacitated was carried out and was

referred to as case C. It consisted of a quartz monolayer that filled the whole cylindrical

working section of the PBER after the sulfide mineral powder lower layer and the sintered-

glass filter disc were substituted by extra quartz particles. Powdered Au (50 mg) and Ag

(20 mg) were dispersed beforehand in the quartz monolayer for the benchmark test. In all of

cases A-C, to minimize adventitious losses of fine PM particles out of the PBER -especially

for the very high leaching levels-, the sintered glass discs (pore sizes < 2µm) were covered

on both sides by means of Whatman filter paper filters (porosity ˂ 1 µm).

Table III-3 Sulfide/PM areal ratios implemented in cyanidation experiments.

Areal ratio MRI-1 MRI-2 MRI-3 MRI- 4 MRI-5 MRI-6 MRI-7

MRI/Au 1476 1095 1390 1335 1138 738 1483

MRI/Ag 1263 936 1189 1142 973 631 1268

In all the experiments, the concentration of dissolved metals was measured using a Perkin

Elmer AA-800 atomic absorption spectrometer (AAS), while a capillary electrophoresis

instrument (CE, Agilent Technologies) was used to analyze the dissolved sulfur and

cyanicides (Azizi et al., 2010). The experimental analytical errors were estimated to vary

between 5% and 10% for the analyzed elements (Petre et al., 2008). Up to 4 repeat tests

under the same cyanidation operating conditions were carried out to judge of the tests’

93

reproducibility. In this work, gold and silver leaching performance was expressed in terms

of leaching conversion (or percentage) of dissolved metal with respect to the initial PM

loading. Residual (unreacted or free) cyanide was quantified using a silver nitrate titration

method with rhodamine as color-change indicator.

Although the PBER metal sulfide/PM weight ratios as used in this work do not reflect those

encountered in an industrial context, our choice of using 50 mg Au/ 4 g sulfide and 20 mg

Ag/ 4g sulfide was motivated by two factors constraints: 1) to achieve very short PBER

residence times by minimizing the amount of sulfide minerals; 2) as a requirement of the

CE protocol and its limited sensitivity, to identify the fastest kinetic steps by tracking more

accurately and in quantitative terms the most problematic cyanicides during gold leaching.

III.2.2.2. Electrochemical campaign

The electrochemical behavior of gold in electrical contact with various minerals was

studied in the same PBER using a gold/silver alloy (96%-4% by weight) electrode (Fig.

SD-1-1, supplementary data, Appendix B). The choice of this specific Au/Ag alloy

composition was explained elsewhere (Azizi et al., 2010). The gold alloy electrode, shaped

as a rectangular rod (14 mm x 3 mm x 1 mm), had an active surface area in contact with the

cyanidation medium of 1.18 cm2. The Au-Ag rod was threaded and attached to an insulated

platinum wire, itself attached to a carbon brush to maintain electrical contact between the

gold electrode and a potentiostat. The alloyed rod was inserted in the upper layer of the

working section that was packed with quartz particles.

The mineral electrodes were prepared by packing the lower part of the PBER working

section with 4 g of the targeted mineral (i.e., MRI-1 to MRI-7). A platinum spring was

inserted inside the mineral packed bed and an insulated platinum wire was used to establish

electrical contact (via a carbon brush) between the mineral layer and the potentiostat. The

PBER top and bottom layers were separated by an insulator sintered-glass disc filter to

prevent interlayer electrical contacts with the Au-Ag rod inserted in the upper quartz-

containing layer in a manner reminiscent of case A depicted above. The reference electrode

consisted of Ag/AgCl in saturated AgCl-KCl solution (+197 mV vs. standard hydrogen

electrode). A Luggin capillary tube housing the reference electrode was placed inside the

94

previously described 250 mL glass reservoir containing the NaCN solution which was

prepared using boric acid, adjusted to pH 11using NaOH, as background electrolyte. In this

three-electrode electrochemical cell configuration, the electrodes were exposed to the same

reacting solution while being recirculated through the reactor in an upward manner. Thus,

the established galvanic potentials and the galvanic corrosion currents flowing between the

upper and lower compartment pairs were recorded using the zero resistance ammeter

(ZRA) technique. A multichannel potentiostat-galvanostat (model VSP-27 from Bio-Logic

SA) was used along with EC-Lab software, allowing on-line data acquisition as time

elapses. Since Au/Ag was used as the main electrode in all our experiments, the current was

positive when the direction of electrons was from Au/Ag electrode, i.e., PM preferentially

corroding, towards the porous 3-D mineral electrode, while negative current values

corresponded to electrons moving in the opposite direction, i.e., PM cathodically protected.

Before each experiment, the Au/Ag electrode was systematically polished with sand paper,

followed by polishing with alumina particles and finally rinsed with nitric acid and distilled

water, while the platinum spring and the platinum wire were washed successively by nitric

acid and distilled water. Before starting the ZRA experiments, the gold and mineral

electrodes were maintained in aerated cyanide solution for about 30 min to allow the open

circuit potential (OCP) to stabilize.

III.3. Results and Discussion

III.3.1. Effect of Pyrite (MRI-2) on Gold and Silver Leaching

A strategy is devised to decouple and assess the contributions of passivation phenomena

(PP) and galvanic interactions (GI) on gold and silver leaching in the case of pyrite. Figs.

III.2a and III.2a` show the evolution of gold leaching as a function of time-on-stream and

of the actual (physical) time, respectively. The time on stream is expressed as multiples of

the PBER space time of the aerated cyanide solution for each circulation within the overall

loop, whereas the physical time represents the actual duration of the experiment (see

appendix C for more details).

95

In the benchmark test (case C), Au and Ag powders were mixed with quartz – that filled up

the whole PBER. Passivation phenomena were singled out by mixing Au and Ag only with

a quartz layer being juxtaposed to the PM-free pyrite (MRI-2) lower-layer (case A). The

global response from both PP and GI was obtained through mixing Au and Ag powders

with pyrite only (case B).

0

10

20

30

40

50

60

70

80

0 2 4 6 8 10 12 14 16 18 20

% A

u l

ea

ch

ed

Time-on-stream (min)

Au; case B

Ag; case C

Au; case C

Ag; case A

Au; case A

Ag; case B

(a)

Au; case B

Ag; case C

Au; case C

Ag; case A

Au; case A

Ag; case B

(a`)

96

2000

2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

-700

-650

-600

-550

-500

-450

-400

-350

-300

0 20 40 60 80 100 120 140 160 180

I (m

A/m

2)

E (m

V)

vs.

Ag

/Ag

Cl

Time (min)

(b)

Figure III-2 Effect of pyrite (MRI-2) on: (a and a`) gold and silver dissolution, (b) the


Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11.

When electrically in contact with pyrite (case B), gold particles exhibited the most intense

dissolution (filled squares) as compared to case A gold leaching (empty squares) where

physical contact between gold and pyrite particles was prevented or to case C alike (Fig.

III.2a). This behavior can be explained by a boosting effect of gold leaching owing to

positive galvanic interactions between pyrite and gold in case B.

The amounts of metallocyanide complexes formed in both cases A and B, as well as and

their cyanide consumptions, were comparable (results not shown). In the absence of PM-

pyrite galvanic interactions (case A), pyrite-induced passivation phenomena inflicted a

severe setback on Au leaching with respect to the benchmark test (Fig. III.2a). In the

presence of PM-pyrite GI (case B), a net promotion of gold leaching is observed largely

offsetting the PP detrimental effect. For instance, after 12 min of cyanidation, 74% of gold

was extracted under the combined PP and GI effects (case B) against only 1.5% under the

exclusive PP effect (case A) and 18% for case C. Hence, the GI-driven enhancement of

gold recovery would amount to 72.5%, whereas, losses on gold recovery due to PP alone

would amount to -16.5%.

97

Finally, Fig. III.2a also shows that Au recovery losses due to passivation (case A) when

compared to the benchmark test kept increasing with time. Similarly, a progressive increase

of gold recovery with time in the presence of GI was achieved, suggesting that positive

galvanic interactions between gold and pyrite outweighed durably the suppressive

passivation phenomena.

To confirm this hypothesis, electrochemical tests were performed according to the

methodology explained in section III.2.2.2 above. An Au-Ag rod (electrode) inserted in the

quartz upper layer was electrically connected through a zero resistance ammeter to the

pyrite lower-layer in the PBER via its platinum spring electrode (Fig. SD-1-1,

supplementary data, Appendix B). The recorded galvanic data are illustrated in Fig. III.2b

(galvanic current densities were expressed with respect to the Au-Ag rod surface area).

Once the electrical contact established, at first the galvanic current density suddenly

increased to a relatively high value (~ 3390 mA/m2) and then slowly decreased to stabilize

at ~ 2370 mA/m2 after 180 min of reaction. The same figure shows that the galvanic

potential between gold and pyrite remained relatively constant (around -400 mV) for the

entire experiment. Since the measured open circuit potential for pyrite (-180 mV vs.

Ag/AgCl) in aerated cyanide solution was drastically higher than that of gold (-600 mV vs.

Ag/AgCl), one can conclude that a galvanic cell was formed with Au acting as anode and

the pyrite surface as cathode. Therefore, this high potential difference between gold and

pyrite (420 mV) can certainly justify the measured high galvanic current (2370 mA/m2) in

Fig. III.2b.

Pyrite was previously shown to exhibit favorable electro-catalytic activity for oxygen

reduction (Rand, 1977; Aghamirian and Yen, 2005a). For example, Aghamirian and Yen

(2005a) reported, using a rotating disc electrode (RDE) system, relatively important

galvanic current densities (ca. 865 mA/m2) when gold and pyrite electrodes were coupled in

the same electrochemical cell. In the present study, the higher surface area provided by the

packed bed pyrite electrode when compared to the surface of Au-Ag electrode (713 cm2 vs.

1.18 cm2) seemed to ensure additional surface sites for oxygen reduction to occur. The

electric current drawn from the pyrite surface was likely to give rise to larger galvanic

currents as suggested by the mixed potential theory (Budruk-Abhijeet et al., 2008). This

98

was confirmed by the important current density recorded in Fig. III.2b. The decrease of

galvanic current density as time elapsed (Fig. III.2b) could be attributed to the formation of

passivating films and/or increase of passivation thickness on gold surface following the

release of several sulfide and iron species from the pyrite layer. The gold OCP shifted with

time to nobler values resulting in subsequent decrease of the potential difference with

respect to pyrite. On the other hand, a possible progressive formation of a protective layer

on the pyrite surface (likely ferric hydroxide) may reduce its activity toward oxygen

reduction which would also be reflected on the galvanic current.

Regarding silver cyanidation, it is clear from Fig. III.2a that when in electrical contact with

pyrite, Ag dissolves much faster (solid diamonds, case B) then when placed in an

electrochemically inauspicious environment (empty diamonds, case A) or when under

benchmark test conditions (solid triangles, case C). Furthermore, Fig. III.2a shows that the

positive galvanic effects between Ag and pyrite play the most important role in the leaching

kinetics of the former. It can be seen that under combined GI and PP effects, silver recovery

after 4 min of cyanidation had already amounted to ca. 90% of that accumulated after 20

min (case B); the latter representing more than 60 % of the total available Ag being leached

out. On the contrary, under the exclusive influence of PP (case A), only 15% of available

silver leached off after 20 min of cyanidation. Hence, as in the case of gold, galvanic

corrosion had a tremendous positive effect on silver dissolution, not only that the

detrimental effect of passivation was neutralized but also an important acceleration of Ag

dissolution rate was prompted. The effect of Au/Pyrite and Ag/pyrite galvanic associations

on the speciation of pyrite reaction products species are illustrated in Fig. SD-1-2

(supplementary data, appendix B).

III.3.2. Effect of Chalcopyrite (MRI-3) on Gold and Silver Leaching

The effect of chalcopyrite (MRI-3) on PM recovery was also investigated in both presence

(case B) and absence (case A) of permanent galvanic contacts. As illustrated in Fig. III.3a,

mixing PM and MRI-3 in the same layer (case B) led to measured gold recovery (solid

squares) higher than the one obtained in the benchmark test (solid triangles). However, Au

recovery remained low (ca. 31% after 12 min) in comparison with the sibling case B for

pyrite (Fig. III.2a). This important setback in gold leaching rate was qualitatively similar to

99

the one diagnosed when a gold-containing rotating disc electrode was immersed in

chalcopyrite slurry (Azizi et al., 2010). It was then interpreted as resulting from conjugated

negative effects of rapid depletion of free cyanide through copper complexation along with

gold surface passivation by hydrosulfide ions whose concentration was indeed objectified

by capillary electrophoresis (ca. 5 mg/L). This same interpretative phenomenology seems

also not to contradict the present findings. Disabling GI effect as highlighted by case A test

(Fig. III.3a, empty squares), led to Au recovery of only ca. 2 % after 12 min. In the present

study, an excess of free cyanide (30 mM CN-) was used and ca. 55% of the initial free

cyanide was still available at the end of experiment. Hence, hydrosulfide-mediated gold

passivation appeared to play an extremely suppressive role affecting gold recovery,

relegating cyanide starvation, in all likelihood, to a minor role. With reference to Fig. III.3a,

gold recovery added by the sole action of galvanic interactions (case B vs. case A) and gold

recovery losses (case A vs. case C) taken away due to the sole action of passivation

phenomena amounted, respectively, to 29% and -16% after 12 min of cyanidation.

0

10

20

30

40

50

60

0 2 4 6 8 10 12 14 16 18

% M

eta

ls l

ea

ch

ed


Ag; Case C

Ag; Case A

Ag; Case B

Au; Case B

Au; Case A

Au; Case C

(a)

100

1000

1100

1200

1300

1400

1500

1600

1700

1800

-700

-650

-600

-550

-500

-450

-400

0 20 40 60 80 100 120 140 160 180

I (m

A/m

2)

E (m

V)

vs.

Ag

/Ag

Cl


(b)

Figure III-3 Effect of chalcopyrite (MRI-3) on: (a) gold and silver dissolution, (b) the



Enabling GI (case B) increased Au recovery by more than 10% comparatively to case C.

Galvanic corrosion not only attenuated passivation but also increased Au recovery. These

findings align qualitatively with those depicted earlier in the case of pyrite (Fig. III.2a) that

the presence of galvanic contacts mitigates the magnitude of passivation.

The results of measured galvanic current and potential between gold and chalcopyrite are

shown in Fig. III.3b. A large value of galvanic current (~ 1650 mA/m2) was measured after

180 min. However, the amplitude and variation intervals of measured galvanic current were

both lesser than those for pyrite (Fig. III.2b) despite a chalcopyrite slightly higher surface

area in comparison to pyrite (900 cm2 vs. 713 cm

2). The measured OCP of chalcopyrite and

gold were, respectively, -344 mV and -600 mV (vs. Ag/AgCl electrode) in aerated cyanide

solution, resulting in a lower potential difference (256 mV) as compared to the Au/pyrite

system (Fig. III.2b). Aghamirian and Yen (2005a) demonstrated that the Tafel slope for

oxygen reduction on chalcopyrite is lesser than that for pyrite; the former hence drawing

less cathodic current than the latter. Furthermore, in the present study, the higher content of

resistive material within chalcopyrite (gangue and sphalerite, Table III.1) in comparison to

101

pyrite (gangue, Table III.1) aligns also with the lower chalcopyrite oxygen reduction

activity. Lower cathodic current would be tantamount to lower galvanic current (Figs.

III.2b, III.3b), resulting in lower galvanic corrosion of gold with chalcopyrite (31% Au

leached in 12 min) than with pyrite (74 % Au leached in 12 min) in the bilayer case B tests.

Unlike gold, the presence of chalcopyrite did not appear to restrict the reactivity of silver

during cyanidation. The results shown in Fig. 3a reveal that both in the presence (case B) or

absence (case A) of galvanic contacts between Ag and chalcopyrite, silver dissolution

profiles remained confounded and coincided, nearly on the dot, with the leach profile for

the benchmark test (case C).

Figs. SD-1-3a,b (supplementary data, appendix B) show that enabling galvanic interactions

between Au and chalcopyrite helped reducing perceptibly the concentrations of sulfate,

thiocyanate and cyanocupro-complex during cyanidation. Also, the effect of galvanic

interactions between gold and MRI-3 on the Fe-cyanide complexes speciation is illustrated

in Fig. SD-1-4a,b (supplementary data, appendix B). A galvanic protection by gold anodic

dissolution appeared to have restricted chalcopyrite reactivity. As a matter of fact, the

measured concentrations of SO42-

and SCN- after 22 min of cyanidation time decreased

from 132 mg/L and 65 mg/L (case A) to, respectively, 65 mg/L and 35 mg/L (case B) by

enabling the galvanic interaction (Fig. SD-1-3a, supplementary data, appendix B). The

same trend was also observed, to a lesser extent though, for copper cyanide concentration,

which decreased from 91 mg/L to 77 mg/L (Fig. SD-1-3b, supplementary data, appendix

B). The kinetics of copper cyanide complex formation revealed to be relatively fast; more

than 50% of the complex final concentration after 23 min appeared to have formed during

the first minute. This suggests that mitigation measures to curb the buildup of such

cyanicides should be efficient over the first moments of cyanidation.

III.3.3. Effect of Sphalerite (MRI-4) on Gold and Silver Leaching

Unlike pyrite and chalcopyrite, the presence of sphalerite (MRI-4) induced a strong

retarding effect on gold leaching kinetics both in the presence or absence of GI (Fig. III.4a).

Since cyanide consumption was relatively low in both cases (no more than 27% of initial

free cyanide), the drastic decrease of gold recovery is likely to be ascribed to surface

102

obstruction by passivating films resulting from sphalerite dissolution in aerated cyanide

solution. As can seen from Fig. III.4a, this time the positive effect of GI between Au and

MRI-4 was not sufficient to overcome the negative effect of PP (case B vs. case A),

resulting in gold recovery much lower than with respect to the benchmark test (case C). As

a matter of fact, gold recovery after 12 min of cyanidation decreased from 18% for the

benchmark test to 4% when Au and MRI-4 were inter-mixed and then to only 1. % when

kept apart. The recovery loss taken away because of passivation phenomena alone

amounted to -17 %; whereas the recovery loss due to the combined GI and PP effects

attained -14%.

Fig. III.4b shows that the galvanic current density between gold and sphalerite in aerated

cyanide solution was generally low (< 40 mA/m2). Once Au (OCP = -600mV vs. Ag/AgCl)

and sphalerite (OCP = -137mV vs. Ag/AgCl) electrodes were electrically connected

through ZRA, the galvanic current density rapidly peaked nearby 40 mA/m2 and then

gradually decreases to ca. 14 mA/m2 after 180 min. These values are much lower that the

0

10

20

30

40

50

60

70

0

5

10

15

20

25

0 2 4 6 8 10 12 14 16 18 20

% A

g l

ea

ch

ed

% A

u l

ea

ch

ed

Time-on-stram (min)

Au; Case C

Ag; Case B

Ag; Case C

Ag; Case A

Au; Case B

Au; Case A

(a)

103

0

5

10

15

20

25

30

35

40

45

50

-600

-580

-560

-540

-520

-500

0 20 40 60 80 100 120 140 160 180

I m

A/m

2)

E (m

V)

vs.

Ag

/Ag

Cl

Time (min)

(b)

Figure III-4 Effect of sphalerite (MRI-4) on: (a) gold and silver dissolution, (b) the



galvanic current densities measured in the same conditions for Au-pyrite and Au-

chalcopyrite systems. Such a low galvanic current is at odds with the large cathode surface

area available for oxygen reduction (about 870 cm2) and the potential difference between

Au and sphalerite (ca. 463 mV) which is larger than with pyrite (420 mV) or with

chalcopyrite (256 mV). One possible explanation could be the low electrical conductivity

of sphalerite that makes it unable to maintain electron transfer on its surface necessary for

oxygen reduction (Mirnezami et al., 2003). Thus, the kinetics of oxygen reduction on the

mineral cathodic surface might be electron-transfer limited with respect to adsorbed oxygen

molecules, with direct effect on gold galvanic corrosion performances.

Fig. III.4a also shows that silver surface passivation by sphalerite led to significant losses in

silver recovery (case A). However, unlike gold, GI (case B) was efficient enough to

overcome the PP detrimental effect resulting in some increase in silver recovery in

comparison to benchmark C case.

104

Figs. III.5a,b show the influence of GI on the speciation of sphalerite reaction products in

aerated cyanide solution. The period that preceded the formation of SO32-

, SO42-

and SCN-

was observed to significantly increase in the presence of galvanic contacts between PM and

sphalerite. This may suggest that some sphalerite intermediate oxidation products were

formed and then their further oxidation resulted in the simultaneous formation of SO32-

,

SO42-

and SCN- as detected by capillary electrophoresis (CE). These explanations are in

agreement with previous findings (Luthy and Bruce, 1979) that indicated that the oxidation

products of hydrosulfide/polysulfide ions in oxygenated cyanide solutions are mainly SO32-

,

S2O32-

and SCN-, along with some SO4

2-. Therefore, the negative effect of sphalerite on

gold cyanidation could be interpreted by the action of the intermediate products (HS- and/or

Sx2-

), which are known to be responsible for gold surface passivation. Another interesting

observation is the formation of a large concentration of SO42-

when GI was set on (case B)

comparatively to the case where GI was set off (case A). Thus, one can interpret that the

micro-galvanic contacts between the PM and sphalerite could contribute to prevent

passivation by accelerating the oxidation of unstable sulfur-bearing ions to inert oxidation

product (i.e., SO42-

). Though, further investigation is needed to confirm this hypothesis.

Finally, the low GI induced only a small cathodic protection of sphalerite, which resulted in

a slight decrease in zinc cyanide formation (Fig. III.5b) when compared to the case without

active GI.

105

0

2

4

6

8

10

12

14

16

18

20

0

2

4

6

8

10

12

14

16

18

20

0 2 4 6 8 10 12 14 16 18 20

SO

42

-le

ach

ed

(m

g/L

)

SO

32

-le

ach

ed

(m

g/L

)


Case B

Case ACase B

Case A

(a)

0

2

4

6

8

10

12

14

16

18

20

0

2

4

6

8

10

0 2 4 6 8 10 12 14 16 18 20

Zn

(C

N) 4

2-

lea

ch

ed

(m

g/L

)

SC

N-

lea

ch

ed

(m

g/L

)


Case B

Case A

Case A

Case B

(b)

Figure III-5 Effect of galvanic interactions between gold and MRI-4 on: (a) oxy-sulfur ions

speciation and (b) cyanicides formation. Reaction conditions: CN- = 30 mM, DO2 = 0.25

mM, pH = 11.

106

III.3.4. Effect of Chalcocite (MRI-5) on Gold and Silver Leaching

The effect of chalcocite on the gold leaching kinetics was also investigated and the results

are shown in Fig. III.6a. Gold dissolution was for all practical purposes quasi-nil (0.65 %)

after 20 min of cyanidation when gold and chalcocite were in contact (case B, solid

squares), whilst in case A when no contacts were allowed gold dissolution was completely

blocked (empty squares). Two striking features accredited to chalcocite were depletion of

free cyanide and surface passivation of gold. Unlike for the previous sulfides tested in this

study, virtually the whole available cyanide was consumed after 20 min in the case of

chalcocite. In addition, case B galvanic corrosion was next to helpless leading to tiny Au

dissolution as compared to case A leaching test.

Regardless of PBER modality, i.e., whether GI disabled or enabled, silver dissolution was

below detection limits (data not shown) which was explained, as will be discussed next, by

HS--driven passivation of silver surfaces. The role of hydrosulfide in reacting with silver

ions to yield stable Ag2S was already discussed in the literature (Licht, 1988).

Fig. III.6b shows the time evolution of the galvanic potential and current for the Au-

chalcocite assemblage. Although with comparable surface area (about 740 cm2) to that of

pyrite, chalcocite showed a very low negative galvanic current, rapidly decreasing from 68

mA/m2 (absolute value) at the beginning of the test to nearly plateauing around 20 mA/m

2

(absolute value). After 40 min, it again suddenly increased to ca. 76 mA/m2 (absolute

value) then gradually decreasing back to ca. 16 mA/m2 (absolute value) towards the end.

With time, the galvanic potential shifted to less negative values indicating ennoblement or

passivation of gold (Blasco-Tamarit et al., 2009). The galvanic current density being

negative in addition to chalcocite OCP (-722 mV vs. Ag/AgCl) being more negative than

Au’s (-600 mV vs. Ag/AgCl) both pointed out to chalcocite taking part in a prevailing

anodic process. However, this appears to be in contradiction with the small beneficial effect

of the Au/MRI-5 galvanic contact on gold recovery observed in Fig. III.6a (case B vs. case

A).

107

0

5

10

15

20

25

30

35

40

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,8

2

0 2 4 6 8 10 12 14 16 18 20 22 24

% A

u l

ea

ch

ed

% A

u l

ea

ch

ed


Case C

Case B

Case A

(a)

-100

-80

-60

-40

-20

0

20

-700

-600

-500

-400

-300

-200

-100

0 20 40 60 80 100 120 140 160 180

I (

mA

/m2)

E (m

V)

vs.

Ag

Ag

Cl

Time (min)

(b)

Figure III-6 Effect of chalcocite (MRI-5) on: (a) gold and silver dissolution and (b) the



108

The dissolution products of chalcocite were also studied to gain further insights on the

associated gold/chalcocite electrochemical behavior during cyanidation. The results showed

that chalcocite dissolves readily in aerated cyanide solution to form substantial amounts of

copper cyanide (around 800 mg/L, copper basis) and hydrosulfide (Fig. III.7), along with

some thiocyanate and sulfate (results not shown). Chalcocite dissolution proceeded much

faster and more massively than gold’s (Figs. III.6a, III.7): more than 80% of leachable

copper versus 0.5% for Au within first 2 min. Hydrosulfide formation peaked especially in

the early reaction stages (Fig. III.7, 0-2 min) subsequently vanishing presumably via

oxidation towards more stable sulfur-oxyanions (i.e., SO42-

and SCN-). Hydrosulfide-driven

formation of a shielding Au/Sx film (Weichselbaum et al., 1989; Jeffrey and Breuer, 2000),

presumably taking effect early in the reaction, combined with rapid depletion of free

cyanide via copper complexation-dissolution worsened even further Au recovery.

0

200

400

600

800

1000

1200

0

10

20

30

40

50

60

70

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Cu

-Cu

(CN

) x(1

-x)-

HS

-le

ach

ed

(m

g/L

)


Cases A and B

Case A

Case B

Figure III-7 Effect of galvanic interactions between gold and MRI-5 on the reactivity of

MRI-5. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11.

109

Chalcocite dissolution in cyanide solution was suggested to be a non-oxidative process

(Fisher, 1994). The sudden consumption of hydrosulfide subsequent to its formation in the

early cyanidation stage (Fig. III.7) coincided quite well with the observed sudden increase

in galvanic current (Fig. III.6b). Such a coincidence could simply be attributed to HS-

oxidation on the surface of chalcocite resulting in a net electron flow from the mineral

electrode to the gold electrode (negative galvanic current). Finally, the galvanic current

gradually subsided (Fig. III.6b) once hydrosulfide was completely oxidized (Fig. III.7).

III.3.5. Effect of galena (MRI-6) on gold and silver leaching

Fig. III.8a demonstrates the beneficial effect of galena on gold and silver leaching in

aerated cyanide solution. In the presence of galvanic interactions (Au, case B), galena

offers a spectacular improvement of the kinetics of gold dissolution as compared to the

benchmark test (Au, case C). The results are showing that more than 90% of gold was

extracted after only 4 min of cyanidation; more than a 10 fold improvement in comparison

with the benchmark test at 4 min time-on-stream. Furthermore, even by disabling GI (case

A) a net improvement in Au recovery was achieved comparatively to the benchmark test

(Au; case A vs. Au; case C) unlike the above studied sulfides. Similar trends were also

observed for silver leaching. Although no attempt was made to monitor the concentration

of Pb in solution, lead cations released from galena when GI was disabled are believed to

be responsible for the improvement of gold recovery.

Galena is known to be soluble to some extent in aerated cyanide solution. Thus,

discharging lead (II) cations into the solution was put forward to explain the higher gold

recovery observed in the presence of this mineral (Jeffrey and Ritchie, 2000; Guo et al.,

2005; Senanayake, 2008). Jeffrey and Ritchie (2000) showed that cementation of lead on

gold surface disrupts the AuCN passivating film and increases significantly both gold

oxidation and oxygen reduction half-reactions, resulting in a net acceleration of gold

dissolution. In the present study, the extra cathodic surface area provided by the porous

packed-bed galena electrode would explain the additional amount of gold recovery

observed when gold and galena were mixed in the same layer. For example, after 12 min of

110

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10 12 14 16 18 20

% L

ea

ch

ed

meta

ls


Ag; Case B

Au; Case B

Ag; Case A

Ag; Case C

Au; Case C

Au; Case A

(a)

1200

1300

1400

1500

1600

1700

-660

-640

-620

-600

-580

-560

0 20 40 60 80 100 120 140 160 180

I (m

A/m

2)

E (m

V)

vs.

Ag

/Ag

Cl

Time (min)

(b)

Figure III-8 Effect of galena (MRI-6) on: (a) gold and silver dissolution and (b) the



111

cyanidation gold recovery soared from 18% in the benchmark test to 68% with the GI

disabled (case A) and then to over 90% with GI enabled (case B). It can be suggested that

galena has a double positive effect on gold leaching: i) activation of gold surface by

dissolved species improving by up to 40% gold recovery; ii) acceleration of gold leaching

rate by positive GI prompting a further 22% gain in gold recovery.

The galvanic corrosion current density and potential registered in aerated cyanide solution

for the gold (OCP = -600 mV vs. Ag/AgCl) and galena (OCP = -110 mV vs. Ag/AgCl) pair

are presented in Fig. III.8b. The galvanic current density rapidly plateaued at a relatively

high value (~1560 mA/m2) and remained constant for all the duration of the experiment.

Lower galvanic current and larger OCP difference between gold and galena contrasted with

those corresponding to the gold-pyrite galvanic couple (Figs. III.2b and III.8b). This is not

unlikely knowing that the surface area used for the galena packed bed electrode (about 480

cm2) is lower than that for pyrite (730 cm

2). Furthermore, Aghamirian and Yen (2005a)

also demonstrated that the reduction of oxygen occurs more easily on the surface of pyrite

than on the surface of galena. In other words, the magnitude of galvanic current is limited

by the extent of the half reaction of oxygen reduction.

Fig. III.8a shows that galena also had a remarkably promoting effect on the kinetics of

silver dissolution. Under combined GI and PP effects, the maximum recorded silver

recovery (90%) was attained after only 4 min of cyanidation. In the absence of GI, a 70%

Ag recovery was recorded after 14 min, in excess of that of the benchmark test (45%).

III.3.6. Effect of stibnite (MRI-7) on gold and silver leaching

Gold extraction profiles in the presence of Stibnite (MRI-7) are illustrated in Fig. III.9a.

The results show that in both the presence and absence of gold-stibnite GI, stibnite severely

hindered gold leaching kinetics. The dissolution of gold was marginally promoted by

enabling galvanic interactions (case B) in comparison to the case with disabled GI (case A),

though in both instances, gold recovery was much less than in the benchmark test (case C).

112

5

6

7

8

9

10

-450

-440

-430

-420

-410

-400

0 20 40 60 80 100 120 140 160 180

I (

mA

/m2)

E (m

V)

vs.

Ag

/Ag

Cl

Time (min)

(b)

Figure III-9 Effect of stibnite (MRI-7) on: (a) gold dissolution, (b) the evolution of galvanic

potential and galvanic current vs. time between Au and MRI-7. Reaction conditions: CN- =

30 mM, DO2 = 0.25 mM, pH = 11.

0

5

10

15

20

25

30

35

40

0

0,2

0,4

0,6

0,8

1

1,2

1,4

0 2 4 6 8 10 12 14 16 18 20 22 24

% A

u l

ea

ch

ed

Time-on-Stream (min)

Case C

Case B

Case A

(a)

113

After 12 min, gold extraction dropped from 18% in the benchmark test to 0.3% under

combined GI and PP effects and down to 0.1% under the sole effect of PP. These results

corroborate those of Liu and Yen (1995), Hollow et al. (2003a) and Guo et al. (2005) who

reported that stibnite, even at a very low concentrations, had a significant retarding effect

on gold leaching. Since less than 20% of the initial free cyanide was ultimately consumed,

the significant drop of gold recovery could likely be attributed to some passivation. This

effect was ascribed to the formation of a passivating film of antimony oxide although the

exact nature of this passivating layer was not yet exactly known.

Since the measured OCP in aerated cyanide solution of gold (-600mV vs. Ag/AgCl) was

much lower than for stibnite (-264 mV vs. Ag/AgCl), a large and positive galvanic current

would be expected between gold and stibnite. On the contrary, Fig. III.9b shows that the

galvanic current established between the Au-stibnite pair was very low, likely resulting

from surface obstructions of gold by an antimony oxide layer. Despite a large surface area

available for oxygen reduction provided by the mineral electrode, cyanide diffusion to gold

surface would be impaired and the magnitude of galvanic current (and that of gold

recovery) would only be controlled by the low anodic current.

Expectably, stibnite also had an important negative effect on silver dissolution (results not

shown). Here, the galvanic interactions did not have any significant effect on the kinetics of

silver cyanidation. Another aspect worth mentioning is that, unlike gold, Ag recovery in the

presence of stibnite was larger (5% after 12 min) than in the presence of chalcocite (0%

after 12 min), suggesting that silver recovery is probably more affected by the presence of

dissolved sulfur than dissolved antimony.

III.3.7. Effect of industrial gold-containing ore (MRI-1) on gold leaching

The experimental results shown in Fig. III.10a illustrate the effect of an industrial sulfide-

rich ore (MRI-1) on the gold leaching. When brought into physical contact (case B), MRI-1

prompted the highest dissolution of gold in comparison with both the benchmark test (case

C) and the Au-MRI-1 contactless test (case A, empty squares).

114

0

10

20

30

40

50

60

70

0 2 4 6 8 10 12 14 16 18 20 22 24

% A

u l

ea

ch

ed


Case B

Case C

Case A

(a)

100

200

300

400

500

600

700

-700

-660

-620

-580

-540

-500

0 20 40 60 80 100 120 140 160 180

I (m

A/m

2)

E (m

V)

vs.

Ag

/Ag

Cl

Time (min)

(b)

Figure III-10 Effect of the industrial gold ore (MRI-1) on: (a) gold dissolution and (b) the



115

After 12 min of cyanidation, 44% of gold was extracted for case B combining both GI and

PP, whereas only 7% was extracted under the exclusive PP effect in case A. Since in this

latter case, plenty of free cyanide, ca. 70% was left unreacted at the end of the reaction, it is

reasonable to conclude that this ore was prone to passivation phenomena. Therefore, the

gain in gold recovery due to galvanic corrosion and gold losses due to PP were,

respectively, 37% and -11%. Galvanic corrosion measurements shown in Fig. III.10b also

confirmed the positive effect on the gold leaching rate of the micro-galvanic contacts

between Au and MRI-1 particles. Once the gold and mineral electrodes were electrically

connected, the galvanic current gradually increased with time to plateau near 635 mA/m2

after 180 min.

Comparative analysis of the results obtained with the industrial ore (MRI-1) and those

obtained with its major sulfide constituents (i.e., pyrite, chalcopyrite and sphalerite) has

been conducted. Under the combined effects of GI and PP, it was observed that sphalerite

brought an important reduction in gold leaching rate whereas chalcopyrite, industrial ore

and pyrite counterbalanced, more or less successfully, the negative effect of passivation and

significantly increased gold leaching. The following upward gradation takes place in terms

of gold dissolution: chalcopyrite > industrial ore > pyrite, as shown in Table III.4.

Moreover, the greatest improvement in gold leaching when gold is in mutual contact with

pyrite is coherent with the highest galvanic corrosion currents achieved (Figs. III.2b, III.3b,

and III.10b). Nevertheless, despite galvanic interactions between gold and industrial ore

were less important than those between gold and chalcopyrite (Figs. III.3b and III.10b),

MRI-1 led to better gold leaching when compared with chalcopyrite. Such ambiguous

results could be explained by a possible interdependence between galvanic interactions and

passivation phenomena in the PM-containing sulfide ores.

It is worth noting that the gradation of the gold recovery in the presence of sulfides as

obtained in the PBER configuration is not completely in agreement with the gradation

established in our previous study (Azizi et al., 2010; chapter II) where the leaching

experiments were carried out using a RDE Au/Ag disc immersed in a slurry reactor of the

same powdered sulfide minerals, see Table III.4 (Au/Ag disc in RDE/slurry vs. PM in

PBER).

116

Table III-4 Relative gold recoveries (normalized with respect to silica base cases) from

Au/Ag disc immersed in slurries of minerals and from PM-mineral mixtures (case B) in

PBER.

Relative dissolution of gold (%)

Au/Ag disc in slurry reactor* PM-mineral PBER mixtures**

Silica 100 100

MRI-1 23 244

MRI-2 31.5 411

MRI-3 2.5 172

MRI-4 12 22

*after 60 min cyanidation, **after 12 min (time-on-stream) cyanidation

In contrast to the results obtained with the PBER, it can be seen from Table III.4 that

chalcopyrite, pyrite and industrial ore in RDE/slurry mode significantly decreased the gold

dissolution. These results can now be explained by the fact that when the Au/Ag disc

electrode was immersed in slurry containing one of the above ores there was no permanent

galvanic contact between the gold electrode and the mineral particles. Therefore, gold

RDE-slurry cyanidation experiments had a tendency to inflate overly the importance of

passivation phenomena over the corrective trend of galvanic interactions inherently present

within the ore grains as emulated in our PBER gold leaching study. All above results and

observations point out that, up to now, the new PBER constitutes a potent electrochemical

tool to assess separately and in quantitative terms the effects of GI and PP on Au leaching

rate.

III.3.8. Effect of silver on gold leaching in the presence of sulfides

Gold dissolution mechanisms proceeding via the formation of adsorbed AuCNads

monolayer or more complex multi-layer were proposed (Cathro and Koch, 1964; Kirk et

al., 1978; Kirk and Foulkes, 1980; Sawaguchi et al., 1995). Jeffery and Ritchie (2000)

showed that trace amounts of silver (1-5%) present in gold-silver alloys significantly

enhance Au leaching by increasing both gold oxidation and oxygen reduction half

reactions. Bimetallic corrosion was proposed to occur with oxygen preferentially reducing

at the silver active sites. Other investigations (Sun et al., 1996; Wadsworth and Zhu, 2003)

showed that dissolved silver cyanide effectively enhances gold dissolution, to a lesser

117

extent though in comparison with gold-silver alloys. Even if the effect of silver on gold

leaching in clear cyanide solutions was widely studied, similar effects in the presence of

sulfide minerals were seldom addressed. Gold cyanidation experiments were thus

performed in the presence of pyrite, which represents the sulfide mineral most commonly

associated with gold (Fig. III.11).

First, in the absence of galvanic interactions (case A) between Au and pyrite, mixing silver

with pyrite promoted higher gold dissolution (empty circles-dashed line) in comparison to

the test where silver was mixed with quartz (empty squares-solid line). These results come

in support of the previous literature findings, as silver cyanide at high concentration is

likely to form as a result of Ag-pyrite galvanic interaction (Fig. III.2a). Thus, it can be

speculated that high concentration of Ag+ would both activate the gold surface and disrupt

the mechanisms of its passivation in the presence of sulfide ions. Alternatively, it may be

likely too that silver cyanide reacting with sulfide and/or hydrosulfide ions (Petre et al.,

2008) lead to Ag2S and Fe(III) hydroxide co-precipitates leaving across the passivating film

some porosity beneficial for gold leaching. The role of silver as highlighted in the PBER

experiment is in support of Wadsworth and Zhu (2003) scheme that the beneficial effect of

silver (I) on anodic cyanidation of gold would come from reactions between AuCN film

and Ag(CN)2- ions to yield AuCN-AgCN film. Such composite film offering increased

diffusion capacity in comparison to a tight AuCN film. Finally, the dissolution of this

heterogeneous AuCN-AgCN film by CN- will produce Au(CN)2

- regenerating Ag(CN)2

- to

become available for another gold surface activation process.

Second, in the presence of galvanic interaction between Au and pyrite (case B), when Ag

was in permanent contact with pyrite (and gold) an important decrease in gold leaching

kinetics during the first 6 min of cyanidation was noticed when compared to the case when

silver was mixed with quartz (filled squares-solid line vs. filled circles-solid line). This

slow down in Au dissolution can be interpreted as resulting from a competition between Au

and Ag for the same cathodic oxygen reduction. After 6 min of cyanidation, once the

largest fraction of Ag dissolved due to Ag-pyrite galvanic contact, the high concentration of

leached Ag entrained higher dissolution of gold in comparison to the case where silver was

packed together with the quartz particles.

118

0

10

20

30

40

50

60

70

80

0 2 4 6 8 10 12 14 16 18 20

% A

u l

ea

ch

ed


case B : MRI-2+Au+Ag//Silica

MRI-2+Au//Silica+Ag

case C : Silica+Au+Ag

MRI-2+Ag//Silica+Au

case A : MRI-2//Silica+Au+Ag

Figure III-11 Effect of silver on gold leaching in the presence of pyrite (MRI-2). Reaction

conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11.

III.4. Concluding remarks

A new experimental strategy using a packed bed electrochemical reactor (PBER) was

developed to assess the relative importance of individual contributions from passivation

and galvanic phenomena on gold and silver leaching in the presence of a wide range of

sulfide-rich ores. The following sulfide ore and minerals were investigated: industrial ore,

MRI-1; pyrite, MRI-2; chalcopyrite, MRI-3; sphalerite, MRI-4; chalcocite, MRI-5; galena,

MRI-6; stibnite, MRI-7. Enabling galvanic interaction (GI) between Au (and Ag) with

respect to each sulfide in the PBER experiments prompted more PM leaching in

comparison to the tests where GI was disabled. In the presence of MRI-4, MRI-5 or MRI-7,

Au galvanic corrosion was not sufficient, whereas with MRI-1, MRI-2 or MRI-3 minerals,

gold dissolution was mainly controlled by galvanic interactions. Not only the detrimental

effect of passivation phenomena was neutralized, but also gold recovery was improved in

comparison to the benchmark test where the PBER was loaded with silica and PM. GI were

also shown to affect both the nature and abundance of the sulfides oxidation products. The

119

low negative galvanic current with chalcocite (MRI-5) was due to rapid oxidation of

hydrosulfides on the chalcocite surface. MRI-6 exhibited an odd behavior, as it was found

to doubly influence gold leaching through: (1) Activation of gold surface by dissolved

species (i.e., Pb(II)) and (2) Acceleration of gold leaching rate by positive effect of galvanic

corrosion.

III.5. References

Aghamirian, M.M., Yen, W.T., 2005a. Mechanism of galvanic interactions between gold


Aghamirian, M.M., Yen, W.T., 2005b. A study of gold anodic behavior in the presence of

various ions and sulfide minerals in cyanide solution. Minerals Engineering 18, 89-102.

Azizi, A., Petre, C.F., Olsen, C., Larachi, F., 2010. Electrochemical behavior of gold



Blasco-Tamarit, E., Igual-Muñoz, A., García Antón, J., García-García, D.M., 2009.

Galvanic corrosion of titanium coupled to welded titanium in LiBr solutions at different

temperatures. Corrosion Science 51, 1095-1102.

Budruk-Abhijeet, S., Balasubramaniam, R., Gupta, M., 2008. Corrosion behaviour of Mg–

Cu and Mg–Mo composites in 3.5% NaCl. Corrosion Science 50, 2423-2428.

Cathro, K.J., Koch, D.F.A., 1964. The anodic dissolution of gold in cyanide solutions.

Journal of The Electrochemical Society 111, 1416-1420.



Fisher, W.W., 1994. Comparison of chalcocite dissolution in the sulfate, perchlorate,

nitrate, chloride, ammonia, and cyanide systems. Minerals Engineering 7, 99-103.







Hollow, J., Deschenes G., Guo, H., Fulton, M., Hill, E., 2003a. Optimizing Cyanidation

Parameters for Processing of Blended Fort Knox and True North Ores at the Fort Knox

Mine. Hydrometallurgy, 2003: Proceedings of the 5th International Symposium Honoring

Professor Ian M. Ritchie. 1, 21-34.

120




of impurities II. The effect of silver. Journal of The Electrochemical Society 147, 3272-

3276.

Kirk D.W., Foulkes, F.R., Graydon, W.F., 1978. A study of anodic dissolution of gold in

aqueous alkaline cyanide. Journal of The Electrochemical Society 125, 1436-1443.

Kirk, D.W., Foulkes, F.R., 1980. Anodic dissolution of gold in aqueous alkaline cyanide

solutions at low overpotentials. Journal of The Electrochemical Society 127, 1993–1997.

Licht, S., 1988. Aqueous solubilities, solubility products and standard oxidation-reduction

potentials of the metal sulfides. Journal of The Electrochemical Society 135, 2971-2975.

Linge, H.G., 1995. Anodic oxidation of pyrrhotite in simulated CIP liquors. Minerals


Liu, G.Q., Yen, W.T., 1995. Effects of sulphide minerals and dissolved oxygen on the gold




Luthy, R.G., Bruce, S.G., 1979. Kinetics of reaction of cyanide and reduced sulfur species

in aqueous solution. Environmental Science & Technology 13, 1481-1487.

Marsden, J.O., House, C.I., 1992. The Chemistry of Gold Extraction. Ellis Horwood,

London.

Marsden, J.O., House, C.I., 2006. The Chemistry of Gold Extraction. 2nd

Edition. Society


Mirnezami, M., Hashemi, M.S., Finch, J.A., 2003. Measurement of conductivity of

sulphide particles dispersed in water. Canadian Metallurgical Quarterly 42, 271-276.




Rand, D.J.A., 1977. Oxygen reduction on sulfide minerals. Part III. Comparison of

activities of various copper, iron, lead, and nickel mineral electrodes. Journal of


Sawaguchi, T., Yamada, T., Okinaka, Y., Itaya, K., 1995. Electrochemical scanning

tunnelling microscopy and ultrahigh-vacuum investigation of gold cyanide adlayers on

Au(111) formed in aqueous solution. Journal of Physical Chemistry 99, 14149–14155.

Senanayake, G., 2008. A review of effects of silver, lead, sulfide and carbonaceous matter

on gold cyanidation and mechanistic interpretation. Hydrometallurgy 90, 46-73.

Sun, X., Guan, Y.C., Han, K.N, 1996. Electrochemical behaviour of the dissolution of

gold–silver alloys in cyanide solution. Metallurgical and Materials Transactions B. 27, 355-

361.

121

Wadsworth, M.E., Zhu, X., 2003. Kinetics of enhanced gold dissolution: activation by

dissolved silver. International Journal of Mineral Processing 72, 301-310.




122

CHAPITRE IV.

CHAPITRE IV. The role of multi-sulfidic mineral binary

and ternary galvanic interactions in gold cyanidation in a

multi-layer packed-bed electrochemical reactor

Abdelaaziz Azizi,1,2

Catalin Florin Petre,1‡

Gnouyaro Palla Assima,1 Faïçal Larachi

1*

1-Department of Chemical Engineering, Laval University, Québec, Canada, G1V 0A6;

2-COREM Research Center, 1180 Rue de la Minéralogie, Québec, Canada, G1N 1X7;

Abstract/Résumé

A multi-layer packed-bed electrochemical reactor (PBER) approach was used for studying

the leaching behavior of associated and free (silica-trapped) gold from a series of synthetic

multi-mineral systems consisting of pyrite, silica, Au, and successively, X = chalcopyrite,

sphalerite and chalcocite. The PBER was filled with sieved powders of sulfidic minerals

(pyrite, X), gold and silica arranged as electrically-isolated three-layer pyrite//X//silica and

two-layer pyrite+X//silica systems. Introduction of gold successively in each layer of the

three- and two-layer PBER systems highlighted the role of Au-pyrite, Au-X and pyrite-X

binary, and Au-pyrite-X ternary galvanic interactions in the leaching of associated and free

gold. The highest Au recovery was achieved within the pyrite layer while the lowest was

within the silica layer. For the pyrite//chalcocite//silica system, depletion of free cyanide

and surface passivation inhibited strongly gold leaching. Cyanidation experiments

performed in the mixed sulfide//silica systems showed that the effect of Au-sulfide galvanic

interactions largely depend on the mineralogical association between phases. The presence

of galvanic effects between the associated minerals in the PBER was confirmed by a

conjunction of electrochemical and chemical speciation studies. Increasing the pyrite-Au

areal ratio increased gold leaching as a result of increased cathodic areas for oxygen

reduction. However, gold leaching from mixed multi-mineral systems where multi-factorial

123

galvanic contacts were enabled was controlled mostly by the relative area ratios between

the associated sulfide minerals.

L’approche du réacteur électrochimique à lit fixe (multicouches) a été utilisée pour l’étude

du comportement de dissolution de l’or associé et libre (piégé dans le quartz) à partir d’une

série de systèmes synthétiques multi-minéraux constitués de pyrite, d’Au, de quartz et

successivement X = chalcopyrite, sphalérite et chalcocite. Le réacteur a été chargé de

mélanges homogénéisés de sulfures (pyrite, X), de poudres d’or et de quartz arrangés sous

forme de systèmes en tri-couches Pyrite//X//quartz et en bi-couches pyrite+X//quartz,

électriquement-isolés. L’implantation de la poudre d’or, successivement, dans chacune des

couches des deux systèmes (en tri- et en bi-couches) a permis de mettre en évidence l’effet

des interactions galvaniques binaires (Au-pyrite, Au-X et pyrite-X) et ternaires (Au-pyrite-

X) sur la lixiviation de l’or libre et associé. La récupération de l’or la plus élevée a été

obtenue à partir de la couche de pyrite alors que celle à partir de la couche de quartz a été

très faible. Pour le système pyrite//chalcocite//quartz, l’effet conjugué de la consommation

de cyanure et de la passivation de surface inhibe fortement la dissolution de l’or. Les

expériences de cyanuration réalisées dans le cas de systèmes mixtes bisulfures-quartz ont

démontré que les effets des IG Au-sulfures dépendent largement des associations

minéralogiques entre les différentes phases. L’augmentation du rapport de surface pyrite-

Au entraine une augmentation de la vitesse de lixiviation de l’or en raison de

l’augmentation de la surface disponible pour la réduction de l’oxygène. Toutefois, la

lixiviation de l’or à partir de systèmes multi-minéraux mixtes, où les interactions

galvaniques multifactorielles sont établies, est principalement contrôlée par l’importance

relative des surfaces des différents sulfures associés.

IV.1. Introduction

Facing virtually depleted gold-bearing oxide deposits around the globe, the mining industry

has turned its attention to gold extraction from less friendly sulfide deposits sailing

cyanidation into uncharted territories. Unlike oxides, which divert marginal quantities of

cyanide and have generally minor impact on gold leaching, most metal sulfides display a

range of reactivities in alkaline gold leaching cyanidation (Marsden and House, 1992;

124

2006). Due to poor cyanide selectivity for gold over the enclosing sulfide mineral matrix,

such parasitic reactions inflict utility extra costs and profit losses resulting from high levels

of cyanide consumption, e.g., formation of thiocyanate, and dissolution of transition (Cu,

Fe, Zn, etc.) metals (Habashi, 1967). In addition, some side-reaction products can directly

inhibit cyanidation by passivating gold surface via insoluble coatings of various chemical

natures (Weichselbaum et al., 1989; Jeffrey and Breuer, 2000; Guo et al., 2005; Dai and

Jeffrey, 2006; Azizi et al., 2010, 2011). Hence, with the increasing occurrence of refractory

and complex sulfide-rich low-grade Au ores in current industrial operations, thorough

understanding of the mechanisms and kinetics of gold leaching from sulfide-rich ores has

become an R&D task of top priority.

Most sulfide minerals exhibit electric conductivity to allow charge transfer at their surfaces.

By virtue of gold-sulfide contacts, galvanic interactions were previously shown to influence

considerably the gold leaching behavior (Lorenzen and Van Deventer, 1992; Aghamirian

and Yen, 2005; Dai and Jeffrey, 2006; Azizi et al., 2010). Given that the factors affecting

gold leaching in aerated cyanide solutions are mostly of mineralogical nature, most of the

previous studies focused on the “one-to-one” effect of individual sulfide mineral phases

whether in contact or not with gold (Liu and Yen 1995; Deschenes et al., 2002; Aghamirian

and Yen, 2005; Guo et al., 2005; Dai and Jeffrey, 2006; Azizi et al., 2010). However,

sulfidic ore deposits are generally complex mixtures wherein gold coexists in association

with various minerals/phases. Attempts to make sense of the refractory behavior stemming

from such complex sulfide gold ores stress on comprehensive determinations of the host

rock mineralogy and gold deportment (gold distribution and associations) to identify

efficient processing routes (Goodall and Scales, 2007).

Representative sampling of low grade gold deposits is recognized to be virtually impossible

(Jones and Cheung, 1988) a consequence of which is that studies of Au-sulfides

associations and how they would influence gold extraction are venturous and very costly

(Goodall and Scales, 2007). Consequently, accurate mineralogical characterization of

complex gold ores is often not undertaken until the first processing problems appear.

Alternatively, diagnostic leaching techniques have been proven to be invaluable for

assessing ore refractoriness and for determining gold deportment (Lorenzen and Tumility,

125

1992; Lorenzen and van Deventer, 1992; 1993). In this technique the least stable mineral

phase in the sample matrix is leached off first acidically/oxidatively before a cyanide

leaching step is enabled to extract gold. The process is repeated successively with more

oxidative acid leaching until gold recovery is maximized. This ingenious diagnostic

leaching strategy was implemented by Lorenzen and van Deventer (1993) to determine the

amount of gold to be liberated from gold ores containing pyrite and pyrrhotite as the main

sulfides constituents. Destruction of sulfides via nitric and sulfuric acid oxidation resulted

in nearly 95% of gold extraction. The same approach was also used to investigate the effect

of various constituents of two complex sulfide ores on gold leaching using a gold disc

immersed in minerals slurries (Lorenzen and van Devanter, 1992). Comparing gold

recoveries before and after each leaching step led to quantify the effect of each eliminated

phase on Au recovery. Such “backward” diagnostics is prone to three weak points: i) it is

difficult to quantify the individual effect of each component due to lack in selectivity of

acid pretreatments, i.e., partial elimination of pyrite by H2SO4 and HNO3, of sphalerite by

H2SO4 and FeCl3 and of galena by HCl and FeCl3; ii) the oxidative acid pre-treatment may

irreversibly affect the surface representativeness of the minerals left over; iii) since

cyanidation is performed after elimination of the targeted sulfide mineral phases, Au-

sulfide and sulfide-sulfide galvanic interactions would have been erased which may give

rise to radically different cyanidation patterns.

Very recently, a packed-bed electrochemical reactor (PBER) technique was developed and

tested to decouple and quantify the individual contributions of passivation phenomena (PP)

and galvanic interactions (GI) on precious metals (gold and silver, PM) leaching rates

during cyanidation of sulfide-rich ores (see chapter 3; Azizi et al., 2011). The PBER was

filled with homogenized mixtures of several minerals (pyrite, chalcopyrite, sphalerite,

chalcocite, galena, stibnite and an industrial ore), PM and sulfide-silica powders arranged

as electrically-isolated segregated bilayer(s) with permanent particle-particle electrical

contacts among all constituents in each layer. Implantation of PM powders successively in

the silica layer and then in the sulfide layer, enabled passivation and binary galvanic

interactions to be singled out and quantified separately for each sulfide. The results showed

that GI ameliorated, to various degrees, the leaching of PM, particularly those due to pyrite,

126

chalcopyrite and the industrial ore were so positive that they largely outweighed the

negative impact of PM passivation.

Most sulfide minerals are known to develop dissimilar rest potentials in aerated cyanide

solutions. Therefore, GI could be initiated among the various ore constituents resulting in

substantially increased or decreased reactivity for some associated minerals. Azizi et al.

(2010) demonstrated that the behavior of an industrial sulfide-rich gold ore during

cyanidation cannot be inferred from the barycentric linear combination obtained from the

individual behavior of its different sulfide mineral constituents. In spite of this, the effect of

multi-factorial GI among sulfides on their reactivity and consequently on Au recovery by

cyanidation is very poorly addressed in the literature. Thus, extending the new PBER

experimental strategy (Azizi et al., 2011) to interrogate mineralogical characteristics in

multi-mineral systems is foreseen to further elucidate the higher-order galvanic interactions

involving contacts between two sulfidic minerals and precious metals. Therefore, this work

aims at the following targets:

1. To sort out, using the new PBER configuration during cyanidation, the effect of

gold distribution and mineral galvanic associations on both Au recovery and

reactivity of sulfide-rich ores.

2. To investigate, using the same PBER configuration, the effects of varying the

cathode-to-anode areal ratio on gold leaching from multi-mineral systems.

IV.2. Experimental

IV.2.1. Materials and Reagents

In this study four sulfide-rich ore samples received from Ward’s Natural Science were

used. These samples consisted of pyrite ( ), chalcopyrite ( ), sphalerite ( ) and chalcocite

( ) by virtue of the dominant proportion of the named sulfide mineral therein. Their

chemical and mineralogical characterizations are given elsewhere (see section II.2.1,

chapter III; Azizi et al., 2011).

127

The samples were sieved to remove particles coarser than 90 µm and finer than 45 µm.

Consequently, the same granulometric fraction was used for the different ores, providing a

uniform total area per unit-weight in all cyanidation experiments. Pure gold (P80 = 39 µm,

99.998%) powder was purchased from Alfa Aesar (USA).

Distilled water was used for all cyanidation experiments. The reagents used in this study,



analytical grade.

IV.2.2. Equipment and procedures

The influence on gold recovery of gold distribution within different mineral phases of a

synthetic ore was studied using the PBER. The aerated cyanide solution feed was

continuously recirculated in a closed loop through the fixed bed using a peristaltic pump.

The time on stream against which all the concentration profiles of this study were plotted

corresponded to the time the aerated cyanide solution sojourned inside the PBER excluding

the remainder of the transit time of flight in the external loop of the liquid circuit. See

chapter III for a thorough description of the setup.

Binary galvanic interactions between gold and each of the above minerals were assessed in

three-layer //X//silica configurations by seeding, one at a time, Au powder (50 mg) in the

pyrite, the other sulfide (X= , , or ) or the silica layers (Fig. IV.1a). First, 4 g of pyrite

(usually the most abundant mineral in Au sulfide deposits) was packed in the lower part of

the PBER working section. Then, 2 g of one of the other minerals (X= , , or ) was

packed on top of the pyrite layer. Finally, inert silica particles completed the remaining

upper part of the bed. The mineral phases were electrically disconnected by means of

sintered glass filter discs (VWR Intern., USA) used as inter-layer insulators depicted as “//”

in the //X//silica syntax. The following modalities were studied Au//X//silica,

//XAu//silica and //X//silicaAu where Au subscript stands for the seeded layer.

To investigate the effect of mineral associations on gold leaching, Au- -X ternary and -X

binary galvanic interactions were assessed in two-layer +X//silica configurations by

128

seeding Au either in the +X (X= , , or ) layer or in the silica layer (Fig. IV.1b). The

following modalities were interrogated ( +X)Au//silica, and +X//silicaAu. Gold dissolution

patterns were registered after Au powder (50 mg) was homogenized with 4 g of pyrite and

2 g of one of the other minerals (X= , , or ); the 3-component mixture being located in

the PBER lower part. As previously, inert silica particles were packed in the remaining

PBER upper part.

In all experiments, the concentration of dissolved metals was measured using a Perkin

Elmer AA-800 atomic absorption spectrometer (AAS), while capillary electrophoresis (CE,

Agilent Technologies) was used to analyze the dissolved sulfur and cyanicides (Petre et al.,

2008). Residual cyanide was quantified using a silver nitrate titration method with

rhodamine as color-change indicator.

(a) (b)

ELECTROLYTE OUTLET

ELECTROLYTE INLET

+X

insulator

silica

ELECTROLYTE OUTLET

ELECTROLYTE INLET

X

silica

Figure IV-1 Sketch of packed-bed electrochemical reactor (PBER) strategies used to study

the effect of gold distribution and mineralogical associations on gold recovery. (a) pyrite

mineral phase: ; chalcopyrite, sphalerite or chalcocite mineral phase: X; silica layer;

insulator sintered-glass disc filter; (b) homogenized mixture of mineral phases and X.

129

IV.2.3. Electrochemical campaign

The effect of galvanic interactions between and each of the , , or minerals on the

electrochemical behavior of these associated minerals was studied separately by means of

the PBER. For each experiment, two-layer packed-bed mineral electrodes were prepared

where a 4 g pyrite powder layer was topped successively by 4 g of one of the , , or

mineral layers. A platinum spring was inserted inside each mineral layer and linked to an

insulated platinum wire that was used to establish (via a carbon brush) electrical contact

between the mineral electrode and a potentiostat. As previously, the two mineral packed-

bed electrodes were separated by using an insulator sintered-glass disc filter to prevent any

direct electrical contact between them. The potentials were monitored using a silver/silver

chloride (Ag/AgCl in saturated AgCl-KCl solution) reference electrode (+197 mV vs.

standard hydrogen electrode). A Luggin capillary tube housing the reference electrode was

placed in a 250 mL magnetically-stirred glass reservoir containing 100 mL solution

prepared by adding NaCN to 0.01 M boric acid (background electrolyte) and adjusted to pH

11 (using NaOH). As described previously (see chapter III; Azizi et al., 2011), this

experimental arrangement enabled building a three-electrode electrochemical cell where the

electrodes were placed in the same medium by recirculating in a closed loop and upflow

manner the cyanide solution through the packed-bed reactor. The galvanic potentials and

corrosion currents established between the electrodes pair were obtained by using the zero

resistance ammeter (ZRA) technique. Since the lower part of the PBER containing the

layer was always used as the secondary working electrode, the current was negative when

the direction of electrons was from to the other mineral electrode (X), while the current

values were positive when the electrons flowed in opposite direction. Before starting the

ZRA experiments, the mineral electrodes were maintained in aerated cyanide solution for

ca. 30 min to allow stabilization of the open-circuit potential.

130

IV.3. Results and discussion

IV.3.1. Gold Cyanidation and the Pyrite-Chalcopyrite-Silica System

Au- and Au- binary galvanic interactions were assessed with gold distributed within

electrically-disconnected three-layer // //silica mineral phases in the PBER as shown in

Fig. IV.2a. The kinetics of gold leaching was compared with Au powder successively

contacted with (A), (B) and silica (C) minerals. A benchmark gold leaching test (D)

was obtained with Au powder added to an all-through silica layer deprived of sulfide

minerals. Embedding Au within inert silica mineral ((C), (D)) mimicked free (or non-

associated) gold. To determine the impact of multi-galvanic effects on gold dissolution

from complex mineral associations, mixed mineral systems were also prepared as shown in

Fig. IV.2b: ( + )Au mixed powder to prompt ternary Au-pyrite-chalcopyrite galvanic

interactions in two-layer ( + )Au//silica system (A’); two-layer + //silicaAu system to

assess the effect of binary - galvanic interactions on the leaching of free gold (B’) from

the silica powder layer.

Also plotted in Figs. IV.2a’ and IV.2b` is the evolution of gold leaching as a function of the

physical time for the same experimental results shown in Figs. IV.2a and IV.2b, see

Chapter III (see appendix C for more details).

Binary galvanic interactions between gold and mineral particles highlighted various gold

reactivity patterns (Fig. IV.2a). The Au- binary galvanic interactions led to the highest

gold recovery (A) followed by the -prompted binary galvanic interactions (B). Pyrite,

whose mass was twice chalcopyrite’s, brought greater galvanic currents, and thus more Au

dissolution (A), than chalcopyrite (B). The worst gold recovery occurred with free gold in

the presence of the synthetic ore of pyrite-contactless chalcopyrite (C) –galvanic contacts

amidst Au and sulfide minerals precluded– with pyrite and chalcopyrite dissolution

inducing species severely obstructing the Au surface (Azizi et al., 2010, 2011). This

resulted in even less Au leaching compared to the benchmark test (D). Also, Au- galvanic

association (B) was sufficient enough (case B vs. case D, Fig. IV.2a) to overcome gold

131

passivation triggered by chalcopyrite and pyrite dissolution. When pyrite and chalcopyrite

were contactless, gold leaching (A) was promoted the most by Au- GI.

0

10

20

30

40

50

60

70

80

0 2 4 6 8 10 12 14 16

% A

u

lea

ched


case A : MRI-2+Au+Ag//MRI-3//Silica

case A' : MRI-2+MRI-3+Au+Ag//Silica

case D : Silica+Au+Ag

case B' : MRI-2+MRI-3//Silica+Au+Ag

0

10

20

30

40

50

60

70

80

0 2 4 6 8 10 12 14 16

% A

u l

each

ed



case B : MRI-2//MRI-3+Au+Ag//Silica


case C : MRI-2//MRI-3//Silica+Au+Ag

(a)

(b)

Silica

+ Au

(B)

Silica + Au(C)

Silica + Au

(D)

Silica

+ Au

(A)

Silica

+ + Au

(A’)

Silica + Au(B’)

Silica + Au

(D)

Silica

+ Au

(A)

(C)

(B’)

+


pyrite, chalcopyrite and silica layers: Au within pyrite (A), Au within chalcopyrite (B), Au

within silica (C,D); (b) electrically-connected pyrite-chalcopyrite//silica systems: Au within

pyrite-chalcopyrite (A’), Au within silica (B’). Inset: Effect of pyrite-chalcopyrite galvanic

association on free-gold leaching from inert silica layer. Reaction conditions: CN- = 30

mM, DO2 = 0.25 mM, pH = 11.

132

0

10

20

30

40

50

60

70

80

0 50 100 150 200 250 300 350 400

% A

u l

ea

ch

ed

Physical time (min)

case A

case B

case C

case D

0

10

20

30

40

50

60

70

80

0 50 100 150 200 250 300 350 400

% A

u

lea

ch

ed

Physical time (min)

case A

case A`

case D

case B`

(a’)

(b’)

Silica

+ Au

(B)

Silica + Au(C)

Silica + Au

(D)

Silica

+ Au

(A)

Silica

+ + Au

(A’)

Silica + Au(B’)

Silica + Au

(D)

Silica

+ Au

(A)

+

Figure IV-2. Effect on gold leaching of Au distribution within (a`) electrically-disconnected


within silica (C,D); (b`) electrically-connected pyrite-chalcopyrite//silica systems: Au

within pyrite-chalcopyrite (A’), Au within silica (B’). Inset: Effect of pyrite-chalcopyrite

galvanic association on free-gold leaching from inert silica layer. Reaction conditions: CN-

= 30 mM, DO2 = 0.25 mM, pH = 11.

133

Fig. IV.2a inset shows that when Au was non-associated (located in the silica layer), the

+ galvanic interactions degraded gold recovery much further than when the and

layers were disconnected, i.e., (B’) < (C). To search for this phenomenon’s origin, pyrite

and chalcopyrite were packed in two separate layers in the PBER. Once electrically

interconnected through ZRA, the galvanic current suddenly increased to a relatively high

value (~ 100 µA) and then slowly decreased to stabilize at ~ 80 µA after 180 min of

reaction, Fig. IV.3. The galvanic potential remained relatively constant nearby -250 mV for

the entire experiment. As pyrite open circuit potential (OCP) (-180 mV vs. Ag/AgCl) in

aerated cyanide solution surpassed that of chalcopyrite (-344 mV vs. Ag/AgCl), a galvanic

cell formed with chalcopyrite as an anode and pyrite as a cathode (positive galvanic

current, Fig. IV.3). This galvanic cell was held responsible for the differences in Au

leaching between (B’) and (C) shown in Fig. IV.2a inset.

0

20

40

60

80

100

120

-400

-350

-300

-250

-200

-150

-100

0 20 40 60 80 100 120 140 160 180

I (µ

A)

E (

mV

) v

s. A

g/A

gC

l

Time (min)

Figure IV-3 Measured galvanic current and potential vs. time for pyrite-chalcopyrite

galvanic couple in the PBER. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH =

11.

134

To substantiate Fig. IV.2a inset observations, the release of metallic cyanide complexes

(Fig. IV.4a) and sulfur-bearing anions (Fig. IV.4b) was monitored via CE as a function of

time. The concentrations of SCN-, SO4

2- and Cu(CN)3

2- after 30 min of cyanidation

increased, respectively, from 350 mg/L, 18 mg/L and 50 mg/L ( + GI disabled (C)) to 600

mg/L, 37 mg/L and 75 mg/L ( + GI enabled (B’)). This was ascribed to chalcopyrite

reactivity promoted anodically (Fig. IV.3) via - galvanic association. Conversely, pyrite

reactivity was partially repressed by the same galvanic interactions. When enabling + GI

(B’), lower amounts of Fe-cyanide were detected in solution in comparison to when

disabling + GI (C), Fig. IV.4a. Chalcopyrite dissolution was shown to strongly inhibit

gold leaching as a result of rapid depletion of free cyanide (through copper complexation)

and gold surface passivation by hydrosulfide ions (see chapter III). Hence, chalcopyrite

electrochemical activation when enabling + GI may explain the worst gold dissolution of

case (B’) in Fig. IV.2b.

Also because chalcopyrite equilibrium potential is lesser than pyrite (see chapter III), -

galvanic interactions attenuated the magnitude of the positive Au- galvanic interactions,

i.e., (A’) < (A), Fig. IV.2b. Such galvanic association drew the + mixture OCP toward

the anodic region, as compared to pyrite OCP, decreasing the driving force for gold

galvanic corrosion. The decrease of gold leaching upon - association, i.e., (A’) < (A),

could also be explained by the fact that oxygen reduction over pyrite surface is far more

active than on chalcopyrite surface (Aghamirian and Yen, 2005). Therefore, galvanic

contacts between pyrite and chalcopyrite particles in the PBER could influence the activity

of pyrite particles toward oxygen reduction. Coherent with the inhibiting effect on oxygen

reduction (Ahlberg and Broo, 1996; Cruz et al., 2001), oxygen reduction on pyrite surfaces

was shown to largely depend on mineralogy and surface properties. Hence, not only the

mineralogy of the host ore should be accurately determined but knowledge of gold

deportment is as vital to decipher the role of sulfide ores on gold cyanidation.

135

Cu(CN)32-

Fe(CN)62-

0

10

20

30

40

50

60

70

80

90

100

0 5 10 15 20 25 30 35

Meta

ls l

ea

ch

ed

(m

g/L

)


MRI-2//MRI-3//Silica+Au+Ag

MRI-2+MRI-3//Silica+Au+Ag



(a)

0

10

20

30

40

50

60

70

80

90

100

0

100

200

300

400

500

600

700

800

900

0 5 10 15 20 25 30 35

SC

N-

lea

ch

ed

(mg

/L)

SO

42

-le

ach

ed

(mg

/L)






(b)

// //SiAu (C)

+ //SiAu (B’)

// //SiAu (C)

+ //SiAu (B’)

// //SiAu (C)

+ //SiAu (B’)

// //SiAu (C)

+ //SiAu (B’)

Figure IV-4 Effect of galvanic interactions between pyrite and chalcopyrite on: (a) Fe-

cyanide and Cu-cyanide evolution and (b) sulfur-bearing anions speciation. Reaction


136

IV.3.2. Gold Cyanidation and the Pyrite-Sphalerite-Silica System

Within electrically-disconnected three-layer // //silica phases, gold leaching decreased in

a sequence order A, D, B and C (Fig. IV.5a). This was unlike the corresponding sequence

order A, B, D and C (Fig. IV.2a) for the // //silica mineral phases. The mixed + //silica

mineral systems led to a similar sequence A, A’, D and B’ (Fig. IV.5b) as its + //silica

analog (Fig. IV.2b).

In A, D, B and C cases, large amounts (ca. 50%) of free cyanide were still unconverted at

run ends. After 10 min cyanidation, Au recovery drastically dropped from 70 %, when

associated to pyrite (A), down to ca. 4 % when associated to sphalerite (B). Free gold was

vulnerable to the presence in ore of sphalerite (C) with virtually no Au recovery. Clearly,

Au- galvanic association (B) was insufficient to overcome gold passivation by sphalerite

and pyrite dissolution. Only when sphalerite and pyrite were contactless that Au- galvanic

promoted gold leaching (A). Unlike - mixtures ((A’) vs. (D), Fig. IV.2b), - mixtures

((A’) vs. (D), Fig. IV.5b) induced larger positive effect on gold recovery compared to the

Au-free benchmark test (D) which was attributed to more negative chalcopyrite OCP than

sphalerite’s.

Sphalerite was shown to exhibit modest promoting galvanic effect on Au dissolution (see

chapter III; Azizi et al., 2011). Thus, the lesser gold recovery after 14 min in ( + )Au//silica

(A’, Fig. IV.5b) vis-à-vis Au// //silica (A, Fig. IV.5b) was ascribed to sphalerite anti-

galvanic effects induced in multi-galvanic systems. Sphalerite in aerated cyanide solutions

was nobler than pyrite (see chapter III) suggesting that - GI could magnify pyrite anodic

behavior. The simultaneous pyrite and gold oxidations would increase dissolved oxygen

demand and thus competition for the same pyrite cathodic sites resulting in lesser gold

leaching, i.e., (A’) < (A), Fig. IV.5b.

137

0

10

20

30

40

50

60

70

80

0 2 4 6 8 10 12 14 16

% A

u l

ea

ch

ed






0

10

20

30

40

50

60

70

80

0 2 4 6 8 10 12 14 16

% A

u l

ea

ch

ed






Silica

+ Au

(B)

Silica + Au(C)

Silica + Au

(D)

Silica

+ Au

(A)

Silica

+ + Au

(A’)

Silica + Au

(D)

Silica

+ Au

(A)

Silica + Au(B’)

+

(C)

(B’)

(a)

(b)


pyrite, sphalerite and silica layers: Au within pyrite (A), Au within sphalerite (B), Au

within silica (C,D); (b) electrically-connected pyrite-sphalerite//silica systems: Au within

pyrite-sphalerite (A’), Au within silica (B’). Inset: Effect of pyrite-sphalerite galvanic

association on free-gold leaching from inert silica layer. Reaction conditions: CN- = 30

mM, DO2 = 0.25 mM, pH = 11.

138

Interestingly, Fig. IV.5a inset shows that the - association improved free gold recovery,

i.e., (B’) > (C), unlike the outcome form - association (Fig. IV.2a inset). The measured

galvanic current and potential between sphalerite and pyrite are shown in Fig. IV.6. Despite

pyrite OCP (-180 mV vs. Ag/AgCl) was lower than sphalerite’s (-137 mV vs. Ag/AgCl),

galvanic current went undetected in accordance with sphalerite low electric conductivity

(Ahlberg and Àsbjörnsson, 1994). In addition, leaching of free gold from + //silicaAu

system (B’) liberated less SO42-

and SCN- than of free gold from // //silicaAu system (C),

Fig. IV.7. This suggests improved galvanic protection of sphalerite decreasing its reactivity

in aerated cyanide solutions in the former configuration in accordance with Cruz et al.

(2005) findings.

-0.2

0

0.2

0.4

0.6

0.8

1

-300

-250

-200

-150

-100

-50

0

0 20 40 60 80 100 120 140 160 180

I (µ

A)

E (

mV

) v

s. A

g/A

gC

l

Time (min)

Figure IV-6. Measured galvanic current and potential vs. time pyrite-sphalerite galvanic

couple in the PBER. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11.

139

0

5

10

15

20

25

30

35

40

45

50

0

100

200

300

400

500

600

700

800

0 5 10 15 20 25 30

SC

N-

lea

ched

(mg

/L)

SO

42

-le

ach

ed(m

g/L

)






// //SiAu (C)

+ //SiAu (B’)

// //SiAu (C)

+ //SiAu (B’)

Figure IV-7. Effect of galvanic interactions between pyrite and sphalerite on SO42-

and

SCN- evolution. Reaction conditions: CN

- = 30 mM, DO2 = 0.25 mM, pH = 11.

IV.3.3. Gold Cyanidation and the Pyrite-Chalcocite-Silica System

Gold leaching decreased in the order D, A, B and C (Fig. IV.8a) within electrically-

disconnected three-layer // //silica phases. Enabling Au- association in Au// //silica,

barely led to 5 % gold extraction ((A), Fig. IV.8a). This is 14 times less what Au-

association gave for Au// //silica ((A), Fig. IV.2a) and Au// //silica ((A), Fig. IV.5a). Gold

dissolution for the // Au//silica (B) and // //silicaAu (C) systems was completely blocked.

Such deteriorations could be explained by the well-known reactivity of chalcocite in

cyanide solutions leading to complete exhaustion of free cyanide and severe obstruction of

gold particle surfaces (see chapter III). Whether free-gold, - or -associated gold, the

presence of chalcocite is overkill to gold leaching. Mixed + //silica mineral systems did

not exhibit so different trends with gold leaching decreasing in the order D, A, A’ and B’

(Fig. IV.8b). These results were expectable as chalcocite OCP is more negative than that of

gold (Aghamirian and Yen, 2005; Azizi et al., 2010) and may therefore alter the magnitude

of Au- galvanic interactions.

140

0

2

4

6

8

10

0

5

10

15

20

25

0 2 4 6 8 10 12 14 16

% A

u l

ea

ch

ed





case C :MRI-2//MRI-5//Silica+Au+Ag

% A

u l

each

ed

0

1

2

3

4

5

6

7

8

9

10

0

5

10

15

20

0 2 4 6 8 10 12 14 16

% A

u l

ea

ch

ed

% A

u l

ea

ch

ed






(a)

(b)

Silica

+ Au

(B)

Silica + Au(C)

Silica + Au

(D)

Silica

+ Au

(A)

Silica

+ + Au

(A’)

Silica + Au

(D)

Silica

+ Au

(A)

Silica + Au(B’)

+


pyrite, chalcocite and silica layers: Au within pyrite (A), Au within chalcocite (B), Au

within silica (C,D); (b) electrically-connected pyrite-chalcocite//silica systems: Au within

pyrite-chalcocite (A’), Au within silica (B’). Reaction conditions: CN- = 30 mM, DO2 =

0.25 mM, pH = 11.

141

Once ZRA contact was set between pyrite and chalcocite, the galvanic current suddenly

increased to a very high value (~ 800 µA) and then steadily diminished to stabilize at ~ 510

µA (Fig. IV.9). Chalcocite leaching was promoted by - galvanic associations featuring

larger galvanic currents compared to the - couple (Fig. IV.3). A strong galvanic cell was

formed with chalcocite acting as anode and pyrite as cathode. Such electrochemically

activated chalcocite by - galvanic interactions inhibited further the leaching of free gold

((B’) < (C), not shown). The - galvanic interactions (Fig. IV.10) promoted Cu-cyanide

complexes while slightly depressing the liberation of Fe-cyanide complexes. Hence,

chalcocite leaching was promoted by the same galvanic interactions that protected pyrite

from dissolution.

200

300

400

500

600

700

800

900

-700

-650

-600

-550

-500

-450

-400

-350

-300

0 20 40 60 80 100 120 140 160 180

I (µ

A)

E (

mV

) v

s. A

g/A

gC

l

Time (min)

Figure IV-9 Measured galvanic current and potential vs. time for pyrite and chalcocite

galvanic couple in the PBER. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH =

11.

142

0

20

40

60

80

100

120

140

160

0

100

200

300

400

500

600

700

800

900

0 5 10 15 20 25 30

Fe

lea

ched

(m

g/L

)

Cu

lea

ched

(m

g/L

)

Time-on stream (min)

MRI-2+MRI-5//Silica

MRI-2//MRI-5//Silica

MRI-2+MRI-5//Silica

MRI-2//MRI-5//Silica

Cu(CN)x(x-1)-

Fe(CN)62-

+ //SiAu (B’)

// //SiAu (C)

+ //SiAu (B’)

// //SiAu (C)

Figure IV-10 . Effect of galvanic interactions between pyrite and chalcocite on: Fe-cyanide

and Cu-cyanide evolution. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11.

IV.3.4. Effect of - -Au Areal Ratio on Gold Cyanidation in the Pyrite-

Chalcocite-Silica System

Investigations addressing the effect of sulfide-gold areal ratios on the magnitude of sulfide-

gold galvanic interactions in aerated cyanide solutions are scarce. It was exemplified first

with gold cyanidation in the pyrite-chalcocite-silica system in presence of -Au binary

galvanic interactions. Gold was associated to pyrite as shown in Fig. IV.11a where the

three-layer Au// //silica system (A) was compared to a four-layer ( /2)//( /2)Au// //silica

system (A*). In this latter, half the pyrite mass (2 g) was placed first in the PBER. It was

topped with the remaining pyrite mass (2 g) mixed with 50 mg Au powder, and then

followed by a 2 g chalcocite layer and finally silica. All layers were isolated from each

other by means of sintered-glass filter discs.

Doubling the pyrite-Au areal ratio almost doubled gold recovery, everything else being

invariant (Fig. IV.11a). Cyanidation experiments were carried out at NaCN:O2 ratio = 120

whereby Au leaching rate is oxygen-diffusion limited (Heath and Rumball, 1998; Jeffrey

143

and Ritchie, 2000). It is likely that pyrite connected galvanically to gold provides further

cathodic areas for O2 reduction; hence, doubling such pyrite areas induced greater cathodic

currents and higher gold dissolution.

Since the factors affecting gold extraction are mostly of mineralogical nature, the observed

relationship between gold leaching and cathode-anode areal ratio is crucial for the design of

optimal metallurgical processes fed with ores neither over-ground nor under-ground. For

gold interspersed within conductive sulfide minerals, e.g., pyrite, finer ore grinding could

result in decreasing the mineral-gold galvanically shared areas. This would not only reflect

in larger oxygen and cyanide consumptions via mineral dissolution, but also in lower gold

recovery as finer grinding liberates new orphaned gold fractions deprived of the

advantageous positive sulfide-Au galvanic interactions.

The effect of pyrite, chalcocite and Au powder areal ratios on gold leaching was

investigated to highlight the role of - -Au ternary galvanic interactions as typified by the

multi-layer systems in Fig. IV.11b. The syntax is self-explaining: three-layer

( /2)//( /2+ )Au//silica system (A&

), two-layer ( + )Au//silica system (A’), and three-layer

( + /2)Au//( /2)//silica system (A#). The corresponding pyrite-chalcocite mass ratios were 1

(A&

), 2 (A’), and 4 (A#). Cyanidation experiments were carried out using invariable

mineralogical composition (i.e., 4 g pyrite, 2 g chalcocite and 50 mg Au) and variable

galvanic associations to modulate gold leaching. For all the experiments, the different

layers were electrically isolated from each other.

The pyrite-chalcocite areal ratio had a strong effect on gold leaching (Fig. IV.11b). When

the - mass ratio was increased from 1 to 2, gold extracted after 14 min cyanidation

noticeably improved from 0.2 % (A&

) to 1.4 % (A’). This corroborates Fig. IV.11a findings

where a doubling of pyrite area in contact with gold improved gold leaching because of

more cathodic areas for O2 reduction. At constant chalcocite concentration in the mixture,

increasing pyrite concentration decreased the probability of chalcocite-gold contacts which

improved the magnitude of GI between gold and pyrite, i.e., (A’) > (A&

). However, a

further increase of - mass ratio to 4 stifled nearly completely gold dissolution (A#).

Pyrite-chalcocite galvanic association was shown in §3.3 to substantially enhance

144

chalcocite leaching. In principle, a small anode surface (chalcocite) connected to a larger

cathode surface (pyrite) would result in a high anodic current density and on rapid

corrosion of the anodic surface. The dramatic decrease in gold recovery when - mass

ratio was increased from 2 to 4 suggests the galvanically accelerated chalcocite dissolution

prevails over the positive pyrite-Au galvanic interactions (Fig. IV.11b).

0

0.5

1

1.5

2

2.5

0 2 4 6 8 10 12 14

% A

u l

each

ed


0

0.5

1

1.5

2

0 2 4 6 8 10 12 14 16 18

% A

u l

each

ed


(a)

(b)

Silica

/2 + Au

(A*)

Silica

+ Au

(A)

/2

(A*)

(A)

Silica

+ + Au

(A’) Silica

/2 + + Au

/2

(A&) Silica

/2

+ /2 + Au

(A#)

(A#)

(A&)

(A’)

Figure IV-11. Effect of (a) pyrite-Au and (b) pyrite-chalcocite-Au areal ratios on gold

leaching from multi-mineral systems containing pyrite, chalcocite, gold and silica. Reaction


145

IV.4. Conclusions

Synthetic gold ores containing pyrite, silica, Au, and successively, X = chalcopyrite,

sphalerite and chalcocite were prepared to investigate galvanic interactions in gold

cyanidation of associated and free (silica-trapped) gold using a new packed-bed

electrochemical reactor (PBER). Au-pyrite and Au-X binary galvanic interactions were

assessed in three-layer electrically-disconnected pyrite//X//silica systems. Similarly, pyrite-

X and ternary Au-pyrite-X galvanic interactions were interrogated in two-layer pyrite +

X//silica systems.

The gold leaching pattern in the pyrite//chalcopyrite//silica and pyrite//sphalerite//silica

systems was determined by the gold-mineral galvanic interactions: the highest recovery

achieved with pyrite-associated Au and the lowest from free Au (silica-trapped). The Au-

pyrite positive galvanic interactions were downplayed by chalcocite irrespective of the

gold-bearing mineral layer. In the mixed sulfide pyrite + X//silica systems, the Au-pyrite

galvanic interactions were found to depend on the mineralogical association between pyrite

and the other sulfides. The pyrite-X galvanic association resulted in a significant decrease

of gold leaching as compared to the case where pyrite was the sole gold-bearing mineral.

Pyrite-chalcopyrite and pyrite-chalcocite had a negative effect on the leaching of free gold

(silica-trapped), while pyrite-sphalerite galvanic association boosted free gold leaching.

Chalcocite and chalcopyrite associated with pyrite led to high galvanic currents and anodic

oxidations of the less noble sulfide paired with pyrite as confirmed by the higher copper-

cyanide and lower gold concentrations measured in the solution with pyrite-X mixtures.

Despite electrochemical evidence of galvanic interactions was not possible for the pyrite-

sphalerite system, their presence was indirectly measured by the change in concentration of

the leached species which suggested possible galvanic protection of sphalerite by pyrite.

The effect of Au-sulfide areal ratios on gold leaching was investigated in the multi-mineral

system containing pyrite, chalcocite and silica. The increase of pyrite-Au areal ratio

promoted gold dissolution via increased cathodic areas for O2 reduction. This was unlike

the pyrite-X areal ratio which exhibited non-monotonic incidence on gold leaching.

146

IV.5. References

Aghamirian, M.M., Yen, W.T., 2005. Mechanism of galvanic interactions between gold


Ahlberg, E., Àsbjörnsson, J., 1994. Carbon paste electrodes in mineral processing: an

electrochemical study of sphalerite. Hydrometallurgy 36, 19-37.

Ahlberg, E., Broo, A.E., 1996. Oxygen reduction at sulphide minerals. 3. The effect of

surface pre-treatment on the oxygen reduction at pyrite. International Journal of Mineral

Processing 47, 49-60.




Azizi, A., Petre, C.F., Olsen, C., Larachi, F., 2011. Untangling galvanic and passivation

phenomena induced by sulfide minerals on precious metal leaching using a new packed-

bed electrochemical cyanidation reactor. Hydrometallurgy 107, 101-111.

Cruz, R., Bertrand, V., Monroy, M., González, I., 2001. Effect of sulfide impurities on the

reactivity of pyrite and pyritic concentrates: a multi-tool approach. Applied Geochemistry

16, 803-819.






Deschenes, G., Pratt, A., Riveros, P., Fulton, M., 2002. Reactions of gold and sulfide

minerals in cyanide media. Minerals and Metallurgical Processing 19, 169-177.

Goodall, W.R., Scales P.J., 2007. An overview of the advantages and disadvantages of the

determination of gold mineralogy by automated mineralogy. Minerals Engineering 20, 506-

517.







Heath, A.R., Rumball, J.A., 1998. Optimising cyanide:oxygen ratios in gold CIP/CIL

circuits. Minerals Engineering 11, 999-1010.



http://cat.inist.fr/?aModele=afficheN&cpsidt=14471351

http://cat.inist.fr/?aModele=afficheN&cpsidt=14471351

http://linkinghub.elsevier.com/retrieve/pii/S089268750700026X

http://linkinghub.elsevier.com/retrieve/pii/S089268750700026X

147



3276.

Jones, M.P., Cheung, T.S., 1988. Automatic method for finding gold grains in ores and mill

products. Asian Mining 88, 73-81.

Liu, G.Q., Yen, W.T., 1995. Effects of sulphide minerals and dissolved oxygen on the gold


Lorenzen, L., Tumilty, J.A., 1992. Diagnostic leaching as an analytical tool for evaluating

the effect of reagents on the performance of a gold plant. Minerals Engineering 5, 503-512.



Lorenzen, L., van Deventer, J. S. J., 1993. The identification of refractoriness in gold ores

by the selective destruction of minerals. Minerals Engineering 6, 1013-1023.

Marsden, J.O., House, C.I., 1992. The Chemistry of Gold Extraction. Ellis Horwood,

London.

Marsden, J.O., House, C.I., 2006. The Chemistry of Gold Extraction. 2nd

Edition. Society






the rate of gold cyanidation. Proc. Austr. I.M.M. Annual conference Perth/Kalgoorlie.


http://www.sciencedirect.com/science/journal/08926875

148

CHAPITRE V. Leveraging strategies to increase gold

cyanidation in the presence of sulfide minerals - Packed-

bed electrochemical reactor approach

Abdelaaziz Azizi,1,2

Catalin Florin Petre,1‡

Faïçal Larachi1*

1-Department of Chemical Engineering, Laval University, Québec, Canada, G1V 0A6;

2-COREM Research Center, 1180 Rue de la Minéralogie, Québec, Canada, G1N 1X7;

Abstract/Résumé

Several leveraging strategies were explored to improve gold leaching during cyanidation in

the presence of pyrite, chalcopyrite, sphalerite and chalcocite. Galena, used a source of

lead, was found to largely neutralize the negative effect of sulfide minerals dissolution on

gold leaching, especially in the cases of pyrite, chalcopyrite and sphalerite. Pyrite-galena

galvanic interactions improved gold leaching while for chalcopyrite, sphalerite and

chalcocite active galvanic interactions with galena were found to be detrimental to gold

leaching. Pre-oxidation of metal sulfides prior to cyanidation significantly increased gold

leaching rate when gold-pyrite galvanic interactions were disabled, while their enablement

identified pyrite pre-oxidation as counter-productive. Pre-oxidation of mixed mineral

systems of pyrite-chalcopyrite and pyrite-sphalerite showed that enabling galvanic

interactions between the associated minerals during pre-oxidation controlled gold leaching

in the subsequent cyanidation process. For pyrite, lead nitrate addition during pre-oxidation

resulted in a significant increase of gold leaching, whereas its addition to subsequent

cyanidation cancelled the positive effects of pre-oxidation on gold dissolution.

Plusieurs mesures ont été menées pour améliorer la cinétique de lixiviation de l’or pendant

la cyanuration en présence de la pyrite, de la chalcopyrite, de la sphalérite et de la

chalcocite. La galène, utilisée comme source de plomb, neutralise largement l’effet négatif

149

de la dissolution des minéraux sulfureux sur la lixiviation de l’or, en particulier dans le cas

de la pyrite, de la chalcopyrite et de la sphalérite. Les interactions galvaniques (IG) entre la

galène et la pyrite entrainent une amélioration de la lixiviation de l’or par rapport au cas où

les mêmes IG sont absentes. Toutefois, lorsque la chalcopyrite, la sphalérite ou la

chalcocite a été connecté à la galène, les IG se sont révélées infructueuses pour la

lixiviation de l’or. La pré-oxydation des minéraux sulfureux ci-dessus a fait

significativement grimper la vitesse de dissolution de l’or, en absence des IG Au-pyrite.

Cependant, en présence des mêmes IG la pré-oxydation a été identifiée comme contre-

productive. La pré-oxydation de systèmes minéraux mixtes, contenant de la pyrite et de la

chalcopyrite ou bien de la pyrite et de la sphalérite, a montré que les IG sulfure-sulfure, se

produisant lors de l’étape de prétraitement, constituent le principal facteur contrôlant la

réactivité de l’or pendant la cyanuration. Dans le cas de la pyrite, Il a été constaté que

l’ajout du nitrate de plomb pendant l’étape de pré-oxydation améliore significativement la

vitesse de dissolution de l’or, contrairement à son utilisation au cours de la cyanuration qui

provoque une réduction considérable sur l’efficacité de la pré-oxydation.

V.1. Introduction

Due to the poor cyanide selectivity for gold over the carrier sulfide mineral matrix and to

the important electric conductivity of both gold and carrier matrix, gold cyanidation is

generally accepted as being complex due to a multitude of factors: (1) gold surface

passivation by some reaction products such as Fe(OH)3 (Guo et al., 2005) or sulfur

(Weichselbaum et al., 1989; Dai and Jeffrey, 2006); (2) galvanic interactions between gold

and sulfide minerals (Lorenzen and van Deventer, 1992; Dai and Jeffrey, 2006, Azizi et al.,

2011); (3) galvanic interactions between conducting sulfide mineral phases present in

multi-mineral systems (see chapter IV) and (4) excess cyanide and oxygen consumptions as

a result of transition metals (i.e., Cu, Fe , Zn) dissolution (Habashi, 1967).

To alleviate the harmful effects of sulfide minerals dissolution in terms of reagents

consumption (cyanide and oxygen) and gold surface passivation, atmospheric pre-oxidation

strategies and addition of lead nitrate to the slurry have been widely investigated in the

open literature (Guo et al., 2005; Dai and Jeffrey, 2006; Senanayake, 2008 and references

150

therein). Therefore, deep understanding of the mechanisms and reactions involved during

pre-oxidation and/or lead (II) addition represent important R&D challenges in today’s gold

processing. It was reported that the main role of lead (II) in the presence of sulfide minerals

is to precipitate dissolved sulfides as PbS, avoiding thus the formation of a sulfide film on

the gold surface (Hedley and Tabachnick, 1968; Deschenes et al., 2000; Breuer et al.,

2008). Deschenes et al. (2000) conducted a systematic study on the mechanisms via which

lead nitrate is affecting gold cyanidation in the presence of various sulfide minerals. Their

results demonstrated that lead (II) salts react with gold to form AuPb2, AuPb3 and some

metallic lead, which in turn activate the gold surface during cyanidation. Furthermore, the

ability of lead (II) to reduce the reactivity of some sulfide minerals as well as their affinity

for gold surface was found to be strongly affected by the nature of the sulfide mineral used.

For instance, X-ray photoelectron spectroscopy (XPS) analysis identified thin Pb layers on

the surface of gold in the presence of chalcopyrite, but no lead was found in the presence of

pyrite or pyrrhotite.

Addition of galena, which can occur naturally in sulfide gold ore deposits, as a source of

lead was also shown to have some beneficial effect on gold leaching kinetics (Deschenes,

2005). However, understanding of the mechanisms and reactions involved during gold

activation by galena is still fragmentary as the effect of this mineral on gold cyanidation in

the presence of sulfide minerals was poorly studied. This could be explained by the fact

that galena frequently occurs in direct association with other conducting mineralogical

phases within complex sulfide mineral systems. In these systems, galvanic interactions

could substantially increase or decrease the leaching of both galena and associated sulfide,

resulting in totally different effects on gold leaching.

Regarding pre-oxidation, investigations focused often on individual sulfide minerals.

Consequently, multi-factorial galvanic interactions resulting from the different

mineralogical phases present in the ore during pre-oxidation were not addressed with the

proper attention. Multi-factorial galvanic phenomena are suspected to give rise to

cyanidation responses unexpected from mono-sulfide studies. Further investigations using

multi-mineral systems are needed to provide reliable feedback for plant cyanidation circuit

where galvanic interactions between associated sulfides may play a significant role.

151

A new packed-bed electrochemical reactor (PBER) technique was developed and tested to

decouple and quantify the individual contributions of passivation phenomena and galvanic

interactions on precious metals (gold and silver, PM) leaching rates during the cyanidation

of sulfide-rich ores (Azizi et al., 2011). The PBER was filled with homogenized mixtures

of several minerals (pyrite, chalcopyrite, sphalerite, chalcocite, galena, stibnite and an

industrial ore), PM and sulfide-silica powders, which were arranged as electrically-isolated

segregated bilayer(s) with permanent particle-particle electrical contacts among all

constituents in each layer. Implantation of PM powders successively in the silica layer and

then in the sulfide layer, enabled passivation and galvanic interactions to be singled out and

quantified separately for each sulfide. The results showed that galvanic interactions

ameliorates, to various degrees, the leaching of PM, particularly those due to pyrite,

chalcopyrite and industrial ore were so positive that they largely outweighed the negative

impact of passivation.

In consideration of above interrogations, this work aims at several targets:

1. To investigate, using the new PBER, the effects of galena (free and/or associated with

sulfides) on gold cyanidation in the presence of several sulfide minerals;

2. To investigate the effects of pre-oxidation of galvanically associated sulfide-sulfide

systems and their individual components on gold leaching rate;

3. To highlight the effects on gold leaching of different strategies of lead addition in the

presence of sulfide minerals.

V.2. Experimental

V.2.1. Materials and Reagents

Five sulfide-rich ore samples received from Ward’s Natural Science were used. These

samples consisted of pyrite ( ), chalcopyrite ( ), sphalerite ( ), chalcocite ( ) and galena

( ) by virtue of the dominant proportion of the named sulfide mineral therein. Their

chemical and mineralogical characterizations are given elsewhere (see chapter III). The

samples were sieved to remove particles coarser than 90 µm and finer than 45 µm.

Consequently, the same granulometric fraction was used for the different ores, providing a

152

uniform total area per unit-weight in all cyanidation experiments. Pure gold powder (P80 =

39 µm, 99.998%) and inert silica powder (P80 ≤ 149 µm) were purchased, respectively,

from Alfa Aesar (USA) and Sigma-Aldrich (Canada). The reagents used in this study,



analytical grade. Distilled water was used for all cyanidation experiments.

V.2.2. Equipment and procedures

Four arrangements for each sulfide mineral (X = , , , ) with galena ( ) were tested in

the PBER (Figs.1a,b) using 4 g of X powder, 2 g of galena, 50 mg of gold and the rest

completed with silica. Gold in all these arrangements was embedded within the inert silica

mineral to mimic free (or non-associated) gold. Bilayer X//silicaAu and three-layer

X// //silicaAu configurations where binary Au-sulfide and X- galvanic interactions were

disabled were referred to, respectively, as cases A (Fig. 1a) and B (Fig. 1b). To investigate

the effect of mineral associations on gold leaching, X- binary galvanic interactions were

enabled in two-layer X+ //silicaAu configurations using homogenized mixtures between X

and galena powders (case C, Fig. 1a). The benchmark gold leaching test (case D, Fig. 1a)

was obtained with Au powder added to an all-through silica layer deprived of sulfide

minerals. The mineral phases were electrically disconnected by means of sintered glass

filter discs (VWR Intern., USA) used as inter-layer insulators and depicted as “//” in the

above multilayer syntax.

The influence on gold recovery of galena distribution within these various synthetic

minerals/phases was studied using the PBER. The aerated cyanide solution feed was


The time on stream against which all the concentration profiles were plotted corresponded

to the time the aerated cyanide solution sojourned inside the PBER excluding the transit

time of flight in the external loop of the liquid circuit. See chapter III for a detailed

description of the setup.

153

(b)(a)

ELECTROLYTE OUTLET

ELECTROLYTE INLET

Case A X

Case C +X

Case D No layer

insulator

Silica + Au

ELECTROLYTE OUTLET

ELECTROLYTE INLET

X

Silica + Au

Case B

Figure V-1. Sketch of packed-bed electrochemical reactor (PBER) strategies used to study

the effect of gold distribution and galena mineralogical associations on gold recovery

with/without pre-oxidations. (a) mineral phase: X (= pyrite; chalcopyrite, sphalerite or

chalcocite) or homogenized mixture of mineral phase X + galena: X + ; silica layer;

insulator sintered-glass disc filter; (b) mineral phase: X (= pyrite; chalcopyrite, sphalerite or

chalcocite); galena: ; silica layer; insulator sintered-glass disc filter.

V.2.3. Electrochemical campaign

The effect of non-associated galena on the electrochemical behavior in the X// //silicaAu

configurations was compared to that in the X//silicaAu configurations within a PBER variant

using a three-electrode cell as described elsewhere (see chapter III). A working (Au/Ag,

96/4 %) rod electrode was inserted in the upper silica layer in the PBER. A platinum spring

(reference electrode) was implanted in the sulfide mineral (X) powder layer placed at the

bottom of the PBER and galena was optionally intercalated between these upper and

bottom layers and separated from them by sintered-glass disc filters to prevent interlayer

electrical contacts. The potentials were monitored using a Ag/AgCl in saturated AgCl-KCl

solution reference electrode (+197 mV vs. standard hydrogen electrode) immersed in a 250

mL magnetically-stirred glass reservoir containing 100 mL NaCN solution in 0.01 M boric

acid background electrolyte at pH 11. The three electrodes were swept by the same medium

154

by recirculating in a closed loop and upflow manner the cyanide solution through the

packed-bed reactor, while the zero resistance ammeter (ZRA) technique was used to

monitor and register the galvanic data (see chapter III). The electrodes were systematically

polished before each experiment, as described elsewhere (chapter III). Before starting the

experiments, the Au/Ag and mineral electrodes were maintained in the aerated cyanide

solution for about 30 min to allow the open circuit potential (OCP) to stabilize. This value

was registered and later used when comparing the sulfide OCP values.

V.2.4. Alkaline pre-oxidation of sulfide ores in PBER

Since pyrite frequently occurs in sulfurous ore deposits associated with other sulfide

minerals, e.g., chalcopyrite or sphalerite, pre-oxidation tests were conducted for the single-

and mixed-mineral systems. Gold powder (50 mg) was interspersed within the upper silica

layer. This was electrically insulated from a lower PBER layer consisting of 1) 4 g mono-

sulfide , and powders, or 2) homogenized binary 4 g + 2 g or 4 g +

2 g mixtures. Oxidative pre-treatments were conducted at room temperature by

sparging pure oxygen through a 0.4 L magnetically-sintered glass reactor containing 100

mL of boric acid solution at pH 11.5. The oxygenated solution was continuously re-

circulated in a closed loop through the PBER with a peristaltic pump for 14 h. After pre-

oxidation, NaCN was added to the solution and gold leaching experiments were carried out

as previously.

In all experiments, the concentration of dissolved metals was measured using a Perkin

Elmer AA-800 atomic absorption spectrometer (AAS), while a capillary electrophoresis

instrument (CE, Agilent Technologies) was used to analyze the dissolved sulfur and

cyanicides (Petre et al., 2008). Residual cyanide was quantified using a silver nitrate

titration method with rhodamine as color-change indicator.

155

V.3. Results and Discussion

V.3.1. Effect of Galena on the Leaching of Free Gold in the Presence of

Pyrite

Figs. V.2a and V.2a` show the evolution of gold leaching as a function of time-on-stream

and of the actual (physical) time, respectively. The time on stream is expressed as multiples

of the PBER space time of the aerated cyanide solution for each circulation within the

overall loop, whereas the physical time represents the actual duration of the experiment (see

appendix C for more details).

Fig. V.2a emphasizes the impact of pyrite-galena interactions during the leaching of free

gold (interspersed in the inert silica layer) in the pyrite-containing PBER. Whether the -

galvanic interactions were enabled (C) or disabled (B), the presence of galena dramatically

counteracted the pyrite passivation effect resulting in net accelerations of gold leaching as

revealed in the bilayer //silicaAu (A) configuration. 58% of gold was dissolved in 10 min

when - galvanic interactions were enabled (C). It was followed by 46% Au recovery

when - galvanic interactions were disabled (B), then 18% in the benchmark test (D) and

2% in the absence of galena (A). Passivation, as shown by Azizi et al. (2011), is ascribed to

the dissolved species, liberated from the pyrite layer, which migrated downstream in the

bed to obstruct the surface of free gold particles mixed in the silica layer. Cyanide

consumption was comparable either in the presence (B and C) or absence (A) of galena.

Therefore, it can be stated that the presence of galena, as a source of lead, enhanced gold

dissolution by preventing passive films to form on the surface of gold grains. This is

supported by Deschenes et al. (2000) and Guo et al. (2005) XPS analysis of gold

conditioned in pyrite-containing slurries where S and Fe surface concentrations were

significantly reduced using pre-treatments with 100 ppm of lead nitrate.

Pyrite OCP (-180 mV vs. Ag/AgCl) in aerated cyanide solution is lower than galena’s (-110

mV vs. Ag/AgCl), suggesting a positive role of - galvanic interactions in promoting

further gold leaching, i.e., (C) > (B). With pyrite behaving anodically and galena acting as a

cathode, galvanic protection of the former by the latter helps mitigating the passivation

156

effect of pyrite (A). Monitoring lead leaching gave 0.2 mg/L and 2 mg/L of dissolved

Pb(II) as measured in (C) and (B) cases, respectively. This suggests that low to very low Pb

(II) concentrations are sufficient to neutralize the negative effect of pyrite during gold

cyanidation.

0

20

40

60

80

100

0 2 4 6 8 10 12 14 16

% A

u l

ea

ch

ed



case B : MRI-2//MRI-6//Silica+Au+Ag

case C : MRI-2+MRI-6//Silica+Au+Ag


//Silica + Au

// //Silica + Au

+ //Silica + Au

Silica + Au

(a)

0

20

40

60

80

100

0 50 100 150 200 250 300 350 400

% A

u l

ea

ch

ed

Physical time (min)

case A:

case B:

case C:

case D:

//Silica + Au

// //Silica + Au

+ //Silica + Au

Silica + Au

(a`)

157

250

300

350

400

450

500

0 20 40 60 80 100 120 140 160 180

I (µ

A)

Time (min)

MRI-2//MRI-6//Silica-Au

MRI-2//Silica-Au//Silica + Au

// //Silica + Au

(b)

Figure V-2. Effect of galena ( ) in the presence of pyrite ( ) on the evolution of: (a and a`)

free gold dissolution and (b) -Au galvanic current. Reaction conditions: CN- = 30 mM,

DO2 = 0.25 mM, pH = 11.

In addition, as stated by Deschenes et al. (2000) and Jeffrey and Ritchie (2000), the role of

lead (II) might not be only to counter the passivating effect of dissolved species from

pyrite, but also to electrochemically activate gold dissolution through the formation of

gold-lead alloy spots on the gold surface.

An Au-Ag rod (electrode) was implanted in the silica upper layer which was electrically

connected via ZRA to the pyrite lower layer (see V.2.3. Electrochemical Campaign). An

intermediate electrically-isolated galena layer was optionally intercalated in-between to

assess galena’s mediation, via lead(II) dissolution, on the electrochemical behavior of the

Au-pyrite galvanic couple. Fig. V.2b compares the evolution of galvanic currents between

gold and pyrite with intercalated galena layer (dashed line) or in the absence of galena

(solid line). Clearly, gold galvanic corrosion is less active in the absence of galena. Gradual

passivation of the gold electrode surface progressively reduced the galvanic current from

ca. 400 A to ca. 280 A (solid line) after 180 min. In the presence of galena, the galvanic

current first dipped to 320 A after 30 min, and then quickly increased with time to

stabilize at a relatively very high value 460 A after 180 min. These results are consistent

158

with those from Fig. V.2a and corroborate the speculation of a positive role of dissolved

lead on gold surface activity during cyanidation. Literature findings described that in

addition to its ability to avoid the formation of a passivating film on gold surface (Guo et

al., 2005; Hedley and Tabachnick, 1968; Dai and Jeffrey, 2006), soluble lead (II) also

modifies the surface characteristics of gold (Jeffrey and Ritchie, 2000) via cementation to

enhance both gold oxidation and oxygen reduction half-reactions boosting gold dissolution.


Chalcopyrite

With less than 2 % Au recovery after 12 min cyanidation, chalcopyrite (A, Fig. V.3a)

behaved similarly to its pyrite analog (A, Fig. V.2a). However, chalcopyrite passivation of

gold surfaces was substantially reduced by galena mediation whether with - galvanic

interactions enabled (C) or disabled (B). In both such instances, gold dissolution

outperformed the benchmark test (D). Unlike with pyrite, the - galvanic interactions led

to lower Au recovery (45 %, C) in comparison to the case when - galvanic interactions

were silenced (70%, B).

The loss of recovery for free gold in the presence of chalcopyrite (A) was due to

passivation by the release of hydrosulfide which formed a sulfur protective layer on gold

surface (Azizi et al., 2010, 2011). No hydrosulfide was detected in the presence of galena in

cases (B) and (C) suggesting that the released Pb(II) could help scavenging dissolved

sulfide ions and prevent formation of a sulfur layer on gold surface. Lead sulfide (PbS)

precipitate was shown to rapidly form in the presence of both sulfide and lead (II) ions

(Hedley and Tabachnick, 1968; Breuer et al., 2008). In addition, acceleration of gold

dissolution in cases (B) and (C) as opposed to (D) could result from possible modifications

of gold surfaces in the presence of galena.

The evolution of galvanic current between chalcopyrite and gold in presence (dashed line)

and in absence (solid line) of galena is shown in Fig. V.3b. Intercalating an electrically-

disconnected galena layer between silica-contained free gold and chalcopyrite layers

resulted in a large increase of galvanic current up to 350 µA vs. 180 µA. The higher

galvanic current in the presence of galena, consistent with the higher gold dissolution, is

159

likely to be attributed to lead (II) activation of gold surfaces (Budruk-Abhijeet et al., 2008).

Coherent with the activation of gold anodic behavior by the released Pb(II), Deschenes et

al., (2000) evidenced through XPS the formation of a thin lead layer on gold foils cyanided

in the presence of chalcopyrite and lead nitrate.

The (B) and (C) different gold leaching patterns (Fig. V.3a) were further investigated by

removing the silica layer and gold rod electrode and inserting a second platinum electrode

in the galena layer while chalcopyrite and galena were kept separate but electrically

connected through ZRA. A weak galvanic current (ca. 4 µA) was registered (data not

shown) by the - galvanic cell: chalcopyrite acting as anode (OCP = - 344 mV vs.

Ag/AgCl) and galena as cathode (OCP = -110 mV vs. Ag/AgCl). Hence the lower gold

leaching response in case (C), Fig. V.3a could be the result of inhibition of galena’s

leaching by chalcopyrite galvanic protection.

0

20

40

60

80

100

0 2 4 6 8 10 12 14 16

% A

u l

ea

ch

ed






//Silica + Au

// //Silica + Au

+ //Silica + Au

Silica + Au

(a)

160

100

150

200

250

300

350

400

0 20 40 60 80 100 120 140 160 180

I (µ

A)

Time (min)

MRI-3//MRI-6//silica-Au

MRI-3//Silica-Au//Silica + Au

// //Silica + Au

(b)

0

20

40

60

80

100

120

140

160

0 5 10 15 20 25 30

Lea

ch

ed

sp

ecie

s (m

g/L

)


MRI-3//Silica+Au+Ag



MRI-3//Silica+Au+Ag

MRI-3//Silica+Au+Ag


SO42-

SCN-

S2O32-

//Silica + Au

// //Silica + Au

// //Silica + Au

//Silica + Au

//Silica + Au

// //Silica + Au

(c)

Figure V-3. Effect of galena ( ) in the presence of chalcopyrite ( ) on the evolution of: (a)



161

Nevertheless, the negative effect of - galvanic interactions on gold leaching (B > C, Fig.

V.3a) was different from the one stemming from the - galvanic associations (C > B, Fig.

V.2a). This could be explained by the higher potential difference between galena and

chalcopyrite (164 mV) than between galena and pyrite (70 mV). The reinforced

chalcopyrite-induced galena protection reduced Pb(II) release hence lowering Au

dissolution.

Fig. V.3c shows that SO42-

and SCN- are the main sulfur by-products formed either in

presence or in absence of galena, while S2O32-

was detected only in the presence of galena.

It is worth mentioning that no S2O32-

was generated when galena and chalcopyrite were

tested individually using the PBER under similar cyanidation conditions (Azizi et al.,

2011). Consequently, as previously suggested by Breuer et al. (2008), galena can accelerate

the oxidation of hydrosulfide generated from chalcopyrite through some complex

heterogeneous reactions on its surface.


Sphalerite

After 6 min cyanidation, 67 % and 40 % of Au was recovered, respectively, in cases (B)

and (C), as compared to only 10 % for the benchmark test (D) and 1 % for (A), Fig. V.4a.

Free cyanide remained aplenty (>70%) at reaction ends for all cases. As for the previous

sulfide minerals, it is thus reasonable to assume galena increased Au recovery in the

presence of sphalerite by avoiding the formation of a passivating sulfur film on Au surfaces

and by activating the anodic behavior of gold. The - galvanic association negatively

affected gold leaching, B > C, Fig.V. 4a. Since sphalerite OCP in aerated cyanide solution

(-140 mV vs. Ag/AgCl) is lower than galena’s (-110 mV vs. Ag/AgCl), interpretations as

proposed for the galena/chalcopyrite system are equally similar.

The measured galvanic current between sphalerite and gold varied from very low (5 µA)

without galena to relatively high with a galena intercalated layer (>130 µA). The high

galvanic current between Au and sphalerite measured in the presence of galena (dashed

line) contrasts with the low electrical conductivity of sphalerite (Mirnezami et al., 2003).

162

Therefore, in addition to the activation of gold surface, it can be speculated that the galena-

leached lead (II) would also be responsible for the formation of a low band gap conducting

layer on the sphalerite surface that will promote its cathodic behavior. Consequently, the

electron-acceptor ability of sphalerite particles would be enhanced with oxygen reduction

occurring preferentially at the active lead sites. This complex aspect was never addressed in

the open literature and further investigations might be needed in the future.

Fig. V.4c shows the speciation of sulfur for Fig. V.3a cases (B) and (C). The concentration

of SO42-

and SCN- significantly increased in the presence of galena (filled symbols vs.

empty symbols). Hydrosulfide and polysulfides were the intermediate oxidation products of

sphalerite found responsible for the passivating effect of this mineral (Azizi et al., 2011). It

is plausible that their heterogeneous oxidation into SO42-

and SCN- -thus explaining these

are present in larger concentrations- on the surface of galena could promote higher gold

recovery as observed in Fig. V.4a.

0

20

40

60

80

100

0 2 4 6 8 10 12 14 16

% A

u L

ea

ch

ed






//Silica + Au

// //Silica + Au

+ //Silica + Au

Silica + Au

(a)

163

0

40

80

120

160

0 20 40 60 80 100 120 140 160 180

I (µ

A)

Time (min)


MRI-4//silica-Au//Silica + Au

// //Silica + Au

(b)

0

10

20

30

40

50

60

70

0 5 10 15 20 25 30

Lea

ched

sp

ecie

s (m

g/L

)


MRI-4//Silica+Au+Ag

MRI-4//Silica+Au+Ag



SO42-

SCN-

SO42-

//Silica + Au//Silica + Au// //Silica + Au// //Silica + Au

(c)

Figure V-4. Effect of galena ( ) in the presence of sphalerite ( ) on the evolution of: (a)



164


Chalcocite

Unlike the other tested sulfides (Figs. V.2a, V.3a, V.4a), the negative effect of chalcocite

on Au recovery was barely remedied by the use of galena (Fig. V.5a). Only case (B)

showed some very low dissolution of gold, while for cases (A) and (C) gold cyanidation

was completely blocked. - galvanic association had a detrimental effect on gold leaching,

B > C, Fig. V.5a. Low galvanic currents were generally registered (Fig. V.5b) either with

galena (dashed line) or without (solid line). However, intercalating a galena layer resulted

in lower galvanic current (absolute value) between gold and chalcocite with respect to the

test without galena. Chalcocite and galena OCP were respectively, -720 mV (vs. Ag/AgCl)

and -110 mV (vs. Ag/AgCl). It can be fairly assumed that a strong galvanic cell was formed

between the two minerals with galena acting as a cathode hence totally blocking lead (II)

leaching and strongly prompting chalcocite leaching.

Fig. V.5c shows that galena strongly influenced the evolution of the sulfur-bearing species

in the leach solution. Large amounts of HS- (empty squares) were detected during the early

reaction stages in the absence of galena, whereas in its presence no HS- was detected and

important amounts of SO32-

(filled diamonds) were generated after 5 min of reaction. In

addition, in the presence of galena a new unassigned species was formed at the beginning

of the reaction and then quickly disappeared after 15 min of reaction (filled triangles). This

species was detected by capillary electrophoresis and based on its separation

characteristics; its CE peak was previously assigned to polysulfides, Sx2-

(Petre and

Larachi, 2006). Investigations by Breuer et al., (2008) in synthetic solutions containing

sulfide ions, lead (II) and free cyanide showed that the oxidation of sulfide in presence of

lead proceeds with the formation of polysulfides as intermediate species. Therefore, based

on the results from Fig. V.5c, it can be suggested that galena and/or dissolved lead(II) act as

catalysts for sulfide oxidation to polysulfides, explaining both the absence of HS- and the

formation of Sx2-

in the presence of galena.

Finally, the sudden increase of galvanic current measured in the absence of galena (solid

line) after 40 min of reaction (Fig. V.5b) was previously interpreted as resulting from the

165

complete oxidation of HS- on the chalcocite surface (Azizi et al., 2011). The results from

the present study also confirm this interpretation, as this peak intensity (Fig. V.5b) was

reduced significantly by the presence of galena. This suggests that the formation of HS-/Sx

2-

is the main factor causing losses on gold recovery in presence of chalcocite and that galena

improves Au recovery by removing these ions from solution via precipitation (likely as

PbS) and/or via oxidation on its the surface.

0

5

10

15

20

25

0

0.2

0.4

0.6

0.8

1

0 2 4 6 8 10 12 14

% A

u l

ea

ched

% A

u l

ea

ched






//Silica + Au

// //Silica + Au

+ //Silica + Au

Silica + Au

(a)

166

-12

-10

-8

-6

-4

-2

0

2

4

0 20 40 60 80 100 120 140 160 180

I (µ

A)

Time (min)


MRI-5//silica-Au//Silica + Au

// //Silica + Au

(b)

0

5

10

15

20

25

30

35

40

0

10

20

30

40

50

60

0 5 10 15 20 25 30 35

Pea

k a

rea

SO

32

-le

ach

ed (

mg

/L)


MRI-5//Silica+Au+Ag



HS-

Sx2-

SO32-

//Silica + Au

// //Silica + Au

// //Silica + Au

(c)

HS

- ,

Sx

2-(a

rb.

un

its)

Figure V-5. Effect of galena ( ) in the presence of chalcocite ( ) on the evolution of: (a)



167

V.3.5. Effect of Alkaline pre-Oxidation of Pyrite, Chalcopyrite and

Sphalerite on Gold Leaching

As shown in Figs. V.6-8, the following pre-oxidation strategies were also tested with the

mono-sulfide X//silica PBER bilayer systems: XAu//silica; X//silicaAu; X(preOx)Au//silica;

(XAu//silica)(preOx); (X//silicaAu)(preOx). The syntax is self-explaining regarding the

location of gold powder and the Au-X galvanic interactions. The first two systems

exemplify, respectively, the sulfide-associated and free gold bilayer PBER without pre-

oxidation. In the third, the sulfide was pre-oxidized individually according to the above

described pre-oxidation (section V.2.4) and then subsequently homogenized with gold

powder (to enable galvanic interactions) before cyanidation. The fourth system corresponds

to an idle silica layer juxtaposed to pre-mixed sulfide-gold where galvanic interactions were

pre-existent to pre-oxidation. The last corresponds to a system with juxtaposed sulfide layer

and gold-silica homogenized mixture. This system featuring free gold leaching underwent

in-situ pre-oxidation before cyanidation in the absence of Au-X galvanic interactions.

The bilayer //silica system was the one that gave the best gold leaching response to pre-

oxidation (Figs. V.6-8). Fig. V.6 indicates that in absence of Au-pyrite galvanic

interactions, pre-oxidation of pyrite improved gold leaching (empty vs. filled triangles, Fig.

V.6). Conversely, galvanic interactions preexisting to pre-oxidation degraded gold

dissolution with respect to Au-pyrite galvanic interactions without pre-oxidation (empty vs.

filled squares, Fig. V.6). Blocking the pyrite surface by deposition of iron hydroxide

(FeOOH) during pre-oxidation (Zhu et al., 1993 and references therein) could explain the

negative effect of pre-oxidation on gold leaching for the Au- system (empty vs. filled

squares, Fig. V.6). Such Fe (III) hydroxide protective layers forming on the mineral

surfaces would impair the positive Au-pyrite galvanic interactions which are known to play

an important role in accelerating gold leaching kinetics (Azizi et al., 2011). To confirm this

contention, pyrite was pre-oxidized individually according to section V.2.4 pre-oxidation

procedure. After pre-treatment, pre-oxidized pyrite was removed from the PBER,

desiccated at 60oC overnight and used in a subsequent cyanidation experiment. The results

in Fig. V.6 (empty circles vs. filled squares) are consistent with FeOOH layer responsible

168

for prohibiting formation of the good Au-pyrite contacts. Interestingly, when targeting the

leaching of free gold (buried in the inert silica layer) pyrite pre-oxidation improves gold

leaching via pyrite surface passivation (empty vs. filled triangles, Fig. V.6).

Chalcopyrite pre-oxidation had no significant effect on gold recovery numbers (Fig. V.7)

regardless of whether Au- galvanic interactions were enabled or disabled. In case of

sphalerite (Fig. V.8), pre-oxidation was counter-productive for leveraging gold leaching in

the presence of Au- galvanic interactions, while in absence of such galvanic interactions

no significant effect was noticed.

Sphalerite was also shown to be oxidized in alkaline media (Peters, 1976) to form zinc and

iron hydroxides and various sulfur-containing species. Therefore, possible accumulation of

hydroxides at the surface of sphalerite particles could also explain the negative effect on

gold leaching of pre-oxidation of the gold-sphalerite system (empty vs. filled squares, Fig.

V.8).

The results from this study tell that the success of pre-oxidation strategies requires a precise

knowledge of gold deportment. As a result, it can be argued that different strategies for

oxidative pre-treatments should be determined not only based on the mineralogy of the

sulfide ore but also on the occurrence and associations of exposed gold. Therefore, pre-

oxidation may improve gold recovery in finely ground ores rich with free gold, while when

exposed gold is intimately associated with conductive sulfide minerals, e.g., pyrite, pre-

oxidation will be unsuitable to improve gold leaching.

The effect of pre-oxidation of mixed mineral systems containing pyrite-chalcopyrite and

pyrite-sphalerite on free gold leaching was also investigated. As illustrated in Fig. V.9, pre-

oxidation induced noticeable retarding effects on gold leaching. Gold recovery after 10 min

of cyanidation decreased from 6.7 % without pre-oxidation to 2.8 % with pre-oxidation for

pyrite-chalcopyrite system (filled vs. empty diamonds) and from 5.7 % to 1 % for pyrite-

sphalerite system (filled vs. empty circles). Since the measured OCP of chalcopyrite (-88

mV vs. Ag/AgCl) and sphalerite (-148 mv vs. Ag/AgCl) are lower than that of pyrite (-36

mV vs. Ag/AgCl), it can be suggested that pyrite-chalcopyrite and pyrite-sphalerite

galvanic associations led to electrochemical activation of sphalerite and chalcopyrite.

169

Consequently, more reactive chalcopyrite and sphalerite during the pre-oxidation period

will favor larger accumulation of oxidation products (i.e., Zn(OH)2, Cu(OH)2, S) on free

gold particles (trapped in the silica layer) yielding therefore lower gold leaching during

subsequent cyanidation. The measured galvanic current (~ 10µA) for the pyrite-

chalcopyrite pair confirmed the presence of active galvanic interactions between the two,

while for the pyrite-sphalerite system a very low galvanic current (~1 µA) was detected.

This was probably due to the lower sphalerite conductivity (Mirnezami et al., 2003), even if

the presence of important galvanic effects was previously demonstrated (Cruz et al., 2005)

for a naturally associated pyrite-sphalerite system.

The above results point out that pre-oxidation of individual sulfide minerals went from

being beneficial for gold leaching for pyrite (Fig. V.6), to effectless for chalcopyrite (Fig.

V.7) and sphalerite (Fig. V.8), and to harmful within mixed sulfide-sulfide mineral systems

(Fig. V.9). For actual plant operation, pre-oxidative strategies based on mono-sulfide

mineral phases are not reliable for predicting the efficiency of pre-oxidative treatment on

cyanidation. This is due to the many synergetic and anti-synergetic galvanic interactions

manifesting between the different mineralogical phases present in the ore.

0

20

40

60

80

100

0 2 4 6 8 10 12 14 16 18 20

% A

u l

each

ed


MRI-2+Au+Ag//silica

(MRI-2;Ox)+Au+Ag//silica

MRI-2+Au+Ag//silica; Ox

MRI-2//silica+Au+Ag; Ox

MRI-2//silica+Au+Ag

+ Au//Silica

(preOx) + Au//Silica

( + Au//Silica)preOx

( //Silica + Au)preOx

//Silica + Au

170

Figure V-6. Effect of pre-oxidation on gold leaching in presence of pyrite. Full symbols:

tests without pre-oxidation; empty symbols: tests with pre-oxidation. Reaction conditions:

CN- = 30 mM, DO2 = 0.25 mM, pH = 11.

0

5

10

15

20

25

0 2 4 6 8 10 12 14 16 18

% A

u l

each

ed


MRI-3+Au+Ag//silica


MRI-3//slica+Au+Ag; Ox

MRI-3//silica+Au+Ag

+ Au//Silica



//Silica + Au

Figure V-7. Effect of pre-oxidation on gold leaching in presence of chalcopyrite. Full



0

5

10

15

20

25

30

0 2 4 6 8 10 12 14 16 18 20

% A

u l

each

ed


MRI-4+Au+Ag//silica


MRI-4//silica+Au+Ag

MRI-4//silica+Au+Ag; Ox

+ Au//Silica


//Silica + Au


171

Figure V-8. Effect of pre-oxidation on gold leaching in presence of sphalerite. Full



0

1

2

3

4

5

6

7

0 2 4 6 8 10 12

% A

u l

each

ed


MRI-2+MRI-3//silica+Au+Ag

MRI-2+MRI-4//silica+Au+Ag

MRI-2+MRI-3//silica+Au+Ag; Ox

MRI-2+MRI-4//Silica+Au+Ag; Ox

+ //Silica + Au

+ //Silica + Au

( + //Silica + Au) preOx

( + //Silica + Au)preOx

Figure V-9. Effect of pre-oxidation of mixed mineral systems containing pyrite and

chalcopyrite or pyrite and sphalerite on free gold leaching. Full symbols: tests without pre-

oxidation; empty symbols: tests with pre-oxidation. Reaction conditions: CN- = 30 mM,

DO2 = 0.25 mM, pH = 11.

V.3.6. Combined pre-Oxidation and Lead Nitrate Addition on Gold

Leaching in Presence of Pyrite

Fig. V.10a shows that when pre-oxidation is employed, adding lead nitrate during the

cyanidation step completely canceled the positive effects of pre-oxidation for gold

dissolution (filled triangles vs. filled squares). In addition, when lead nitrate is added during

cyanidation, using or not pre-oxidation had similar effects on gold recovery (Fig. V.10a,

filled squares vs. crosses). On the contrary, when lead nitrate was added during pre-

oxidation, a noticeable increase of gold leaching was observed when compared to the test

when pre-oxidation proceeded with no lead nitrate addition (filled circles vs. filled

triangles). The results also shows that choosing the optimum lead addition strategy is very

important, as after 19 min of cyanidation 4 times more Au was recovered when lead nitrate

was added during pre-oxidation as compared to the test when it was added during

172

cyanidation. Moreover, Fig. V.10a shows that all employed strategies successfully

increased gold recovery with pre-oxidation being slightly more efficient than lead nitrate

addition.

Speciation of the reaction products released into the leach solution was also examined.

When pre-oxidation was performed without lead nitrate (Fig. V.10b), important amounts of

S2O32-

(up to 45 mg/L) and SCN- (up to 25 mg/L) were generated from pyrite dissolution.

On the contrary, neither S2O32-

nor SCN- was detected when lead nitrate was added during

pre-oxidation. Fig. V.10b also shows that without lead (II), mainly S2O32-

formed during

the pre-oxidation stage and that no further increase of its concentration was observed

during cyanidation (filled squares). Pyrite pre-oxidation in alkaline solutions is believed to

lead to accumulation of large polysulfide concentrations as intermediates (Azizi et al.,

2010). Therefore, it is possible that part of polysulfides generated during pyrite pre-

oxidation were further oxidized into S2O32-

(filled squares), while the remaining part

reacted with free cyanide to form thiocyanate once cyanide was added at the end of the pre-

oxidation period (filled triangles). These explanations are consistent with previous findings

(Hewitt et al., 2009) which indicated that in cyanide-free solutions long polysulfide chains

react with oxygen to form mainly thiosulfate, while in the presence of cyanide thiocyanate

is the major reaction product. On the contrary, when lead nitrate was added during pre-

oxidation, polysulfides rapidly react with lead (II) to form lead sulfide precipitate (Hedley

and Tabachnick, 1968; Breuer et al., 2008), preventing thus the formation of both

thiosulfate and thiocyanate (Fig. V.10b, empty triangles and squares). These explanations

are also coherent with the positive effect on gold leaching of lead nitrate addition during

pre-oxidation (Fig. V.10a, filled circles vs. filled triangles), as a possible chemisorption of

polysulfides on gold particles (Lustemberg et al., 2008) could entrain Au surface

passivation. However, as comparable profiles for sulfur speciation were obtained when lead

nitrate was added during cyanidation with (Fig. V.10b, empty symbols) or without pre-

oxidation (data not shown), further investigations are required to understand the adverse

effects on the efficiency of pre-oxidation observed when lead nitrate was added during

cyanidation.

173

0

5

10

15

20

25

30

0 2 4 6 8 10 12 14 16 18 20

% A

u l

ea

ch

ed


Preoxidation+Pb followed by cyanidation

Preoxidation followed by cyanidation

Preoxidation followed by cyanidation+Pb

Direct cyanidation+Pb

Direct cyanidation

(a)

0

5

10

15

20

25

30

35

40

45

50

0 5 10 15 20

Lea

ch

ed

sp

ecie

s (m

g/L

)


(S2O32-);Ox

(SCN-);Ox

(S2O32- and SCN-); Ox+Pb

(b)

Figure V-10 Effect of lead nitrate addition strategy in presence of pyrite, on: (a) gold

leaching and (b) pyrite oxidation products speciation. Reaction conditions: CN- = 30 mM,

DO2 = 0.25 mM, 5 mg/L Pb-Pb(NO3)2, pH = 11.

174

V.4. Concluding remarks

Investigations aiming at improving gold leaching during cyanidation in the presence of

various sulfide minerals were conducted. A conjunction of electrochemical and chemical

speciation studies using a PBER was employed to provide valuable information about the

mechanisms and reactions involved.

Addition of galena, as a source of lead, was shown to bring important positive effects on

gold recovery with pyrite, chalcopyrite and sphalerite. Not only the detrimental effect of

passivation was neutralized but also an important acceleration of Au dissolution rate was

prompted. The mechanisms involving galena are believed to be similar to those proposed in

the literature for lead nitrate. The effects of galena-other sulfide mineral galvanic

associations on gold leaching showed galena-pyrite galvanic interactions induced a positive

effect on gold leaching, whereas significant negative effects were observed when galena

was successively associated with chalcopyrite, sphalerite and chalcocite.

In absence of galvanic interactions between gold and sulfide minerals, pre-oxidation was an

effective tool to improve gold leaching in the presence of pyrite while, no beneficial effect

was observed in the case of chalcopyrite and sphalerite. However, when galvanic

interactions were enabled, the influence of pre-oxidation on gold dissolution went from

having no effect for chalcopyrite to being detrimental for pyrite. Furthermore, pre-oxidation

of mixed mineral systems of pyrite-chalcopyrite and pyrite-sphalerite showed that enabling

galvanic interactions between the associated minerals during pre-oxidation controlled gold

leaching in the subsequent cyanidation process.

When pre-oxidation was used in the presence of pyrite, lead nitrate addition during pre-

oxidation resulted in a significant increase of gold leaching, whereas its addition during

subsequent cyanidation step canceled the positive effects of pre-oxidation for gold

dissolution.

175

V.5. References




Azizi, A., Petre, C.F., Olsen, C., Larachi, F., 2011. Untangling galvanic and passivation

phenomena induced by sulfide minerals on precious metal leaching using a new packed-

bed electrochemical cyanidation reactor. Hydrometallurgy, 107, 101-111.

Breuer, P.L., Jeffrey, M.I., Hewitt, D.M., 2008. Mechanisms of sulfide ion oxidation during

cyanidation. Part I: The effect of lead (II) ions. Minerals Engineering 21, 579-586.

Budruk-Abhijeet, S., Balasubramaniam, R., Gupta, M., 2008. Corrosion behaviour of Mg–

Cu and Mg–Mo composites in 3.5% NaCl. Corrosion Science 50, 2423-2428.






Deschenes, G., Lastra, R., Brown, J.R., Jin, S., May, O., Ghali, E., 2000. Effect of lead

nitrate on cyanidation of gold ores: progress on the study of the mechanisms. Minerals


Deschenes, G., 2005. Advances in the cyanidation of gold. In: Adams, M.D. (Ed.),

Developments in mineral processing. Elsevier, Amsterdam 15, 479-499.







Hedley, N., Tabachnick, H., 1968. Mineral Dressing Notes No 23, Chemistry of

Cyanidation. American Cyanamid Company, New Jersey, USA.

Hewitt, D.M., P.L. Breuer, P.L., Jeffrey, M.I., F. Naim, F., 2009. Mechanisms of sulfide

ion oxidation during cyanidation. Part II: Surface catalysis by pyrite, Minerals Engineering

22, 1166-1172.



3276.



Lustemberg, P.G., Vericat, C., Benitez, G.A., Vela, M.E., Tognalli, N., Fainstein, A.,

Martiarena, M.L., Salvarezza, R.C., 2008. Spontaneously formed sulfur adlayers on gold in



http://www.sciencedirect.com/science/article/B6VDR-4WPS9HX-1/2/8bde789d63c0139f06252675a98cd9a2

http://www.sciencedirect.com/science/article/B6VDR-4WPS9HX-1/2/8bde789d63c0139f06252675a98cd9a2

176

electrolyte solutions: Adsorbed sulfur or gold sulfide? The Journal of Physical Chemistry

C. 112, 11394-11402.

Mirnezami, M., Hashemi, M.S., Finch, J.A., 2003. Measurement of conductivity of

sulphide particles dispersed in water. Canadian Metallurgical Quarterly 42, 271-276.

Peters, E., 1976. Direct leaching of sulfides: Chemistry and applications. Metallurgical

Transactions B 7, 505-517.

Petre, C.F., Larachi, F., 2006. Capillary electrophoretic separation of inorganic sulfur –

Sulfide, polysulfides and sulfur-oxygen species. Journal of Separation Science, 29, 144-

152.




Senanayake, G., 2008. A review of effects of silver, lead, sulfide and carbonaceous matter

on gold cyanidation and mechanistic interpretation. Hydrometallurgy 90, 46-73.




Zhu, X., Li, J., Wastworth, M.E., 1993. Kinetics of transpassive oxidation of pyrite. In:

Hager, J.P. (Ed.), EPD Congress. The Minerals, Metals & Materials Society 355-368.

177

CONCLUSION GENERALE

Une veille bibliographique détaillée nous a permis de conclure que le caractère réfractaire

de l’or exposé durant le traitement des minerais aurifères sulfureux par cyanuration serait

imputable, sans que les mécanismes ne soient clairement élucidés, aux phénomènes de

passivation (PP) de surface et d’interactions galvaniques (IG), qui impliqueraient, tout au

long de la cyanuration, un éventail de processus redox complexes empêchant une

dissolution poussée du métal visé. Par ailleurs, la forte réactivité des différents minéraux

sulfureux dans les solutions aérées de cyanure entrainent une surconsommation de cyanure

et d’oxygène en raison de la lixiviation inopportune de leurs métaux de base comme le

cuivre, le fer et le zinc...etc. Parallèlement à la lixiviation des métaux nuisibles, la

cogénération de sulfures est soupçonnée d’influencer directement la dissolution de l’or à

travers la formation d’une couche protectrice à la surface du métal.

Pour ce qui est de l’avancement de la partie expérimentale du présent projet, les faits

saillants suivant ont été mis en exergue :

Une étude visant à évaluer l’importance relative sur la dissolution de l’or des phénomènes

de passivation et d’interactions galvaniques a été réalisée. Pour se faire, des électrodes d’or

(en alliage Au/Ag) et minérales, préparées à partir d’un minerai industriel de ses sulfures

majoritaires, ont été utilisées, et les expériences de cyanuration ont été conduites sur un

montage électrochimique à une et à deux cellules. Les résultats de cette étude montrent que

les phénomènes de passivation réduisent l’oxycyanolixiviation de l’or alors que les

interactions galvaniques Au-sulfure présentent globalement un effet positif sur la

dissolution du métal d’intérêt.

Pour comprendre le comportement des minerais sulfureux dans les solutions de cyanure,

une étude systématique des effets du même minerai industriel et de ses sulfures métalliques

prédominants, scrutés individuellement, sur la cinétique de lixiviation de l’or a été réalisée

en mode RDE/slurry. Les résultats de cette étude montrent que tous les sulfures étudiés ont

un effet inhibiteur sur la vitesse de dissolution de l’or. L’analyse des lixiviats des tests de

cyanuration par électrophorèse capillaire nous a permis de nous rendre compte que la

178

chalcopyrite et la sphalérite, bien que minoritaires dans le minerai industriel testé,

majoritairement constituée de la pyrite, semblent influencer fortement sa réactivité et par

conséquent la vitesse de dissolution de l’or. Ceci a été confirmé quand les mêmes sulfures

ont fait l’objet de tests de pré-oxydation avant les expériences de cyanuration. Cette étude a

permis de montrer que la pré-oxydation améliore nettement la vitesse de dissolution de l’or

en présence du minerai industriel, cependant, elle n’a aucun effet en présence de ses

constituants, testé individuellement.

Pour une meilleure évaluation des phénomènes de passivation et d’interactions galvaniques

dans des conditions de contact simulant le minerai industriel ainsi que pour une

compréhension pratique des mécanismes de dissolution de systèmes multi-minéraux

complexes, où les différents sulfures constitutifs sont naturellement associés, nous étions

amenés à penser à une autre stratégie expérimentale. Cette stratégie consiste en une

disposition Au-sulfure et sulfure-sulfure en lit fixe dans un réacteur tubulaire. Une telle

disposition reflète de façon plus réaliste la topochimie de surface des minerais industriels

en termes de contacts galvaniques à l’échelle intra-particulaire. Ainsi, un réacteur

électrochimique à lit fixe (PBER) a été mis au point pour des tests de cyanuration

originaux.

Le PBER a été utilisé pour évaluer la contribution individuelle des phénomènes de

passivation et d’interactions galvaniques sur la dissolution de l’or et de l’argent (PM) en

présence d’une large gamme de sulfures métalliques (pyrite, chalcopyrite, sphalérite,

chalcocite, galène, stibine et un minerai industriel). Les résultats obtenus montrent que la

dissolution de l’or est plutôt contrôlée par les IG Au-sulfure. Ces dernières ne se contentent

pas seulement de neutraliser l’effet négatif de la passivation, mais entraînent une sur-

lixiviation de l’or par rapport aux tests témoins (en absence de sulfure), surtout en présence

de la pyrite, de la chalcopyrite et du minerai industriel. Les courants galvaniques mesurés

aux interfaces Au-sulfures confirment les résultats obtenus en termes de lixiviation de l’or.

Corollairement, l’effet des IG Au-sulfure s’illustre bel et bien également sur la réactivité de

certains minéraux sulfureux (i.e., pyrite, chalcopyrite et la galène), qui se retrouvent

cathodiquement protégés, en partie, par les poudres de métaux précieux (Au & Ag).

179

L’effet des associations galvaniques Au-sulfure et sulfure-sulfure sur la lixiviation de l’or à

partir de plusieurs systèmes multi-minéraux a été également investigué en utilisant le même

réacteur (PBER). L’étude de systèmes synthétiques de type pyrite-chalcopyrite-quartz puis

pyrite-sphalérite-quartz, électriquement-séparés, a montré que la vitesse de dissolution de

l’or est surtout contrôlée par les IG entre les particules d’or et celle de la phase minérale

que les renferme. Cependant, lorsque la chalcopyrite ou la sphalérite ont été substituées par

la chalcocite, la forte réactivité de cette dernière en présence du cyanure inflige une baisse

draconienne de la dissolution anodique de l’or. Des investigations faites sur des systèmes

mixtes de type pyrite/chalcopyrite, pyrite/sphalérite et pyrite/chalcocite,

galvaniquement/associés, ont montré que la réactivité de l’or est fortement influencée par

les IG multifactorielles entre les particules d’or et celles des différentes phases minérales du

mélange. Par ailleurs, les IG sulfure-sulfure se sont révélées infructueuses pour la

lixiviation de l’or libre (non associé aux sulfures conducteurs) dans le cas de mélanges

pyrite/chalcopyrite et pyrite/chalcocite, cependant, elles ont un effet bénéfique dans le cas

du mélange pyrite/sphalérite.

Afin de mettre en évidence l’effet des rapports de surface anode/cathode à l’échelle inter-

particulaire sur la dissolution de l’or, plusieurs rapports massiques ont été étudié dans le cas

du système mixte préparé à partir des poudres de métaux précieux (Au & Ag, PM) et

minérales (pyrite & chalcocite). Les résultats de cette étude ont montré que la dissolution

de l’or est contrôlée aussi bien par le rapport PM/pyrite que par celui pyrite/chalcocite.

Pour remédier à l’impact inhibiteur d’une série de sulfures métalliques sur la lixiviation de

l’or, plusieurs traitements ont été mis en œuvre sur le PBER :

Premièrement, l’utilisation de la galène, comme source de plomb, a permis de montrer que

cette dernière améliore nettement la vitesse de dissolution de l’or en présence, surtout, de la

pyrite, de la chalcopyrite et de la sphalérite. Les IG galène-sulfure ont un effet promoteur

sur la dissolution de l’or en présence de la pyrite, alors qu’en présence de la chalcopyrite,

de la sphalérite ou de la chalcocite elles se sont montrées défavorables.

Deuxièmement, des tests de pré-oxydation ont été réalisés en présence et en absence des IG

Au-sulfure et sulfure-sulfure. Cette étude a permis démontrer que l’efficacité de la pré-

180

oxydation dépend strictement du statut minéralogique des particules d’or et minérales

(pyrite, chalcopyrite, sphalérite). À titre d’exemple, la pré-oxydation a abouti à une baisse

de la dissolution de l’or en présence des IG Au-pyrite, alors qu’en absence de ces dernières

la corrosion du métal d’intérêt a été améliorée. Par ailleurs, la pré-oxydation de mélanges

homogénéisés de type pyrite/chalcopyrite et pyrite/sphalérite, galvaniquement associés, a

montré que le rôle des contacts galvaniques inter-particulaires entre les sulfures associés est

loin d’être simple. Ces derniers donnent lieu à des flux d’électrons par contact direct solide-

solide, durant la période de pré-oxydation, influençant l’importance des dépôts, résultants

de l’oxydation des sulfures, à la surface des particules d’or. De tels dépôts entravent l’accès

aux réactifs (CN- & O2) à la surface active des particules d’or pendant l’étape de

cyanuration et par conséquent, ralentissent leur dissolution.

Finalement, il a été également démontré dans cette étude que l’efficacité du nitrate de

plomb pour pallier aux phénomènes de passivation dépend étroitement des conditions de

son utilisation. En effet, lors de test de pré-oxydation de la pyrite, l’ajout du nitrate de

plomb pendant la période de pré-oxydation a été bénéfique pour la lixiviation de l’or,

contrairement à son ajout pendant la période de cyanuration qui a été révélé infructueux.

PERSPECTIVES

Les résultats trouvés au niveau de la partie portant sur l’effet de la pré-oxydation de

sulfures individuels, de leurs mélanges mixtes ainsi que du minerai industriel (mélange

natif des mêmes sulfures) sur la récupération de l’or, nous ont permis de démontrer que la

réactivité de ces derniers dans les solutions de cyanure est fortement influencée par les

contacts galvaniques permanents entre les différents sulfures. Par conséquent, il serait très

intéressant de tenir compte des rapports de surface sulfure/sulfure dans des conditions qui

seront les plus possible représentatives des abondances des différents sulfures dans le

minerai industriel. Ceci permettrait d’élaborer une stratégie de pré-oxydation tenant

compte des spécificités minéralogiques du minerai industriel en question.

Il serait également intéressant d’élaborer un modèle statistique descriptif en mesure

d’interpoler le comportement de minerais industriels multi-minéraux (lixiviation de l’or et

de cyanicides, consommation de cyanure, etc.). En effet, la mise en œuvre de tests de

181

lixiviation de l’or en présence de mélanges synthétiques de plusieurs sulfures

galvaniquement associés, au moyen du lit fixe, dans des rapports plus réalistes est très utile.

L’analyse des ions résultant de la dissolution des minerais, jumelée à l’information qui sera

obtenue de la caractérisation des dépôts sur la surface d’or, par microscopie électronique à

balayage (SEM) et par spectroscopie des photoélectrons (XPS), permettraient, nous

l’espérons, de contribuer à une interprétation plus élaborée des mécanismes de

développement de la couche de passivation à la surface de l’or en tenant compte des effets

des interactions possibles entre les minéraux sulfureux sur leurs réactivités respectives.

Ainsi on pourrait commencer à dégager des règles de représentativité du réacteur

électrochimique à lit fixe (PBER) à émuler un véritable minerai industriel.

Des essais menés durant les dernières semaines de cette thèse ont montré que l’utilisation

des complexes tricyanocupreux (Cu(CN)32-

) et teracyanozincates (Zn(CN)42-

) pour la

lixiviation de l’or pourraient faire baisser la lixiviation du cuivre et du zinc à partir des

minerais sulfureux tout en gardant des taux de récupération d’or acceptables. De ce fait, il

serait crucial de mener des études permettant de contribuer tant au plan de la

compréhension fondamentale des mécanismes de la cyanuration à partir des complexes

cyanométalliques qu’au plan de la mise au point de nouvelles stratégies de cyanuration

permettant d’améliorer les conditions de dissolution préférentielle de l’or au détriment des

métaux indésirables. Autrement dit, de telles stratégies seraient en mesure d’améliorer la

sélectivité de la cyanuration et de réduire la consommation de cyanure.

Vu la robustesse du protocole développé par électrophorèse capillaire, il serait également

intéressant d’élargir le champ d’application de la technique pour couvrir d’autre éléments

(i.e., Sx2-

, Zn, Co, Ni, As, Sb, CN-). Ceci permettrait d’aider pour une meilleure

compréhension des réactions des différents minéraux sulfureux, problématiques pendant

l’extraction de l’or par cyanuration.

182

APPENDIX A. Capillary electrophoretic analysis of

sulfur and cyanicides speciation during cyanidation of

gold complex sulfidic ores

183

184

185

186

187

188

189

190

191

APPENDIX B. Untangling galvanic and passivation

phenomena induced by sulfide minerals on precious

metal leaching using a new packed-bed electrochemical

cyanidation reactor - Supplementary Data SD-1

Abdelaaziz Azizi, Catalin Florin Petre, Caroline Olsen, Faïçal Larachi

2.2.2 Electrochemical campaign

Potentiostat

ZRA

V

A

4

1

1

2

3

6

5

78

9

11

1012

Figure SD-1-1 Scheme of the ZRA electrochemical cell within the PBER. Legend as

follows: 1, insulated platinum wire; 2, platinum spring; 3, MRI packed bed; 4, insulator

sintered-glass disc filter; 5, quartz packed bed; 6, Au-Ag rod; 7, peristaltic pump; 8,

magnetic stirrer; 9, oxygen probe electrode; 10, air bubbling system; 11, pH-meter

electrode; 12, reference electrode.

3.1 Effect of Pyrite (MRI-2) on Gold and Silver Leaching

Fig. SD-1-2 shows the evolution of released sulfur-bearing anions as a function of time-on-

stream. The results indicate that the presence of gold and silver in direct contact with pyrite

play an important role in defining the reactivity of the latter in aerated cyanide solutions.

As PMs act as electron donors, the observed decrease in SO42-

and SCN- concentrations in

the case B, as opposed to the case A, could also be explained by the effect of galvanic

192

protection offered by these metals to pyrite which acts as an electron acceptor thus

exhibiting less reactivity.

0

2

4

6

8

10

0

100

200

300

400

500

600

0 2 4 6 8 10 12 14 16 18 20

SC

N-

lea

ch

ed

(m

g/L

)

SO

42

-le

ach

ed

(m

g/L

)


Case B

Case A

Case B

Fig. SD-1-2 Effect of galvanic interactions between gold and MRI-2 on: (a) SO4

2- and (b)

SCN- formation. Reaction conditions: CN

- = 30 mM, DO2 = 0.25 mM, pH = 11.

3.2 Effect of Chalcopyrite (MRI-3) on Gold and Silver Leaching

0

20

40

60

80

100

120

140

0 2 4 6 8 10 12 14 16 18 20 22 24

SO

42

-le

ach

ed

(m

g/L

)


Case A

Case B

Figure SD-1-3a Effect of galvanic interactions between gold and MRI-3 on SO4

2-

formation. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11.

193

0

20

40

60

80

100

0

20

40

60

80

100

0 2 4 6 8 10 12 14 16 18 20 22 24

SC

N-

lea

ch

ed

(m

g/L

)

Cu

(CN

) 32

-le

ach

ed

(m

g/L

)


Case A

Case B

Case A Case B

Figure SD-1-3b Effect of galvanic interactions between gold and MRI-3 on cyanicides

(SCN- and Cu(CN)3

2-) formation. Reaction conditions: CN

- = 30 mM, DO2 = 0.25 mM, pH

= 11.

Figs. SD-1-4a,b show electrophoregrams of ions speciation in the conditions

corresponding to the PM conversion profiles of Figs. 3a. Iron-cyanide complexes are

present as ferrocyanide (Fe(CN)64-

) when Au-chalcopyrite galvanic interactions (GI) are

enabled (case B), Fig. SD-1-4a. Conversely, disabling Au-chalcopyrite GI (case A) gave

rise to ferricyanide (Fe(CN)63-

), Fig. SD-1-4b. Au-chalcopyrite galvanic interactions

promote gold leaching (Fig. 3a). This translates into a more pronounced consumption of

dissolved oxygen in case B than in case A. It could be possible that ferricyanide, resulting

from chalcopyrite oxidation, may turn into the electron acceptor ferrocyanide to

compensate for the GI-promoted demand on dissolved oxygen to help gold anodic

dissolution and/or further oxidation of chalcopyrite. These observations seem to be

coherent with those of Xie and Dreisinger (2007) who reported that ferricyanide is a

powerful oxidant that can effectively dissolve silver sulfide in cyanide solutions.

Xie, F., and Dreisinger, D.B., 2007. Leaching of silver sulfide with ferricyanide-cyanide

solution. Hydrometallurgy 88, 98-108.

194

0

10

20

30

40

50

60

2.5 3 3.5 4

Ab

so

rp

tio

n (

mA

U)

Migration time (min)

Fe(CN)64-

SO42-

Cu(CN)32-

SCN-

a

0

20

40

60

80

100

120

2.5 3 3.5 4 4.5 5

Ab

so

rp

tio

n (

mA

U)

Migration time (min)

Fe(CN)63-

Cu(CN)32-

SO42- SCN-

b

Figs. SD-1-4a,b Effect of galvanic interactions between gold and MRI-3 on the Fe-cyanide

complexes speciation. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11. a) GI

enabled, i.e., case B, b) GI disabled, i.e., case A.

195

APPENDIX C. PBER: time-on-stream vs. physical time

For all the experiments carried out with the PBER, the aerated cyanide solution feed was


At different time intervals, samples of the solution were taken from the magnetically-stirred

glass reactor and used to measure the concentration of dissolved metals, dissolved sulfur

and cyanicides. For each sampling time interval (physical time), the corresponding

residence time of the aerated cyanide solution in the gold containing layer (time on-

stream) was computed after the bed porosity was determined for every powder loading in

the PBER and the prevailing cyanide solution volumetric flow rate. This latter was kept

constant at 10.4 mL/min for all the experiments. Consequently, The time on stream against

which all the concentration profiles of this study (chapters III. IV and V) were plotted

corresponded to the time the aerated cyanide solution sojourned inside the PBER excluding

the remainder of the transit time of flight in the external loop of the liquid circuit.

The following formula was used to calculate the time-on-stream:

(1)

TOS(ST) = TOS/pass x N(ST) (2)

Where:

TOS/pass: time on-stream (per-pass residence time of the cyanide solution in the packed

bed)

V: PBER volume, mL

Q: volumetric flow rate (mL/min)

TOS(ST) : time on-stream at each sampling time interval

N(ST): solution’s turnover in the PBER at each sampling time interval

For the sake of clarity some figures from chapters III. IV and V were duplicated in the

present appendix (see below).

196

III.3.1. Effect of Pyrite (MRI-2) on Gold and Silver Leaching

0

10

20

30

40

50

60

70

80

0 2 4 6 8 10 12 14 16 18 20

% A

u l

ea

ch

ed


Au; case B

Ag; case C

Au; case C

Ag; case A

Au; case A

Ag; case B

(a)

Au; case B

Ag; case C

Au; case C

Ag; case A

Au; case A

Ag; case B

(a`)

Figure III-2 Effect of pyrite (MRI-2) on gold and silver dissolution: (a) as a function of

time-on-stream; (a`) as a function of the physical time. Reaction conditions: CN- = 30

mM, DO2 = 0.25 mM, pH = 11.

197

IV.3.1. Gold Cyanidation and the Pyrite-Chalcopyrite-Silica System

0

10

20

30

40

50

60

70

80

0 2 4 6 8 10 12 14 16

% A

u

lea

ched






0

10

20

30

40

50

60

70

80

0 2 4 6 8 10 12 14 16

% A

u l

each

ed






(a)

(b)

Silica

+ Au

(B)

Silica + Au(C)

Silica + Au

(D)

Silica

+ Au

(A)

Silica

+ + Au

(A’)

Silica + Au(B’)

Silica + Au

(D)

Silica

+ Au

(A)

(C)

(B’)

+



within silica (C,D); (b) electrically-connected pyrite-chalcopyrite//silica systems: Au within

pyrite-chalcopyrite (A’), Au within silica (B’). Inset: Effect of pyrite-chalcopyrite galvanic

association on free-gold leaching from inert silica layer. Gold leaching as a function of

time-on-stream. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH = 11.

198

0

10

20

30

40

50

60

70

80

0 50 100 150 200 250 300 350 400

% A

u l

ea

ch

ed

Physical time (min)

case A

case B

case C

case D

0

10

20

30

40

50

60

70

80

0 50 100 150 200 250 300 350 400

% A

u

lea

ch

ed

Physical time (min)

case A

case A`

case D

case B`

(a’)

(b’)

Silica

+ Au

(B)

Silica + Au(C)

Silica + Au

(D)

Silica

+ Au

(A)

Silica

+ + Au

(A’)

Silica + Au(B’)

Silica + Au

(D)

Silica

+ Au

(A)

+

Figure IV-2. Effect on gold leaching of Au distribution within (a`) electrically-disconnected


within silica (C,D); (b`) electrically-connected pyrite-chalcopyrite//silica systems: Au

within pyrite-chalcopyrite (A’), Au within silica (B’). Inset: Effect of pyrite-chalcopyrite

galvanic association on free-gold leaching from inert silica layer. Gold leaching as a

function of the physical time. Reaction conditions: CN- = 30 mM, DO2 = 0.25 mM, pH =

11.

199

V.3.1. Effect of Galena on the Leaching of Free Gold in the Presence of Pyrite

0

20

40

60

80

100

0 2 4 6 8 10 12 14 16

% A

u l

ea

ch

ed






//Silica + Au

// //Silica + Au

+ //Silica + Au

Silica + Au

(a)

0

20

40

60

80

100

0 50 100 150 200 250 300 350 400

% A

u l

ea

ch

ed

Physical time (min)

case A:

case B:

case C:

case D:

//Silica + Au

// //Silica + Au

+ //Silica + Au

Silica + Au

(a`)

Figure V-2. Effect of galena ( ) in the presence of pyrite ( ) on free gold dissolution. Gold

leaching (a) as a function of time-on-stream; (a`) as a function of the physical time (a`).


Documents

SULFIDIC ORE MIXED/MULTILAYER FIXED BEDS